No matter how many times you repeat this it is still falsehood based on your lack of understanding. The difference is not caused by phase (i.e. phase shift) but by change in phase modulation. There is nothing extraordinary in that the resulting AM modulation is audible.
What I said is the files were produced by phase rotation. Why the brain hears them as different could possibly be because the phase rotation in frequency domain changes the time domain waveform. It could be that the time domain waveform is more salient to brain auditory processing in this situation than is the associated phase shift as viewed in the frequency domain. What some people seem to be forgetting is that the time domain view and the frequency domain view are two different ways at looking at the exact same thing. They are equivalent.
EDIT: In case not everyone understands this: Not all frequencies were phase shifted, only some of them. That's why the time domain waveshape changed.
Last edited:
Time domain and frequency domain are means to analyse signals. They have nothing to do with hearing. According to studies hearing system is sensitive to amplitude and pitch but not to absolute phase. So there is nothing extraordinary that amplitude modulation caused by phase modulation may be audible.
It was long ago postulated that the brain processed sound much like an FFT, in the frequency domain. That probably turns up in at least some of the psychoacoustics literature. We now are learning that brain auditory processing is more complex than once thought. One can search for the term "Auditory Scene Analysis" to get more recent information.
EDIT: Regarding old studies of phase perception, IIUC it was done with simple sine wave signals, not more complex sine wave combinations like in lrisbo files. Again IIUC it was assumed that hearing was linear and time invariant enough to that simple sine wave signal research should extrapolate to more complex combinations of sine waves. The assumption now appears to be not perfectly correct.
EDIT: Regarding old studies of phase perception, IIUC it was done with simple sine wave signals, not more complex sine wave combinations like in lrisbo files. Again IIUC it was assumed that hearing was linear and time invariant enough to that simple sine wave signal research should extrapolate to more complex combinations of sine waves. The assumption now appears to be not perfectly correct.
For the Nth time: hearing the difference in Purifi files is not about phase perception but amplitude perception as the other file has amplitude modulation.
Okay, but the effect was still produced by phase rotation in the frequency domain of some frequencies, and not others. Apparently that can be equivalent to modulation in the time domain.
Remember, all we have in the frequency domain are sine waves with properties of amplitude, frequency, and phase. In the time domain we have amplitude and time. Producing amplitude modulation in the time domain has to be equivalent to adjusting some of the frequency domain properties.
https://www.dspguide.com/ch10/5.htm
Remember, all we have in the frequency domain are sine waves with properties of amplitude, frequency, and phase. In the time domain we have amplitude and time. Producing amplitude modulation in the time domain has to be equivalent to adjusting some of the frequency domain properties.
https://www.dspguide.com/ch10/5.htm
Last edited:
Who are these people? Are you once again confusing IMD spurs 20dB below the carrier from those 100dB below? You are trying to claim you should be able to feel a virus land in your nose because you can detect a brick landing on your head.
Bill, it is uncontroversial that some people can hear quantizing distortion nominally down around -93dBFS. Not a big deal. Use a resistor to return feedback from the output to to the inverting input and its in a very sensitive spot. Also, ESS claims the ear is "exquisitely" sensitive to signal correlated noise. If some people claim to hear the sound of some audiophile resistor, that seems more plausible to me that feeling a virus landing on your nose. Besides I'm not saying the audible effect of a resistor, if any, is more than subtle. In addition, I did an experiment and heard a difference (compared 4.7k Dale RN55 to run of the mill Yageo; preferred the latter). To my perception the difference in that application was obvious, no need for DBT. Doesn't mean you will believe it too, unless perhaps you someday have a chance to hear it for yourself.
Also, I didn't listen to lrisbo files. Don't know if there is any obvious difference. Why should I trust you?
Also, I didn't listen to lrisbo files. Don't know if there is any obvious difference. Why should I trust you?
Last edited:
which is marketing BS until some evidence arises to say otherwise.
Why an audiophile resistor? why not just a different well designed industrial transistor. Read any precision analog guide and you'll see that a 90degree rotation of one resistor can make a measurable different. Tomchr showed a simulation of a cm of pcb making a difference to distortion. All Engineering and no flooby dust. One day you might come up with something with a basis in the real world hopefully?
I've NEVER said my ears and brain can be trusted. You are the one claiming feats of listening skill.Also, I didn't listen to lrisbo files. Don't know if there is any obvious difference. Why should I trust you?
Just guessing audiophile resistors may worth taking a look at since that's what most resistor swappers seem to think is good. OTOH I just used some resistors I had on hand, nothing special.
Regarding engineering measurements, reasonably within grasp given a well-equipped lab and careful technique. The harder part is showing convincing evidence of audibility, which is mainly because of the time, effort, and other resources already described by jaddie. Not how I want to spend my retirement days.
