|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
160.62.4.10
I have been thinking for a long time about what design/measurements that are readily available to use as selection criteria for an amplifier.
I have distilled the ideas down into these guidelines. However, the final judgement is always in the listening, I am afterall and empirical scientist and don't rely on what people "know" to be good sound. Keep in mind that I have reached these guidelines by an empirical approach. When I heard an amp that sounded REALLY good I went and found as much info on the design and/or measurements as possible. After doing this for a long time I now have the following rules of thumb:
1) The amp must be Class A or high bias AB (around 15-20 watts A)
2) The amp must have little to no negative feedback or at least no global feedback
3) The amp must have a large, low impedance power supply
4) If the amp is tube output it must have LARGE output transformers that do not easily saturate, or be an OTL without much negative feedback
5) It should be simple in design, having only 2 or 3 total stages from input to output
6) The distortion should increase linearly (or nearly so) with power (also indicates a low feedback design)
7) The distortion should remain constant as a function of frequency, i.e. no rise at high frequencies as is often seen in SS and some tube amps and no rise in bass frequencies as is often seen in tube amps. This distortion is often far worse, harmonically, than the 1Khz typical measurent.
8) The damping factor should remain constant as a function of frequency
9) The harmonic distortion spectrum (usually taken at 1Khz) should be monotonic with a significant drop in level with increasing order and harmonics higher than about 5th should be down in the noise floor of the FFT.
10) Power supply ripple should be minimized so that distortion products do not produce intermodulation distortion products with the power supply harmonics. This results in a "dirty" noise floor.
11) IMD products should be low order like the harmonic distortion products.
If an amp meets most or all of these criteria then it probably has a good chance of being an excellent sounding amplifier. If it fails most then it will likely be mediocre at best and often quite "hifi" sounding rather than musical.
Again, this is an empirical observation and is open to reinterpretation if an amplifier that fails most of these criteria actually sounds really good to me. I haven't found it so far...
Follow Ups:
morricab, further our earlier discussion, and since you have more experience interpreting this sort of test result, I wonder what you make of these frequency spectra. They are the the same for the same amp, first in stereo mode, second bridged. The first especially has spurious peaks at various frequencies, though atlow-levels.
They have a quite different profile from the Nelson Pass sample result I referenced in our earlier discussion, (repeated below), and wonder what we can infer from that.
~~~~~~~~~~~
... 99%'er
Is it just me, or does this type of post just suck all of the fun out of listening to music? There was a time when I would pour over the graphs printed in every issue of Stereophile, pretending that I could figure out exactly what made for a good component. It's like chasing down the Holy Grail, and you're not Indiana Jones.Find an amp that makes you want to listen to MORE music, and then stop looking at more amps.
Edits: 07/05/12
"Find an amp that makes you want to listen to MORE music, and then stop looking at more amps."
Well, that was kind of the point. If you have found that then read other posts or leave the forum altogther. If you haven't and want a guide to which amps might be that then this post is for you. I know its one man's opinion and its not complete, which is why I state "potentially" good sounding amp on the post.
"Find an amp that makes you want to listen to MORE music, and then stop looking at more amps."
Something very important is missing from the above.
Find what makes you happy, then:
- BLATHER ABOUT IT ALL YOU WANT ON ANY FORUM AND AT ANY SHOW BUT DON'T PUT OTHERS DOWN OR ATTEMPT TO DRAW CONCLUSIONS ABOUT THEM FOR THE CHOICES THEY MAKE.
That kind of behavior has been common practice for pushy, obnoxious salespeople who are hell bent on making someone unload a pile of cash on snake oil and lies. It shouldn't be tolerated by anyone whether they just broke into the hobby or have vast experience. Insulting people for the equipment they own is bad for the hobby and in the final analysis doesn't help the industry sell more product. It just drives a lot of normal people away towards their Iphone and ear buds.
I guess it begs the question, are you a music lover, or an equipment lover? It's not meant as an insult, but a legitimate question. I can understand, to a degree, a fascination with negative feedback, circuit design, etc. Some of these amps are achingly beautiful to look at, and built to a standard that would make NASA proud.
Still, if an amp has all of the qualities that you treasure but sound sterile, uninvolving, and flat, what difference would it make? I think we both know of amps that fit that description. At that point, doesn't that put your list of criteria into question? If an amp looks like crap, measures like crap, but sounds beautiful, what's more important?
I'm sure you'll just dismiss me as some sort of troll looking for a fight, but I really just want to know if there really is an amp out there that can make you happy long term. An equipment lover, by definition, can never really be satisfied as long as amp designers keep coming up with the latest and greatest.
I love gear that gets me closer to a realistic sound of music. I go to lots of live, unamplified concerts and want that as close as possible at home, even if it isn't at realistic levels always.
"I think we both know of amps that fit that description. At that point, doesn't that put your list of criteria into question?"
Look again at my criteria and I think the question answers itself. The point is that I STARTED with what sounded good and then found out how they measured.
Ever spend 5 minutes listening to or having to make small talk with a typical "audiophile" at an audio trade show? It's like finger nails scratching a black board. They can suck the fun out of anything. That's not to say all audiophiles are this way. Some professionals/audiophiles like Gary Koh, Andrew Jones, and Jeff Joseph are not only intelligent, but entertaining and fun to be around.
The hyper focus on negative feedback and "tube sound" is a total red herring. I was listening to Kenwood and Accuphase high speed amps way back in the day when Dr. Ottala's work was still the "holy grail" of audio. In the real world, amps that have a flat response out to 500 khz are useless. The claimed audible improvements, if they even exist, are slight. What is far more problematic is the instability and lack of output current capability that wide bandwidth designs create. Moving coil loudspeakers (the vast majority in existence) are low impedance current driven devices that are typically not compatible with tube voltage amplifiers that possess high output impedance. Wide variations in a speaker's load impedance will always produce audible anomalies in the net response of a high output impedance amp/loudspeaker combination. And no matter what tube fanatics say, that is not a good thing. Their suggestion that speakers which possess wide load fluctuations are not "properly designed" is ludicrous.
Moderate use of negative feedback in a good Class AB design has no audible signature. And because transistors typically have much higher gain bandwidth products than tubes, some feedback is necessary. Threads like this are full of uninformed blanket statements that have little or no real world value. If negative feedback induced "time smear" were such a catastrophic problem, Kenwood would still be making and selling a version of its L07MkII - an amp I once owned that experienced wide band melt down. While it was a clean, noise free amp - I could not tell an audible difference between it and the Class AB Threshold and Crown that ultimately replaced it.
But what is moderate? As I understand, it's demonstrable that "moderate" of say 10 dB), reduces 2nd and 3rd harmonics moderately but introduces higher order harmonics. However substantial feedback, say 40 dB, substantially reduces ALL orders harmonics including the higher order ones.Here's a graph curtesy of Nelson Pass (of all people, and advocate for low feedback) that illustrates this point ...
In a similar concept, global feedback consistently reduces distortion across all stages of amplification. The "time smear" argument is invalid because the feedback propagation is typically a small number of micro seconds that put delays orders of magnitude above the audio frequency band.
~~~~~~~~~~~
... 99%'er
Edits: 07/01/12
What you have to remember is that this very old plot, which originally comes from Baxandall, doesn't take into account orders higher than the 6th nor IM distortion. Why is that important? It is those two things that are the most detrimental to the sound. It is relatively easy to find FFT plots today of amps with a 1Khz tone (by far the simplest example) that have harmonics all the way to 20kHz!! And that is with just one frequency. Now, factor in all the IM distortion as well and you begin to get a nasty picture like in Nelson Pass's white paper with large concetrated peaks of distortion.
It is also interesting to look at an amp that has a very clean FFT plot but then to notice that the noise floor is 10 to 100x higher than another amp that has more discrete peaks. This is your high feedback amp most likely that has suppressed all of the lower order harmonics but has created a signal modulated "noise" floor that is really not noise(see the work of Norman Crowhearst for this one). This artificial "noise" floor is signal correlated and this means that it is not true random noise and will obscure low level signals, which has the effect of truncating instrument and room decay and soundstage. This is also partly why a good no feedback amp does the whole sense of space/soundstage thing better than most feedback amps.
... but a poor case, (or no case at all), for low to moderate feedback.
Referring to the aforementioned graph, (below), note that higher order harmonics enter the picture progressively, 6th at about 15 db, and presumably higher mostly by 20 dB. Thereafter all orders decline. Intermodular distortion is, of course, related to the levels of harmonic distortion. Sorry, but I can't help suspecting that there is some level at which distortion becomes inaudible.
So his following graph looks UGLY but note that the highest feedback is 15 dB -- just about the the peak for 6th order distortion per the preceding graph! And even in this here the 4th, 5th, and 6th seem to peak at 10 dB.
Pass avoids the logical conclusion that higher feedback might work better than low or moderate feedback. Pass does suggest that a lot of feedback might require additional amplification stages but doesn't deal with the likely positive effect of global feedback vs. local.
As for space/soundstaging, transparency, and resolution, my own (albeit limited) experience is that my Class D amps are better any Class AB amps I've owned, including the high-bias, low feedback Monarchy SM-70 Pro.
BTW, I'm not privileged to listen to really high-end equipment in my system, much less own it. So I have to work out what's best for me in cheap equipment. And for me resolution, transparency, and accurate instrument timbres trump warmth, "harmonic richness", and the echoy false ambience you get with cheap (or not so cheap) tube equipment.
~~~~~~~~~~~
... 99%'er
You are looking at these graphs without really thinking about them. The first plot you show does make it look like in fact that adding more and more feedback will lower everything. However, take a look at where the "knee" is for each harmonic and you will see that it is shifting further along the x-axis (the amount of feedback). This means as you keep adding more you CREATE higher and higher harmonics, which is BAD. With enough feedback you have made quite a lot of higher harmonics that extend throughout the entire audible range (and that is just the harmonic distortion nevermind the various IM components). Also, the practical reality is that there is a limit to how much feedback you can add to a circuit and it remain stable.
You can see this very clearly in the second plot you posted. The 3db of feedback has peaks out to 6th harmonic but the 15db now has peaks out to at least 10th (its off screen but the 10db feedback has a peak at 10Khz) and the 15db 7th harmonic is much bigger than the 10db, which is much bigger than the 6db etc.
What was gained in the process? A 10db drop in the 2nd harmonic and that is it. So you traded the most harmless harmonic peak for many not so harmless peaks.
You guys have to learn that 2nd and 3rd harmonics are not an issue even up to several percent, its the other crap that is the problem. Up to at least a couple % 2nd harmonic is INAUDIBLE even with pure tones let alone real music.
Norman Crowhurst wrote about this in the late 50s, which at that time meant tube amplifiers, and he basically concluded from the math that with high feedback you will end up with a signal correlated "noise" floor that is in fact an infinite number of distortion peaks. It differs from true noise because it IS correlated with the signal and so rises and falls with the signal level.
