http://www.wickeddigital.com.au/index.php/...eally-necessary (http://www.wickeddigital.com.au/index.php/2011-09-04-22-32-08/industry-interviews/189-high-resolution-is-it-really-necessary)
Sets out what is described as a White Paper by Marco Manunta of M2Tech. It contains the following passage (my emphasis)
"But there is something more. It’s known by signal processing experts, and absolutely not popularized amongst music lovers, that converting an analog signal into a discrete-time one (as it happens when converting from analog to digital) destroys the phase information in the two top octaves of the resulting spectrum. In a CD-standard digital recording, all phase information are lost from 5.5kHz up to 22kHz,"
Can any signal processing experts comment on whether this is true or not, and if so why. As a mere layman I am not able to understand why it even might seem be true.
It's complete nonsense. But hey, it's the internet. People can say anything and someone will believe it.
Can any signal processing experts comment on whether this is true or not, and if so why. As a mere layman I am not able to understand why it even might seem be true.
Hes basically just saying that he doesn't know what phase means.
http://www.wickeddigital.com.au/index.php/...eally-necessary (http://www.wickeddigital.com.au/index.php/2011-09-04-22-32-08/industry-interviews/189-high-resolution-is-it-really-necessary)
Sets out what is described as a White Paper by Marco Manunta of M2Tech. It contains the following passage (my emphasis)
"But there is something more. It’s known by signal processing experts, and absolutely not popularized amongst music lovers, that converting an analog signal into a discrete-time one (as it happens when converting from analog to digital) destroys the phase information in the two top octaves of the resulting spectrum. In a CD-standard digital recording, all phase information are lost from 5.5kHz up to 22kHz,"
Can any signal processing experts comment on whether this is true or not, and if so why. As a mere layman I am not able to understand why it even might seem be true.
The Nyquist theorem (which is mathematically proven) says that the exact waveform can be reproduced if the original signal is frequency limited to less than half the sampling frequency. The word "exact" gets a little shaky if the initial assumptions aren't met (example: each sample is taken exactly on time.)
"But there is something more. It’s known by signal processing experts, and absolutely not popularized amongst music lovers, that converting an analog signal into a discrete-time one (as it happens when converting from analog to digital) destroys the phase information in the two top octaves of the resulting spectrum. In a CD-standard digital recording, all phase information are lost from 5.5kHz up to 22kHz,"
Can any signal processing experts comment on whether this is true or not, and if so why. As a mere layman I am not able to understand why it even might seem be true.
Of all the nonsense I've read about the sampling theorem, this has to be close to the top of the list! I would also add that I wish there *was* such an easy way to trash phase information in a signal. There would be plenty of audio applications for that (starting with better lossy stereo/surround coding).
Their website is filled with various other nonsense, too. No surprise, since they're in the business of selling under-performing hardware at over-the-top prices (http://en.wikipedia.org/wiki/Snake_oil).
[OT]Thank you adamdea, now we all need a 5.6MHz/1mbit iPod.[/OT]
http://www.wickeddigital.com.au/index.php/...eally-necessary (http://www.wickeddigital.com.au/index.php/2011-09-04-22-32-08/industry-interviews/189-high-resolution-is-it-really-necessary)
]...destroys the phase information in the two top octaves of the resulting spectrum. In a CD-standard digital recording, all phase information are lost from 5.5kHz up to 22kHz,"[/i]
Can any signal processing experts comment on whether this is true or not, and if so why. As a mere layman I am not able to understand why it even might seem be true.
I'm not claiming it was audible, but it is true that the
analog low-pass filter used in some very early 44.1 digital devices, introduced considerable phase shifting in the top octave (nearing 180 degrees at 20 kHz). However this was the fault of an analog circuit, not a digital one, and is no longer the case.
Humans aren't very sensitive to phase in the very top frequencies anyways, so it wouldn't matter.
However this was the fault of an analog circuit, not a digital one
What?! You mean digital is not teh eeeivil?!
Gosh, imagine being so afraid of positive progress in technology.
Thanks everyone, I thought it seemed like nonsense but I wanted to check unless it referred to some little known effect. I can understand that analog brickwall filters could mess with phase, but am relieved that there does not appear to be an insurmountable problem.
I may take this up with m2tech.
http://www.wickeddigital.com.au/index.php/...eally-necessary (http://www.wickeddigital.com.au/index.php/2011-09-04-22-32-08/industry-interviews/189-high-resolution-is-it-really-necessary)
Sets out what is described as a White Paper by Marco Manunta of M2Tech. It contains the following passage (my emphasis)
"But there is something more. It’s known by signal processing experts, and absolutely not popularized amongst music lovers, that converting an analog signal into a discrete-time one (as it happens when converting from analog to digital) destroys the phase information in the two top octaves of the resulting spectrum. In a CD-standard digital recording, all phase information are lost from 5.5kHz up to 22kHz,"
Can any signal processing experts comment on whether this is true or not, and if so why. As a mere layman I am not able to understand why it even might seem be true.
