Audio Science: Does it explain everything about how something sounds?
- Thread starter PeterA
- Start date
- Status
Better than THD but unfortunately most measurements even from manufacturers is THD as it is easy to read and been with us for decades.
Also it is compounded by number of harmonics involved (requires multiple ones), frequency, and their level for IMD (it only really stands out at higher frequencies) - not disagreeing just that unfortunately it would be incredibly complex data and test to be able to correlate and model.
How many loudspeaker manufacturers release their very advanced laser and vibration-Q measurements in relation to its performance-behaviour-spec?
Cheers
Orb
I don't know how cross-over distortion relates to IMD, but it is known since decades that even minuscule amounts of IMD can have nasty aural effects -- in the low range of where that link that I posted,
http://www.lindos.co.uk/test_and_measurement/SOURCE=Articles/SOURCE=Articles|VIEW=full|id=5
claims that crossover distortion is audible.
Edit: just checked, IMD may not be as bad as often reported (?).
Actually I think there are more SACD's being released now (the past 2-3 years) than at any time since the first couple of years of its introduction. Not a lot, but some excellent music.
Amir, You continue to misrepresent and correct what I never said. Please point to where I dismissed Dr. Toole's research or denigrate his work. In fact I specifically put this in my post #1067 hoping to avoid this kind of misinterpretation.
And I'm repeating myself here with you, all my comments are related to that one particular video of his speech and whatever college it was at many threads ago when we had this back & forth. If you want to pretend fine, in many people's book spending 45 minutes talking about one particular company that he has ties to, knocking other manufacturers, throwing out numbers calling their products over priced and closing with this marvelous Harman giant killer for only $2k is way beyond just appearance of bias.
While you're at it, aside from my comments about the relative value of the ABX speaker test shown on that video did I oppose any of his work.
david
By the way, this is the perfect reply to those that claim that audio science is able to explain everything, and that we can measure everything.
If we really knew how to measure everything that's important then we would know, based on that, how to build equipment, make recordings, and reproduce those recordings to yield perfect copies of the live event.
That's your opinion of the video which unfortunately based on incorrect statement of what it is, and what was said in there. As I said, I have gotten so tired of repeating over and over again the simple facts about the research and researchers and proceeded to give you regardless, 100% benefit of the doubt that these are a bunch of corrupt corporate shills. Now just present to us some technical research that is devoid of that so that we know where you get your scientific knowledge. Don't keep rehashing this stuff when the facts are not even correct. See example below.
There was no bloody ABX test in the video. No loudspeaker testing done by Dr. Toole whether at NRC or at Harman is ABX. ABX testing is for identification of any difference existing in sound of two devices. Loudspeakers obviously sound different from each other so there is no reason at all for putting them through an ABX test. How can you be such a vocal critique of the work when the ABCs are formal testing is not understood???
Not an opinion! Everything I mentioned is directly from that video. Brand, money, value, comparisons with competitors and product promotion are all there. Do I have to post minute by minute play?
You should be getting tired and hopefully tired enough to stop misrepresenting what I write, and I certainly don't need a lecture on research or researchers and I don't intend to expand to the entire research community, let's stay with Dr. Toole. You're making stuff up to come at me Amir, instead point to where I called him a shill, fraud or whatever name. You like facts, show me where was his research and work opposed and denigrated by me. Seeing this particular video as a marketing effort isn't a put down of the man or his work neither is mistaking terminology a distortion of facts.
From my understanding ABX test is a double blind test, which I see Sean Olive's polling as. If it makes you happy we can call it Olive's Hide The Speaker Behind The Curtain Test or OHTSBTC to clarify the ABCs.
I critique because I have a brain and the experience not to be blinded , impressed or intimidated by titles or anyone screaming at me!
david
Nelson Pass pops up a lot because he is up there as a well respected engineer and builds great products (both high end and 'budget' or diy orientated), it is not just him and his designs/transistor selection but also others and goes beyond a simple summary for thermal compensation (already mentioned in the past some aspects that would differentiate to that).
Cheers
Orb
Last edited:
One thing I have harped on about is noise & noise modulation in audio. I recently came across this loudness standard from following one of ack's links (thanks ack).