It contributes strongly to loss of low level resolution information as this "noise" floor masks the low level signals. True noise to a large extent will not do this because it is not correlated to the music signal (this matters to your brain as it "extracts" the signal out of the noise based on this correlation). That is why an old recording with a lot of tape noise still can have superb low level information in the highs and you can "tune it out" and hear down to very low levels even though the noise is louder. With a correlated low level "noise" the information gets burried.
The only positive effect of negative feedback is for static measurements and even then only the most basic ones like THD and IMD. Once you put an FFT on it you can easily see the damage. Adding more gain stages only makes the problem WORSE not better. Adding non-linearity is not a way to gain linearity unless you think it will all be straightened out by the feedback, which we see just mulitplies the problems by shifting the harmonic distortion and Pass shows this as well, CONCENTRATING the distortion to some very nasty and unpredictable peaks.
What you say about the class D vs. the Monarchy may be true but perhaps you were just pushing the little Monarchy too hard? I am sure the tone of the Class D amps is not as good, even though I wouldn't call the Monarchy world class in that regard. Never forget though that the Monarchy SM70 uses an opamp as an input stage, which is a high feedback device. I have come to loath the sound of opamps.
Ultimately it is whether a person can hear certain types distortion AND whether that person prefers one type over another.
There is just as much evidence that the "tonally rich" sound of tube or zero-feedback class A designs caused by 2nd order HD as there is that the "sterile" sound of feedback s/s designs is caused by higher order HD.
You assert that 2% of 2nd order HD is inaudible but that .01% of 4th or higher is, (even if only as noise floor). It's acknowledged that some level of 2nd order adds an agreeable "warmth" or "tonal richness" to the sound. Your own argument that a level of distortion might be undetectable in a sine wave but apparent with complex signals might apply as much to 2nd order HD as it does to higher order.
Designers willing to use liberal feedback don't argue that it is a panacea; linear devices and a reasonable number of gain stages are everyone's ideal. However it's doubtful that low, local feedback is better than more, global feedback.
I'm skeptical that the Monarchy's short-fall can be entirely attributed to its use of opamps. In fact I think opamps get a worse rap than they deserve. My DAC uses OPA2314's and an OPA2604 and is very transparent and free of grain and etch.
~~~~~~~~~~~
... 99%'er
"There is just as much evidence that the "tonally rich" sound of tube or zero-feedback class A designs caused by 2nd order HD"
Such as??? References please...
" Your own argument that a level of distortion might be undetectable in a sine wave but apparent with complex signals might apply as much to 2nd order HD as it does to higher order."
Your comment tells me you haven't understood the concept of masking if you had you would realize that low order distortion is different because of masking and high order distortion, especially at lower levels, is not masked.
"Designers willing to use liberal feedback don't argue that it is a panacea; linear devices and a reasonable number of gain stages are everyone's ideal. However it's doubtful that low, local feedback is better than more, global feedback."
And you know this how? By making a survey of designers?
"In fact I think opamps get a worse rap than they deserve. My DAC uses OPA2314's and an OPA2604 and is very transparent and free of grain and etch."
And I think that you have no idea what freedom from grain and etch is. I also think that the opamp has proabably done more damage to the music industry and hifi than you can imagine.
Look up the article by Russel O. Hamm on microphone preamps and the switch from tubes to transistors and opamps (they categorize them separately because of their performance difference to discrete transistors). VERY interesting and its no longer so surprising that many of my best sounding recordings predate the widespread use of transistors.
I'm being earnest: I am substantially deaf above 10 kHz. That might make a difference in my evaluations. Despite involvement in hi-fi for 40 years, I know I haven't heard the range of equipment that you have, especially high end. On these accounts I recognized that I don't have the credibility that you do.Morricab said, " You can see this very clearly in the second plot you posted. The 3db of feedback has peaks out to 6th harmonic but the 15db now has peaks out to at least 10th (its off screen but the 10db feedback has a peak at 10Khz) and the 15db 7th harmonic is much bigger than the 10db, which is much bigger than the 6db etc.
I can't see what's off the graph, but I can see that the higher order harmonics are progressively lower in amplitude and that all are lower in amplitude with higher feedback. The HOs become signal-tracking noise: not good, but at some point so small as to be irrelevant
Morricab adds further, " You guys have to learn that 2nd and 3rd harmonics are not an issue even up to several percent, its the other crap that is the problem. Up to at least a couple % 2nd harmonic is INAUDIBLE even with pure tones let alone real music.
... Your comment tells me you haven't understood the concept of masking if you had you would realize that low order distortion is different because of masking and high order distortion, especially at lower levels, is not masked. "But so I'm clear, are you saying the 2nd order HD is masked? Even at 2,3,4,5% distortion? Or rather, is it the 2nd order that's doing the masking? There is nothing new about the assertion that 2nd HD is pleasant sounding. I continue to suspect that 2nd order HD masks the grain and etch that some s/s amps undoubtedly have: in my mind this could the explanation for the popularity of tube preamps -- and even tube buffers -- used ahead of s/s amps.
~~~~~~~~~~~
... 99%'er
Edits: 07/06/12 07/06/12
"I can't see what's off the graph, but I can see that the higher order harmonics are progressively lower in amplitude and that all are lower in amplitude with higher feedback. The HOs become signal-tracking noise: not good, but at some point so small as to be irrelevant"
It's easy enough to see the trend in the plots, no? It's clear there will be a 7th, 8th etc. etc. as the feedback increases because that is how the first 6 harmonics behaved. It won't stop there.
Ah, but how small is small and at what listening level?? Read Cheever for an idea about the audibility and then read the Pass paper to see how small small is. Hint, its a lot more complicated than those simple graphs you found.
No, 2nd order is masked and so is the third to some extent. The higher the level of the fundamental the wider the masking. Its because your hearing mechanism makes it's own distortion was it's all vibrating. Geddes and CHeever both show that its level dependent with masking the lowest at low levels, which means your ear is going to be most sensitive to extraneous distortion (like crossover distortion) at low levels. Crowhurst shows that a signal correlated and modulated "noise" floor is generated with high feedback that will mask low level signal information, ruining low level resolution. You can easily hear many db below a true noise floor, such as tape hiss, because it is not correlated with the signal.
Of course a very high level of 2nd order will tend to make things sound more open and brassy but as long as it doesn't exceed the masking threshold it should be inaudible. When you don't hear the distortion it will always sound more pleasant than audible distortion.
There is an interesting trial done by Keith Howard that you can find in Stereophile where he simulated distortion patterns with software modified music. He found that NO additional distortion sounded better than ANY additional distortion. However, he found that a monotonic or only even order harmonic pattern sounded LESS distorted than one with a lot of odd and high order harmonics. I suggest you read that one too.
Now, obviously if you can't hear what's wrong then live long and prosper but you found a reason to prefer your Class D. All I can say is that I haven't found one yet that I can live with (even though I have tried to live with three different ones and heard at least a dozen others at length) and they most definitely don't fit my criteria.
Meanwhile my class D sounds better than what I've heard before. Albeit it's a different sound than, say, the Monarchys.
In the range of equipment I can afford to listen to, the class D has the cleanest, airiest sound I've heard. That's all I can say -- except that with live music instruments and voices cans sound strident: violins squeaky, brass shrill, etc. It's valid that recorded music might sound that way too. On the other hand I feel there is a lot of bad recording practice. My theory is that there's too much close-mic'ing in a misguided or inept attempt to capture all instrument timbres. However the sound you hear at typical audience distances is not what you hear a couple of feet from the instrument.
So like I said, on the best recording my latest class D amp sounds best; while with some crappy recordings the Monarchy was better precisely because it masked the accurate recorded sound in aid of a "warmer" presentation.
~~~~~~~~~~~
... 99%'er
Not really a good idea, you might miss an outstanding performer.
Never attribute to malice that which can be adequately explained by stupidity.
These criteria are not biased towards a device or topology. However, certain topologies will never result in the desired measurements that I think usually result in at least good sound if not great sound.
You want minimal crossover distortion in class B amps, class AB amps show worse crossover distortion which can be audible.
Sure higher bias sounds better, and you may not leave class A operation depeding on your listening habits, but.............
IME and IMO, there is no comparision between class A and class AB or B, Class A amps do sound superior.
You also want a high PSRR overall as well as for each individual stage. This can make or break an amp's performance often making power supply quality less of an issue.
You want good open loop performance, meaning gain and frequency response. Feedback should be used as/if/when needed and should not be shunned, IMO.
I would chose lateral mosfets as output devices. In my limited desgin/building expereience they do sound superior to BJTs, all things not being equal.
Never attribute to malice that which can be adequately explained by stupidity.
"You want minimal crossover distortion in class B amps, class AB amps show worse crossover distortion which can be audible."
This usually shows up in the FFT at 1Khz as high order products.
"IME and IMO, there is no comparision between class A and class AB or B, Class A amps do sound superior"
Generally, I agree but there are some important exceptions.
"You also want a high PSRR overall as well as for each individual stage"
This you can also see in an FFT as the power supply harmonics will show up there and it also let's you see if there is IM distortion products between the signal harmonics and the power supply harmonics. Amps with almost no power supply noise also have very clean FFT spectra(at least with regard to the IM distortion).
"You want good open loop performance, meaning gain and frequency response. Feedback should be used as/if/when needed and should not be shunned, IMO."
Some of the amps I have heard that sounded really good had SOME feedback but I don't know of one that I heard that I really likeed that had a lot.
"I would chose lateral mosfets as output devices. In my limited desgin/building expereience they do sound superior to BJTs, all things not being equal."
I have heard both sound good so but I would say that BJTs are the least linear amplifying devices. However, you need to run them both as much Class A as possible (tubes too for that matter).
"You want good open loop performance, meaning gain and frequency response."
CONGRATULATIONS! ! !
You 'get it'. Few do...
Rick
... and test results at least as a starting point.
Your criteria will eliminate a lot of amps early in the process. I think I've heard a few good amps that would have gone unheard.
~~~~~~~~~~~
... 99%'er
No, I did not say look at specifications, measurements yes, specs no not really. Although if the spec for distortion is extremely low then I would say beware and the same if damping factor is given. Obviously, information about the design would help to judge (e.g. class A or not, negative feedback or not, power supply design etc.).
nt
~~~~~~~~~~~
... 99%'er
Not really.
Specifications are numbers; frequency response range (+-), input sensitivity, output impedance, THD, IMD, power (RMS), power (peak) etc.
Design description is just that; transistor or tube (or both), Class A, AB etc., one stage, two, three or more, direct coupled etc. etc. etc.
Given you extensive discussion of test results in this thread, what you're saying is that published specifications are simplistic, not the objective measurements are irrelevant.
I applaud your efforts to connect measurements with underlying design, e.g. with use of feedback. I'm not so sure I agree with your "all feedback" is bad philosophy. You know, of course, the some designers believe that global feedback is better than local, and that fairly large amount of feedback is better than a little. They point out that their approach results in lower distortion of all types, including high order .
