 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
58.153.30.166
In Reply to: RE: Thorsten would have to explain it...but posted by Tony Lauck on March 17, 2014 at 08:42:59
I suppose it's comforting to fall back into believing that people are deluded by distortion artifacts when they see the measurements. Too many good ears unfortunately to pass it off as just euphonic distortion.
Then again euphonic means pleasing - and we can;t possibly have any of that in our stereo systems now can we. We must have brutally accurate, not the least bit pleasing and completely banal - let me get my credit card.
Follow Ups:
One can demonstrate the euphonic effects of second harmonic distortion by deliberately adding it. This can isolate the effect of this distortion from artifacts of amplifier design. A similar test can be done by inducing jitter, but this is much harder to do since jitter is a complicated phenomenon. Other possible "improvements" include deliberately adding noise to the signal for its masking effects. (This is not to be confused with proper use of noise as part of dithering, which is a recording process.) Experiments can be done to sort out these individual psychoacoustic effects.
I might consider an SET system if someone can show that it is possible to produce undistorted peak SPL of 115 dB without using colored transducers such as horns. I listen to Mahler symphonies. SETs would probably be fine if I confined my listening to solo guitar and harpsichord. By the way, I find the vast amount of my music collection to be pleasing and many recordings sound quite realistic. Only a minority of recordings are brutally ugly or banal, and reflect mistakes in my purchase decisions. It is usually possible to weed out poor music or performances on download purchases from sites that allow listening to a good selection of sound clips.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
RE DACs
Art Dudley has a preference for tubes (Shindo Labs - which you've probably never bothered to try for yourself) so I can somewhat understand a preference for a valve CD player.I brought up Martin Colloms because he designs or attempts to design Accurate speakers and is published on them. Looking at his system of top of the line Krell and top of the line Wilson Martin is not a "tube guy" or a "I love distortion guy". He ran the AN CD player with his SS and "accurate attempts" loudspeakers. Art's view is somewhat expected, Martin would be in the objective "if it measures better it should be better" camp. But unlike some he actually bothers to test the stuff before opining on it.
The second and curiously irrelevant attack on SET amps seems strange. SET amplifiers, unlike most SS amplifiers, have a direct correlation of distortion (any distortion) with level.At low level SET amplifiers (and class T for that matter) have very low distortion - when pushed to their limit (beyond their limit) is where they distort - and the distortion is second harmonic. So umm if you don't push them to their limit there is no 2nd harmonic distortion. Therefore, any claims for a preference of SET amplifiers is NOT due to any distortion artifacts because the vast majority of SET owners use speakers that never need to use more than about 10-15% of the amplifier's power. Sure when driven to the max a SET will exhibit soft mushy nice sounding distortion but it is only nice sounding in comparison to SS which goes poppity pop pop and hopefully not taking the tweeter out.
Most SS amplifiers work exactly the opposite way - at 80-100% of their maximum level they have their lowest distortion characteristics - when reduced down to where most people listen is where they exhibit their highest distortion (which of course is never measured cause it doesn't look pretty and won't help them sell).
Amplifiers that are class A/B and operate up to say 5 watts in A and switch to B which is where most even good ones operate will be shifting between A and B constantly. Class A the music plays at 3 watts and everything is fine but the bass at 40hz and treble at 3khz hits a peak due to an impedance drop - the amp has to come up with 10-15 watts and yup there is the switch to class B and then the transient drop and the amp shifts to class A - oops then a drum - amp has to switch back to B - Switch switch switch switch - fluorescent light bulb syndrome. There are good and bad variations of them but none of them are natural light.
Notch distortion occurring continuously with virtually every album all the time. Notch distortion (crossover distortion) is audible and fatiguing and constantly on display with all A/B amplifiers.
Getting a true sense of distortion measurements of any amplifier using feedback is pointless as distortion is not measured "before" the feedback loop is engaged.
SET is a side issue. Why is it that people think they can trot out the old Horns Suck, I listen to Mahler at 115dB as if that's SS for the win?