On the question of marketing BS, wasn't lrisbo accused of that a few posts back? Guess it depends on person making a judgment. For myself, I am biased because like you I am human. My bias is a little different though since I have already heard various effects I was once very skeptical of. That's why I wonder if you won't change you mind if you ever have an opportunity to hear some of the same stuff. Then again, maybe its like quantizing distortion, not everybody hears it.
BTW didn't you already say in some other thread you have heard a few systems where it didn't sound like the sound was coming from the speakers? Seems hard to believe.
Regarding engineering measurements, reasonably within grasp given a well-equipped lab and careful technique. The harder part is showing convincing evidence of audibility, which is mainly because of the time, effort, and other resources already described by jaddie. Not how I want to spend my retirement days.
On the question of marketing BS, wasn't lrisbo accused of that a few posts back? Guess it depends on person making a judgment. For myself, I am biased because like you I am human. My bias is a little different though since I have already heard various effects I was once very skeptical of. That's why I wonder if you won't change you mind if you ever have an opportunity to hear some of the same stuff. Then again, maybe its like quantizing distortion, not everybody hears it.
BTW didn't you already say in some other thread you have heard a few systems where it didn't sound like the sound was coming from the speakers? Seems hard to believe.
Last edited:
Was able to dig the script out and added a few comments. have checkked the files produced against the ones on our web. The difference in spectrum amplitude is vanishingly low: 0.005dB maximum at the 1500Hz side band already being at -70dBr, easily explained by the 16 bit quantization noise. The mean value is almost 100dB down. To for all practical purposes: same amplitude spectra only differing in the phases of the sideband components. Feel free to ask questions if you are not familiar with Matlab code
% script for AM/FM Doppler blog post, BP Nov 2019
clear;
N=65536; % length of signal
fs=44100;
t=(0:N-1)'/fs;
fmod=fs/1024; % modulation freq
fcar=fs/32; % carrier frequency
G=0.05; % modulation index
ymod=cos(t*2*pi*fmod)*G; % fcar cosine
ang=t*2*pi*fcar+2*pi*ymod; % modulated angle
PureFM=cos(ang); %FM signal
% calculate the AM signal with the same amplitude spectrum
% by taking the abs() of the FFT spectrum and inverse transform
PureAM=ifft(abs(fft(PureFM)));
% write audio files
audiowrite('PureFM.wav',[PureFM PureFM]*0.5,fs)
audiowrite('PureAM.wav',[PureAM PureAM]*0.5,fs)
Yes, exactly. It depends more on honest presentation and clear answers.
From the beginning of this x-files case, there were some smoke and mirrors present and clear answers not given. Why I have this impression? Ah, yes, it must be a “confusion of cognitive dissonance”.
It can’t be accepted before trained authority listens and issues a seal of approval. Travel expenses + lunch and you get lot of free advices. Mark, you seem qualified for job. What you say?
I do really think you are qualified but don’t accept all your advices. No joke here.
Wow, I thought I was reading Jakob(#)'s post, one who is in audio business.
Say, someone named Mickw4 claims that he can detect sound variations less than 0.5 db with his ears, he can either prove it or be debunked by demonstrating or not in listening test.
Uh oh, there's that word again.
The proper word is "perceive". Whether we actually hear it or not needs to be evaluated to confirm. Just because there is a perceivable difference doesn't mean it's audible difference. People perceive speaker cable difference just by being told that it's changed without physically changing it.
a listening test cannot prove as in mathematics but provide evidence, possibly so strong evidence that the question at hand is determined beyond any reasonable doubt. Depends on how strong the test is (number of independent repeats, panelists etc)
You are barking up the wrong tree. 0.5 dB is not something very exceptional. I have Muses volume control adjustable in 0.25 dB steps. My totally untrained perception believes 1 dB is clear difference in volume or balance control, but 0.5 dB is debatable. Someone devoted to listening techniques surely can do better.
I have measured resistors, published the results and had a few folks even follow up on my technique for measuring them. Since the new technique was shown some manufacturers are now offering products with as much as 5 times better performance. The down side is the IEC acknowledged my method, but stuck to the old standard to measure resistor distortion. Quite simply they didn't really understand it, calling it complicated! You can use my method with basically a computer, sound card and a small amplifier. It does not require special extremely engineered oscillators, amplifiers and filters.
Not surprisingly my measurement results frequently correlated with many experts opinions. More interestingly some folks had exactly the inverse rating opinions.
Now many here have offered their opinions apparently without doing any research or even scanning the existing literature. Why do informed folks bother arguing with them?
Not surprisingly my measurement results frequently correlated with many experts opinions. More interestingly some folks had exactly the inverse rating opinions.
Now many here have offered their opinions apparently without doing any research or even scanning the existing literature. Why do informed folks bother arguing with them?