I agree! The people who claim measurement equipment is more sensitive than hearing have obviously never used such measurements or thought about any measurements they have done. There is no such thing as a universal audio measurement that shows everything possible - all measurements are limited by the test signal used & the limitations of the measurement. FFTs are a wonderful tool but what is the test signal used with them - typically a single or dual tone sine wave. I remember Scott Wurcer saying that he started using multitone tone signals (40 or more sine waves at different but specially selected frequencies) on DAC modules & it sorted out the wheat from the chaff. Maybe he wants to talk about this?
The point is that measurements have to be designed with a goal in mind - using the appropriate test signals & appropriate measurement to reveal what is being searched for - why did the multitone tests reveal more than the previous tests in Scott's case?
In the above example, the premise is that low level signals are being masked by low level noise which is only present when a complex, dynamic signal is being processed by the electronics - it will not be seen when a single or dual tone signal is present (hints of it may be seen in a multitone test?)
So who is testing for this condition & what tests are they using? Those who say audio measurements show everything that we can hear could answer this
Masking.. Subtle signal with frequency very near to strong signal. It will create low level (-40dB) "beat" with 1Hz frequency . Impossible to hear due to masking.
These curves were derived way back from single tone testing - in the presence of complex signals with dynamically changing noise floor (as Dan & planet10 suggest is what's going on), one cannot make the bullish claim "impossible to hear due to masking"
I mentioned comodulated masking release (CMR) on the other thread which is a phenomena that occurs in nature with complex signals & might be worth checking out
Last edited:
But it is than problem of poor acoustic in bad enviroment, too high levels of reflected sound compared to direct sound, and change delays inter ears. It can do worsen localization, and cause changes in "timbre" (spectral content) due variuos reflection coefficient for various frequencies. It is depending on speakers directivity and acoustic conditions.
Any proof for such claim? Or only feeling (untrue..)? In presence of complex signal (multiple tones with changing frequency and level and also noise backround, it is only worse. In presence of complex signal (multiple tones with changing frequency and level and also noise backroung, it is much worse to differentiate than for "simple" two tones (masker + masked) signal..
"comodulated masking release (CMR)" - other name for intermodulation. Known, measurable, no technical problem.
But believe what you want.
Last edited:
I think he is talking about the room acoustic captured on the recording, not the playback room acoustic? When present & audible on recordings, such low level audible cues present our auditory perception with a more realistic illusion as that is what auditory perception is accustomed to perceiving in the natural world.
Again, I refer back to the basic understanding needed of auditory perception - it is an innate sense that develops from birth using the sounds of the real world as it's learning space & as a result has many internal models & rules for what is realistic (sounds like the real world would) sounding. It's ongoing task is to then analyse & interpret the waveforms impinging on the tympanic membrane with this blueprint as it's basis
Once , this is grasped & fully absorbed, lots of mistakes about hearing are clarified - our ears don't work like microphones & our auditory memory doesn't work like a recording - it is a dynamic processing engines working on an ill-formed problem because of the paucity of the signals & making the best sense out of this.
I probably edited/add to my post while you were replying but I gave the example of CMR which I went into in the other thread & gave a link to an interactive example you can try yourself. Essentially CMR reduces the masking thresholds that you show in those plots. You also have to realise that single tones are not handled by our auditory mechanism in the same way as broader band noise (which was what was being talked about here) - it's all got to do with the well known & studied hearing mechanism involving critical bands or ERBs & the signal power in each
BTW, your the one that made the claim - I said you can't be so bullish about such a claim
Than you need nothing more than accurately (amplitude, phase , spectra..) reproduce captured signal. What is relatively simple task . Psychoacoustic ( changes from person to person ) is quite different discipline and has nothing to do with audio components technical quality.
Such claim is simple implication of results of all known hearing perception ability studies..No one study (I do not know any..) prooved something what is here presented as possible for our perception.
Last edited:
Yes, that was what was being said - the accurate reproduction of low level signals is what is needed. But to your second point, no, this is not simple & is what differentiates the better playback systems from the crowd.
Psychoacoustics is the study of the ear/brain mechanism which gives us our auditory perception. It's funny that people try to make the claim that psychoacoustics (auditory perception) is different for everybody. If you measure the frequency spectrum at different peoples eardrum you will find huge differences between these spectra due to the differences in the shape of their pinnae (we all have unique pinnae shapes). Does this mean we all hear a hugely different version of real world sounds? Come on!
Yes, I know, to some it's really a shame & (they consider it a weakness) that we actually use our auditory system to judge the auditory illusion that our playback systems try to create - they want to just use a limited set of measurements to determine 'perfection' & ignore what our far more capable auditory perception is telling us.
