soongsc said:
I think it would make no difference whether the information was presented or not. Since it would be necessary to go through material more than 20 years old, and a pile of data. If someone has recently conducted tests to prove the contrary, it would be interesting to hear about. Then we can figure out whether its a problem with the tests, or if this generation is gradually suffering hearing lose in some way.
Sorry, if I read you right, you have it precisely backwards. If someone claims to have a perpetual motion machine, he cannot justify his claim by saying, "Well, nobody has published explicit proof of the law of mass/energy conservation in the past 20 years, so you can't claim that as an objection."
soongsc said:
Those are certainly enormous. Don't forget the enormous approximation in sampling the sound field at a few discrete locations, them mixing them together to be reproduced by discrete channels in a non-anechoic environment that is not specified.
Maybe that's why I don't lose much sleep worrying about 30kHz components to a waveform.
More information on the subjects of sampling frequency and resolution:
Several interesting papers by Julian Dunn. Of particular interest being '96 kHz Digital Audio' and 'High Dynamic Range Digital Audio' discussing theoretical benefits of increased sampling frequency and bit depth respectively.
In a similar vein, papers by Dan Lavry concerning the same issues: 'Do you need 20 Bits? ' and 'Sampling, Oversampling, Imaging, Aliasing'. He is selling something, but seems respectable.
There do seem to be several studies showing that human hearing may not be quite so limited to 20kHz, similar to the 'Inaudible High-Frequency Sounds Affect Brain Activity' article posted earlier. Frustratingly though, I can find virtually no good tests comparing sampling rates directly.
Likewise, there is plenty of information regarding the dynamic range of human hearing, but very little on the audibility of different bit depths. However, given that human hearing can be considered to have a dynamic range of 120dB (the difference between quietest audible sound and threshold of pain), then it does seem that the 96dB of CD audio might be considered inadequate to capture the maximum desireable range. It just so happens that 120dB is the theoretical dynamic range from 20bit, which is what was suggested as the optimum value in the link I posted earlier.
Several interesting papers by Julian Dunn. Of particular interest being '96 kHz Digital Audio' and 'High Dynamic Range Digital Audio' discussing theoretical benefits of increased sampling frequency and bit depth respectively.
In a similar vein, papers by Dan Lavry concerning the same issues: 'Do you need 20 Bits? ' and 'Sampling, Oversampling, Imaging, Aliasing'. He is selling something, but seems respectable.
There do seem to be several studies showing that human hearing may not be quite so limited to 20kHz, similar to the 'Inaudible High-Frequency Sounds Affect Brain Activity' article posted earlier. Frustratingly though, I can find virtually no good tests comparing sampling rates directly.
Likewise, there is plenty of information regarding the dynamic range of human hearing, but very little on the audibility of different bit depths. However, given that human hearing can be considered to have a dynamic range of 120dB (the difference between quietest audible sound and threshold of pain), then it does seem that the 96dB of CD audio might be considered inadequate to capture the maximum desireable range. It just so happens that 120dB is the theoretical dynamic range from 20bit, which is what was suggested as the optimum value in the link I posted earlier.
thank you janneman for answer
guys
I think this thread starts crossing the thin line beetween science and philosophy. We should divide these two things. A silly comparison comes to my mind:
DIYcooking member says:
it has been proven, that humans can only taste: sour, bitter, sweet and salt. It is all that sense is able to detect.
another DIYcooking member says:
OK. so let's eat cold pork and drink hot beer for a dinner. Does it taste the same? No!!!
and quick answer:
give a proof that it affects the taste itself
tell me where does it end, please?
regards
guys
I think this thread starts crossing the thin line beetween science and philosophy. We should divide these two things. A silly comparison comes to my mind:
DIYcooking member says:
it has been proven, that humans can only taste: sour, bitter, sweet and salt. It is all that sense is able to detect.
another DIYcooking member says:
OK. so let's eat cold pork and drink hot beer for a dinner. Does it taste the same? No!!!
and quick answer:
give a proof that it affects the taste itself
tell me where does it end, please?
regards
Mr Evil said:
Julian Dunn has a very interesting paper. It echos the points made earlier, that higher sample rates allow easier and more relaxed implementations of the necessary anti-aliasing (or reconstruction) filters, which in turn cause less (audible) artifacts in the audio band.
Quote:
"A direct effect of the higher sampling rate is that for an identical filter design the time displacements will scale inversely with sample rate. Hence an improvement can be made just from raising the sample rate - even for those who cannot hear above
20kHz."
