proof only exists in mathematics. Elsewhere we at most have evidence which is more of a statistical concept: how unlikely is the outcome assuming a null hypothesis to be true. Eg null hypothesis: drug A is just as effective as drug B. Hypothesis can only be invalidated not proven. But they can the strengthened by existence of accepted causal relations. An extreme example is the question of a parachute prevents death: we have no double randomised controlled trials but accepted laws of physics tell us that the chute reduces the speed to a survival levels
pureAM and pureFM have the same amplitude spectra within 0.005dB. Only phases differ. Is this still questioned?
pureAM and pureFM have the same amplitude spectra within 0.005dB. Only phases differ. Is this still questioned?
But Mark thinks it should be because there really are "audio gods" living among us, as he tries to implicated over and over again.
Then please tell us what the setup detail would be in such listening session. Without it, there is no listening that can be talked about because some would listen from the kitchen, some from living room, some from a dedicated music room, some through headphones, ...etc. It needs a standard or it's not debatable.
How is proof of audibility produced without listening?
What was the standard listening setup for lrisbo files? Maybe we could use the same for resistors?
Last edited:
audibility is not a mathematical assertion and therefore cannot be proven. Listening tests are experiments and must be analysed statistically. there is always the possibility that the outcome is due to random uncertainty
Tell that to millions judicial cases going on throughout this world.
Audibility to human has been around long before mathematics was born.
By the way, how is your audio business doing this year?
Absolutely is questioned!
If we are to believe textbooks and relevant scientific research, human auditory perception is insensitive to phase but is very sensitive to amplitude.
Phase modulated (PM not FM designation as used) is perceived as ordinary constant tone because we don’t hear phase modulation. On the other hand, other tone is amplitude modulated, changing its power or amplitude if you prefer, 43 times per second, but on average maintaining the same average power or average amplitude. Every person not completely deaf will notice this amplitude change, perceived as tone warble.
Now, you present this as a mystery as “both tones have the same amplitude spectrum”. For what purposes? I’m surprised to find that Purifi is using such smoke and mirrors show to boost marketing claims.
There is no claim of mystery and never has been. There was an observation made that phase-free FFT analysis is widely used and commonly interpreted in terms of the assumption of a slight curvature in a fixed transfer characteristic. None other than Scott Wurcer pointed out that there is overreliance on the one measurement technique. That was after Scott retired as ADI VP of Engineering and ADI Fellow, so no audio business was involved in his remark.
Did you bother to read what the Purify article said about popular FFTs?
EDIT: Regarding some other misunderstanding seen in this thread, legal proof is a separate concept from mathematical proof.
Last edited:
Are you questioning the mathematical statement about the signals having the same amplitude spectra? that would be interesting
Who said it is a mystery and what marketing claims do you refer to?
Amplitude spectrums are the same only on average. You ignore fact that spectrum analyzer displays only average value while AM signal is changing 43 times per second and that is what makes audible difference.
For Judging speakers. On speaker forums. Nothing about resistors, certainly nothing about DACs. You appear to be trying to use a blog by Lars to prove all sorts of things which it just doesn't. When the time domain signal is clearly different to the naked eye there is little to argue about. this really does not support your glom and listen reports into DACs and the like. Or the gloop filled benchmark etc.
For a laugh have a look at the modulation on a test tone on a record courtesy of B&K back in the 70s. And some people whose opinion on resistors you trust say vinyl is better...
Attachments
You might want to read the blog again. I know its more fun to leap in without checking the background, but all the same...
Irisbo's reply quoted above was posted immediately after your post quoted above that.
Audio or legal court, the word is the same in its definition, which you and other audio business people are trying to play the game of.
As for post editing, what's with your frequent editing?
I thought he was pointing out that the spectrum analyzer therefore leaves ambiguity about the time-domain nature of waveforms, which therefore can be misinterpreted? In other words, it sounded to me like a warning not to be too complacent about common FFT measurements.
To use computer parlance, words are often 'overloaded' by having multiple definitions. That applies to the word "proof" as well as many other words.
Why than insisting it this thread that only difference between the same amplitude spectrum AM & PM signals is phase and nothing else, and yet they sound different?
IME the best of vinyl can be better in some ways that much of common digital reproduction, at least to human ear/brain system. From a typical measurement perspective vinyl is not very good. Gets back to what we were discussing earlier about measuring the wrong thing(s).
Good question. IMHO the interesting thing is that if they files sound different then then some well known and commonly accepted psychoacoustic principles are violated: 1) Phase rotation is inaudible, 2) masking theory always holds true.
legal proof:
In a criminal case, the prosecution bears the burden of proving that the defendant is guilty beyond all reasonable doubt. This means that the prosecution must convince the jury that there is no other reasonable explanation that can come from the evidence presented at trial.
Again, this is the same as evidence, ie a probabilistic assessment not a proof in the mathematical sense. eg some other person with a DNA match could be the murder
In a criminal case, the prosecution bears the burden of proving that the defendant is guilty beyond all reasonable doubt. This means that the prosecution must convince the jury that there is no other reasonable explanation that can come from the evidence presented at trial.
Again, this is the same as evidence, ie a probabilistic assessment not a proof in the mathematical sense. eg some other person with a DNA match could be the murder
No, my question was why that claim while phase is not the only difference. There is no question why AM signal sounds different.