As others correctly observe, this is total nonsense. I've encountered this myth before many times. There are many convincing arguments and one simply involves looking at real world signals using a digital editor. You can see that the timing and phase of signals > 5 KHz are preserved.
There are some variations on it that are equally mythical. One has to do with the timing of pulses that appear on both channels but at slightly different times. The myth says that only time differences that are on the order of a sampling period can be recorded and played.
BTW there is an online store that is part of this site. It specializes in audio gear that allegedly has "High resolution". Everybody who is surprised should put on a pointed cap and sit in a corner! ;-)
Thanks everyone, I thought it seemed like nonsense but I wanted to check unless it referred to some little known effect. I can understand that analog brickwall filters could mess with phase, but am relieved that there does not appear to be an insurmountable problem.
I may take this up with m2tech.
Above some frequency in what we call midrange, the ears lose their ability to discern phase. This is because our ears are built like spectrum analyzers, but only what we call the real portion or amplitude is conveyed to the brain. If you wish to obtain phase information from a spectrum analyzer you need two independent kinds of (quadrature) data for each frequency band. Our ears only pass one kind of information to the brain for each frequency band above medium frequencies.
One consequence of this loss of phase information in our ear/brain interface is that massive amounts of phase shift (e.g. 1,000 degrees or more) can be applied to critical high resolution audio signals, with no discernible change in perception. The only caveat is that the phase shift applied to both channels must be essentially the same or else the phase shift will turn into response changes that will be audible. Above 5 Khz this situation dominates with total supremacy.
So even the old CD players with analog filters did have fairly well matched channels, and while what they did to phase was not numerically pretty, there wasn't any serious effect.
Another ugly thing that old CD players sometimes did is share the same DAC between the 2 channels so that their outputs were 1/2 sample time apart. If you electrically summed the two channels this led to a minor frequency response roll-off that on a really good day might be mildly audible.
massive amounts of phase shift (e.g. 1,000 degrees or more)
Erm, wouldn't that just be a 1000 mod 360 = 280 degrees phase shift?
Or maybe I should get me one of those pointed caps.
massive amounts of phase shift (e.g. 1,000 degrees or more)
Erm, wouldn't that just be a 1000 mod 360 = 280 degrees phase shift?
Or maybe I should get me one of those pointed caps.
For a pure tone yes, but if your waveform isn't periodic, phase does not wrap (and is usually expressed in seconds instead of degrees).
...One consequence of this loss of phase information in our ear/brain interface is that massive amounts of phase shift (e.g. 1,000 degrees or more) can be applied to critical high resolution audio signals, with no discernible change in perception. The only caveat is that the phase shift applied to both channels must be essentially the same or else the phase shift will turn into response changes that will be audible. Above 5 Khz this situation dominates with total supremacy.
[emphasis mine]
If I understand you correctly, you mean when listening
through speakers. Through headphones, where we get nearly perfect channel separation, the massive amount of phase shift you speak of
could be different, per channel, and we
still wouldn't notice it, since the two channels have no opportunity to alter the overall response (due to the waveform being reinforced/canceled when combined
in the air, at certain points). Did I get that right?
Detection of phase differences between channels does not go away by virtue of using headphones. If anything it is likely enhanced.
While I entertain getting to the truth on this specific issue, I fear it's getting off-topic. Not like there is much more to say on the matter that is on-topic.
massive amounts of phase shift (e.g. 1,000 degrees or more) can be applied to critical high resolution audio signals, with no discernible change in perception. The only caveat is that the phase shift applied to both channels must be essentially the same or else the phase shift will turn into response changes that will be audible.
That's what I would have said if you didn't beat me to it. The only other time phase shift is audible, even when the channels are matched, is if the amount of shift is currently changing. My AES Audio Myths Workshop (http://www.youtube.com/watch?v=BYTlN6wjcvQ) video demonstrates this. You can jump ahead to that part of the video at 47:45.
--Ethan
Detection of phase differences between channels does not go away by virtue of using headphones..
What goes away is the ability of the L and R waveforms to interact and constructively and destructively collide with each other in the air, causing a ripple effect in the overall frequency response of peaks and nulls, which he is pointing out
may theoretically in some circumstances, he stated, be audible. At least that's what I got out of it, but I'm sure he'll explain if I was right or wrong.