It's the ITU-R 468-weighting curve into the perception of noise in audio which came out of BBC research dept ''The Assessment of Noise in Audio Frequency Circuits''
So what we have here is a measurement that is much more useful as a measurement for distortion in audio that has a good subjective correlation - a much better correlation than THD + N gives - we see quite a greater sensitivity between 1Khz & 10Khz to noise than we have to tones - peaking at 12.2dB difference @ 6KHz . Why isn't it widely used? Maybe this excerpt from the article answers this "Engineers in the USA never 'caught on' to 468-weighting, probably because for many decades they were part of a strong independent manufacturing economy that tended to import little from abroad. For the same reason they never adopted the PPM (Peak programme meter), which also came out BBC Research. Nevertheless, 468-weighting is still demanded by the BBC and many other broadcasters, and knowledge of its existence and validity needs to spread. It is superior in allowing fair comparison of specifications for all types of equipment, which A-weighting cannot do because of differing noise characteristics. "
Secondly, it highlights something else I have been saying for a while - the JNDs or thresholds need to be revisited (I didn't know about this work). If we invert the graph below, we get the loudness contours as related to noise perception.
I'm reminded of Amir's graph where he showed that 16bit dynamic range of 96dB was just slightly shy of the full range needed to handle the dynamics of music - 20bits being acceptable. With this new (to me) knowledge of audibility (12dB lower than the Fletcher-Munson equal loudness graphs he used) should this not be revisited & the whole concept of dither noise opened up?
Some exerpts:
"In fact the human ear responds quite differently to noise, and it is this difference that gave rise to the 468-weighting, which arguably is the only valid weighting to be used for all noise measurements, whether on audio equipment or in the assessment of low-level environmental noise."
"The CCIR curve differs greatly from A-Weighting in the 5 to 8 kHz region where it peaks to +12.2 dB at 6.3 kHz, the region in which we appear to be extremely sensitive to noise.....the difference probably relates to the way in which our ears analyse sounds in terms of spectral content along the cochlea. This behaves like a set of closely spaced filters, which, if they had constant 'Q' would have bandwidths proportional to their centre frequencies. High frequency hair-cells would therefore be sensitive to a greater proportion of the total energy in noise than low frequency hair cells."
"468-weighting is also used in weighted distortion measurement at 1 kHz. Weighting the distortion residue after removal of the fundamental emphasises high-order harmonics, but only up to 10 kHz or so where the ears response falls off. This results in a single measurement (which Lindos refer to as Distortion Residue measurement) which corresponds well with subjective effect even for Power Amplifiers where crossover distortion is known to be far more audible than normal THD (Total harmonic distortion) measurements would suggest."
Edit: BTW, there are also many good articles on this site but one in particular relates to the above use of 96dB for CD dynamic range "Signal-Noise and Dynamic Range"
"The 96dB Myth
Audiophiles may talk in terms of 96 to 120dB dynamic range, but they often fail to refer to any measurement standard, making the figures meaningless. Attempts to calculate the dynamic range of digital audio on the basis that 16 bits represents a ratio of 65000:1 or 96dB are invalidated by the fact that the full digital count represents the peak possible level, rather than the rms equivalent of the maximum possible sinewave, while the minimum count of one has little to do with the noise level, which depends on the type of dither (or noise-shaping) used. They also fail to take any account of weighting for subjective validity. "
It's the ITU-R 468-weighting curve into the perception of noise in audio which came out of BBC research dept ''The Assessment of Noise in Audio Frequency Circuits''
So what we have here is a measurement that is much more useful as a measurement for distortion in audio that has a good subjective correlation - a much better correlation than THD + N gives - we see quite a greater sensitivity between 1Khz & 10Khz to noise than we have to tones - peaking at 12.2dB difference @ 6KHz . Why isn't it widely used? Maybe this excerpt from the article answers this "Engineers in the USA never 'caught on' to 468-weighting, probably because for many decades they were part of a strong independent manufacturing economy that tended to import little from abroad. For the same reason they never adopted the PPM (Peak programme meter), which also came out BBC Research. Nevertheless, 468-weighting is still demanded by the BBC and many other broadcasters, and knowledge of its existence and validity needs to spread. It is superior in allowing fair comparison of specifications for all types of equipment, which A-weighting cannot do because of differing noise characteristics. "
Secondly, it highlights something else I have been saying for a while - the JNDs or thresholds need to be revisited (I didn't know about this work). If we invert the graph below, we get the loudness contours as related to noise perception.