As you said, you can believe what you want but I gave you the evidence you asked for and even an example to try for yourself - guess you didn't bother looking into these?
How do you define accurate - by measurement? What is being said is that no known measurement is showing what is claimed to be the audible difference that makes for a better playback system.
I make such systems for more than 28years. And accurate reproduction (to speaker terminals) is no problem for complex signals in dynamic range about 110dB, supposing good HD signal source (recording). This can be done much more accurate than anybody is able distinguish. What is not simple (or quite impossible..) is to beat many incorrect prejudications and feelings present as premise for conclusions in audio society..It is big task for technical community.
Maybee problem is in words "what is claimed to be audible". Measurements (no only one, isolated measurement) can show differencies in orders of magnitude bellow audibility treshold.
Last edited:
The idea of DDR still sounds to me like a form of Fourier denial, a small signal is added to a larger one, there is no mystery. Then the aim is to reproduce that signal with all the cues it contains, accurately. Anything else that is going on is in the listener's head, and I'm not interested in what goes on in people's heads, usually because they don't know, so how can I?
That does not mean I'm not interested in psychoacoustics, I am, I just don't think it's particularly relevantI make such systems for more than 28years. And accurate reproduction (to speaker terminals) is no problem for complex signals in dynamic range about 110dB, supposing good HD signal source (record). This can be done much more accurate than anybody is able distinguish. What is not simple (or quite impossible..) is to beat many incorrect prejudications and feelings present in audio society..It is big task for technical community.
I see you never bothered to look into CMR or the link I gave - I can give you far more technical links that suit your mindset
but if you prefer to remain in denial, that's fine tooThe idea of DDR still sounds to me like a form of Fourier denial, a small signal is added to a larger one, there is no mystery. Then the aim is to reproduce that signal with all the cues it contains, accurately. Anything else that is going on is in the listener's head, and I'm not interested in what goes on in people's heads, usually because they don't know, so how can I?
Have you ever performed an FFT?
What test signal & FFT parameters did you use?
What might be the limitations of yoru choices & of FFT itself?
If you prefer to deal in simple mantras like "Forier denial" & not do the scientific questioning needed to search for truth then there's no need to reply
I bothered, but found only this
and from incorrect premise can result only incorrect conclusion.. So give please more technical links.
Or try to do test with musical recording with added tone(s) with level about 40-60dB below signal ( modulation by envelope of music). It is not about the mindset, but about facts.
Last edited:
Care to explain what "incorrect premise" you are refrring to?
I will find the reference papers but first of all, try this working example which is just one example & type of CMR
Comodulation Masking Release | Auditory Neuroscience
The thing to remember is that this has been researched scientifically in animals, humans & the findings show that masking (which your plots for single tone tests show) is released when certain types & combinations of other concurrent tones are part of the soundscape - the sort of sound combinations heard in nature, where single tones are seldom if ever presented on their own. In any complex soundscape, including music, amplitude comodulation occurs very often between separate frequencies
My statement stands - you bullish statement about masking is simplistic & indefensible, based on known scientific research
Here's a starter research paper
"Comodulation masking release (CMR) enhances the detection of signals embedded in wideband, amplitude-modulated maskers. "
Physiological Correlates of Comodulation Masking Release in the Mammalian Ventral Cochlear Nucleus | Journal of Neuroscience
Last edited:
It always strikes me that what people mean who think psychoacoustics is irrelevant to this audio hobby, is that psychoacoustics is what people 'imagine' they hear
You could be wrong of course, it's not always easy to tell. To be clear I meant I thought it was irrelevant to this thread (I was wrong) not this audio hobby, far from it, that should be evident from my previous posts.
Last edited:
Psychoacoustics gives us subjective result (feelings) based on data (sound) and highly subjective processing (brain). So it has very little to do only with reproduction (data) accuracy, which is (should be..) base for audio hobby . And you can fool it so easy..
e.g "a tone of a certain frequency is masked by an amplitude modulated masker at that same frequency"
You can ofcourse stand by your statement, but this way it will not become true.
If it occurs with music, it must occur also for any two or multitone signal . It is produced same way.
Last edited:
OK, I was wrong in your case, sorry about that. It seems I was correct in BV's case, however
BTW, I asked you earlier about how you defined "accurate reproduction" of a signal. I also asked about "fourier deniers" & FFTS but I may have missed your answers?
Wrong - see Jakob's post
As I said, CMR applies to a tone at another frequency whose amplitude is modulating in unison with the amplitude modulation of the masking tone - this set of circumstances (comodulation) causes the masking to be greatly threshold seen in your plots to be greatly reduced.
Did you try the working example?
- Status
- Not open for further replies.