Head-Fit: Software Crossfeed and EQ

[donunus]

Quote:

Some things Ive noticed with headfit...

I just noticed when tuning headfit that the LF and HF ms needs pairing for the tonality of the crossfed signal to be right. I came up with 3 setting pairs with good tonality where the hf doesn't noticeably bleed too much to the opposite ears separate from the lows... theres .23,.18 .25,.20 and .32,.27 at the stock ILD levels. Changing the ILD levels may need different ms adjustments but I am still content with the stock levels at the moment.

The difference between .25,.20 and .32,.27 for example are that the former is more subtle sounding and more natural to my ears for example while the latter sounds slightly more processed and The doors recordings for example sound more dare I say REMASTERED with the latter setting. For now the former setting is what I prefer. There is a better sense of coherence from left to right with that one.

Now looking back at an old post of mine and my two settings here that I thought were great had the same difference in phasing of .05. I guess I liked a difference of .05 before and now .09. Maybe it was due to different cans I can't confirm because I don't have my dt150s anymore.

 

[ironmine]

Quote:

The .32, .23 setting gives a little better fill though to my ears. I ended up with this. Sorry ironmine
 

I would certainly try your settings, but I have two problems with that;
 
1) My ears are - anatomically and, therefore, acoustically - different from yours.
 
2) Tomorrow, as sure as the sun will go up, you will post some other "better and wiser" settings which will override your "inferior and outdated" today's settings. The same would happen the day after tomorrow, etc., etc. So, I'll think I'll wait for a year or two until you... eh... equilibrate yourself completely.
 
The Audio Evangelist walks to binaural perfection with too frequent and long strides, I cannot keep pace with him...

 

[ironmine]

Lee Perry,
 
I've found the DDMF Metaplugin (it's a VST-chainer similar to Console, it compensates for latencies automatically) http://www.ddmf.eu/product.php?id=3.
 
I tried to oversample the DDMF Metaplugin with Christian Budde's oversampler, but it does not work this way. The other VST Oversampler works though - http://sites.google.com/site/chrisrwalton/oversampler
 
So, it put all my plugins inside the 2x oversampled DDMF Metaplugin (e.g., crossfeed + equalizer, etc.) and it works (even inside the console).
 
By the way, remember you said that upsampling is not possible in the VST architecture, only resampling? But how comeAlgorithmix says about their equalizers: "To avoid bell filter asymmetry at high frequencies, typical for many digital equalizers, we have applied reference-quality upsampling techniques, automatically switchable if the sampling frequency of the input signal is 44.1 or 48 kHz." (http://www.algorithmix.com/en/classic_peq_blue.htm)
 
It means it is possible to upsample (not just resample) with VST plugins?

 

[J-a-k-e]

 
Quote:

Jake, you are wrong! There is a step-by-step guide suggested by the author of Head-Fit:
 
www.ohl.to/about-audio/audio-softwares/head-fit

Yea, I did try following that guide by enabling just the left channel and playing with the crossfeed sliders with both pink noise and music but I found it didn't have near enough of an in depth description of what I should be listening for. I need something to the effect of
 
"adjust this LF slider until you get exactly this kind of sound and placement and use that other LF slider there to focus, now adjust this HF slider  so that the sound blends to the same position as the LF and provides a depth of height and use that other HF slider to focus. Now adjust the f central slider so that a sine wave test tone provides seamless integration between high and low frequencies.
 
That's what I liked about the Isone-Pro manual, it wasn't over complicated and it provided complete instructions on what each slider does and what to listen for when you go the through the initial adjustment process.

 

[donunus]

Talking about concepts and procedures of how to do things, I was wondering if anyone here knows what exactly happens with the ILD and F Central settings in headfit. All I know is that the ILD LF changes the lows in the crossfed signal and the ILD HF changes the highs but to get into more detail I scanned a basic sketch here...
 