You like a certain sound -- no arguing with preferences -- but that's subjective. You like Monarchy for affordable equipment. But I swapped by a pair of SM-70 Pro's for a Class D Audio SDS-258. The latter is a different sound to be sure -- and better to my ear. (More resolved, transparent, grain-free, and dynamic.)
~~~~~~~~~~~
... 99%'er
Well, I have never made any attempt to hide that these are criteria that I think are necessary to have a chance for really good sound. Again, I use it as a starting point for deciding to audition something or not and this comes from a lot of negative listening experiences for which I have then found measurements of the device in question. Since there is quite a bit of literature out there with measurements (thanks to Stereophile and Soundstage) and I have heard a lot of amps both really good and really bad I began to connect the dots to see where it led.
If the results had led me to a conclusion that the absolute level of distortion was important but harmonics not and that a rising distortion with frequency results in good sound and that a flat distortion with power increase is best then I would obviously posted different criteria. Alas, the devices that provided the best sound could be argued often to have the worst meausrements from an absolute distortion perspective. Why? I think it is because our hearing is not an oscilloscope or even FFT plot and that the way distortion is perceived is complex.
Nevertheless, it seems that if the distortion is predominantly low order and in a distinct decreasing pattern and that distortion is constant regardless of frequency then there is a good chance it will sound at least good. Once I decided this (and this was actually observed by the French reviewer Jean Hiraga 25 years ago or so) from looking at LOTS of curves of amps I knew pretty well, I think started to look at how they were designed.
Tube, transistor, Class A or AB, hybrid, no feedback, no global feedback, big power supplies etc. That is where my criteria and the kind of amp that makes them comes from. The Monarchy SM70 is unlikely to meet all of these criteria but the SE160 and SE250 appear to meet most of the criteria and sound darn good. The SM70 sounds, ok, especially for the price.
Regardless of what you say about how you like your class D I haven't found one yet that sounds to my liking and I have owned 3 different ones and listened to at least a dozen others. I find them to sound quite unnatural from the mids up and the measurements show that they have problems up there. If they sounded really good I think I would have to find out what made them sound that way but so far its consistent with my hearing and measurement observations. I don't expect them to sound good now when I look at a new ones measurements and see the same old same old from the numbers.
I recently heard the newest Nuforce reference monos on a pair of highly revealing Piega planar/hybrid speakers (2nd to top model) and using a top of the line Moon cd player and the top Pass preamp. The sound was terrible. Simply terrible. The amp made the speaker sound completely disintegrated between the planar and the bass drivers and the mid/highs were thin pinched and forward. No tone in the strings of violins and everything a bit steely. Not nice. The Piegas gave the amps nowhere to hide.
My friends also like Monarchy Audio products. Two of them now have the Monarchy Audio NM24 (I have the original M24) DAC with DIP. It has gone toe-to-toe with the Resolution Audio Cantata and done quite well and sounds substantially better to me than the DCS Elgar.
You are relating design and measured results to the sound you prefer -- that's a good thing in itself -- and very handy if you get to listen to a lot of equipment (which, unfortunately, I don't).
My Monarchy SM-70 Pros sounded better than the Class D Audio with some recordings (about 30%), but they happen to be the worst recordings (based on my long-standing evaluation of particular recordings). The Class D Audio is better on 70% AND they happen to be the best recordings.
The Monarchys were better than my former Bel Canto eVo2i (Tripath) amp for 70% of recordings but they were still the less-good recordings .
I am not looking for amps that sugar-coat my recordings.
~~~~~~~~~~~
... 99%'er
If you are a careful listener of live, unamplified music, and attempt to match what you have heard live at home then the preference is only a small part of the details but some gear is fundamentally closer to allowing the real thing through than others. In other words, two somewhat different sounding amps can both be GOOD and capture the basic Gestalt of live music. However, most amps do not capture this and as a result are just different flavors of noise generators.
.
Perhaps a good idea to have criteria for potentially good sounding speakers along with the amp.
Cheers
Bill
Ok, but I will have to work hard on this because it is harder to generalize as I have heard successful designs coming from radically different designs/technologies. I would have to have criteria for types of speakers I think.
I am not sure since the desired end result, whatever mode of design is used for the speaker, is good sound or lets say the musicality. Like saying a very well designed class A amp is better than a very well designed Class AB amp, perhaps we can conclude that a well designed sealed speaker is better than a well designed ported design. Well, there is great scope for discussion on the Speaker forum if you post there.
Cheers
Bill
I will start then by saying that for ultimate transparency, truth in timbre and most accurate reproduction of the input waveform (within dynamic limits) you cannot beat a good electrostatic or planar magnetic speaker. What they lack; slam, headroom, sometimes bass etc. are things that are often not really material to great sound. For some full-range models there is no better coherence and faithfulness to the recording to be found in all the most important ways. The delineation of soundstages is equally impressive.
THat said, there are also extrememly good conventional and conventional speakers and hybrids.
The best all-around pure ESLs I have heard were from Infinity, Genesis and Wisdom Audio.
I am in total agreement to your conclusions.
Regards
Bill
I totally disagree, I have no interest whatsoever in tube amps , planar & stat speakers or subwoofers.
Edits: 06/22/12 06/23/12
maybe you should change your name from disbeliever to disinterested???
I am very interested in musically accurate hi-rez multi-channel sound. Do not know of any recording studios who use antique tubes, planar or stat speakers for monitoring.
Edits: 06/23/12 06/23/12
I know of some that use tubes for making the recordings at least. Don't know about the monitoring but I know that Phillips classics used to use Audiostatic electrostatic speakers for monitoring.
" used to" are the operative words, name one major label that still uses antique tubes.
Edits: 06/25/12
Why would I care what "major" labels are doing? That would be like saying Bose makes the best speakers because they are a "major" label.
So you are not interested in Hi-res.mch SACD from all the major labels. which minor labels are still using antique tubes for recording ?
Edits: 06/25/12
Tacet and Water Lilly (at least they were until recently).
I'm curious...how would Pass lab amps, say the x250.5 or the like, fare with your criteria?
Not sure about the newer X250.5 but here are the measurements from the X1000 monos.
http://www.stereophile.com/content/pass-labs-x1000-monoblock-power-amplifier-measurements
IMO, they fail in a few ways. 1) there is a large rise in distortion at higher frequencies. 2)Distortion vs. power first goes down with increasing power, which indicates a moderate amount of feedback. 3)The distortion measured at 50Hz still shows quite a few harmonic components (mostly odd harmonics) and that doesn't account for the HF distortion rise. 4)The IM distortion has quite a few harmonics but this might not matter much
The Pass Aleph 1.2 has a similar problem with distortion vs. frequency but I don't think the FFT in this one is fair becuase it was done at 201W and this is about clipping for this amp. I don't think TJN really knows what to measure to be honest.
http://www.stereophile.com/content/pass-aleph-12-monoblock-power-amplifier-measurements
Bizzarely, the XA30.5 has a very bad distortion vs. frequency plot.
http://www.stereophile.com/content/pass-labs-xa305-power-amplifier-measurements
I haven't heard this amp but I would bet that it has not the best sounding highs despite being class A.
The 1Khz FFT shows a number of peaks and a lot of power supply related IM products with the signal (all that fuzz in the "noise" floor). The poor IMD results are largely due to the distortion vs. frequency problem.
Now, I have heard the original X250 and thought that is was a quite good amp...for SS with some negative feedback and not class A but that it can be bettered by a number of other competitors. I remember a review in Bound for Sound where the X350, which sounded worse to me than the X250, was not so favorably compared to a pair of Monarchy Audio SE250 monos, which sound very good to me and fit most of my criteria.
So, in answer, I think the X250 is an ok amplifier but not really special and I see that its brothers (none of which i have been terribly impressed with) fail a few criteria but that doesn't necessarily mean a fail sonically. Just means its less likely to be stellar.
http://www.stereophile.com/content/ayre-mx-r-monoblock-power-amplifier-measurements
Seems to meet all your requirements ... with no tubes to boot.
Ok, I had a look at the Ayre and my initial impression is that while it does some things right it does others wrong and as a result will also be a bit of mixed bag sonically (note: I have not heard this amp from Ayre but I have heard others and while not bad, I would not own one).
1) The distortion vs. frequeny at 8 ohms shows a modest rise at high frqeuencies but is better into lower impedance but higher across the board, which may be masking the rise. At least it doesn't rise even more into 4 or 2 ohms.
2)They claim zero feedback but the distortion vs. power curve suggests that there is at least some global feedback because it is far too flat with power. A good example of a no feedback curve can be found here:
http://www.soundstagemagazine.com/measurements/monarchy_se160/
See Chart 2 and chart 3 and 5. This is how a no feedback amp measures, although I would like to see less power supply garbage in Chart 5 as it is making the baseline a bit "fuzzy".
Also, the Lamm ML 1.1:
http://www.soundstagemagazine.com/measurements/lamm_ml11/
See Charts 2, 3 and 5 again.
I can tell you from experience that both of these amps sound very good. The Monarchy is an amazing bargain, the Lamm is not a bargain but arguably worth the money.
3)Having some feedback has consequences then on the FFT distortion spectrum. Figure 8 of the stereophile review:
http://www.stereophile.com/content/ayre-mx-r-monoblock-power-amplifier-measurements
Does not show a monotonicity to the distortion spectrum with high order harmonics at about the same level as the 2nd and 3rd harmonic. This usually occurs when there is feedback and amps that do this often sound mediocre. You also see this in Figures 9 and 10.
Compare this with the Monarchy and Lamm curves (Chart 5)
4) The IMD is pretty good
I am sure the power supply is more than capable as well, However, there are some circuit design choices that I think led to better "objective" numbers but might lead to sonic compromise. I am surprised to see the distortion vs. power and FFT spectra looking more like a typical feedback ampifier given their claim of no feedback.
For my money I would still audition but likely choose something else.
"I know just enough to get into trouble."
Or something like that.
I believe it's, "A little knowledge can be dangerous." True, indeed.
It must be true, because I read it in The Audio Poseur!:-)
Edit: Yes, I know you would never do that!
Edits: 06/22/12
After all, they wouldn't print it if it weren't true.
Heard you moved. Austin? Nice town.
So then, are you using local feedback around each stage or only around the input and driver and leaving the output out of the loop? You get in some ways similar measurements to darTZeel, who claims to use only local feedback loops, but they have higher distortion products and they also get a pretty flat distortion vs. power plot.
might even sound good... I liked the Theta Dreadnaught that was designed by Charles Hansen. Haven't heard these new ones from Ayre though.
This is not based on my knowledge of amp design, but my knowledge of morricab's posts :-)
Sure, if I were to commission some new custom amps, I would probably order up a set of requirements similar to your list. However, I have ENJOYED music on amplifiers that fail your list, and I still think the most important thing is for an audiophile to use his experience to match up the sympatico speakers to the amp on hand and create an ancillary environment conducive to enjoyable sound, all with the type of music one listens to.
Because as far as I am concerned most amps suck, big time. Most don't remotely capture an accurate sonic portrayl.
