[_gl]

  I have something similar with my HD650 - a simple 40hz low shelf filter really helps recover subbass in my opinion.

 
Interesting, so even though they're bassier they still rolloff sub-bass.  I only have the HD600s, but the bass rolloff really annoyed me (I'm used to a pretty flat studio setup - with a subwoofer : ).  It's amazing how much you can get back with EQ.

 

[Bob A (SD)]

  I've done the impulses for Foobar2k (see the first post for download & instructions).  Right now they're 44.1k only, so will probably only sound correct with 44.1k files.  I'll look into adding other sample rates later.  I like to upsample my CD rips 2x, if you do that then put the resampler after the Convolver (for now).
 
(the impulses might also work with other convolvers, but I'm not yet sure what a 'standard' source pulse is.  These use the 'Unitpulse2k' pulse which the foobar plug uses, there's also a Dirac pulse I can try if anyone wants it).
 
It's the first time I've done my own impulses, but they sound right compared to GlissEQ.  As I mentioned, I like Curve4 best, but the bass may be slightly over 'flat'.  If it's too much try Curve 3.
 
Let me know how it works for you.  I'm really into foobar now and have done my own touchscreen friendly skin with custom graphics & scripts.  Steep learning curve, but great for total customisation.

 

Got it  

   Thank you very much!
 
Trying your curve #4 first.  Obviously it will take some time to evaluate the curves and with Easter tomorrow....   I'm running a mix of 44.1/16 FLAC files ripped from my CDs and 96/24 FLAC files from my digitized vinyl/tape collection.  There's also a few 192/24 sample files.   I will definitely get back to you!

 

[_gl]

Sure, same here, have fun.

 

[_gl]

Quick Convolver plugin tip:
 
Due to the curves' heavy bass boosting, bass-heavy tracks can overload.  So my impulses are 4db quieter than normal, ie. the output volume is also quieter (seems to work quite well to avoid overloads).
 
However when the Convolver plug's "Auto level adjust" is ticked, it boosts them back up.  That's great for comparing with/without curves as it's volume matched.  But when it then overloads it seems to apply strong limiting to avoid clipping - I just heard it on a track and it sucks (if you're a purist like me).
 
So once you're done A/B'ing curves / no curves, I suggest you untick this to avoid the limiting.

 

[_gl]

I'm running a mix of 44.1/16 FLAC files ripped from my CDs and 96/24 FLAC files from my digitized vinyl/tape collection.  There's also a few 192/24 sample files.   I will definitely get back to you!

 
The impulses are only 44.1k (for now), so they probably won't sound right at higher sample rates.  Haven't tried yet though, the convolver might upsample them automatically (that might work but wouldn't be ideal, I'll do some native files when I get time).

 

[Bob A (SD)]

 

I'm running a mix of 44.1/16 FLAC files ripped from my CDs and 96/24 FLAC files from my digitized vinyl/tape collection.  There's also a few 192/24 sample files.   I will definitely get back to you!

 
The impulses are only 44.1k (for now), so they probably won't sound right at higher sample rates.  Haven't tried yet though, the convolver might upsample them automatically (that might work but wouldn't be ideal, I'll do some native files when I get time).

Played some 96/24 files and didn't detect and degradation at all.   Using the convolver and your Curve #4 (couldn't hear that 0.1dB difference betwen it and curve #3) beats the graphic equilizer approach three ways from Sunday! 

    Anyone using foobar2000 withn HD600s really ought to try this.  Again thank you for sharing!

 

[_gl]

Glad they're working for you Bob.
 
I'm doing about 5 things at once right now, when I get time again I'll figure out the extra sample rates.  I might even extend the foobar SOX resampler to do intelligent 2x/4x upsampling only when needed, automatically choosing the correct sample rate impulse.  If I get time ... : ).

 

[MorrisL]

Can anyone use this technique to make HD600 sound identical to the HD650? I've asked this elsewhere, but would really appreciate more feedback, since most who said they tried it also said that they did it years ago and had to rely on memory. I think this would be a wonderful test for those who own both (I only have the 600s) and would make us stop craving for the other so much if we already own either the 650 or the 600.