 

Questions
Does the F Central behave more like the bottom pic where it is like a crossover point where to start rolling off the mids through the highs? Or Is it like the top pic where it separates where the ILD LF and HF settings affect?
 
If it behaves like the bottom pic for example, does lowering the ILD HF setting just make the crossover setting less steep like the #2 setting? In case you don't understand what I'm trying to ask with that picture, what I meant was that the dotted line is closer to the 0 setting on the HF because it only has a setting of -1 which takes 1 db away from the original curve and a -3, -3 setting will have the same curve as a 0,0 setting but with a more intense crossfeed volume going to the opposite ear. Do you guys think this is how it works? The website really doesn't explain this like the bs2b site does.

 

[SP Wild]

I intsalled Headfit on my big computer - it sounds so much better than when it was installed on my Netbook - I think this one is CPU instensive and the Netbook screwed it up.  Anyway - this is the most transparent crossfeed - more so than JR MC.  It doesn't expand the soundstage like JR does, but it doesn't shrink it either.  Tonal balance is unaffected.  I started to tire of the JR MC expanding the soundstage and making everything sound distant - Headfit retains that headphone intimate feeling.  Isone pro is good for simulating speakers - but is not as transparent and I don't really want speaker sound.
 
I have yet  to play around with the sliders - but I think I'm gonna settle down with head-fit for a while - it's superb and free.

 

[SP Wild]

Just playing around with the sliders - this is definitely the best crossfeed implementation I've come across.  I don't know why it was awful when it was on my netbook.

 

[J-a-k-e]

Quote:

I intsalled Headfit on my big computer - it sounds so much better than when it was installed on my Netbook - I think this one is CPU instensive and the Netbook screwed it up.  Anyway - this is the most transparent crossfeed - more so than JR MC.  It doesn't expand the soundstage like JR does, but it doesn't shrink it either.  Tonal balance is unaffected.  I started to tire of the JR MC expanding the soundstage and making everything sound distant - Headfit retains that headphone intimate feeling.  Isone pro is good for simulating speakers - but is not as transparent and I don't really want speaker sound.
 
I have yet  to play around with the sliders - but I think I'm gonna settle down with head-fit for a while - it's superb and free.

 
Yea it is fairly cpu intensive. My desktop machine is an Amd Phenom II dual core@3.5ghz and enabling head-fit kicks my cpu usage from 2% up to 14%. I agree the headphone dsp in J. river is not quite right and while definitely better than some I've found I'm not so keen on the boxy out of the head effect.

 

[gecolin]

Thanks!

 

[donunus]

I went back again and am using almost stock LOL
-8
-15
1996
.32
.27

 

[cooperpwc]

The fun is in the journey.

 

[leeperry]

@ironmine: yeah, Christian Budde's software is always quite buggy..
 
Algorithmix prolly meant oversampling, as the VST architecture doesn't allow to have a different sample rate on the input and output AFAIK. I tried their EQ, not my thang.
 
@J-a-k-e: yeah, the instructions to use the pink noise sound like broken english to me...I can't make any sense out of them. I wish Jean-Luc could explain me in french, and I could try to write a clearer tutorial.
 
@donunus: glad you like them! personally, I've gone back to my good ole' settings...they just "work" on all tunes, and don't color/make the sound phasey-sounding.

 

[donunus]

Did you try that very last one? Its the least phasey setting but may be a little smooth for you I predict 

 

[leeperry]

 

Did you try that very last one? Its the least phasey setting but may be a little smooth for you I predict 

I've been using the same settings for about a month now, my brain rings the alarm when I mess w/ them. It feels like it's gotten use to how the sound is projected. These were my favorite settings at the time for a reson I would guess.

 

[J-a-k-e]

Quote:

 

I've been using the same settings for about a month now, my brain rings the alarm when I mess w/ them. It feels like it's gotten use to how the sound is projected. These were my favorite settings at the time for a reson I would guess.

Those are the settings you posted on page 7 overlaying quake 3?