It is on topic to the opening post's query regarding phase anomalies in the top octaves:
destroys the phase information in the two top octaves of the resulting spectrum. In a CD-standard digital recording, all phase information are lost from 5.5kHz up to 22kHz,"
Which ABK and I are in agreement was "not pretty numerically" in some of the very early CD players:
So even the old CD players with analog filters did have fairly well matched channels, and while what they did to phase was not numerically pretty, there wasn't any serious effect.
And which I also originally pointed out wouldn't be expected to be audible anyways, before he even joined the conversation:
Humans aren't very sensitive to phase in the very top frequencies anyways, so it wouldn't matter.
...One consequence of this loss of phase information in our ear/brain interface is that massive amounts of phase shift (e.g. 1,000 degrees or more) can be applied to critical high resolution audio signals, with no discernible change in perception. The only caveat is that the phase shift applied to both channels must be essentially the same or else the phase shift will turn into response changes that will be audible. Above 5 Khz this situation dominates with total supremacy.
[emphasis mine]
If I understand you correctly, you mean when listening through speakers. Through headphones, where we get nearly perfect channel separation, the massive amount of phase shift you speak of could be different, per channel, and we still wouldn't notice it, since the two channels have no opportunity to alter the overall response (due to the waveform being reinforced/canceled when combined in the air, at certain points). Did I get that right?
I dunno. I certainly hear phase inversion in just one channel with a 2 KHz pure tone using my ATH M50 headphones.
The listening tests I am referring to were also done with headphones.
@Ethan:
I hope this isn't, "I tried it on a 1khz sine wave, therefore it applies to all signals."
I did not take the OP as meaning phase relation between stereo channels since it wasn't specified.
Again, phase differences between channels can quite easily be detected with headphones, perhaps even moreso than with speakers. Incoherent reflections from the listening environment that do not exist when using headphones are the very reason why.
I hope this isn't, "I tried it on a 1khz sine wave, therefore it applies to all signals."
LOL, no. Did you play that part of the video? I've tested this a lot, and found that phase shift in audio gear is a non-issue. Further, in typical gear phase shift occurs mostly at the frequency extremes, versus in the midrange where it's potentially more audible.
Related, this article offers some files to test the audibility of absolute polarity with explanations:
Dispelling Popular Audio Myths (http://www.ethanwiner.com/myths.html)
Scroll about 3/4 of the way down the page.
--Ethan
That's what I would have said if you didn't beat me to it. The only other time phase shift is audible, even when the channels are matched, is if the amount of shift is currently changing. My AES Audio Myths Workshop (http://www.youtube.com/watch?v=BYTlN6wjcvQ) video demonstrates this. You can jump ahead to that part of the video at 47:45.
Actually, I did a bit of research on that topic, for the purpose of decorrelating channels to help stereo acoustic echo cancellation. What I found was that you can make phase shift inaudible
even when channels aren't matched and while varying the amount of phase shift. What's important is only matching the channels below about 2 kHz, where the ear is sensitive to interaural phase difference (IPD). Beyond 2 kHz, you can get away with unmatched phase differences (which was the useful part of my research). As for changing the phase, the main difficulty is to do that without causing glitches or spectral distortion. With the help of some overlap-add, I was able to vary the phase shift without problem. Interestingly, when I tried varying the phase shift below 2 kHz, I could induce motion sickness just from listening to music with headphones.
See details in:
J.-M. Valin, Perceptually-Motivated Nonlinear Channel Decorrelation For Stereo Acoustic Echo Cancellation (http://jmvalin.ca/papers/valin_hscma2008.pdf), Proc. Joint Workshop on Handsfree Speech Communication and Microphone Arrays (HSCMA), 2008.
I dunno. I certainly hear phase inversion in just one channel with a 2 KHz pure tone using my ATH M50 headphones.
Try the exact same thing but at 15 kHz and I bet you won't.
You're being mean with old men (like me)
Actually, I did a bit of research on that topic, for the purpose of decorrelating channels to help stereo acoustic echo cancellation. What I found was that you can make phase shift inaudible even when channels aren't matched and while varying the amount of phase shift. What's important is only matching the channels below about 2 kHz, where the ear is sensitive to interaural phase difference (IPD).
Interesting, you've done more types of testing than I have. Also, when I said "the phase shift is changing" I should have been clearer. The type of phase shift effects I've tested vary the center frequency of the phase shift, not the amount of shift at a fixed frequency. So by "changing" I mean sweeping the frequency. Though I imagine changing the amount of shift at a fixed frequency would also be audible at midrange frequencies.
