-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Measurable aspects of sound perception
- Thread starter pozz
- Start date
solderdude
Grand Contributor
What if the sinc filter is at say 96kHz or 192kHz ? What if the errors are below audible thresholds ? What if the brain cannot distinguish in a blind test ?
Is it not arrogant or naive to think subjective sighted listening is the best way to evaluate ?
There is a way to exactly measure (or chart) what distortions our brain can distinguish.
It is called listening to and comparing music (or signals) in lab conditions to determine thresholds. Of course this must be done with trained listeners, equipment that can (re)produce the signals accurately, must have statistically valid trials and must be done blind under tightly controlled conditions.
You see, there is no sighted and subjective home listening involved here. If it were and would get objectively more accurate results that way science would be all over it.
- Joined
- Apr 24, 2019
- Messages
- 1,501
- Likes
- 2,822
Think this site is overcrowded with members believing that when we (in analogy) measure the chemical composition of a wine we can conclude how it will taste....................I would suggest: Mix some chemicals and send this advice to the winemakers
Sorry horrible flawed analogy, if you don't comprehend information theory or have no interest in it that's fine. We've been over this before no point in revisiting it.
Hi BDWoody,
Did not mean anything personal with this remark.
It's my personal conviction that we (us human beings) are far from understanding nature.
Indeed I have ideas I clearly believe in and others will have different ones.
I don't think this is a problem and just a fact of live.
While it is true that we may not know everything, we do know a whole lot more than nothing.
I'll just have to conclude that you are wronger than wrong.
Hi Scott Wurcer,
Agree it was too silly
I'm very also interested in information theory; so please explain your point.
- Joined
- Dec 5, 2019
- Messages
- 3,843
- Likes
- 3,790
Hi Scott Wurcer,
Agree it was too silly
I'm very also interested in information theory; so please explain your point.
why not start with a wiki on it?
OP
- Joined
- May 21, 2019
- Messages
- 4,036
- Likes
- 6,827
- Thread Starter
- #209
Claude Shannon's original paper: http://people.math.harvard.edu/~ctm/home/text/others/shannon/entropy/entropy.pdfHi Scott Wurcer,
Agree it was too silly
I'm very also interested in information theory; so please explain your point.
I don't understand it in the mathematically rigorous way that some of our other members do, but it's a foundational, enlightening text nonetheless.
Credit to @j_j for the link.
xr100
Addicted to Fun and Learning
Think this site is overcrowded with members believing that when we (in analogy) measure the chemical composition of a wine we can conclude how it will taste....................I would suggest: Mix some chemicals and send this advice to the winemakers
Blind testing IS used in sensory evaluation testing.
I have no interest or knowledge about wine but it is a dreadful example as a complex "moving target" (aging, volatile compounds, etc.) Nevertheless,
I think you'll find that chemical analysis is used to inform development.
Moreover, wine is a "creation" as a piece of recorded music is.
Completely different to digitally recorded audio where there is, was, and always will be a reference signal. More akin to testing the just noticeable difference level of adding, say, salt to food or something, than comparing different wines.
100% of serious winemakers and wine evaluators use double blind methods for serious tasting. When I was judging competitions, not only was the judging done double blind with replications and hidden controls, but judges themselves had to qualify by performing in double blind tests with spiked samples.
(nb: "double blind" is colloquially used in the wine world with a different meaning, but nonetheless, serious analytical evaluation is ALWAYS done double blind in the sensory science meaning)
xr100
Addicted to Fun and Learning
The very basic idea is straightforward, e.g...
"The recent development of various methods of modulation such as PCM and PPM which exchange bandwidth for signal-to-noise ratio has intensified the interest in a general theory of communication. A basis for such a theory is contained in the important papers of Nyquist and Hartley on this subject. In the present paper we will extend the theory to include a number of new factors, in particular the effect of noise in the channel, and the savings possible due to the statistical structure of the original message and due to the nature of the final destination of the information."
So if that paragraph were appended with #=the, then...
There are 15 "the" instances = 3x15 = 45 characters.
"#=the" is 5 characters.
So, 40 fewer characters would be needed to code. And, indeed, this is (crudely) how ZIP and other entropy (lossless) coding systems work.
EDIT: My brain must be totally frazzled... channel capacity not entropy coding...
Last edited:
xr100
Addicted to Fun and Learning
One of U.C. Davis' big tools is a gas chromatograph.
NMR, FTIR, HPLC... along with a deep level of biochemical and genetic measurement (e.g., DNA sequencing for ampelography).
Ditto Bordeaux, the other major school emphasizing a science-based approach to winemaking. It is funny, though, to see the reaction of a Davis-trained winemaker to a, ummm, rustic sort of Rhone.
They are very, very similar, but kind of opposite as well.
Entropy coding assumes a clean channel, and attempts to reduce the (bits/second) necessary to send a signal by reducing redundancy. One can, for a given message and transmission medium, figure out that minimum, that's the signal's entropy. (There's more than that, but the qualifications and explanation would be a book or two.)
Channel coding ADDS REDUNDANCY in order to INCREASE a channel's transmission capacity, but there is a fundamental limit (Shannon Bound) to that channel's bit rate.
But the overall idea is the same in both, address the signal entropy, and get it from here to there.
I haven't seen it but yes, that might be true.
Up here in Washington State, we do pretty well, too, you know, but a lot of our stuff doesn't make it past the end of Oregon or the Dakota's before the truck is empty.
That's true of most of the best stuff, in fact.Ever hear of deLille, Bookwalter, Matthews, Hedges, Terra Blanca?
- Joined
- Apr 24, 2019
- Messages
- 1,501
- Likes
- 2,822
deLille, "The Pride of Woodinville."
We'll have a few bottles together at some point... my cellar is heavy on aged Nothern Rhones.
Some Juge for my fly? We're partial to Dundee Hills pinot, direct shipping has sort of ended the you can't get it here myth. And then there's the no such thing as terroir discussion that has no where good to go except court.
That was too private of a joke.
And sadly, Jan, Cynthia, and I polished off our last bottle of '04 Olga Juge Cuvee C a few nights ago.