Sounds very subjective and unsubstantiated to everyone but your ears. I'm so relieved I don't live with your misery.
The funny thing is that you think because I am super selective that I must be miserable because I can't be happy with "any old amp". Well, nothing could be further from the truth. It has been a quite enjoyable quest and now that I have found several amps that do meet these criteria (in total or mostly) then I also have choices for good sound.
"... I have found several amps that do meet these criteria (in total or mostly) ..."
Please, and these are?
I have never heard any of these except for the Atmasphere OTLs (which I liked very well driving Classic Audio Loudspeakers), though have heard of nearly all of them. Nice that not all of them are too awfully expensive (for me). Thank you again.
Anything from KR Audio
CAT JL-2
Lamm ML1.1, ML2.1, M1.1 and 1.2
Monarchy Audio SE160 and SE250
Jadis Defy 7 mkIII (don't know about the measurements but sounds good)
VAC PHI 70, 30/30, 70/70
Antique Sound Labs Hurricane DT and Cadenza DT
Einstein Light in the Dark, The Absolute Tune and Final Cut OTLs
Atmasphere OTLs
Joule Electra OTLs
Vaic 52B reference (later Ayon 52B reference)
New Audio Frontiers 845 monoblocks and Stereo 845
Wyetech labs 572 stereo amp
AcousticPlan Santor
Sphinx Project 16
Sphinx Project 14
Silvaweld OTL reference monoblocks
Audio Note UK, P4SE monos
Bat VK75 and VK200/500 (good but not great)
Edge NL Reference monos (best SS I have heard)
NAT SE1,SE2SE, Symbiosis hybrid
Blue Circle Audio BC2,BC2.1, BC2000, BC200 series
the list varies greatly, no? varies in topology, design, ss/tube/ etc. [forgetting money for the moment as that is not germane in your o.p.]
H.F.N.
Yes it does but I never said that there was ONE right topology that delivers (potentially) great sound. However; all the ones that I have heard which sound great do many things more similar than different.
that is what you may end up with.
THe bottom line in audio is proper matching. And of course auditioning a piece. Numbers and designs are great. But the end all is "how does it sound?"
Listen first
charles
You want minimal crossover distortion in class B amps, class AB amps show worse crossover distortion which can be audible.
Sure higher bias sounds better, and you may not leave class A operation depeding on your listening habits, but.............
IME and IMO, there is no comparision between class A and class AB or B, Class A amps do sound superior.
You also want a high PSRR overall as well as for each individual stage. This can make or break an amp's performance often making power supply quality less of an issue.
You want good open loop performance, meaning gain and frequency response. Feedback should be used as/if/when needed and should not be shunned, IMO.
I would chose lateral mosfets as output devices. In my limited desgin/building expereience they do sound superior to BJTs, all things not being equal.
Never attribute to malice that which can be adequately explained by stupidity.
They went belly up
D
That was kinda my point.
Never attribute to malice that which can be adequately explained by stupidity.
My criteria would be not to use speakers which require more than 20W and especially pannel speakers which are for alliens with their strange properties;) Morricab ,since you favor those strange, unnatural emiters your need for beefy transformers which have to loose something in the HF perfomance (there is no other way around) is totaly understandable ;)
Truth to be told I have no clue what my criteria would be.There is always an example which doesn't fit and plays better that the ones which fit perfectly. I listen to opera for a week than I go to a worst country club and ask my girl not to shower for three days... , how can I do with one amp ;?) b rgsd W
Well that may be true about the HF roll off, although my KR measures flat to 25Khz and the distortion at 20Hz at 10 watts is less than 1%. That is wide enough bandwidth and low enough distortion for me. The larger Kronzilla models (mine is a VA350i) do have a bit more rolloff in the highs (-2db or so at 20Khz) but mostly this is inaudible and is more than a fair tradeoff to have deep natural sounding bass and clear, uncolored mids.
I once asked two very good designers what measurement correlated best to what they heard since it wasn't the classic things being measured. The reply was one, linearity. If an amp takes in say one volt and puts out 5 volts then 2 volts in should be 10 out. The aplifier that does that most closely sounded best to them.
Great post!!! It cuts through most of the crap that has been posted here.
This spec is very easily measured with the most basic of equipement. It ain't rocket science. Linearity demonstrates that an amp will be able to amplify a signal without causing any anomalies in the frequency domain. Extremely important...more important than SNR IHO.
Linearity is also very important in speaker response too. A linear speaker will ALWAYS sound better than a non linear speaker.
Sure, that's the goal.
But the devil is how to measure deviations from linearity which is psychoacoustically important.
Linearity is just a ratio..very easily measured.
Just get an SET tube amp and a pair of Klipsch Hi-efficiency speakers and be done with it! ;)
Good SETs meet most if not all of my criteria. Not so good ones do not. Most fail with the output tranformer quality and the bass distortion (i.e. distortion vs. frequency).
Ok. Then get an SET made by either Bottlehead or by Decware.
That coupled with Klipsch Hi Eff speakers should meet any audiophile's requirements.
Horrible sounding speakers. Who cares what amp drives them.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
While I certainly do not disagree with anything you write, I think it
would be constructive to hear from experienced designers. I mention
Charles and Victor only because they are heard from here on a semi-
regular basis.
And them's the best sort! Thanks.
I'm trying to boil it down, so far I think it means that you can stand very little crossover distortion and need a lot of gain-bandwidth if you use feedback.
Regards, Rick
I have only heard a handful of good sounding amplifiers with small amounts of feedback and none sounding good with a lot of feedback.
Well done and not just because I agree.
I remember the first time I heard the Sugden A21a. This was before I ever heard a tube amplifier. It was immediately the best sounding thing in the shop but I purchased an Arcam because as a beginner I put faith in reviews over my ears. Mistake.
But then over the next several years I put some of these things together - the best amps were pure class A little to no feedback, single ended or at least class A with low feedback. As for some of the power demands it depends on the requirements of the speakers. And of course from your list you have to make certain accommodations with budget. Obviously you can't expect the same from a $2000 amp that you would expect of a $25,000 amp. But like with Sugden or many tube amps - you try to check as many off the list as you can.
The idea is that the amp have linearity without the need for feedback.
The issue then becomes, how do you drive speakers that are clearly made for amps with a very low 'output impedance' (I use the quotes for a reason, which is that the term is not entirely factual)?
What I have found is that if the speaker has these requirements of the amp then it will never sound like real music on account of such amps in and of themselves can't sound like real music because of the limitations imposed by feedback and the like.
This then requires that the speaker is able to accommodate an amplifier with a higher output impedance. This requires different box design (if it is a box) and also different crossover design. With respect to the latter, if you use the wrong amps with the wrong speakers, the crossover might be a couple of octaves off! This is why some horns will 'honk' with some transistor amps. There is more at the link.
Essentially for the most part the industry has been headed down the wrong road for about the last 45 years. What I mean by this is that the human hearing rules are pretty much ignored in favor of specs on paper. The former is why some amps sound good with "poor specs", the latter is why some amps sell better, as the specs on paper can make a bad amp look good.
Well Ralph, I have heard some amps that can drive difficult speakers and still sound great. Some of those speakers even sound great (such as the Apogee Scintilla) when driven by a great amp that can handle them.
For example, I have some friends who are Apogee nuts (I also owned Apogee) and several of them have Scintillas. The hunt was on for a great sounding amp that could handle those bad boys. What we found was this:
Sphinx Project 16 sounds great with the Scintillas. It is a 100 watt Class A hybrid beast that has adjustable power supply depending on the load.
Lamm M1.1 also drives Scintillas well and sounds good in the process
VacuumState prototype (not sure if it went commercial before Allen died). Allen brought this over to my friends and we demoed it on the Scintilla. 150 watts into 1 ohm! A tube amp designed specifically for Scintillas and sounded great.
For somewhat less demanding speakers there are a number of good hybrid amps that don't use any or only a little negative feedback. The Einstein Light in the Dark is a good example and I think its even a circoltron design. The AcousticPlan Santor is another very good sounding hybrid and drove Apogee Divas beautifully. For a low budget there is even the excellent sounding Monarchy Audio SE250. And if you have a big budget then the KR Audio Kronzilla SX will give just about anything (yes even your amps Ralph) a run for their money.
This may be going back a stretch but in one of the first reviews on any of our amplifiers the reviewer used a set of Apogee Full Ranges, which were also one ohm. We used a set of our Z-Music autoformers between the speaker and our (early) set of MA-1s and the result worked really well! IOW they were easy to drive other than their impedance.
However I should point something out. While an amplifier *may* be able to drive a difficult load, that is not the same as saying it is sounding its best when doing so- there is a reason why a difficult load is considered difficult- its harder for the amp to drive. While this logic is not rocket science by any means, it is a simple truth that lower impedances do not benefit high end audio reproduction at all, nor do difficult phase angles in the load.
With respect to impedance, its a simple fact that you will get more power, more bandwidth and lower distortion if a tube amp with an output transformer is driving a higher impedance (IOW it will sound better). What most people don't realize is that while they will not make more power, otherwise this is also true of any transistor amp. You can see it in their specs. Four ohms came in as a way to get amps to make more power since efficiency had taken a back seat to the cost of the driver itself. This is a historical situation; these days many audiophile accept four ohms and don't think much about it but there was a time when it was unusual and for good reason.
IOW four ohms has no benefit to reproduction if sound quality is your goal . If on the other hand sound pressure is your goal then there is a slight advantage if you have a transistor amp that can manage it. I've not gone into speaker cables but they are critical on four ohms where they are far less so on eight or sixteen.
These things that I have mentioned here are not only audible but also measurable. FWIW I don't agree that the Lamm you mentioned is serious competition although I am sure its a very competent amplifier. Someday maybe we should have shoot-out :)
Would you consider Einstein as serious competition to your products Ralph? I would...also their much less expensive hybrids are right up there.
They are using circuits that we pioneered- yes, they are competition but they've yet to break the price barrier. So we have amps for about 1/4 the cost of theirs that are more reliable and sound better.
Of course, they like to use negative feedback and don't use class A.
There is a sort of 'trap' with OTLs where the designer tries to make the amp do things that it really shouldn't be, in an effort to show off what the amp can do. If you just accept the fact that an OTL might have a higher 'output impedance' (I put the term in quotes as under the Voltage Paradigm, the term is misused) then the thing to do is to look at whether that is important and what are the consequences.
Using negative feedback makes the amplifier violate a fundamental rule of human hearing, IOW it exacerbates odd-ordered harmonic distortion. The violation is that the ear uses odd-orders to determine how loud a sound is, so such an amp will sound artificially louder and brighter than the actual music.
This means that any amplifier built in this way cannot sound like real music!
So if a speaker demands this of an amplifier it in turn cannot sound like real music either. Since the Einstein commits this sin, it will never sound right. That, other than the price, is the reason we beat them on a level playing field.
"Using negative feedback makes the amplifier violate a fundamental rule of human hearing, IOW it exacerbates odd-ordered harmonic distortion."
