 
|
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 |
149.22.84.9
In Reply to: RE: So IOW, it's just as I suggested? posted by morricab on June 13, 2024 at 01:22:02
. . . to have deleterious effects on the sound quality. I also believe that this approach to design has ebbed quite a bit over the decades and is thus not as relevant to the discussion today as it may have been in the past.
As for "bad" recordings, it would help if you could name actual recordings which suffer from the problems you list, rather than stating generalizations.
BTW, is this the Geddes you mean (at the link below)? This particular Geddes says things like the following:I have always thought that if someone's measurements do not "tell the whole story" then they are the wrong measurements. Technology has simply come too far to believe that "there are things that we cannot measure." I have also never believed that all that matters is "how it sounds," because this is such an unstable and personal opinion. Sound quality opinions can and will differ from person to person, system to system and most importantly even within the same person on different days (as I said before, I have personally witnessed this in well regarded "reviewers"). Personal preferences have such a low stability as to be an almost completely pointless thing to stake a claim to. "Hi-Fi" does not mean "pleasant" - it means "accurate"; accuracy, as opposed to preference, is absolutely quantifiable and extremely stable - as stable as I care to control in my lab from day to day or test to test (but in any case its uncertainty is easy to quantify and understand). Decisions based on accuracy are therefore much more likely to be valid than decisions based on "how it sounds." I do not see how one could ever support a position that "preference" trumps "accuracy." That's simply taking a giant step backwards in the evolution of Hi-Fi.
Follow Ups:
Please look at the two papers he published on the subject at this link
http://www.gedlee.com/Papers/papers.aspx
The first is Auditory Perception of Non-Linear Distortion- Theory
The 2nd is Auditory Perception of Non-linear Distortion
There is also a slide deck that summarizes the two papers.
that's the one. Do continue with what he says...
"That's the problem with THD, it just does not show what we need to know."
I learned that lesson in the 70s as a teenager.
And yet, SINAD is the sole metric by which ASR judges gear with its color coded chart - "higher better".
No negative feedback at all for you? Or do you like a bit more distortion and noise in your listening?
Correct me if I'm wrong, but doesn't SINAD include more than just the THD component?
I am pretty much an absolutist...none of my gear has negative feedback other than degenerative feedback (like a cathode follower would use...even then I am not sure I have any cathode follower circuits in my gear...maybe my phonostage...)
The problem with only using a little bit of feedback is that you still have relatively high distortion but you have created a whole bunch of high order distortion, which remains unmasked and is likely audible.
Empirically though I have found that amps with a little bit of feedback often still sound more dynamic and less "gray" tonally than those that use massive amounts.
As Geddes and others point out, you can tolerate a fairly high amount of low order (meaning primarily 2nd and 3rd order...although Cheever allows for more at higher SPL because of the ears own distortion generation acting as a mask) due to masking. In fact, if the harmonics are effectively masked, it sounds to the listener as if there is NO distortion at all.
Cheever posits that as long as the distortion components are falling below the ears own distortion pattern and intensity that the component will sound pure and undistorted...even if it has measurably high distortion compared to other components that don't sound as pure. The pattern has to match (and at a given SPL) or remain below the ears own harmonic distortion generation.
As the ear makes essentially nothing above 9th harmonic...even at very high SPL (at normal listening levels it makes nothing really above 5th harmonic), ANY distortion harmonics higher than that will be audible even at exceedingly low levels.
As to why high feedback amps can sound "dead", "gray" and lacking in perceived dyanmics, especially on high sensitivity speakers where all is revealed, I think it has a lot to do with the "noise" floor. Crowhurst demonstrated in the 1960s that high feedback makes a lot of distortion produts...essentially an infinite number with a complex music signal. These are mostly at every low levels and essentially generates a signal correlated "noise" floor. True noise is not correlated with the music signal and as such it is possible to hear below that noise floor which helps with space, dynamics and low level detail. A correlated noise floor acts as a wall that one cannot hear below and as its modulating with level, it probably has very detrimental effects on the perception of the music. This might be why a minimal amount of feedback is not so bad as it would not create a very strong effect on this signal correlated noise floor.
High feedback also means that a lot of the back EMF from the speaker will find it's way back into the signal path, as some of it is injected back to the input stage (for amps with global feedback). That signal of course looks nothing like what went into the speaker and besides it is signal that doesn't align with what is currently passing through the amp.