I have heard tons of systems than can do massive scale at massive levels that truly suck donkey balls on female vocals at low levels. Yes system A may do massive scale and symphony better but they often get simple girl/piano wrong and most people most of the time listen to smaller scale stuff and "occasionally" listen at 115dB to big music. Indeed, I auditioned Avantegarde speaker in Hong Kong and it does massive quite well but was unlistenable with Eva Cassidy.
And there are a lot of other speakers besides horns that can easily be driven with SET amplifiers - and there are, like anything else, good and bad examples of SET amplifiers and a variety of power levels that will drive most speakers quite well including suicide inducing stuff like Mahler.
Lastly, people tend to move FROM solid state amplifiers to tubes to SET. People don't start with the likes of Audio Note CD players - they typically end up there or with something like it AFTER they owned Bryston/Krell.
And while I agree that most of the horns I have heard are tough to take - you need to get out more and listen to some of the good ones properly set-up with appropriate gear. There is more to the world than Klipsch and Altec and Avantgarde. But there are speakers like the Teresonic Ingenium with over 100dB sensitivity if one just can't do a typical horn. There are also speakers like the AN E and DeVore Orangutan - add a sub if you must.
Edits: 03/17/14 03/17/14
When it comes to most of these sonic arguments I often wonder why people gravitate to one thing or the other.
The prices of good audio equipment has dropped. Yes the best costs money but today's affordable stuff is quite doable for audiophiles. So why not have our cake and eat it?
One is not limited to one CD player or one DAC or one kind of amplifier. As a reviewer I am trying to buy a SS based stereo systems as well as a valve based one. I would also like a variety of speakers on hand to compare. But anyone can do that.
Fred Crowder on our staff owns some big Acapella Triolon speakers and directly compared $60,000 SS power amps (arguably some of the best SS in the business from Pass Labs) VS 300B SET power amps. One does some things better the other does some things better.
But heading on down to those of us on meager finances you can get a pretty good 300B SET amplifier from the likes of Audio Note for about $2200. Kit One.
You can then go out and buy a second hand Rotel RB 1090 (1KW into 1ohm) ($2000 new in 2006) weighing in at 100lbs or perhaps a Parasound A21 power amp.
Or newer and cheaper Emotiva XPR-2 which pumps out 1000 watts at 4ohms
And hey to many if its distortion is below the threshold of human hearing JND then they all sound the same so for $1799 it will drive anything and the measurements indicate no one can tell it apart from any more expensive exotic name brands. (measurements is all that matters in the world of SS since no one can hear the difference in controlled level matched tests).
Assuming Emotiva is reliable then here you go - Mahler is covered with pretty much any speaker
Now you have your "precious pretty" SET amplifier as well as your slam your "slam bang" amplifier and you're under $5k for both and at nearly the flip of a switch. Best of all you have a back-up amp should one of the two ever fail.
Check out this power amp!
I have no interest in filling up my limited living space with redudent components. I can deal with an old system and its replacement, but not various mix and match components. Also, having many choices means that I can't just listen to music without first finding some components that make a recording sound good. IMO the result of this approach is constant fiddling and worrying about the technology, rather than just enjoying the music. I also use my system for remastering old analog recordings and without a constant reference of how these recordings sound vs. the range of recordings on the marketplace I have no good way of positioning the sound of these recordings in such a way that they stand a chance of sounding good on a variety of systems. Here the problem is the same as just listening: one's mental confusion about the system leaves little space "in one's head" to think about important things. Similar arguments pretty much cover tube equipment that drifts, has to be rebiased periodically, tube rolling, or many other things associated with retro audio technology. Just IMO, as I realize that some people like to fiddle with equipment. I don't. I like to listen to music or make (i.e. fiddle) with recordings.
Measurements can work when a complete set has been developed that adequately covers the known and unknown aspects of a given technology's limitations and the known and unknown aspects of psycho-acoustics. They are a useful tool for dealing with known factors, but usually useless for dealing with unknown factors. After some of these unknowns become known then it is possible to come up with new measurements that can capture this new knowledge, but this may take years or even decades. The general rule is that new technology creates new unknowns with new distortions that can be heard but not captured by existing measurements. Later, one understands these new distortions and how to measure them as well as how to relate these measurements to what some people can hear. Still later one may get products that sound better than the old technology. Still later, if we're lucky, manufacturers' spec sheets and reviewer's product reviews include these new measurements. In this paragraph the key words are "unknown" and "lucky".
