Headphone CSD waterfall plots

[purrin]

Quote:

Way harder to compare FRs now (at an angle and some is chopped off up top)

The stuff chopped up on top isn't that important for these purposes! The angle is to get a better sense of the speed of the drop so I wouldn't have to make two graphs. Still experimenting. May still go back to the old ones. Part of it was because I wanted to animate the new ones to go from -30db floor to -50db floor.

 

[redwarrior191]

 
Quote:

Way harder to compare FRs now (at an angle and some is chopped off up top)

i agree with this.. i think it's easier to understand the old plot..

 

[purrin]

Here are the old-style ones:
 

Revision 2

 
Nice fast decays, but driver matching not as good as LCD3.
 

Revision 1

 

 

[morks]

@purrin: Is there any way for you to acquire a T1 again? I stumbled over http://www.head-fi.org/t/531401/ade-tuned-beyerdynamic-t1 and on a whim equalized the 8 kHz band to -4 dB (build-in 10-band-equalizer of Banshee on Ubuntu 11.10). It suddenly sounded more natural and balanced, the tinnyness and piercing shapness was greatly reduced! I would be VERY interested in measurements of ADE-T1s and equalized T1s. Would that be possible?

 

[rhythmdevils]

Someone has to send him a tampon modded pair of T1's then.  He's not getting paid for this.  I also agree that would be really fascinating to see what would happen in the csd plot. 
 
Send him yours! 

 

[purrin]

 
Quote:

@purrin: Is there any way for you to acquire a T1 again? I stumbled over http://www.head-fi.org/t/531401/ade-tuned-beyerdynamic-t1 and on a whim equalized the 8 kHz band to -4 dB (build-in 10-band-equalizer of Banshee on Ubuntu 11.10). It suddenly sounded more natural and balanced, the tinnyness and piercing shapness was greatly reduced! I would be VERY interested in measurements of ADE-T1s and equalized T1s. Would that be possible?

Actually I did acquire another T1. This is a good one sounding one too. What you did on reducing the 8kHz lever on your 10 band EQ is perfect. Good ears!
 
The CSDs below have the floor moved down to -40db. With the older measurements I took, the floor was only down -30/-36, so these graphs will look considerably messier. There is quite a lot of stored energy between 7-8k.
 

 

 

 

[morks]

Uh, I'm confused, is that with equalization or just the baseline to equalize against? Because the 8 kHz area still looks as aggressive as the first measurement you posted

 

[purrin]

That's baseline - what you should EQ against.

 

[Anaxilus]

Looks like they need a bigger magnet.

 

[rhythmdevils]

I miss seeing new CSD plots every day. 

 

[CEE TEE]

Thanks, @purrin...it is nice to have the 2D along with the CSD so that it is easier to understand which FR spots have the residual energy.
 
(Hope not too much more work to provide both together...)

 

[purrin]

Quote:

Thanks, @purrin...it is nice to have the 2D along with the CSD so that it is easier to understand which FR spots have the residual energy.
 
(Hope not too much more work to provide both together...)

I may be going back to the old one.
 

 

[slwiser]

Quote:

I may be going back to the old one.
 

 
Not to sound like a five year old but WHY???? Old single line charts? or some other water fall chart?

 

[morks]

@purrin: Could you slap a 10-band-Eq with -4 dB at 8 kHz on that T1 and measure it again? I'm curious what the effect on the FR and resonances is. Hum... then again, I tested 2 equalizers in maybe not equal circumstances, but they sounded different with the same setting... mh. maybe I'm doing it wrong.

 

[plin]

Quote:

@purrin: Could you slap a 10-band-Eq with -4 dB at 8 kHz on that T1 and measure it again? I'm curious what the effect on the FR and resonances is. Hum... then again, I tested 2 equalizers in maybe not equal circumstances, but they sounded different with the same setting... mh. maybe I'm doing it wrong.

In theory, I believe, cone resonances (what purrin measures with his top secret method 

 ) are minimum phase phenomena. So, in theory, if you know the exact frequency, dB level and Q factor of a resonance you can completely eliminate its effects by applying an equal value minimum phase parametric filter . Audyssey (and other similar products) work in this way.  Linear phase and/or band equalisation will not produce the same results. 
They are differences in the implementation of minimum phase filters in the various software. One software that, I think, has very good minimum phase parametric filter implementation is FABfilter Q.
 
One final note: if you equalize based on published plots, remember that your headphone maybe has a little different plot than the one measured. So, maybe it is better to apply a bit less dB level than the one shown in plot.