Finally, Nelson Pass also weighed in on feedback (although his conclusion was a bit wishy washy). You can read that here:
https://www.passlabs.com/technical_article/audio-distortion-and-feedback/
You call his conclusion "wishy washy" but I would characterize it as less dogmatic.
You know, this whole discussion started when a few posters seemed to get triggered when Amir and Audio Science Review were brought up (not by me I might add!). As I've said all along, people have the right to their beliefs and perceptions - even though these perceptions may (and do!) change from day to day (as Geddes pointed out!).
People can believe what they want but the listening data strongly suggests that there are preferences that probably don't align with some people's belief. What is clear, is that THD and IMD don't correlate well with those preferences.
he doesn't bother following links to actually understand what's being discussed.
I found the listening correlation data comparing THD, IMD and G m compelling. You might as well flip a coin with the first. And really enjoy the selection chosen!
You're trying to pretend that all I care about is THD. Once again, not true.
you don't pay attention. You don't follow links. You don't follow observations. You don't even read the manuals for your gear.
Let's review what my last post observed of your behavior:
"As previously observed, he doesn't bother following links to actually understand what's being discussed."
It is clear from other questions you've made that you didn't download and read the two references found a day earlier here.
Hint: blue text in my posts signals a hyperlink. Click Chris, click!
I'm convinced this will be the very first time you've seen the referenced Powerpoint slides like this:
. . . studies which have only the most peripheral relationship to to my posts. Let me remind you of what the main point of my posts was: that measurements (and NOT just THD!) can help us in describing the sound of a given component. You admitted that that was true yourself. But then you turned around and started in with a bunch of bunk you somehow divined from my previous posts.
So I don't follow observations? I think that's MORE true of you. If distortion measurements are so useless, why do almost ALL audio manufacturers provide them? Are they trying to deceive customers? Seems to me that you and Morricab are out in the cold with you own contentions on this point. (And, BTW, I'm NOT disputing that masking is a thing. After all, I prefer to listen to Dolby Atmos these days - which I listen to mostly via Apple Music's LOSSY streaming! Talk about masking! LOL!)
And just because you posted a picture of your listening environment which CLEARLY showed your speakers placed perpendicularly to each other doesn't make it incumbent on me to follow YOUR observations - or anyone else's "observations" for that matter.
which CLEARLY showed your speakers placed perpendicularly to each other.
How they radiate into the room, however, is symmetric. Per the manual, each is aimed 30 degrees to listening area.
I can recommend a good oculist for you. ;)
Indeed. Cheever also came up with a metric that included SPL because the ear/brain masking extent (i.e. what harmonics and how much) changes with SPL.
This is why low SPL is more critical than high SPL because masking is less effective, particularly for high order harmonics even at very low levels. They also tend to be about the same regardless of amp output in a feedback type amplifier. It is also probably why low level resolution and soundstage/imaging are most affected by the negative effects of high order distortion and negative feedback.
. . . was that even INDIVIDUAL subjective perceptions of sound can change from day to day or even hour to hour FOR THE SAME PERSON perceiving the same sound.
How are you going to account for that - and then "quantify" it? It seems to me that the only option you have is to do so in a faux scientific way - by arbitrarily assigning "values" to reactions which cannot by their nature be quantified.
The two of you try to denigrate listeners who set some store by measurements (and not just THD!), and yet you then turn around and quote these "scientific" study measurements which are IMHO pretty dubious.
You clearly haven't heard of statistics...think about it or read about it and it will perhaps come to you eventually how you can get some useful data out of individual preferences.
As to the industry, they provide the measurements they kind of have to, not the ones that would be useful (if fitted with the proper model). They publish THD and maybe IMD...almost no one provides a detailed FFT and model fitting. Aries Cerat provides the level of H2 and H3 at a given power...but that is an exception rather than the rule and the assumption they make is that there is an exponential decay in the harmonics with increasing harmonic order.
. . . that equipment can have as much distortion as possible, and you would still claim there's no correlation? I think one appeal vinyl has for some listeners is that vinyl distortion is euphonic - and they find that pleasing. Euphonic does not equal accurate.
Where do you draw the line? How much distortion is acceptable. Is it anything goes? I doubt that many listeners would agree you in that case.