 

[_gl]

In theory, probably.  The Sonarworks plugin actually has an HD650 emulation that you can use on the HD600.  You can download their trial and try it.
 
I've never heard an HD650 so I can't say how accurate it is, but it should be ballpark as they custom-measure their cans for their correction curves.  Saying that, they're only +-3dB accurate (they can do better with custom calibration where you send in your cans to get them measured).
 
If someone already has both cans, try my curves and let me know how close the bass response gets.

 

[_gl]

BTW Morris, from what you said in the HD600 Impressions thread, it sounds like you just want a gentle treble rolloff?  I would experiment with a basic EQ and see if it gets you there.

 

[MorrisL]

  BTW Morris, from what you said in the HD600 Impressions thread, it sounds like you just want a gentle treble rolloff?  I would experiment with a basic EQ and see if it gets you there.

Thanks. I absolutely would want a nice 650-like treble roll-off. I was also curious to hear what 650 sounds like, and since they are more similar than different I'm wondering if one can eq one to sound exactly the same as the other and indistinguishable in a blind test (easy to conduct since they pretty much feel the same on the head). Imagine how cool that would be!

 

[_gl]

  Thanks. I absolutely would want a nice 650-like treble roll-off. I was also curious to hear what 650 sounds like, and since they are more similar than different I'm wondering if one can eq one to sound exactly the same as the other and indistinguishable in a blind test (easy to conduct since they pretty much feel the same on the head). Imagine how cool that would be!

 
Yes it would be, and that's what Sonarworks actually does.  I suspect it's easier for the HD600 than the other way around, as the 600 is bright and detailed.  ie. it's probably easier to take away some high-end, than to boost it on the HD650 (though maybe not).  I have tried the Sonarworks HD650 emulation on my HD600 and it was interesting, but as I said I don't know how accurate it is (probably pretty good).
 
But if you don't know what the 650 actually sounds like, then how do you know that's the exact sound you want? : ).  Try EQing for less treble, that might be all you need.

 

[448624]

Thanks for this. I'm still testing the EQ with Equalizer APO. Does anyone know why I can't change the gain frequencies to increments of 0.1? The least I can do is 0.5.

 

[_gl]

  Thanks for this. I'm still testing the EQ with Equalizer APO. Does anyone know why I can't change the gain frequencies to increments of 0.1? The least I can do is 0.5.

 
You're welcome.  I hadn't heard of Equalizer APO, so this is a plugin that works right in the Windows audio pipeline?  Google says that it now supports convolution, then you could use my impulses rather than trying to match with an EQ.
 
Do you know what type of source pulse E-APO expects from its impulses?  DIRAC or something else?

 

[Bob A (SD)]

  Do you know what type of source pulse E-APO expects from its impulses?  DIRAC or something else?

_gl,
 
   Found this for you
 

Convolution (since version 1.0)

Syntax:
Convolution: <File name>
Description:
Adds a convolver that processes the signal using the impulse response contained in the specified file. The file must be in one of the formats supported by libsndfile (e.g. wav, flac or ogg). If the file contains multiple channels, the channels are assigned to the selected channels in round-robin order (e.g. a stereo file is assigned to 4 channels as L->1, R->2, L->3, R->4). The sample rate of the file must match the sample rate of the device, otherwise the convolver can not be created. Latency and CPU usage depends on the length and the phase behaviour of the impulse response (linear-phase will have a latency of half the file length while minimum-phase has a lower, but inconsistent latency). The specified file name is relative to the current configuration file's path. While impulse response files can be opened from any directory with sufficient access rights, if the files reside in Equalizer APO's config path or a subdirectory, the configuration will be reloaded automatically if the files are changed so that the change is applied immediately.
 
https://sourceforge.net/p/equalizerapo/wiki/Configuration%20reference/#convolution-since-version-10