Accustic Arts AMP II (global feedback amp), the high odd-order harmonics can be clearly seen.
1kHz signal at 10W into a 4-ohm load
Dartzeel NHB-108 (zero global feedback amp), no high odd-order harmonics present.
1kHz signal at 10W into an 8-ohm load
Despite your apparent reversal of the graphs with the descriptions (assuming you are suggesting minimal harmonics with the "zero feedback" case), your graphs and your post proves absolutely nothing. Feedback does in fact improve linearity and reduce distortion. It is an effective way of sacrificing gain for linearity and bandwidth by using a form of electrical damping. Most novices don't understand the true nature of feedback and its real world impact on performance. And unfortunately, snake oil salesmen and factually challenged "reviewers" love to capitalize on this misunderstanding or knowledge deficit to gain some level of control or power.
A very readable, well thought out, sensible, and accurate description of negative feedback can be found here:
http://sound.westhost.com/articles/distortion+fb.htm
You don't have to have an electrical engineering degree or background to grasp what is being said.
Also, as noted earlier, one cannot reliably attribute a set of distortion measurement findings to the use of feedback or lack thereof without attempting to operate the device with feedback and then without - taking distortion measurements in both instances. Citing the distortion patterns of two different amps and automatically attributing the difference solely to the use of feedback is idiotic since distortion has a multitude of sources - potential feedback oscillation being only one.
I was just being ironic towards Ralph's (over-repeated) statement that feedback increases the harmonics order in the distortion spectrum. So no, I don't suggest minimal harmonics on zero feedback amps.
I've read Rod Elliott's article long time ago...
IMO has more to do with feedback than it does the devices, although I can safely say its *easier* to do it with tubes than transistors, that is not the same as saying that you can't do it with transistors. You can.
And this coming from one who arguably makes the 'most tube' amplifiers on the planet.
However, just to be clear, the differences in odd-ordered harmonic content I am talking about need only be a few 100ths of a percent to be audible.
More to the point though the big deal is understanding human hearing perceptual rules, and then building equipment that follows *those* rules rather than the rather arbitrary 'look good on paper' rules that we have now. I am sure any audiophile will agree that it is our ears that are the most important thing in audio. So I argue for a pragmatic approach where the stereo follows as closely as possible the human perceptual rules- if it does that who cares what it looks like on paper- that is secondary (I like it to look good on paper too but I won't push it if that messes with the sound, which is what feedback *generally* does).
Question for Ralph:
Does any other manufacturer make amplifiers for home audio use that have a 600V/microsecond slew rate?
Do you see this as necessary and if so, why?
We didn't set that as a goal, its just how it worked out. Speed is an indication of bandwidth. Wide bandwidth means correct phase relationships in the audio passband (20Hz-20KHz). We *were* interested in good phase relationships. The 600V/mS figure is a bit overboard but that is just how it panned out. A lot of people previously thought that such speeds were the preview of transistors, but with no output transformer, tubes are just as fast if not faster.
Ralph said:
"Speed is an indication of bandwidth. Wide bandwidth means correct phase relationships in the audio passband (20Hz-20KHz)"
Really Ralph? I never would have guessed that! LOL!
Who do you think you're talking to - another clueless audiophool?
Once again, like Morricab and many others who have a clear agenda to sell, compared to the phase problems and distortion problems (of all types) produced by speakers in the audio chain, the distortion and phase problems produced by reasonably well built amplifiers are MINISCULE. No rational, reasonable person is going to hyerfocus on "hundredth's of a percent" of distortion from an amplifier when the loudspeakers the amps are connected to are generating distortion on the order of a few percent.
The claimed audible differences between tube and solid state amps is related to the interaction between the loudspeaker's impedance with that of the amplifier. In the case of a typical tube amp, this can be significant. Extremely wide bandwidth in the context of audio is a fool's game that I learned about 25 years ago. There is no audible advantage to an amp that has a half megahertz bandwidth. And the disadvantage is huge - high susceptibility to oscillation and self destruction when components age and their values shift.
I would challenge any of the audiophools participating in this or any other thread to put their money where their big fat mouths are. If you think you can hear the third, fourth, fifth, or even up to the 15th harmonic of a 1khz signal that is 60 db down from and masked by a moderately loud fundamental, I'd like to buy you an Atma-Sphere tube amp. If you can't, you can donate all of your system to a charity of my choosing - excepting the tube amp - you can keep that.
New Audiophool Rule:
TUBE AND ANTI FEEDBACK FANATICS PUT THE PSYCHO IN PSYCHO-ACOUSTICS.
Who do you think you're talking to - another clueless audiophool?
Your demeanor is suggesting that may be the case. But maybe not...
I run a recording studio and in it is an LP mastering lathe. In setting it up, we ran it with several different amplifiers, just to see how they interact with the load, which is benign. We used the same cut and ran the same current on the head with each amplifier.
The stock amplifier is a traditional transistor circuit. Also used was a MOSFET power amplifier and a stock Dynaco ST-70 that was refurbished but otherwise stock. Of course we used a set of our own amplifiers and also a higher-power SET.
The same track was recorded with each with a lockout groove as obviously we could not cut the whole lacquer at once. Essentially what we created a recording that demonstrates the differences between the amps.
This media is literally created from the amplifiers themselves as the cutterhead is a load similar to a loudspeaker. You can clearly hear the differences between the various amplifiers on any system. IOW it does not matter the playback- the difference between each of the amplifiers is readily audible, even on a lowly HK reciever from the 1970s with cheap speakers.
So we can hear the difference between two amps that measure exactly the same on the bench in terms of bandwidth, and they do so in a predicable fashion, based on the kind of distortion they make. So yes, the slight amounts of odd ordered harmonic distortion that are characteristic of 99% of solid state amplifiers does indeed cause a harshness and a brightness. You can hear it on the lacquer. That we can hear trace amounts of this distortion is because our ears happened to be tuned for this sort of thing through evolution.
Understanding that fact is likely what separates us here; I have been looking into human hearing rules and from your demeanor it sounds like you are more focused on the specs. The issue, IOW is that the specs have no meaning if they don't also have something to do with human hearing/perceptual rules.
As the engineer at Scott put it, to paraphrase, "If the measurement is not audible you are measuring the wrong thing."
Ralph said
"from your demeanor it sounds like you are more focused on the specs."
Ralph also said earlier:
"What we are talking about here in terms of distortion is differences of only a few 100ths of a percent."
I'm focused on science - not subjective blather. If a difference can be heard, it should be measurable. When people on forums like this start proclaiming that differences are audible but can't be measured, huge red flags go up. Just as you ASSumed from "my demeanor" that I'm "focused on specs", Morricab ASSumed that I had no experience with, understanding, or appreciation for the "superiority" of tube amplification and low negative feedback amps because among many amps, I happen to own a Crown Macro Tech 2400. The root of the words assume and assumption are sometimes suggestive of the disposition or knowledge of the person making assumptions. I'm very familiar with most of the landmark studies on human hearing pertaining to localization, sensitivity to phase aberrations, sensitivity to distortion, and the like. I didn't just fall off the turnip truck.
Both you and Morricab keep avoiding the obvious elephant in the room by refusing to acknowledge that loudspeakers produce far more distortion - harmonic, amplitude, phase, intermodulation, etc... than modern amplifiers. And the claimed differences in sound from one amp to the next are practically non existent compared to the differences between the speakers they're connected to. The diversion to your lp cutterhead "experiment" is just more subterfuge.
By definition, an experiment is a scientific procedure undertaken to make a discovery, test a hypothesis, or demonstrate a known fact. Contrary to your "beliefs", listening to different recordings done with different amplifiers is not a scientific procedure. To be a "scientific procedure", data has to be collected and that set of data must be able to be duplicated under the same conditions by anyone else. Without the data, it's just another anecdotal flim flam. I know it's a shock to your system, but your personal subjective opinion is not science - it's just blather.
Loudspeakers are nothing like lp cutterheads and similarly are not like moving coil cartridges. Yes, all three can use moving coils but this is essentially where the similarity begins and ends. The damping, energy storage, and losses associated with each system are vastly different and it is precisely these factors that determine the distortion signature of each. A near zero output impedance amp is most definitely going to have a different cutterhead system response than an amp with several ohms of impedance. If you actually intended to produce a meaningful conclusion from your "experiment", you should be presenting data that was collected - not total conjecture that "a few hundredths of a percent" of odd ordered harmonic from an amplifier are somehow going to be evident from sound produced by a loudspeaker that contains hundred's of times that amount of odd ordered harmonic as a result of loudspeaker generated distortion.
Your assertions about audibility of odd ordered distortion generated by amplifiers are neither reasonable nor plausible. It is akin to detecting a pin dropping while a massive 10 car pileup is happening - the loudspeaker being the pileup and the amplifier generating the pin drop. Perhaps this explains why you decided to throw out the lp cutterhead non sense to prove a point that has no logic or truth.
And the claimed differences in sound from one amp to the next are practically non existent compared to the differences between the speakers they're connected to. The diversion to your lp cutterhead "experiment" is just more subterfuge.
No, its not- its very real. I certainly concede that there are huge differences in speakers- I deal with that all the time in my work! But this thing with the mastering lathe? One of the speakers we heard that on was a set of boom box speakers . Even there, with nothing for tweeters, the difference could be heard. And just so we are clear, measurable too (unless you are trying to say that the measurements of odd ordered harmonics I was referring to earlier don't exist...). When it is measurable and audible I treat that as real.
By definition, an experiment is a scientific procedure undertaken to make a discovery, test a hypothesis, or demonstrate a known fact. Contrary to your "beliefs", listening to different recordings done with different amplifiers is not a scientific procedure. To be a "scientific procedure", data has to be collected and that set of data must be able to be duplicated under the same conditions by anyone else. Without the data, it's just another anecdotal flim flam. I know it's a shock to your system, but your personal subjective opinion is not science - it's just blather.
You are making a lot of assumptions here- the first being that I don't use instruments or something. I don't go on listening alone. I won't even start if the measurements indicate there is no point. I can also give the lathe cut to someone else and they can hear the same things on **any system**. That is pretty 'repeatable'. When we started the tests with the different amps, we didn't know what they would do but we found out pretty quick. Anyone could repeat the same tests, and I can also say that the distortion characteristics of each amp can be correlated by anyone. 'Repeatable' is in fact the name of the game- on this we agree.
Loudspeakers are nothing like lp cutterheads and similarly are not like moving coil cartridges. Yes, all three can use moving coils but this is essentially where the similarity begins and ends. The damping, energy storage, and losses associated with each system are vastly different and it is precisely these factors that determine the distortion signature of each. A near zero output impedance amp is most definitely going to have a different cutterhead system response than an amp with several ohms of impedance. If you actually intended to produce a meaningful conclusion from your "experiment", you should be presenting data that was collected - not total conjecture that "a few hundredths of a percent" of odd ordered harmonic from an amplifier are somehow going to be evident from sound produced by a loudspeaker that contains hundred's of times that amount of odd ordered harmonic as a result of loudspeaker generated distortion.