Jan Didden
SY said:
I don't quite understand your first paragraph, but I am not in this thread just to prove what I say to be correct, but only provide information that possibly would help audio systems be better. How others spend time in the forum, it is up to them.
It is true that mixing will create distortion in many ways, espetially if the equipment is not optimum.
Good man sleep well.
darkfenriz said:
Mr Evil said:
You could have chosen a better analogy there, since there are actually five tastes! I'm not making this up either; a fifth taste really does exist, but just doesn't have a name in English. The Japanese do have a name for it...
Could you call"HOT" a taste? or is there something else?
For the die-hards amongst you:
Quoting from EDN magazine:
"Turning Nyquist upside down by undersampling
Analog columnist Bonnie Baker explains how you can intentionally
force a system configuration that aliases or folds back higher
frequency signals that occur at values greater than the converter's sampling rate. Using the Nyquist theorem in this way suits applications such as wireless-communication receivers, radar
instrumentation, infrared instrumentation, and video.
http://email.ednmag.com/cgi-bin2/DM/y/emBC0FWR8O0Gcq0CUhD0Es&rid=243880419
...and for more background on undersampling, read this 2-page PDF on Pentek's site: Putting undersampling to work
http://email.ednmag.com/cgi-bin2/DM/y/emBC0FWR8O0Gcq0CUhE0Et
In addition, Baker wrote about the foldback phenomenon in this
article: Filtering? Before or after?
http://email.ednmag.com/cgi-bin2/DM/y/emBC0FWR8O0Gcq0CUhF0Eu&rid=243880419 "
Jan Didden
Quoting from EDN magazine:
"Turning Nyquist upside down by undersampling
Analog columnist Bonnie Baker explains how you can intentionally
force a system configuration that aliases or folds back higher
frequency signals that occur at values greater than the converter's sampling rate. Using the Nyquist theorem in this way suits applications such as wireless-communication receivers, radar
instrumentation, infrared instrumentation, and video.
http://email.ednmag.com/cgi-bin2/DM/y/emBC0FWR8O0Gcq0CUhD0Es&rid=243880419
...and for more background on undersampling, read this 2-page PDF on Pentek's site: Putting undersampling to work
http://email.ednmag.com/cgi-bin2/DM/y/emBC0FWR8O0Gcq0CUhE0Et
In addition, Baker wrote about the foldback phenomenon in this
article: Filtering? Before or after?
http://email.ednmag.com/cgi-bin2/DM/y/emBC0FWR8O0Gcq0CUhF0Eu&rid=243880419 "
Jan Didden
soongsc said:
I don't think you understand what I am saying. From what I understand, you are claiming that some people can detect transients with 50kHz component in them. I am asking, is 50kHz the lowest component? If it's not (and it most likely isn't, such as the case for a pulse or a step), then it is the lower (<20kHz) frequencies that the person is hearing; in this case nothing is proved.
Ok, great, cymbals produce waves with frequency components > 20kHz; so do heaps of other instruments (and many other things too). This doesn't mean that you need these frequency components in them to accurately reproduce the sound.
real said:
From the link that scott wurcer posted (nr 78):
http://www.cco.caltech.edu/~boyk/spectra/spectra.htm
real said:
50K Hz is not the lowest component. If it was, then I think it would not be detectible. What really makes sense is testing whether you can detect lack of higer frequency content when it is with lower frequency content, otherwise you will not be able to prove what benefit preserving higher frequency content provides when listening to music.
I think the issue would be how accurate you need to reproduce, since you will always be able to find differences in reproducing signals after they go through the AD/DA process. So different people have different opinions as to how well the preservation should be untill differences becomes audible. Since digital recording is becoming more and more popular, anyone can just do their own recording to find out what differences they can detect.
Like I mentioned before, I friend of mine could not hear an ultrasonice mosquito repelling device even when stuck in her ear, and I could hear it inches away from my ear (when I was young).
percy said:
Probably could be the same situation for some listening tests.
I actually did a test using the MP3 recorder to record whatever I hear by sticking two mikes each in one ear. Then I had people listen to the recording. The results were interesting. The image is very realistic, not so precise as lots of audio systems are, but real! You can hear coloration of the sounds, and background noise that normally you would not be ware of. Didn't try to figure out why though.
sorry janneman, but I don't see any point in undersampling an audio signal? Did you mean that? It makes not much sense as audio band starts from nearly 0Hz, it is worth to undersample band signals, where center freq. is much higher than lowest freq. in signal. HF rubbish won't alias into sub-sonic range.
regards
regards
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.