There is audible and then there is perceived as distorted.
Look at some of the levels of distortion applied in the Geddes paper. Some are very low and some are pretty darn high.
No negative feedback at all for you? Or do you like a bit more distortion and noise in your listening?
Some works fine to stabilize the circuit. Noise has nothing to do with it.
Yes, SINAD is noise+THD. Key being (useless) THD.
Maybe you're like Feanor and listen by looking at graphs. It that works for you, fine. :)
Feanor himself will tell you that I'm not like him an any significant way. So do you totally ignore graphs? If so, I think that kind of isolates you to your own little corner of the audio universe, along with a few other true believers. ;-)
And BTW, which is it? Negative feedback is OK, or noise and distortion are OK? Maybe a little of both for you? LIKE FOR MOST OF US? Remember, I listen to most Dolby Atmos recordings in LOSSY 24/48. ;-)
Feanor himself will tell you that I'm not like him an any significant way.
Ok, so your shared enthusiasm for the value of SINAD is not significant.
So do you totally ignore graphs?
Only the ones that don't correlate to what we hear.
I do, however, use measured graphs to fine tune speaker placement and EQ for the HT. Here's upstairs which took some work along with a small forest of bass traps to achieve.
Negative feedback is OK, or noise and distortion are OK?
Why not quote something I posted and frame a question based upon it?
"Ok, so your shared enthusiasm for the value of SINAD is not significant."
Why not quote something I posted and frame a question based upon it?
[I] ignore the [graphs] that don't correlate to whatweI hear.
There! Fixed it for you, Mr. Self-Appointed Golden Ears.
Why not quote something I posted and frame a question based upon it?Question unnecessary.
Calls to mind this . Like Feanor, you're also not in the Geddes camp where correlation is the objective.
Edits: 06/14/24
"Calls to mind"? Not to most intelligent people.
Sometimes, the correlation exists, but you simply deny that it does.
Geddes papers show a correlation does not exist.
I brought up the question of measurements in general, and you're trying to confine what I said to THD only.
from the Geddes interview:
"That's the problem with THD, it just does not show what we need to know."
I learned that lesson as a teenager.
I'm convinced Chris didn't follow and thus doesn't understand my "calls to mind this" embedded link where another inmate asks why he makes snide and scornful conclusions. Seems he rarely follows embedded links I provide and later asks question answered by them.
His reply is all the more ironic. ;)
. . . you act as if I'm totally pushing THD as the be all and end all - which I NEVER have done. Check this thread. I just say that, in my experience, measurements (in general - not just THD) DO usually correlate with listening experience. Even YOU have admitted that when you say you use frequency response measurements to improve the bass response in your room.
What do frequency response measurements (linear distortion) have to do with THD (non-linear distortion)? What studies have you done to demonstrate that you can correlate your preferences with THD level? I guess the answer there would be none, right? It's just your gut feeling that you would prefer a low distortion amplifier over a higher distortion one.
. . . I quoted?
If you bothered to read the papers, he explains what he is on about and that he developed a metric to process distortion in a more meaningful way than THD and IMD. His metric takes the THD and IMD data and establishes a much better correlation with what the listeners reported than just plotting that versus THD or IMD.
So, what he said in the interview doesn't contradict the papers.
I'll quote the relevant part once again:I have also never believed that all that matters is "how it sounds," because this is such an unstable and personal opinion. Sound quality opinions can and will differ from person to person, system to system and most importantly even within the same person on different days (as I said before, I have personally witnessed this in well regarded "reviewers"). Personal preferences have such a low stability as to be an almost completely pointless thing to stake a claim to.That does not sound to me like he takes "what listeners reported" very seriously, metric or no metric. Contrary to your assertion, what he said in the interview DOES contradict your summary of the paper(s). How recent was his paper? Maybe he changed his mind over time.
Read the papers and slide presentation and answer the questions for yourself.
He is talking about an individual response. Taken in aggregate you will build a statistical preference for most listeners (assuming you have a good correlation to the measured data) for particular types of distortion vs. other types.
Geddes papers (if you don't read them then we have nothing more to discuss) indicate that THD and IMD had no correlation (even a slightly negative slope) to the distorted sounds they were subjecting listeners to. When he applied his metric, which takes into account some kind of weighting for harmonic content, he got a much stronger and positive correlation between preference and his metric.