Sorry, on this matter its obvious to anyone with LP mastering experience you don't know what you are talking about. If I were you, I would spend some time around a cutterhead if you can, try some different amps and see if you still say that :) There is a very practical reason why amplifiers designed to drive speakers are also used to drive cutterheads. The results we recorded on the lathe cut can be duplicated on any loudspeaker.
Your assertions about audibility of odd ordered distortion generated by amplifiers are neither reasonable nor plausible. It is akin to detecting a pin dropping while a massive 10 car pileup is happening - the loudspeaker being the pileup and the amplifier generating the pin drop. Perhaps this explains why you decided to throw out the lp cutterhead non sense to prove a point that has no logic or truth.
I'm not the one that discovered the facts above that you want to refute; that work was done by General Electric in the 1960s. Maybe go talk to them. I just read the results of the study. There is more from it that I have not revealed in this thread as well- I recommend that you look it up (which, BTW, is a pain in the rear as it never got posted to the web, but its out there). However, since you are no 'audiophool', I will give you an exercise:
Get a sine/square generator, an amplifier, a speaker and a VU meter. The type of amp or speaker is not important but they should work properly. Run the amp into the speaker using the VU, set the sine wave output so the meter reads 0VU. Cover the meter, turn the volume of the generator down and switch to square wave. Run volume up until it *sounds* as loud as the sine wave did. Uncover the meter and observe the reading.
Now squarewaves are composed of odd-ordered harmonics. What GE learned is that the ear uses odd orders to determine how loud a sound is and you will see this in this experiment. Its apparent to me that you were not aware of this aspect of human hearing so you need to learn it first hand and this is a simple way. As you see from the test, the squarewave output will be a fraction of the original sine wave level.
So what I am talking about is this: that the **measurable** odd ordered harmonic distortions of very slight amounts are indeed also audible to the human ear because that is one of the things that to which it is the most sensitive . Much more so than human vocal ranges, image location, etc. It is arguably the most fundamental rule of human hearing. Don't believe me- then you didn't do the experiment above...
It also sounds to me like you are expecting the odd ordered distortions to be somehow 'buried in the mix'. Distortion does not work that way. The loudest sound in any recording will have this distortion on top of it. The idea that its buried is really just fantasy (just look at an oscilloscope...). If we go back to that sine wave, we already know it will be slightly distorted as a result of the generator, the amp, the speaker... so where do you suppose that distortion is? On top of the waveform or somewhere else??? Honestly, if you work on amps, preamps, speakers and do measurements on them we would not be having this conversation! Go do the test and see for yourself. The ear treats harmonic distortion of all types as tonality. Its something anyone can hear (its the source of the tube vs transistor debate for example), and yes, its quite measurable and repeatable too.
Ralph said:
" Honestly, if you work on amps, preamps, speakers and do measurements on them we would not be having this conversation!"
That's a pretty ignorant and presumptuous statement - but I expect no less from a tube amp shill.
In this thread you've provided nothing in the way of concrete data or measurements to back up your outrageous assertions concerning the alleged superiority of tube amplification or the failings of designs employing significant amounts of negative feedback. Every single post you've produced has focused on pure conjecture and anecdotal BS. There really is no mystery concerning the BS that the "high end" amplifier industry relies on for its survival. Claimed differences in sound were debunked long ago. And tube fanatics refuse to conduct blind tests or endorse the results of such testing like this famous one:
http://www.hometheaterfocus.com/receivers/amplifier-sound-quality.aspx
wherein this is stated:
"When it came down to the blind A-B testing of the 25 testers, only 3 participants scored 60% or greater correct when they guessed which amp was which, when comparing between two. Nobody got higher than a 63% score. Most amazing was that in testing between the Pioneer amp and the Futterman array, only 114 of 212 listeners could tell the difference -- thats a 54% correct guess! This is the most extreme example: audiophiles not able to tell the difference between a $200 Pioneer receiver and a $12,000 separate mono-block tube amp array with separate power supplies."
Sadly, your claims and thus you have devolved to the level of fraud and I see no point in further communication on the subject. "Honestly", you obviously derive a livelihood from the sale of this equipment so it is highly unlikely that you would admit to something that would hurt sales - even if it were true which in this case - it is.
I'm not saying any of this to try to convince anybody who has fallen in love with glowing amplifier tubes that they are wasting their money. I'm merely responding to condescension and ridicule by obnoxious snake oil peddlers who hurl insults at people (like me) that don't buy into their BS. I don't mind criticism as long as it is valid and based on science rather than audiophool religion or pure conjecture.
In this thread you've provided nothing in the way of concrete data or measurements to back up your outrageous assertions concerning the alleged superiority of tube amplification or the failings of designs employing significant amounts of negative feedback. Every single post you've produced has focused on pure conjecture and anecdotal BS. There really is no mystery concerning the BS that the "high end" amplifier industry relies on for its survival. Claimed differences in sound were debunked long ago. And tube fanatics refuse to conduct blind tests or endorse the results of such testing like this famous one:
http://www.hometheaterfocus.com/receivers/amplifier-sound-quality.aspx
Obvious troll comments aside, I for one welcome double blind testing, nor have I anywhere asserted the superiority of tubes- that is something that you made up in your mind. I do not feel that tubes are the only way to go by any means. What I am interested in is not generating the distortions to which the ear is most sensitive: IM and odd-ordered harmonics. That is perhaps easiest done with tubes but can be done with transistors as well, though most designers prefer not to jump through the hoops needed. IOW, you are the one being presumptuous and insulting with your comments above.
It seems that you are shilling for a bit a paper with meaningless specs, hoping that because it looks good on paper it must therefore be state of the art, which is almost never the case as the specs used by the industry do not have much to do with human hearing/perceptual rules.
"When it came down to the blind A-B testing of the 25 testers, only 3 participants scored 60% or greater correct when they guessed which amp was which, when comparing between two. Nobody got higher than a 63% score. Most amazing was that in testing between the Pioneer amp and the Futterman array, only 114 of 212 listeners could tell the difference -- thats a 54% correct guess! This is the most extreme example: audiophiles not able to tell the difference between a $200 Pioneer receiver and a $12,000 separate mono-block tube amp array with separate power supplies."
Actually, I would have said that this was perfectly predictable, as the Futterman amplifier employs a large degree of feedback like the Pioneer. Both are trying to be Voltage Sources and the feedback needed becomes the primary distortion source, which is to say that nearly all distortion is gone save for trace amounts of odd ordered distortion, which the human ear uses as loudness cues. I would IOW, expect them to sound very similar, which is to say both bright and harsh.
You did not address the human hearing rules in your post so I have to assume that you did not run the test . This unfortunately appears to have been predictable as I pointed out in my prior post. The issue here is you have more invested in making me wrong rather than dealing with the facts, so much so that you are willing to embarrass yourself to do it.
No one else that I know of in the audio industry uses the phrase "human hearing rules". Only a jackass like you would do so because it takes a jackass to make up his own terminology. Every human being hears things differently with the possible exception of identical twins. Ever heard of a "head transfer function", jackass?Moreover, only a jackass would conclude that the comparison of apparent volume between a square wave and sine wave has any value or significance. Musical waveforms look nothing like square waves so the "experiment" is totally irrelevant like 99% of your post content.
When you begin to treat others with a modicum of respect, you will receive respect in kind. I've been attempting to stick to facts - avoiding making this too personal. Since you see fit to label me a troll - I'll dispense with the niceties and present another FACT.
YOU ARE AN OBNOXIOUS, IGNORANT, JACKASS AND YOUR ATTITUDE/POSTS DON'T DESERVE TO BE RESPONDED TO.
Edits: 07/05/12
"YOU ARE AN OBNOXIOUS, IGNORANT, JACKASS" - pot & kettle!
Seriously, are you out of your mind? You honestly sound mentally unstable. Ralph was nothing but perfectly polite to you throughout the entire exchange, while you respond with schoolyard-level name-calling and bullying.
In fact, you are a troll. You clearly have nothing but disdain for audiophile principles, which defines you as a troll in this space.
And it is entirely obvious, as Ralph pointed out, that you do not understand not only human hearing but even the nature of distortion at the level of the electronics. Indeed, look at a scope, look at what kind of IMD products amps with large amounts of negative feedback can produce (on peaks, where it matter), and maybe think a bit beyond your innate hostility.
The fact that you simply dismiss out-of-hand controlled, scientific studies performed by entirely objective entities (GE) because, apparently, you don't care for the results (or are too angry to look at them) says enough.
I won't respond to anything you say but I did have to get my two cents in here.
I guarantee that if you had no ears, neither would you mess with audio products. Apparently you still did not run the test...
If you don't know any others that talk about human hearing/perceptual rules, its high time to get going with it. Equipment design to honor those rules offers a great deal greater musical performance. Conversely those that ignore the research into human hearing of the last 40 years in the audio industry also make inferior sounding equipment. The trick is to develop specs that agree with what we hear. to put it bluntly.
If you are not trolling, best to prove it.
Technically I don't think a manufacturer of tube amps can be a shill for tube amps.
There is plenty of literature out there dealing with this subject if you didn't have your head burried so far up...er, in the sand.
Read the writings of Norman Crowhurst in the late 1950s regarding negative feedback. Read the latest writings from Nelson Pass http://www.firstwatt.com/articles.html. Click the Audio distortion and Feedback article.
Read the master's thesis of Cheever: "A NEW METHODOLOGY FOR AUDIO FREQUENCY POWER AMPLIFIER TESTING BASED ON PSYCHOACOUSTIC DATA THAT BETTER CORRELATES WITH SOUND QUALITY" You can google it and read the .pdf of it.
Read the papers published by Earl Geddes:http://www.gedlee.com/distortion_perception.htm
Distortion perception and Distortion perception II
"Distortion by‐products that lie closer to the excitation are less likely
to be perceived (they are masked) than those that lie farther away
(masking is a localized effect – it mostly occurs in the vicinity of the
masker).
3. Distortion by‐products of any kind are likely to be more perceptible
at lower signal levels than at higher signal levels. (Less masking
occurs at lower signal levels)"
"The masking effect of the human ear will tend to make
higher order nonlinearities more audible than lower
order ones.
2. Nonlinear by‐products that increase with level can be
completely masked if the order of the nonlinearity is
low.
3. Nonlinearities that occur at low signal levels will be
more audible than those that occur at higher signal
levels."
These are some fundamental problems with high feedback (like your Crown) amplifiers. A non-feedback amplifier, if otherwise designed properly, will not have high order harmonics or if they are produced at higher volumes then become masked as the the higher levels increase the masking range. Clearly an amp with only low order distortion will often sound completely UNdistorted at all levels short of clipping because of the ear/brain masking.
Further:
"A sound systems quality is often judged (but not
exclusively) by its THD and/or IMD numbers.
In the context of the perception of distortion, it is not
unreasonable to question the validity of these numbers
for several reasons."
