Head-Fi.org › Forums › Equipment Forums › Headphones (full-size) › Orthodynamic Roundup
New Posts  All Forums:Forum Nav:

Orthodynamic Roundup - Page 1436

post #21526 of 23451
Quote:
Originally Posted by vid View Post

^ I'm not sure what your point is, but if it was to question what I've done, the decency would be welcome to put more effort into it, since you admit to not know what it is I do.

 

 I must have missed some posts, but have you actually explained how your digital convolution achieves an effect similar to physical damping ?


Edited by gurubhai - 4/28/13 at 10:41pm
post #21527 of 23451

^ You've missed lots of graphs and subjective impressions on this subject over the last six months or so in this thread.

post #21528 of 23451

I haven't been following the forum with the same zeal that I used to but I still have kept an eye on this thread. But still its possible that I might have missed out on some posts and discussions in the recent past.

I know that you have been playing with digital convolution for achieving better channel balance but I can't recall a discussion on how this affects the transient response.

If it isn't too much of an effort, can you please point me to where you explain all this?

post #21529 of 23451

I can't explain to you how it happens. For that you'll need to consult an expert in DSP.

 

For its effects on the transient response I refer you to the following three waveforms that I plotted six months ago.

 

Test signal with transients

 

Test signal recorded from phones X

 

 

Test signal with convolution applied and recorded from X

 

First (black) and third (blue) waveforms overlaid on each other.

 

Apart from the ruggedness, which is likely noise in the recording due to low input level, the transients in the convolved version line up perfectly in attack and decay with the test signal. That is, the data say the transient response not only improved but was perfected. For the sake of improving everyone's understanding of why this happens and whether it can be exploited, the counterargument should be more than 'it dun feel right'.

post #21530 of 23451

I must have missed this too as this looks pretty impressive , I did try and load your F2K config ( jumping to conclusions here - the one that posted a few days ago , might not have been yours )  but it foobarred my foobar, I just need time to tinker it back together again. Since I retuned a pair of NADs recently, my interest has been rekindled in the vintage drivers and I have been having fun when time permits. This was my next step in exploring the sound. Thanks for taking the time to repost this..dB

post #21531 of 23451

Because time and frequency response are, in short, inseparable.

 

See this article on Inner Fidelity:

 

http://www.innerfidelity.com/content/headphone-measurements-explained-square-wave-response

 

Quote:

Originally Posted by gurubhai View Post

I haven't been following the forum with the same zeal that I used to but I still have kept an eye on this thread. But still its possible that I might have missed out on some posts and discussions in the recent past.

I know that you have been playing with digital convolution for achieving better channel balance but I can't recall a discussion on how this affects the transient response.

If it isn't too much of an effort, can you please point me to where you explain all this?

post #21532 of 23451

@Vid : Thanks, that does look interesting. I looked up for some of your earlier posts as well, unfortunately the images you linked up are down & so its pretty hard to make out what you were saying.

So, how do yo arrive at the convolved impulse response. Do you have a reference headphone or do you just physically damp one driver to get a target curve?

post #21533 of 23451
Quote:
Originally Posted by vid View Post

I can't explain to you how it happens. For that you'll need to consult an expert in DSP.

 

For its effects on the transient response I refer you to the following three waveforms that I plotted six months ago.

 

Test signal with transients

 

Test signal recorded from phones X

 

 

Test signal with convolution applied and recorded from X

 

First (black) and third (blue) waveforms overlaid on each other.

 

Apart from the ruggedness, which is likely noise in the recording due to low input level, the transients in the convolved version line up perfectly in attack and decay with the test signal. That is, the data say the transient response not only improved but was perfected. For the sake of improving everyone's understanding of why this happens and whether it can be exploited, the counterargument should be more than 'it dun feel right'.


 I believe that a while back we discussed the convolver you were using and it was the foobar one.

that being the case have you used Garf's impulse response files at all?

 

http://www.sjeng.org/foobar2000.html

 

Which curiously enough seem to use the exact same headphones that MIT's media lab did with their dummy head test.

 

http://sound.media.mit.edu/resources/KEMAR.html

post #21534 of 23451
Quote:
Originally Posted by gurubhai View Post

@Vid : Thanks, that does look interesting. I looked up for some of your earlier posts as well, unfortunately the images you linked up are down & so its pretty hard to make out what you were saying.

So, how do yo arrive at the convolved impulse response. Do you have a reference headphone or do you just physically damp one driver to get a target curve?

 

The old graphs are offline for clarity. Get a microphone and measure the response of whatever it is you want to convolve to. Then measure the response of whatever you're convolving from. Take the difference and convolve with its impulse response.

post #21535 of 23451
Quote:
Originally Posted by vid View Post

I can't explain to you how it happens. For that you'll need to consult an expert in DSP.

For its effects on the transient response I refer you to the following three waveforms that I plotted six months ago.

Test signal with transients



Test signal recorded from phones X




Test signal with convolution applied and recorded from X



First (black) and third (blue) waveforms overlaid on each other.



Apart from the ruggedness, which is likely noise in the recording due to low input level, the transients in the convolved version line up perfectly in attack and decay with the test signal. That is, the data say the transient response not only improved but was perfected. For the sake of improving everyone's understanding of why this happens and whether it can be exploited, the counterargument should be more than 'it dun feel right'.

Anyone remember Quad's feed forward amps and Philip's Motional Feedback speaker line?
post #21536 of 23451
Quote:
Originally Posted by MDR30 View Post


Anyone remember Quad's feed forward amps and Philip's Motional Feedback speaker line?


Aha. If you are referring to that masterpiece of British audio the legendary 405-2, sure do, intimately :)

post #21537 of 23451

and on the 8th day, God invented felt: http://www.engadget.com/2013/04/25/marshall-monitor-headphones/ basshead.gif

 

Ftf systeme(1)

 
Ftf systeme1
 
"F.T.F. system"(Felt Treble Filter), oh wow etysmile.gif

Edited by leeperry - 5/1/13 at 3:49am
post #21538 of 23451

I hate it when they come up with complicated marketing terms for everyday things. Dishonest and stagnant.

post #21539 of 23451
Quote:
Originally Posted by vid View Post

I hate it when they come up with complicated marketing terms for everyday things. Dishonest and stagnant.

Haha...

Like the D.D.D. I saw elsewhere. Dual Dynamic Drivers.

No... just no.

Seems the only proprietary thing to it is the name.

post #21540 of 23451

Let me just quickly copyright S.H.A.T. (Stereo High Accuracy Transducers) while we're at it.   -_-

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Headphones (full-size)
Head-Fi.org › Forums › Equipment Forums › Headphones (full-size) › Orthodynamic Roundup