I scanned through the two AES papers and the conclusion is convincing that THD and IMD doesn't match the subjective scoring of their test group whereas their Gm metric (bad choice of symbol as Gm means something else to a lot of people) does correlate much better. There is much to pick over - in the first part they don't quantify masking. Yes it is a psychoacoustic thing but how much effect does it have and why choose a cos^2 weighting factor? I also question one of their precepts that distortion will be more apparent at low signal levels than high signal levels - that needs some explaining as distortion grows quicker than the signal level that causes it (i.e. if a signal increases by 1dB any second harmonic will grow by 2db, third harmonic by 3dB etc).We aren't told what the 21 transfer functions are nor their correspondance to trial number. We do know one/some/many have severe discontinuities - are those really relevant for high-end audio? And they remove 6 of those transfer functions at the end because they don't fit the scatter plot so well (at least they were open enough to admit this as data manipulation is often a big problem in academic papers).
But, the big unanswered question is whether a low distortion system is somehow worse? We don't know because they don't detail the transfer functions! We can probably conclude that many non-linearities are not that audible, or audible at all, but we can't conclude that low distortion devices are somehow subjectively inferior. So it is a bit of a shield for audiophiles to stand behind and an excuse for high-end audio designers to not work too hard but not the full story.
Edits: 06/20/24
Indeed! Rush rules!
Cheever added an SPL factor into his model because there is also a change in the ear's own generated harmonics. It also then would depend on the sensitivity of the speaker and how much distortion/power at a given SPL...
So, a high sensitivity speaker that only needs mW to a couple of watts will be an easy, low harmonic dominated (which are basically masked) distortion pattern that will sound clean.
Once you get beyond a certain SPL it will really depend on how rapidly the devices distortion, particularly the high order harmonic distorion rises and how that still compares to the ear/brain masking...complicated but can go a long way to explaining a lot of the subjective varaiabilty observed.
"But, the big unanswered question is whether a low distortion system is somehow worse? We don't know because they don't detail the transfer functions! "
It is not low distortion per se, it is the fact that there are high order harmonics without lower order harmonics to provide masking. The audibility seems to be down to very low levels. Achieving this with negative feedback has it's issues with regard to artificial noise floor (read Crowhurst and Pass white paper) and IMD that gets generated. Just seeing absent harmonics in a static test does at all tell the whole story.
:)
I guess simplistic metrics work for simplistic folks.
Denialism of measurements and graphs will only get you so far! In fact, it could even be revealing of a. . . uh. . . simplistic mind set. ;-)
Belief in simple interpretation of measurements will only get you so far.
But, in general, I believe that there IS a correlation between the sound and most measurements.
Believe what you want. Unless the data is processed with an algorithm like the Gedlee metric or what Cheever came up with, you will not find a correlation between what people like and measurements. This was known as far back as the 1930s and 40s, because D.E.L Shorter from the BBC was trying to find such an equation to correlate the listening impressions with the measurements.
Raw THD and IMD numbers will not give you a good agreement with listening. Even looking at a raw FFT will not tell you too much. IMD also plays a big role but not the amount...the composition and that is very complex.
There is a program that allows you to add specific amounts of distortion to an audio file and then you can listen, if so inclined. I dabbled and found I could not distinguish what I thought were high levels of harmonics so I went back to just enjoying music :) And, as I've noted a few times recently, I have become primarily a vinyl listener in the past 2-3 years and LP reply is rife with non-idealities (that puts some of arguments about digital artifacts to shame) but it sounds really nice (because of vs. in spite of?)
Measurements in high-end audio are a moveable feast - a designer can use them if they show the product in a good light or declare them fake if they don't. I have the 'comfort' in my design day-job of having specs so I know when I am done. My head would like to solve all objective-subjective related discrepancies but my heart likes to think there are unsolved mysteries.
One final point, measurements - whether you think them valid or not - are usually taken with calibrated equipment. Audiophiles' opinions are never calibrated :)
Very interesting software. Keith Howard made something simpler many years ago and I tried and found I could clearly hear differences between different distortion patterns.
What would be really interesting would be to simulate existing amp profiles and see if one can hear the difference and then if one has access to those amplifiers, see if there is a similar response with the real thing rather than a simulation. My guess is that there are other things the software doesn't account for, like back EMF of a speaker for example.
| 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 ]
