Folks,
(Sorry for the cross-posting.)
I'm looking for someone who knows how to write DirectX plugins and who is willing to devote a small amount of time to help advance scientific knowledge.
The goal is to settle for once and for all whether phase shift alone is audible, and whether having a frequency response beyond the 20 Hz. to 20 KHz. boundaries is really audible. One plugin I'm looking for will implement an all-pass filter in DSP, and allow varying the amount of phase shift / corner frequency with a slider. This is essentially how a phaser (not flanger) stomp box works, but without the part that sums the input and output to create the dips and peaks.
The other plugin will have two sliders to control at what frequencies the upper and lower extremes are rolled off. The slopes need to be sharp (or better, adjustable), and the sliders must of course be calibrated. Further, this plugin needs to work at a 96 KHz. sample rate so the test audio source (pink noise, whatever) can have frequencies higher than 20 KHz. yet avoid the additional sharp filtering inherent in a 44.1 KHz. sample rate.
Both of these pluginswill then be distributed freely so everyone can try these experiments.
Any takers? If you are willing to help me with this experiment, please go to my web site at http://www.ethanwiner.com and click the "email Ethan" button near the top of the screen.
Thanks!
--Ethan Winer
http://www.ethanwiner.com/articles.html
"The truth is in there"
Comments
audiokid, > Aren't harmonics effected from hp,lp filters, rollo
audiokid,
> Aren't harmonics effected from hp,lp filters, rolloff, eq etc. and if they are then it would be an audible difference. if a freq is moved more than 3 db phase incoherence begins to happen right? <
Yes, harmonics are obviously affected by EQ. But that's not the issue. What I want to test is whether phase shift alone is audible. A lot of folks seem to believe it is, so I'd like to make a way to test that independant of everything else. Then I can make the plug-in, and some test source material, available for free public download.
> If a 96 KHz sample rate is nessesary for the test then it seems to me that you can hear a difference. <
Not at all! The reason I want the bandwidth testing plug-in to work past 20 KHz. is so people running the tests can start at a higher frequency, and then adjust the HF cutoff downward until a difference is audible. If the highest frequency testable were only 19 or 20 KHz., there wouldn't be a way to start higher than that.
> I have always been able to hear and feel sub harmonics. I hear what the subs do to the total bandwith when they are moved to far (rolled off or boosted). <
As far as I know there is no such thing as a subharmonic. I suppose you could define it in theoretical terms like the square root of a negative number. But in practice the fundamental pitch of a musical instrument is the lowest tone it generates.
You may hear a change as you play with the crossover frequency of a subwoofer, but how do you know for sure that what you are hearing involves frequencies below 20 Hz.? This is what I am proposing be tested.
--Ethan
http://www.ethanwiner.com/articles.html
"The truth is in there"
Hi, sorry, didn't word this sentence right: I have always bee
Hi, sorry, didn't word this sentence right:
I have always been able to hear and feel sub harmonics. I hear what the subs do to the total bandwidth when they are moved to far ( rolled off or
boosted)
Not that it matters but here's what I meant to say:
I have always been able to hear, feel or sense subs and hear the harmonics in general. I hear what the subs do to the total bandwidth when they are moved to
far, either rolled off or boosted to a point where phasing occurs.
If we suck out 50, doesn't it effect 200, 400, 1600 etc.?
I'm not going to try and decipher what's going on in Pro Talk but I do believe I hear things that I've been told aren't audible from a scientific approach.
The more I use digital the more I hate the sound and don't trust the specs ver. real life. I love how it make editing simple but as far as sound goes.......it's a
long ways off from human ears. Maybe that's what we are talking about here but can't find the words to prove the scientific theory wrong so we remain in
quagmire. Kind of like arguing with doctors ver. naturopathic medicine.
The deep dimensions of harmonics that digital audio can't reproduce or measure is what I'm understanding is the question here.
That being said, I don't believe the scientific theory when it comes to my ears anymore "been there , done that".
I think the ear is allot more advanced than the scientific world has ever measured it. Maybe that's what a golden ear is?
I support your quest in this post and look forward to reading the results. I hope things simmer down and we get back on track and hopefully some good will
come out of all this.
audiokid, > I have always been able to hear, feel or sense subs
audiokid,
> I have always been able to hear, feel or sense subs and hear the harmonics in general. I hear what the subs do to the total bandwidth when they are moved to
far, either rolled off or boosted to a point where phasing occurs. <
I'm not sure where "phasing" enters the equation, and this is one of the things I'd like to determine. My feeling is that things we hear can be explained and measured using normal methods.
> If we suck out 50, doesn't it effect 200, 400, 1600 etc.? <
If you set a filter to be 3 dB. down (or whatever) at 50 Hz, it will affect higher frequencies. But the rolloff has nothing to do with the harmonic series. Filtering 50 Hz. will affect 100 Hz. to some degree, but by much above that the filter's affects should be out of the picture.
> ... but I do believe I hear things that I've been told aren't audible from a scientific approach. <
As far as I know "science" has never stated you don't hear things you obviously do hear. Science usually addresses learning why things are as they are, and empirical evidence is just as valid as any other evidence. If a lot of people say they can hear something, then it is science's job to try to understand and explain it.
> can't find the words to prove the scientific theory wrong so we remain in
quagmire. <
No words are needed! But I think the only true way to know what is audible and what is not is to perform scientific tests. Preferably blind or even double blind if possible.
> I support your quest in this post and look forward to reading the results. <
Thanks.
> I hope things simmer down and we get back on track and hopefully some good will
come out of all this. <
Me too! I feel really bad that my questions have been escalated by the name callers into so much more than the frank discussion I was seeking!
--Ethan
http://www.ethanwiner.com/articles.html
"The truth is in there"
I'm just beginning to follow this saga so bare with me as I catc
I'm just beginning to follow this saga so bare with me as I catch up.
Aren't harmonics effected from hp,lp filters, rolloff, eq etc. and if they are then it would be an audible difference. if a freq is moved more than 3 db phase incoherence begins to happen right?
If a 96 KHz sample rate is nessesary for the test then it seems to me that you can hear a difference. This is why we are constantly trying to design higher bit rates. more bites, more clarity, more of the harmonic? smoother sound, less phase and distortion because the print becomes more unique and has less of a chance to cancel out when split and so on.......right?
I have always been able to hear and feel sub harmonics. I hear what the subs do to the total bandwith when they are moved to far ( rolled off or boosted).
Is this what we are talking about?
:cool:
audiokid