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
I dunno but it seems to me a lot of computer audio guys keep fiddling with bitrates - checking which level is right and spending copious amounts of time agonizing over whether it should be DSD/DXD/PCM whether to get the SACD or 24/96 or 24/192 or 32/384.
Not sure why two amps would cause much fiddling and SET amplifiers from Audio Note are all self biasing. The tube in the AN CD player has a 100,000 hour life expectancy or 11 years 24/7. Their amps will get you 8000 hours before you replace a tube - no biasing plug and play.
I was not born in the tube era or the vinyl era - and I concur - I find biasing and dealing with turntables to be a pain. Perfectly fine reasons not to bother with them.
The bias in my Citation II had to be set by hand. I also had a MAC-275 which was very stable and reliable but it never sounded good because of the class B crossover distortion.
I don't recall having to replace the 12AU7s in my Dynaco PAS-2, but some 12AX7's definitely got noisy. Unfortunately, I had to replace all these Mullards after the FBI confiscated them because I had borrowed them for a suspect purpose. :-)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
No question tube amps are a pain - I can't believe the FBI took your tubes LOL.
I think like SS - tube amps have come a long way over the years. After hearing some of the vintage models - ST70, a few McIntosh amps, ARC, and for that matter several new tube amps I'd probably buy a class A SS amp instead.
It was the same with turntables - everyone saying that $200 second hand duals were better than elite CD players. Nope. They can sound excellent but it seems to cost a significant sum to really impress me and most new LPS are just direct transfers from CD and don't sound any better. Still I have some LPS with CD versions that just suck.
Hence why no own both and play the one where it sounds better - if album A sounds better on vinyl - use turntable - if album B sounds better on CD play the CD.
Fortunately the FBI took only the small tubes that were used in my Blue Box, probably Mullard ECC82's (12AU7). They didn't take any of the tubes in the Citation II, including the 4 Gold Lion KT88's, despite the fact that the amp was being used to drive the phone line. More details in the book that I linked in the earlier post. Also, they didn't take the audio oscillator that was used to generate the 2600 Hz tone, which belonged to the college radio station. We had borrowed it to use it to run my turntable at a variable speed, as needed to dub a recording of Wagner's Siegfried at the correct pitch, as the bootleg recording had been made at the wrong speed. (At the time there was no recording of this opera in the Schwann catelog.) The radio station later broadcast a complete Ring cycle.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
"2600"One of my favorite publications! :)
Fond memories of says gone by... Did you actually use your Blue Box?
Edits: 03/20/14
Our group were the first people east of the Mississippi to build a blue box. We thought we were the first in the world, but it turns out that one lone wolf in the Pacific NW had figured out how to build one a year or two before us. The term "blue box" came from the color of the chassis that we built our hardware on, which came from an electronics surplus supply house called Eli's, later acquired by Fry's.
We used our blue box to demonstrate that it worked. We called the weather in Nome, Alaska, etc... We did not use it to cheat ATT out of phone calls that we would conceivably have paid for, nor did we sell any products or services. As it turned out, this was fortunate, because had we done so, then the Deputy Federal District Attorney for Boston would not have told the Boston ATT security head that we had not committed any crime and hence could not be prosecuted. (This letter in files obtained by FOIA by the book's author, with our release.) ATT dropped all civil charges against us as part of a deal we negotiated in return for writing up a detailed report on what we had done, how we had done it, how they should design future networks to be less hackable, and how they could detect fellow blue boxers. When we handed in our report we had to sign a contract and were paid $1.00 for the report and given promises of jobs at Bell Labs after college graduation.
N.B. Were someone to do the same thing today (or otherwise hack any computer) they would be sent up for hard time. A lot of fascist laws have been passed since 1963. No thinking person can continue to claim that the USA is still a free country.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Very cool story. :)
" one lone wolf in the Pacific NW"
John Draper "Captain Crunch"
Yes fun times. Build Blue, Red, and Back boxes also. Lots of fun. Never did anything crazy with them, but they were fun. Old pre digital Telco was interesting with loops, busy signals etc... The good old days.