"They (THD and IMD) are purely mathematical relationships without
consideration for the characteristics of the receiver – the
human ear.
The recent application of psychoacoustics to problems in
audio data compression clearly indicates that masking
plays an almost dominate role in hearing acuity."
"THD and IMD have no correlation to the perception of
the distortion that they are intended to represent.
Correlation is possible with a metric that takes into
account the way the ear actually functions
One of the most important implications is that (non-linear)distortion
in loudspeakers could well be an insignificant factor"
This last point is interesting because generally loudspeakers only make low order harmonic and IM distortion. They DO NOT make the same kind of non-linear distortion that can be generated in amplifiers.
"Loudspeakers tend to have very low order nonlinearities
Higher orders would require very large forces in a
mechanical system
Loudspeaker distortion tends to increase with output,
but is generally low at lower levels
These factors imply that loudspeaker distortion may be
masked in the ear"
The Cheever thesis says much the same thing and he comes up with a new metric as well for distortion perception: Hint, your Crown and it's ilk don't fare to well...
So, even though it is clear you lack the perceptive qualities to actually hear what is wrong with most amplifiers at least you can follow some intellectual discussion on the matter as long as you have not blinded yourself to scientific investigation.
BLAH, BLAH, BLAH
Morricab said:
"So, even though it is clear you lack the perceptive qualities to actually hear what is wrong with most amplifiers at least you can follow some intellectual discussion on the matter as long as you have not blinded yourself to scientific investigation."
I would rather "lack the perceptive qualities to actually hear what is wrong with most amplifiers" than in your case and that of Ralph - lack all credibility. Clearly, blind testing has revealed little or no audible difference between most competently built amplifiers. All of the garbage you wasted pixels posting above amounts to nothing more than he said she said trash. The proof is in the listening and most people can't tell the difference between expensive tube garbage and the average modern day amplifier.
Moreover, you wouldn't know "scientific investigation" if it came up behind you and bit you on the ...
http://eandt.theiet.org/magazine/2011/11/believe-in-better.cfm
http://www.head-fi.org/t/486598/testing-audiophile-claims-and-myths
Perhaps freaks, psychos, and insecure audiophools are the only people with special acoustical perception powers to detect .005% 9th harmonic distortion from an amplifier that's being masked by 1.5% ninth harmonic from the speaker they are listening to. If that's the case, then great. The audiophools and freaks that are plagued by this malady will have to spend many thousands more for their equipment than the average person to avoid hearing this horrible distortion that no one else can hear.
Now that we seem to have this all sorted out, I will stop referring to tube and anti feedback fanatics as freaks and phools if you stop referring to the rest of us as incompetent.
" being masked by 1.5% ninth harmonic from the speaker"
And you claim to know something about speakers, what a joke. Show me a speaker making 1.5% ninth harmonic and its either severely broken or playing 130db. Either way it wouldn't matter even if you had a Crown on it then. Speakers make predominantly LOW order harmonic distortion or didn't you realize this?
Whatever you want to call me you are still incompentent.
Seeing as I do scientific investigation professionally, and I am well respected at it, let's just say I am more observant than you and leave it at that instead...you can go on blissfully ignorant.
Moroncab said:"And you claim to know something about speakers, what a joke. Show me a speaker making 1.5% ninth harmonic and its either severely broken or playing 130db."
The 9th harmonic was used as a theoretical example because everything you and Ralph have been blithering about has been theoretical. Most harmonic distortion testing stops at the 5th harmonic because higher harmonics are typically damped more fully - which undercuts all the BS you and Ralph have been slinging in this thread. Be that as it may, most speakers of decent quality will produce around 1% THD at a moderate sound level (around 90db measured at one meter) or 40 db down from the fundamental. At higher levels, this can jump significantly depending on the frequency and the level of damping (worst near fundamental resonance). The theoretical example given is not at all far fetched. If you had any experience measuring speakers, you'd know this. But clearly, with all the crap you've been slinging - you're just a fraud that has a great deal of nerve suggesting that others who have actual experience here don't know what they're talking about. The Dayton DA135 driver for example produces a 5th harmonic peak at around 3khz that is roughly 40 db down from the fundamental (around 1% THD) - and that is at a moderate drive level. At higher drive levels, it can easily triple.SEE, THE DIFFERENCE BETWEEN YOU AND I IS - I GIVE CONCRETE DATA BASED EXAMPLES DERIVED FROM ACTUAL EXPERIENCE WHILE YOU JUST SLING MUD AND ANECDOTAL BS.
Have a nice day Moroncab!
Edits: 07/06/12
No, you made a stupid statement to try to prove a ludicrous point. NOTHING I have been talking about is theoretical. It is all based on listening experience and test experience of others. READ the papers.
"
SEE, THE DIFFERENCE BETWEEN YOU AND I IS - I GIVE CONCRETE DATA BASED EXAMPLES DERIVED FROM ACTUAL EXPERIENCE WHILE YOU JUST SLING MUD AND ANECDOTAL BS."
No, I gave you good valid references about the effects of feedback and their sonic impact and you just want to pooh pooh it away because it is "inconvenient" to your audio world point of view. You are a narrow minded fool. Your experience with amps I can safely disregard... Your experience with speakers, while perhaps extensive, is also clearly as biased and misguided as well as uninformed (1.5% 9th harmonic...LOL)
So, I see now you have stooped to the last resort of a drowning man, name calling. I am Moronocab and Ralph is a jackass. What does that make you in this debate??? A LOSER! Simple as that.
Morricab, read the books of Bob Cordell and Douglas Self on audio amplifer design, they have chapters that deal with the NFB debate in high end audio; can be found in .pdf as well.
http://www.amazon.com/Designing-Audio-Power-Amplifiers-Cordell/dp/007164024X/ref=sr_1_1?s=books&ie=UTF8&qid=1341583056&sr=1-1&keywords=bob+cordell
http://www.amazon.com/Audio-Amplifier-Design-Handbook-Edition/dp/0240521625/ref=pd_sim_b_2
I meant that I had read some of the stuff by Douglas Self but not Bob Cordell. I had been following many of the discussions with him on DIY audio though.
Thanks I have read them but some time ago. What they write makes sense when you look at it a certain way but it definitely doesn't fit with my listening experience and it doesn't fit well with things that have been put out by Pass and Crowhurst.
'Blah blah blah' is not an impressive argument. The simple fact of the matter is that the list that morricab gave above is pretty much required reading if you want to have an intelligent conversation on matters such as this.
So:
*) a denial of the known lexicon of the last 60 years.
*) A resort to name calling ('audiophools' 'shill', etc.) when faced with facts that don't fit the uneducated world view.
*) A refusal to set up a simple test (likely because the simple gear needed is not on hand; you wanted numbers and figures, I handed them to you on a silver platter); likely based on denial.
Conclusion: troll.
.
Exactly.
With a company like Spectral - that's been making some of the most advanced amplifiers in the world for decades - you'd think by now it would be the holy grail of amplifiers. Clearly, it is an oddity - always has been and probably always will be.
As noted earlier, I've owned amps with 200V/microsecond slew rates and rise times under a microsecond. Along side those, I've owned amps with slewing rates intentionally limited to 25V/microsecond - as with most SS amps. And after many hours of comparison, I could not hear a difference between the two.
Why? Because the loudspeakers we connect these amplifiers to produce harmonic and intermodulation distortion that is orders of magnitude higher than the "hundredths of a percent" comments put forth by folks like those in this thread. The science and the listening confirm over and over and over again that amplifiers aren't the weak link in the audio chain and haven't been for decades - loudspeakers are and to some extent, human hearing is. Most of us can barely hear past 17 or 18khz yet we seem to need bandwidth out to 500 khz from our amps for "accurate reproduction" - RUBBISH!
The preoccupation with TIM and "time smear" that corporations and their sales reps capitalized on many years ago has in the ensuing time been exposed as a fraud - along with various other snake oil schemes in the audio business. Very few people are willing to buy an expensive Spectral amp - as good as they are, because there's no real payoff in sound improvement.
With tube amp manufacturers, the added even ordered harmonics and soft clipping "features" of tubes are still used as "audiophool sweetners" today- literally decades after it was discovered that SS could outperform them in just about every category - most notably with low output impedance and high current delivery.
The Spectral 360 is a great example of a SS amp that "out tubes" most tube amps. On that basis, it should be extremely popular among the audiophool illuminati. But it isn't largely because of marketing, poor profit margin, and the lack of "audio sweetners" present in its performance - and most importantly - it doesn't take several minutes to warm up nor does it glow in the dark very well. In terms of speed, quietness, and strength, the Spectral 360 is arguably one of the most capable amplifiers ever made for home use. But very few people know about it or own it essentially because at that level, the market doesn't know what it wants. Those who have that kind of money to spend typically rely on the advice of pushy snake oil salesmen to whom marketing and profit margin mean everything. Having an amp that can slew at 600V/microsecond is like saying I have a V-12 powered Renault F1 car in my garage. It's not street legal so I can't drive it anywhere but the engine sure sounds nice...
When loudspeakers reach down into the "hundredths of a percent" region in harmonic and intermodulation distortion, I might actually consider listening to the tube fanatic reasoning or claims that some audible difference exists between SS and tube other than that which is readily attributable to tube "issues" like high output impedance or losses associated with output transformers for models that are so equipped. Until then, it will be talk to the hand and keep your insulting, baseless snobbery to yourself. I reserve the right to remain highly doubtful without referring to the tube faithful as wackjobs or audiophools as long as the tube faithful refrain from labeling me ignorant or lacking in sound perception.
With tube amp manufacturers, the added even ordered harmonics and soft clipping "features" of tubes are still used as "audiophool sweetners" today- literally decades after it was discovered that SS could outperform them in just about every category - most notably with low output impedance and high current delivery.
This is quite incorrect. Transistors will make a lot of the same even orders if they are also operated single-ended. What if a tube amplifier is designed to be fully differential and balanced like most transistor amplifiers?? The answer- no even ordered distortions, as they are canceled not just in the load, but at every stage in the amp, just like solid state.
So no 'sweeteners'... yet we get plenty of comments about the improvement that our amps make, even over Spectral but I leave that one up to you.
I think you will find as well that even though you have 'high current delivery' that it does you no good. Its a simple fact that the amplifier may be able to drive lower impedances, but that is not the same as saying it is sounding its best into those loads. You can see it in the specs- distortion is lower driving higher impedances like 8 or even 16 ohms. You hear this as the amp sounding smoother and more detailed- more 'tube like' as many reviewers are prone to comment.
Now if simple sound pressure is your goal, maybe you have a benefit to those low impedance speakers (but you would be better off with horns- there really is no good argument for speakers 4 ohms or less). But if *fidelity* is your goal, think about an easier speaker to drive. There is an instant payoff.
You are like a kid who never got out of the Junior High of hifi. Your lack of sophistication with regard to listening puts you squarely in the "high power rocknroll" camp of "hifi" (I would call it more the loud and don't care what it sounds like crowd).