I'm reminded of Amir's graph where he showed that 16bit dynamic range of 96dB was just slightly shy of the full range needed to handle the dynamics of music - 20bits being acceptable. With this new (to me) knowledge of audibility (12dB lower than the Fletcher-Munson equal loudness graphs he used) should this not be revisited & the whole concept of dither noise opened up?
Some exerpts:
"In fact the human ear responds quite differently to noise, and it is this difference that gave rise to the 468-weighting, which arguably is the only valid weighting to be used for all noise measurements, whether on audio equipment or in the assessment of low-level environmental noise."
"The CCIR curve differs greatly from A-Weighting in the 5 to 8 kHz region where it peaks to +12.2 dB at 6.3 kHz, the region in which we appear to be extremely sensitive to noise.....the difference probably relates to the way in which our ears analyse sounds in terms of spectral content along the cochlea. This behaves like a set of closely spaced filters, which, if they had constant 'Q' would have bandwidths proportional to their centre frequencies. High frequency hair-cells would therefore be sensitive to a greater proportion of the total energy in noise than low frequency hair cells."
"468-weighting is also used in weighted distortion measurement at 1 kHz. Weighting the distortion residue after removal of the fundamental emphasises high-order harmonics, but only up to 10 kHz or so where the ears response falls off. This results in a single measurement (which Lindos refer to as Distortion Residue measurement) which corresponds well with subjective effect even for Power Amplifiers where crossover distortion is known to be far more audible than normal THD (Total harmonic distortion) measurements would suggest."
Edit: BTW, there are also many good articles on this site but one in particular relates to the above use of 96dB for CD dynamic range "Signal-Noise and Dynamic Range"
"The 96dB Myth
Audiophiles may talk in terms of 96 to 120dB dynamic range, but they often fail to refer to any measurement standard, making the figures meaningless. Attempts to calculate the dynamic range of digital audio on the basis that 16 bits represents a ratio of 65000:1 or 96dB are invalidated by the fact that the full digital count represents the peak possible level, rather than the rms equivalent of the maximum possible sinewave, while the minimum count of one has little to do with the noise level, which depends on the type of dither (or noise-shaping) used. They also fail to take any account of weighting for subjective validity. "
Last edited:
John, why are you citing all that research, which is based entirely on scientific quick switch blind controlled testing? The very thing you abhor and completely reject over "long term" uncontrolled non-blind experience.
Isn't that a bit odd? ITU audio standards are blind test derived.
To a small boutique manufacturer, such a presentation may hope that a couple of retailers in the room sign a contract. To a company like Harman, it is reluctantly allowing their top research scientist to present an overview of their research methodology to a roomful of industry insiders. The hope, I suspect, is that those insiders will continue to believe that pretense and prestige will continue to sell more product than a better understanding of what drives consumer preference. So far so good...
YouTube videos? Please. Marketing, in Harman's league, is about message, reach and frequency. None of which are properly achieved in a YouTube video of research presentation. You seem to desperately want to believe this is hucksterism, not legitimate research. I'm sure no amount of reality is going to change your mind.
Tim
Tim
Plackback systems don't capture anything. Recordings do. Playback systems can't play back what wasn't captured in the first place, especially stereo systems that audiophiles try with.
For the third time: http://www.aes.org/e-lib/browse.cfm?elib=9136
And an audiophiles subjective experience of this science based system: http://www.onhifi.com/features/20010615.htm
Your understanding is incorrect. In an ABX test, you have two known samples, A and B. You are then presented one of them at random ("X") and you are asked to tell whether it is A or B. All you are doing there is showing whether you can identify X as being A or B. If there is an audible difference to you between A and B, then you can identify X as being one of them. If not, you can't.
Loudspeakers are different sounding to everyone. Therefore if you put two of them in an ABX test, everyone would be able to guess what X is, confirming what we already know. That is the fact that they sound different. And at any rate, the results of ABX tests is binary. Either you could reliably identify X as A or B, or not.