No, Captain Crunch came later. You have to read the book to learn about the first guy.
Captain Crunch came to visit me at one time during the 1970's. Ron Kessler from the Washington Post who wrote the first articles on us for the Boston Herald, The Wall Street Journal and later Fortune Magazine, had contacted me regarding a story that it was possible to bug any phone in the US by keying in various tones on a random pay phone. I had a brief meeting with John Draper at my home in this regard. Later, I did some digging and asking various people who know (ex Bell Labs people, spooks, etc.) and found out that this could have been done for most phones if someone could have done the appropriate computer hacking. I conducted this research in such a way that the people involved thought it was casual questions and no one of them saw more than a piece of the puzzle so they had no idea what I was up to. I declined to come out of the shadows at that time and told Ron Kessler that I did not want to talk to a Congressional Committee.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
"They are a useful tool for dealing with known factors, but usually useless for dealing with unknown factors." (re. measurements)
Tony... In my experience the only way to chase down unknown factors in a system is to first "get a handle on the problem". That means first finding things such as actions, environment or measurements that affect or correlate with it. Ultimately both. That's the view from the trenches anyhow.
So rather than measurements being "useless", they are far more commonly "essential".
Regards, Rick
I didn't say that measurements weren't necessary. I said that they aren't sufficient. Actually, there are some genius tinkerers who can produce working systems without understanding any knowledge of electronics, just cut and try, but these talents are rare.
There are some serious puzzles where measurements don't explain what people hear. One of them is 24 bit PCM files sounding different (to some people on some systems) where the only difference is the addition of 1 lsb to every sample, creating a minute DC offset that is lost in the analog circuitry. Another, which I've experienced, is where a 24 bit sound file was converted to 32 bit floating point and then converted back by adding dither noise at 24 bits, so that many samples got a random addition of +1 or -1 and some rare ones +2 or -2. When I sent back this file a notoriously vocal well heeled audiophile who has his own web forum told me that this file was "completely trashed". Both of these examples can not be explained by known models of human hearing, but they can possibly be explained by artifacts in the playback chain.
A classical example concerns the SABRE chips sigma-delta modulator that was described in a (video) talk at RMAF. Here the measurements looked fabulous, but a few audiophiles heard artifacts. Later these were measured and reduced and the resulting design sounded better.
Most practical engineers work with averages and safety margins. They do not do worst case design. Therefore most all product claims must be taken with a grain of salt, and it's not just a case of dishonesty. As an example, the lead designer of DECs first Ethernet controller came to me with a question of a possible failure mode where random numbers in the backup algorithm would match 16 times in a row and cause a failure. A brief analysis showed that the "chance" of this happening was such that if one million devices were used this would happen less than once in 1000 years. After the engineer left, my fellow employee and I laughed that we didn't trust our probability calculation, but we didn't really worry because the engineer was extremely careful. In three days he came back to us and told us that our calculations had been wrong, because the failure happened in his lab on one network with three devices every few minutes. As it turns out we had a bad model of the frequency distribution of clock oscillators and didn't understand how the Ethernet backup algorithm actually worked, the latter involving an error of at least 21 to the power 12. Later there were some proprietary hardware designs to address correlated clocks and some compatible algorithms to the Ethernet backup algorithms that were the subject of several patents as a result of this amusing escapade.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
"Most practical engineers work with averages and safety margins. They do not do worst case design."
Well, often they do, hopefully always. But that just means that unless a part or process is out of tolerance that the product should meet it's specifications. And to improve the odds, especially those of catching process failure, there is usually a final test of key parameters.
Doesn't mean it WORKS for Pete's sake, let's not get carried away! But the odds become acceptable that it will...
Measurements are GOOD! It's reason that has failed audiophiles and manufacturers alike. How many factors are there that can be a little out of kilter and be audible under certain conditions? Humans are really really good at sensing small differences.
Especially now, in the digital era, I think we could come up with an adequately reliable electric listener that could do as well as humans at noticing little factors. But what are the odds that there would be one judgement setting for it that everyone would agree with?
Regards, Rick
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:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