--Ethan
http://www.wickeddigital.com.au/index.php/...eally-necessary (http://www.wickeddigital.com.au/index.php/2011-09-04-22-32-08/industry-interviews/189-high-resolution-is-it-really-necessary)
Sets out what is described as a White Paper by Marco Manunta of M2Tech. It contains the following passage (my emphasis)
"But there is something more. It’s known by signal processing experts, and absolutely not popularized amongst music lovers, that converting an analog signal into a discrete-time one (as it happens when converting from analog to digital) destroys the phase information in the two top octaves of the resulting spectrum. In a CD-standard digital recording, all phase information are lost from 5.5kHz up to 22kHz,"
Can any signal processing experts comment on whether this is true or not, and if so why. As a mere layman I am not able to understand why it even might seem be true.
It is false. It is utterly false, absurd, and insane. Your cell phone, MP3 player, and cable modem could not work, for instance, if there was any truth to this utterly preposterious foolishness.
Above some frequency in what we call midrange, the ears lose their ability to discern phase.
Allow me to be more specific.
The ability to discern signal phase starts to degrade around 500Hz, and is almost completely gone by 2kHz, and completely gone by 4kHz.
HOWEVER, the ability to discern time delays in a non-stationary signal envelope at 15kHz is very good. So, time cues matter above 2kHz. A lot.
Actually, I did a bit of research on that topic, for the purpose of decorrelating channels to help stereo acoustic echo cancellation. What I found was that you can make phase shift inaudible even when channels aren't matched and while varying the amount of phase shift. What's important is only matching the channels below about 2 kHz, where the ear is sensitive to interaural phase difference (IPD).
Interesting, you've done more types of testing than I have. Also, when I said "the phase shift is changing" I should have been clearer. The type of phase shift effects I've tested vary the center frequency of the phase shift, not the amount of shift at a fixed frequency. So by "changing" I mean sweeping the frequency. Though I imagine changing the amount of shift at a fixed frequency would also be audible at midrange frequencies.
--Ethan
So, try delaying a 10kHz gaussian impulse with a 4khz bandwidth by half a sample in one channel. Tell me you don't hear that in headphones, eh?
Interesting, you've done more types of testing than I have. Also, when I said "the phase shift is changing" I should have been clearer. The type of phase shift effects I've tested vary the center frequency of the phase shift, not the amount of shift at a fixed frequency. So by "changing" I mean sweeping the frequency. Though I imagine changing the amount of shift at a fixed frequency would also be audible at midrange frequencies.
Actually, I was changing what you call the "center frequency" as well. If you remove the part of my filter that was designed to leave the lower frequencies alone, you have a "comb allpass" filter:
A(z) = (a + z^-N)/(1 - a*z^-N)
I was actually changing not only a, but also the order N to avoid the "nulls" in the phase change.
So, try delaying a 10kHz gaussian impulse with a 4khz bandwidth by half a sample in one channel. Tell me you don't hear that in headphones, eh?
If only I understood what that meant, or how to delay something by half a sample!
I agree that left-right phase differences are audible. I haven't tested this enough to be more specific, but I'm glad you and others have.
--Ethan
Delay by a half sample: up-convert to a higher samplerate, perform the delay and then down-convert.
Pretty simple!
So, try delaying a 10kHz gaussian impulse with a 4khz bandwidth by half a sample in one channel. Tell me you don't hear that in headphones, eh?
If only I understood what that meant, or how to delay something by half a sample!
I agree that left-right phase differences are audible. I haven't tested this enough to be more specific, but I'm glad you and others have.
--Ethan
Actually, I've no idea what a "10kHz gaussian impulse with a 4khz bandwidth" would be, but delaying something by half a sample (or any fractional number of samples) is very easy to do, e.g. using a sinc filter. One thing I should have been more precise about in my earlier comment... you can get away with modifying the phase as long as it's above ~2 kHz
and as long as the group delay (derivative of the phase response wrt frequency) is small enough. But as I said, as long as you're careful with how you do it, it's amazing how much you can get away with when playing with the phase. The real difficulty when the phase is involved is really how messing it up can cause other artefacts. e.g. changes in the actual power spectrum.
Delay by a half sample: up-convert to a higher samplerate, perform the delay and then down-convert.
Pretty simple!
Even easier, it's a gaussian impulse, it has no energy outside Fs/2 to speak of, so you can shift it analytically.
it's of the form
e^(((t-t0)^2))/c)*sin(2*pi*f*(t-t0))
So you just change t0.
Oh, you math guys...
Though I do understand up-sample, delay, down-sample.
--Ethan