Audio Asylum Thread Printer

Get a view of an entire thread on one page

FAQ / News / Events

Computer Audio Asylum

Foobar 0.8X and 0.9X - totally different animals for ASIO

207.216.246.51

[ Follow Ups ] [ Post Followup ] [ Thread: Display Email Next ] [ Computer Audio Asylum ]

Hey gang:
Just doing some more experimentation and stumbled across something unusual.
I have both Foobar 0.8.3 and 9.2.4 running on my PC audio server. With 0.8.3 you can get the Otachan ASIO output plugin. This worked well with my M-Audio Revolution in ASIO mode, and I never even thought about channel mapping - the channels all seemed to map "sequentially" across the M-Audio mixer, so no problem.
Enter Creatives X-Fi. In ASIO mode, you get a full channel mapping matrix at your finger tips. You can not only map any input channel to any output channel, but you can MIX any combination of input channels to a given output channel! (Very useful for recording people I am sure). Anyways, one would think that channel mapping would be intuitive using Foobar 0.8.3 and Otachan ASIO plugins. Think again. The channel mapping is totally bazaare. Only two out of four stereo pairs will map to a physical output (Front L/R and Surround L/R, but not C/LFE or Side L/R). Here is the goofy part - you can access all of the channels from the Otachan plugin using the Center/Top input, but they are no good to you if all mapped to this single (virtual) input. They're basically mixed somehow.
So, I switched over to Foobar 0.9.2.4 and used the "new" ASIO output plugin (developed by Peter Pawlowski himself). This plugin has an amazing attribute - CHANNEL MAPPING! So of course, I was able to make it work flawlessly with the X-Fi.
Right now I have three excellent sounding crossovers at my disposal - Foobar's 4-way "Xover" by Francois Bourdon (IIR based - 16/44.1 only - fixed filter Q's), Aedio Japan's "foo_dsp_channeldividerf3.dll" - an amazing FIR based (linear phase and transient perfect) 3-way crossover with delay taps. The most flexible of the bunch is my latest aquisition: Jan Luszczek's 4-way "Allocator" with 6-42db/octave adjustable Q IIR filters, and "Arbitrator" section which does phase-correction using 'forward-reverse' processing. (This gives you the advantage of IIR filters, but can make them transient perfect as well with phase correction.)
http://xover.sourceforge.net/
http://www.aedio.co.jp/download/
http://www.thuneau.com/allocator.htm
http://www.thuneau.com/arbitrator.htm
Cheers,
Presto

Follow Ups:

Re: Foobar 0.8X and 0.9X - totally different animals for ASIO - aljordan 07:42:52 03/01/07 (0)
Hi Presto,
You said you were using the Foobar ASIO driver channel mapping capability, and then you mentioned the Thuneau Allocator. Are you actually using Foobar's ASIO output mode at the same time you are using the Allocator? If so, I would interested in knowing how. I can't have more than one ASIO application running at a time. When I use the Allocator I have to use DS or Kernel Streaming output.
Thanks,
Alan
Re: Foobar 0.8X and 0.9X - totally different animals for ASIO - Dawnrazor 11:19:16 02/22/07 (3)
Presto,
You are totally right about the mapping of the current version. I did find it a bit counter intuitive for those of us using stereo crossovers.
To get it to work, channels 3 and 4 have to be CENTER and LFE, with channel 5 and 6 being left and right surround speakers.
Anyhow, on the thuneau product, is it resource intensive?
I am thinking of buying it, but I am concerned my under 1ghz processor will not be happy.
The Foobar crossover by Bourdin doesn't seem to have too much of a problem running as long as I dont upsample.
- Re: Foobar 0.8X and 0.9X - totally different animals for ASIO - dwk 16:26:46 02/22/07 (1)
  I have run Allocator on a 1GHz PIII machine running Win2k, and if it's dedicated to playback you should be OK. The xovers themselves don't really use too much cpu. The Arbitrator algorithm does make it slightly more heavyweight, but not overly so, if I remember correctly.
  I'm pretty sure that Jan has a demo download that you can try to be sure, though.
  - Re: Foobar 0.8X and 0.9X - totally different animals for ASIO - Dawnrazor 17:33:05 02/22/07 (0)
    Thanks,
    It is dedicated, and only runs foobar. My concern stems from my desire to use a better front end than Foobar, so another program will be running to.
    From the sounds of it, i should be OK, and I'll try the demo to see.
    Thanks.
- Re: Foobar 0.8X and 0.9X - totally different animals for ASIO - Presto 14:13:37 02/22/07 (0)
  Dawnrazor:
  The thing to be aware of with the Thuneau Allocator/Arbitrator is that it is a VST plugin (it comes with it's own simple and very effective vst host). VST plugins apparently share resources with some video functions, and in some older PC's, sometimes video hardware acceleration needs to be reduced.
  I am running a PIV 2.93GHz Hyperthreading processor and the Allocator, doing 3-way stereo with arbitrator enabled for both filter poles, uses about 15% of CPU.
  I am not a big upsampling fan, but when I did use realtime upsamplers (like secret rabbit code) I found them to have occassional surges of CPU load - they seemed to run into math troubles from time to time.
  The Thuneau products are nice because they accept everything from 16/44.1 to 24/96 and spit out 24bit+input_sample_rate. All this with 64bit internal processing. This crossover costs a little money for good reason...
  So. How will it behave on your PC? Too many variables to tell. I would download the demo and put it through its paces.
  Cheers,
  Presto

This post is made possible by the generous support of people like you and our sponsors:

Analog Engineering Associates

You can not post to an archived thread.