Loudspeaker preference tests as run by Dr. Toole, Olive and crew are preference tests. They play each loudspeaker and ask the listener to score how good they think they are. The testing is double blind. But the mere fact that it is double blind in no way or shape makes it ABX.
What makes me happy is if you are going to consider yourself an authority as to be able to criticize this work, to spend some time learning the science and properly reading and understanding the research so that I don't have to keep explaining the simplest concepts in this field as I had to do yet again above.
Your experience is not of scientific nature. If it were, after asking you three times you would be sharing that with us. Sighted ad-hoc evaluation has been shown to mislead in this field. So believe in your experience and brain but please don't ask me to put any value on it. The research community does not. And so I am not either.
Yes you do. If you are going to accuse research as being all of those things, then you better dot every i and cross every t. I would imagine you would expect that to happen if this was done to you. This is not it:
Here is the snapshot of the Youtube page at the start of the video:
This is what we learn from that just one look:
1. Dr. Toole is an invited lecturer.
2. The invitation is from "CIRMMT." Who is CIRMMT? This: http://www.cirmmt.org/
"CIRMMT is the Centre for Interdisciplinary Research in Music Media and Technology, housed at the Schulich School of Music at McGill University." MacGill University is one of the few schools that teaches audio science. It is not just some random "audio engineering college."
3. The video is posted by CIRMMT, not Harman. Indeed Dr. Toole was surprised when I asked him a question about the Infinity Loudspeakers in the thread where we discussed this video. He did not know the video was online.
4. Dr. Toole's presentation was considered authoritative, and sufficiently non-commercial for a major college to invite him as a distinguished lecturer. But somehow, on a forum, we are to think that our standard is higher and this is some marketing peace? Give me a break, please.
5. Dr. Toole is a consultant to Harman, not an employee any more (retired a few years ago). If you listen to his lectures in person, you see that he actually goes against some of the Harman products like Mark Levinson.
No. You are the one not reading what I post. I said that *I* am tired of dealing with these pedantic non-technical posts where I have to go and spend time taking snapshots of Youtube pages just to demonstrate their wrongness. That if you want to cast doubt on motives of Dr. Toole and crew, let's assume the worst case of that and move on. Show where that commercial motivation has corrupted the research. After all, that is all we should care about because the lesson is not to go and buy Harman products but to believe in Harman research. That you want a loudspeaker with good direct and indirect response. That you want a loudspeaker that is free of resonances. That specific measurements of loudspeakers highly correlate with listening preferences. If you can show how corporate motivation has led to these being incorrect statements, let's see you demonstrate that rather than hoping that the implication of commercial motive gets you there. You have done that three times so far. Let's hope the next post is technical.
Hi Amir,
Can I ask you a question about your comments to David, above?
I'm not an expert at anything, nor any particular sort of authority. I have no technical training, nor a degree in anything related to the hard sciences. You keep asserting your expertise and background, and seem to continue to be disparaging of those of us who cannot match your experience. It seems your patience is wearing thin with those of us who believe - for whatever reason - there may be things current research doesn't tell us (the unknown unknowns), and would rather just have discussions with peers you consider your equal.
But it seems kinda arrogant, dismissive and unwarranted. This is just an impression I'm getting. I could be wrong.
So my question is, on this forum - and in particular, this thread - is there room for people like myself who do not fully understand Harman's research, have questions about it, have indirectly-related experience in other fields, and are investigating a heuristic line of enquiry despite the fact we may have little else to offer?
Of would you rather us just **** off?
'Cause that's the vibe and getting, and to be honest, part of the reason this forum is becoming less and less interesting to me.
If you look back at your poll results, 41% of us "Like participating in discussion of audio science even though (we) mostly rely on (our) ears." That's the majority. That's me. But I'm more and more getting the feeling that you'd prefer to keep company with those who ascribe to your own views, and take on only those whom you dignify worthy of your time.
Again, I could be wrong, but I think I may not be alone on this.
Looking forward to your response.
Cheers.
There is extant theory that says that linear systems can be explored by sine waves, but that non-linear systems can not. I mentioned this in my previous post. I will confess that my reference was somewhat cryptic. This was deliberate on my part, as I was attempting to calibrate the knowledge of parties to the discourse.
Find your own counter examples. If you are not able to do so, I am not going to argue with you, since you probably wouldn't understand the argument.
- Status