Read the essay by Nelson Pass, read Cheever, read Boyk and Sussmann and even read the work by Geddes, which has nothing to do with his speakers.
I find it almost comical that you can't realize that its not the amount but the kind and pattern of distortion that, at least in part, makes an amp good or not. You keep saying it measures low enough you can't hear it, but people do...consistently. You are the fool who has closed his mind because of what he "knows". I run into many silly people who think they "know" something because that is how they were trained (programmed?). A GOOD scientist starts with the open mind and the power of observation. Engineers are generally NOT good scientists and will follow "good" engineering practice because that is how they were trained. There are exceptions (Nelson, Ralph and Charles for example) but most are happy to stick to the mantra "If I make it low enough surely its inaudible...right??".
Even IF we were fooling ourselves, don't you find it strange that we consistently fool ourselves in only one direction? Its not like I made a conscious decision to say "Negative feedback and Class AB amps suck". I started with the usual suspects of amps (receivers from big name Japanese companies) and then some pro type amps because they had high power and low THD. But after HEARING better I drifted towards better sound because I wanted realistic sound and not just loud sound.
THat is how I eventually arrived at these criteria. Try them out and borrow or demo amps that meet them and see if they don't sound better. I dare you!!!!!
I have been through all the great measuring, bad sounding amps I ever care to hear so I won't go back there again to the land of sound but no music.
"The preoccupation with TIM and "time smear" that corporations and their sales reps capitalized on many years ago has in the ensuing time been exposed as a fraud"
Oh? Where? I'd just love to read all about it...
Thanks, Rick
Ah but don't you use some feedback on your smaller amps, Ralph? Also, I have looked at the measurements from the Einstein "The Absolute Tune" on Soundstage.com and it looks like it has very good measurements from a potential sonic standpoint. They may use some feedback but not too much as the distortion still rises smoothly with power. The distortion with frequency if flat and the harmonic distortion looks very good and completely free from power supply artifacts. Not bad I would say.
I have heard their products at length and find them to be very good sounding (but pricey as you have noted). I have heard their all tube OTLs and found them also to be very good but I haven't seen any measurements of them so I can't comment on what is happening there.
I have heard some of your products and also found them to be very good sounding (and less pricey as noted). Which is better would require a proper shootout with compatible speakers of course.
Compatible speakers is indeed the issue. Since we have stayed away from any large amounts of feedback (only found in 1 or 2 db amounts in the M-60 and S-30) the amplifiers don't have a voltage response compatible with some speakers built using the Voltage Paradigm for reasons I stated earlier.
But that does not mean there is a shortage of speakers they can drive.
BTW, loop feedback is not responsible for the type of distortion response where you see a dip at low power and a rise in distortion as output power continues to decrease. That has more to do with how the 'phase splitting' is done.
I don't doubt that the Einstein is a good sounding amp. It had better be for the price and power! and the fact that its an OTL :)
Years ago we were faced with the issue of feedback and I found that whether or not it was beneficial had a lot to do with the speaker. We built a variety of amps that had switchable or variable feedback and it was through that experimentation that we heard how it could foreshorten the sound stage, mess with dynamics and create artificial brightness. I would love to use it (for that matter I would love to use transistors...) but getting the equipment to sound like real music is the goal.
Ralph said:
"...getting the equipment to sound like real music is the goal."
That is your goal. The goal of others (myself included) is:
" to build systems that add or subtract nothing from the sound that is fed into them - in essence unadulterated reproduction of the original recorded signal."
To me, that means driving BOTH odd and even ordered harmonics into the range of inaudibility. For most "normal" people, with respect to amplifiers, that has already been accomplished using solid state equipment, tube equipment, and hybrid equipment. There will always be people who argue whether or not a sonic artifact that's 40 db down in volume from the fundamental is audible. For those folks, reason and rational discourse will always be a struggle. For the rest of us, the contribution of audible even harmonics or a rising top end due to increased impedance is never a "good thing" and is not what we consider to be in the category of equipment "sounding real". Whether or not real musical instruments produce even ordered harmonics is irrelevant. A tube amp that adds more of these harmonics is no better than a solid state amp that adds more odd ordered harmonics. Both need to stay below the threshold of audibility for a system to sound "real".
You can't get it to sound like real music unless the recording does. I've got no trouble with that.
But as to the matter of '40 db down...', I am not talking about something that has to do with religious fervor. The fact is that the human ear is more sensitive to the odd-ordered harmonics than just about anything else; 40 db down- no problem, as long as those harmonics are showing up in a certain range.
Most people think that the human ear is tuned to vocal frequencies but it is not. Its tuned to bird song frequencies. Add to that the way the ear/brain system detects volume, by listening for the trace amount of odd-ordered harmonics, and your 40 db figure is easy to attain.
What we are talking about here in terms of distortion is differences of only a few 100ths of a percent. The tragedy of this is that the industry had this information back in the 1960s- that is when some very basic studies by GE demonstrated that the ear/brain system does indeed operate this way. For the most part it has been ignored by the audio industry, as a sort of inconvenient truth; any generation of trace amounts of odd-ordered harmonics is going to separate the men from the boys, so to speak.
"back in the 1960s- that is when some very basic studies by GE demonstrated that the ear/brain system does indeed operate this way."
Ralph, do you know if this information is on-line?
Thanks, Rick
I think it is far more complicated than even that, Ralph and that it is not just the presence or absence of something but also the PATTERN of what is there or is not there. This is something Jean Hiraga noticed about 25 years ago that a logarithmicly descending distortion pattern containing both even and odd harmonics was consistent with good sounding electronics. Many things in nature have this kind of signature and our ear/brain is wired around this and the masking of low order harmonics and the absence of high order ones, except at loud volumes.
-
nt
One of the most prominent is how the human ear detects the volume or sound pressure of a sound. It is done by processing the odd ordered harmonics of that sound. If you know the effects of loop negative feedback on an amplifier, then you also now know how this runs counter to how we hear!
.
It clearly meets some of your criteria, at least #1, #3 and #9.
Edits: 06/18/12
Umm actually they are the poster child for wrong measurements. Look here:
http://www.stereophile.com/content/soulution-710-power-amplifier-measurements
I will explain. 1) downward sloping distortion with increasing power = high negative feedback design. 2)Increasing distortion with increasing frequency. The best sounding amps have a flat distortion vs. frequency plot. 3) FFT spectrum of 1Khz shows significant distoriton products to at least the 9th harmonic with odd harmonics dominating, so it is not monotonic, i.e. decreasing sharply with increasing order.
So, it is violating most of my measurement "shoulds" but I guess it must have a big power supply but runs only a bit in Class A I suppose.
Have a listen to the new Quad QSP and avoid tubes. Review on Positive feedback Online or for a lot of money try the Devialet Premier,best according to HFN & RR however I do not believe them.
Edits: 06/18/12
Heard it on rebuilt Quad 63 and 57 and, as always with Quad amps, not impressed at all. Nice try, not!
Haven't heard the Devialet yet but based on the reviews and the subtle remarks regardin tone in them, I think I will come away as I do with all Class D and that is, thanks but no thanks. Everyone was touting the latest NuForce reference monos. I heard them with a very revealing pair of Piega speakers and the result was close to fingernails on the chalkboard. Rubbish.
i hate class d with a passion
Most of them are just plain horrible.
.
Don't know, haven't seen any measurments or heard it.
1. Current dumper runs class B
2. Power supply transformer for 140 watts/channel only 600 VA
3. Use of op amps involves lots of feedback
QSP 600VA ? only 469VA according to IAG service but it still sounds very good, far better than any Class D amp. only problem is lack of proper relay protection now being sorted or so they say.
Edits: 06/19/12
As someone who has auditioned "any" and every class D amplifier. Could you give us your impression of the Mola-Mola compared to the Nuforce Reference 18 as well as the circumstances in which you were able to audition them?
http://www.mola-mola.nl/index.htm
I have not auditioned any & every Class D amplifier. However from what I have heard so far I would not waste my time in auditioning anymore Class D amps. I am very happy with Class AB thank you.
Same here
on this website .
That is half the size of my thirty year old Threshold for 100 watts / channel. Click the radio button for the QSP.
only problem is lack of proper relay protection now being sorted or so they say.
Like the 405. Maybe one day, they'll build the output stage strong enough such that it won't need protection.
Have a listen to the new Quad QSP and avoid tubes.
.
"In this land right now, some are insane and they're in charge. To hell with poverty, we'll get drunk on cheap wine."
Because if you want accuracy and to be able to drive any speaker properly avoid tubes.
It would be a mistake to assume that just because there are tubes, that it is thus euphonic. What if the tube design is fully differential? There will be no '2nd harmonic sound' that tubes are so 'famous' for.
Plus as you will see if you read the link below, that there are two different ways of driving loudspeakers; not all speakers are the same!
Hybrid tube amps are just as "accurate."But since there's not an universal definition, - it's left to individuals to decide for themselves..
"In this land right now, some are insane and they're in charge. To hell with poverty, we'll get drunk on cheap wine."
Edits: 06/18/12
.
"In this land right now, some are insane and they're in charge. To hell with poverty, we'll get drunk on cheap wine."
....instead of wasting your time with this guy. The results will surely be the same.
Oz
Don't worry about avoiding temptation. As you grow older, it will avoid you.
- Winston Churchill
N/T
So many dumb people with smart phones...
I wouldn't slip too far down the subjectivist slope, Sordidman. There is something to compare against and that is the real live thing. My criteria are based on observations of gear that gave me more a sense of the live than others.
to the real live thing.....
Sadly, - almost all recordings, (in my experience), are not so close and do not correlate well to the real live thing. And every recording, (even a small chamber ensemble), makes a choice of recording equipment.
In your position, you can say, - "this recording, played back through this system, was pretty OK and close compared to what I heard."
I respect that, and having heard probably more than a few of the same components that you have, - I am reluctant to disagree. I would just say that you are largely right, and I think that people can get close to a decent definition of accuracy, and come to similar conclusions through experiences, and numerous comparisons.
The choices that the recording engineer makes though, do affect the final outcome.
careful....
"In this land right now, some are insane and they're in charge. To hell with poverty, we'll get drunk on cheap wine."
Let me know when you find a stereo amp worth listening to . Whilst I quite like the sound of the Quad QSP into transmission line speakers, I have rejected two of them due to their lack of a proper protection system. IAG agree there is a problem and they say they have a solution, I will believe it when I experience it. I gave up on the 63's long long ago they can not be used now to judge the sound of the latest amps.
Which transmission speakers ? I have Clements that sound good.
Acoustic Insight not distributed in North America. I imported some Clements speakers some years ago but did not like them. I was misled by a rave review which turned out to be totally incorrect. It is very dangerous to buy any audio component without having an audition. Never Never believe ANY review by anybody today. The great reviewers from the past who you could rely on 100% are unfortunately all gone.
Edits: 07/03/12
More Euphonic than accurate, very old tech.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: