Foobar2000 Dolby Headphone config - Comment & discuss!
Oct 21, 2015 at 5:10 AM Post #796 of 868
 
If have written a short tutorial how to setup convolution reverb with BRIRs. Unfortunately the tutorial is not written in English yet and I have not enough time in the moment to  translate it.
 
So here is a very short explanation:
Convolution reverb is a well known method for creating    an “acoustic-fingerprint” of different sound stages.  Normally a frequency sweep is played through speakers and is recorded at one or several positions at a specified distance.  After deconvolution of this recording you get an “impulse response (IR)”.  This impulse response contains the whole information of the room and used speakers.  If you use the IR in a convolution reverb dsp the incoming audio will be filtered with the IR and playback will sound like the audio was recorded in the room where the IR was originally measured.  When a dummy head (or a real head) is used for recording the IR you will get the so called Binaural Room Impulse Responses (BRIR).  If you use a BRIR in a convolution reverb dsp incoming audio will be filtered with the BRIR and playback will sound like the Audio was recorded with a dummy head in the room where the BRIR was originally measured.
 
Using BRIRs with foobar2000:
Download:
VST-Convolver: http://convolver.sourceforge.net/
Resampler:  http://www.hydrogenaud.io/forums/index.php?showtopic=67373
VST-Host: http://www.hydrogenaud.io/forums/index.php?showtopic=84947
 
Impulse responses:
The BRIRs are recorded at 5 different rooms with a dummy head (full 360° rotation, 720 channels) and there are filters for some headphones. You should choose the headphone filter which comes closest to your own headphone model. The BRIRS are published under a  CC BY-SA 3.0  license (http://creativecommons.org/licenses/by-sa/3.0/legalcode)
 and can be found here:
http://www.audiogroup.web.fh-koeln.de/wdr_irc.html
They can’t be used directly with the convolver VST as it can only handle wave and PCM.  But I already have created wave files of the BRIRS which can be used with the convolver.
 http://www.xup.in/dl,13480571/Binaurale_Wiedergabe_ueber_KH24022015.zip/
 
Setup:
1 Install foobar
2 install the VST-Host/ resampler component
3 Install the Convolver
4 unzip the Impulse responses and put the file “BinauralAudio” to c/programs
5 load the resampler into the active component list and resample to 48 kHz
6 load the convolver VST into the active component list and load one of the txt-files you find in the file “BinauralAudio” and choose a useful attenuation.
 
That’s all :)
 
Informations and tutorials  how the convolver works  can be found on the download site.
The linked BRIRs are just a example. You can use every IR as long it is  wave or PCM. You can find a lot other  IRs for free and if  you have the rigth hardware you can make your own IR or BRIRs.

 
When I try adding the convolver VST, it says this: 
 
"Could not load VST DLL. You may have tried to use x64 DLL or there is an internal error in the plug-in.
C:\Program Files (x86)\Convolver\Convolver\convolverVST.dll "
 
Any idea how to solve this?
 
EDIT: Never mind, it works now. 
 
Oct 26, 2015 at 3:06 AM Post #797 of 868
It's been a while and a lot of posts since the original post here.  In this time, anyone added any further meaningful mods or recommend alternatives to this (the original)?  If so, why?  Thanks!
 
Nov 12, 2015 at 4:19 AM Post #799 of 868
I initially just wanted to upmix stereo sources in MPC-BE (video player), I tried the mixing function in LAV Audio Decoder but it doesn't upmix. Figured how to do with FFDshow Audio Decoder with the channel mapping function (although it's just copying the channels to rear),

The FFDshow filter also can use Dolby Pro Logic 2. You can find it under the  ffdshow options under "Dolby Decoder".
 
Nov 13, 2015 at 10:57 PM Post #800 of 868
I have updated my plugins chain once again.

@musicreo

After some testing, I think you're right that I shouldn't use both Dolby Headphones and TB Isone together :3

I experimented with the channel mixer in foobar and figured that stereo sources have to be upmixed to 5.1 in order to sound lively through Dolby Headphones.

That's not a problem with foobar since there's the channel mixer plugin. I couldn't find the VST version of it and thus anything stereo from other apps sound as if coming from front.

I wanted the sound to come out from the rear channels as well because it just sounds more immersive.

I found some VST plugins that upmixes stereo to 5.1 such as:
Soundfield UPM-1
DTS Neural Upmix
Halo Upmix
Iosono Anymix Pro
TC Unwrap
Auro 3D
Penteo
Waves UM225/226

But I wasn't willing to shell out money for something like this, especially since foobar's channel mixer was free :frowning2:

After more searching, I finally found "V.I Stereo to 5.1 Converter VST Plugin". It is provided for free so I tried it immediately.
Link: http://stevethomson.ca/vi/

So, my new chain for stereo sources:
1. VB-Audio Cable (Input)
    v
2. VST Host
    v
3. V.I Stereo to 5.1 Converter
    v
4. DH Wrapper  <- 3. Normalizer (To prevent denormalization)
    v
5. Relife
    v
6. FabFilter Pro-Q2
    v
7. Output to Headphones (ASIO)

For 5.1 sources, I just omit the V.I Stereo to 5.1 Converter
I have 2 banks for both situations in VST Host.

Extra edit:
I initially just wanted to upmix stereo sources in MPC-BE (video player), I tried the mixing function in LAV Audio Decoder but it doesn't upmix. Figured how to do with FFDshow Audio Decoder with the channel mapping function (although it's just copying the channels to rear), but since I've found out about V.I, I might as well use it, since it also adds ambience, and other apps could benefit from it. (YouTube for example)

I could barely notice the latency, suitable for real-time applications such as games (although most games already have 5.1 output configuration), youtube, etc.

CPU usage wise, it's around 1-3% for my Intel Xeon E3-1230V3. On i5-4200U, around 3-5%. Without the normalizer, it would spike to 6-8% when no sound is playing.

Extra extra edit:
After more than 12 hours, the sound will start crackling, I suspect it's DH Wrapper. So I bind a shortcut key to a script that restarts VST Host. I also use it for when explorer restarts and the icon disappears.

Batch script:
taskkill /f /im vsthost.exe

start "" /min "C:\vsthostx86\vsthost.exe"
Wow, VI's channel upmixing is great! I prefer it to the channel mixer and FreeSurround--better sound placement and separation, and a bit of a crisper sound.

Thanks for sharing! What are your VI settings, btw?
 
Jan 21, 2016 at 8:47 AM Post #807 of 868
Does anybody of you guys have experience with the Haas effect?
I have downloaded this vst called SX36 which I am preferring to DH.
It does not have reverb, so it also does not create a wide room. But it can enhance the feeling of space and distance of objects effectively.
Although it is a bit tricky to get the right point. I had to spend hours for that. Luckily it should be universal (not related to ears and device) so I can share my settings for those interested. Just ask.
I use it together with Xfeed for a more natural and centred sound, plus EQ, plus Real Bass Exciter. And SoX, of course.
I have used DH for a while but, too unnatural.
I noticed that I had never used the files of @musicreo, I am gonna do it now.
 
Does anybody know if the DH 1.4.1 is the latest dll? I see @Joe Bloggs speaks of DH2 and DH1...
 
Jan 21, 2016 at 1:01 PM Post #808 of 868
 
Setup: 1 Install foobar
2 install the VST-Host/ resampler component
3 Install the Convolver
4 unzip the Impulse responses and put the file “BinauralAudio” to c/programs
5 load the resampler into the active component list and resample to 48 kHz
6 load the convolver VST into the active component list and load one of the txt-files you find in the file “BinauralAudio” and choose a useful attenuation.

Doesn't work for me. When I open Foobar and I go to install a VST, I select the Convolver and I get this:
"Could not load VST DLL. You may have tried to use x64 DLL or there is an internal error in the plug-in.
C:\Program Files (x86)\Convolver\Convolver\convolverVST.dll"
 
I have tried copying in my components but does not appear (of course I have restarted Foobar).
Suggestions?
 
Jan 21, 2016 at 2:55 PM Post #809 of 868
  Doesn't work for me. When I open Foobar and I go to install a VST, I select the Convolver and I get this:
"Could not load VST DLL. You may have tried to use x64 DLL or there is an internal error in the plug-in.
C:\Program Files (x86)\Convolver\Convolver\convolverVST.dll"
 
I have tried copying in my components but does not appear (of course I have restarted Foobar).
Suggestions?


Put all files from the convolver into your foobar folder and then try again installing the vst. Sometimes this helps.
 
Jan 21, 2016 at 10:00 PM Post #810 of 868
 
Put all files from the convolver into your foobar folder and then try again installing the vst. Sometimes this helps.

It worked. I like it more than DH. It gives a nice and natural room, dramatically enhancing the Soundstage in a much more realistic and less reverbering way than DH.
Now, do you use a crossfeed with it? I feel that it is not "in front of you" like when you listen to speakers. It is more around you. And I want to put it a bit more forward.
With DH and xfeed, or with the SX36 and xfeed, I could. But somehow with this convolver the xfeed does not do much. I have had better results with the meier crossfeed on 6.
 
And what about the partitions? what are they for?
And you mention the possibility of adding a "stereo to 5.1" effect in the chain. What for?
 
Anyway I think you should edit your post with the setup instructions to say people the fact about copying the Binaural folder in C/Programs Files. Otherwise it does not work. (btw, why Program Files and not Program Files X86, if the VST is in the X86 folder?)
 
I also think you should say people that they absolutely need to put a limiter in the chain after the Convolver BEFORE playing around with it, because the slider to set the gain is VERY tricky and can jump from -11 to +37 easily, and it is NOT a pleasant experience...
 
And, in the pdf you say to choose the text file, but when I open the convolver it is set to search for a wave file. As far as I see, if I choose the wave file it sets input and output channels to 16 and the panning is all left. Not good. If I chose the text file (config file) it sets input at 8 and output at 2 and all is fine.
 
Thanks for the plugin, very impressed.
 

Users who are viewing this thread

Back
Top