Results 1 to 15 of 15

Thread: more speakers channels for GrandOrgue ??

  1. #1
    Midshipman, Forte
    Join Date
    Dec 2011
    Location
    Bogotá
    Posts
    42

    more speakers channels for GrandOrgue ??

    Hello VPO-friends.
    I would like to have the output of my computer sending the output differently if it is first manual, second m or pedal. I know that GO is "only" stereo 2.0 but this would be a great enhancement to have for example the sound directed to a 5.1 system. Smthg like the grand-orgue in center and front speakers, recit in rear-speakers, pedal in center and woofer.
    Or using multiple soundcards...
    Is this possible? How to do it?

  2. #2
    Captain of Water Music
    Join Date
    Nov 2009
    Location
    Sweden
    Posts
    275
    Hi!

    This (multi-channel support and routing) is already suggested and is currently "under research" even though no official feature request is registered. From Martins answers on the development mailing list it seems to be somewhat complicated and require lots of changes which makes it likely to be postponed until next version. But amazing things has happened before!

    Kind regards

    Lars P

  3. #3
    Midshipman, Forte
    Join Date
    Dec 2011
    Location
    Bogotá
    Posts
    42
    Hi Lars, thanks. Where can I can register a request?

  4. #4
    Captain of Water Music
    Join Date
    Nov 2009
    Location
    Sweden
    Posts
    275
    Hi!

    At GrandOrgues Sourceforge page. Under Tracker you'll find the Feature Requests, but you'll have to be logged in to register one.

    Kind regards

    Lars P

  5. #5
    Rear Admiral Appassionata (Ret) Ghekorg7 (Ret)'s Avatar
    Join Date
    Mar 2010
    Location
    Athens, Greece
    Posts
    1,098
    Hi !

    I'll keep this phrase of Lars : "But amazing things has happened before!"

    INDEED !! :-)

    Best
    Panos
    *It's like a fight with women, which always ends in .... bed.*
    F.Kafka, Aphorisms.

  6. #6

  7. #7
    Rear Admiral Appassionata (Ret) Ghekorg7 (Ret)'s Avatar
    Join Date
    Mar 2010
    Location
    Athens, Greece
    Posts
    1,098
    Thanks for this prototype !! :-) works OK here (x64) - wasapi tested fine. But still no Asio as on super workin' TDM64.909.
    Any news ?
    *It's like a fight with women, which always ends in .... bed.*
    F.Kafka, Aphorisms.

  8. #8
    Commodore con Forza
    Join Date
    Dec 2011
    Posts
    523
    Quote Originally Posted by Ghekorg7 View Post
    Thanks for this prototype !! :-) works OK here (x64) - wasapi tested fine. But still no Asio as on super workin' TDM64.909.
    Any news ?
    How is the performance, when using more than two output channels?

  9. #9
    Ensign, Principal
    Join Date
    Oct 2009
    Posts
    58
    Quote Originally Posted by e9925248 View Post
    How is the performance, when using more than two output channels?
    Thanks for providing this and previous versions of GrandOrgue v 3.

    A first report on using the multichannel version briefly.

    Test platform: W 7 64 bit Q6600 8GB Ram (already becoming obsolescent)
    Focusrite Saffire with latest W 7 drivers
    Organ used: PiteaMHS 3 manual + Pedal loaded with all loops, all attacks, all releases, no lossless compression.

    1) Using a single stereo channel.
    No sound breakdown with any of the different sound drivers.
    Lowest latency with WDMKS (PortAudio used) : latency 10 ms with buffers at 64
    WASAPI (PortAudio used) : with buffers at 64: still 30 ms, uses Focusrite Saffire WDM driver
    with buffers at 1024: 46 ms
    Windows Direct Sound (PortAudio used) : 14 ms, uses Focusrite Saffire WDM driver
    with buffers at 256; lower buffers gives sound breakdown.
    Directsound : 20 ms, uses Focusrite Saffire WDM driver
    with buffers at 448; lower buffers gives sound breakdown.
    MME (PortAudio used) : 29 ms
    with buffers at 640; lower buffers gives sound breakdown.

    2) Using 3 stereo channels (Front, Surround, Bass)
    With 3 stops drawn in each of the 4 divisions.
    WDMKS (PortAudio used) :
    with buffers at 384 latency for each stereo channel is 10 ms
    lower buffers gives distorded sound
    but only manual 1 and 3 play without sound breakup
    manual 2 and pedal give stuttering sound
    increasing buffers up step by step up to 1024 gives no improvement

    WASAPI (PortAudio used) :
    with buffers at 512 latency for each stereo channel is 30 ms
    lower buffers gives distorded sound
    but only manual 1 and 3 play without sound breakup
    manual 2 and pedal give stuttering sound
    increasing buffers above 512 makes manual 1 and 3 to give stuttering sound
    increasing buffers up step by step up to 1024 gives no improvement

    Windows Direct Sound (PortAudio used) :
    with buffers at 896 latency for each stereo channel is 21 ms
    no sound breakup or stuttering sound playing manuals 1 and/or 3
    stuttering sound playing manual 2 and/or pedal
    with buffers at 960 latency for each stereo channel is 22 ms
    no sound breakup playing manual 2
    stuttering sound playing manual 1 and/or 3 and/or pedal

    Some questions:
    - Is there any logical explanation why playing manual 1 and 3 works and manual 2 and pedal not?
    The midi channels used are: manual 1 = 1, manual 2 = 2, pedal = 3, manual 3 = 4.
    Channels 1 to 3 come from the midi output of the Focusrite Saffire, channel 4 comes from M-Audio 61 usb keyboard.

    - Is it logical that Windows Direct Sound using PortAudio gives better latency than WASAPI?

    - Any specific tests that could help?

    All the best.

    Erik.
    Last edited by erikds; Mar-23-2012 at 19:11.

  10. #10
    Commodore con Forza
    Join Date
    Dec 2011
    Posts
    523
    Could you try increasing the concurrency options?

  11. #11
    Ensign, Principal
    Join Date
    Oct 2009
    Posts
    58
    Quote Originally Posted by e9925248 View Post
    Could you try increasing the concurrency options?
    I did that (using the same platform as given earlier, 3 audio groups, Pitea 3 manuals plus pedal).
    I have glitchfree sound under following conditions:
    - using WDMKS; latency indicated remains 10 ms independent of buffer setting
    no tests done with other driver yet
    - Concurrency level 127
    - Release concurrency level 6 or 7
    - Buffers 960 or 1024
    Most important factor to obtain glitchfree sound is that each time in Audio Settings the Settings tab is opened,
    - the release concurrency is set to 6 or 7 as one before last parameter
    - the concurrency is set at 127 as last parameter
    - clicking OK is postponed long enough to do this while processor load is at its minimum (2% in my case)
    If this is not followed strictly, stuttering sound results but only when playing manual 2, other 2 manuals and pedal give glitchfree sound.

    Remark: There seems to be no difference in behavior between 32 bit and 64 bit version of GrandOrgue latest prototype versions.

    Please explain what the concurrency parameters actually do.
    (Testing haphazardly "in the blind" does not seem very efficient.)
    Any idea why only one manual, and always the same, causes stuttering sound?
    Any idea why WDMKS latency does not follow buffer settings?
    Having to set the upper limits for all 3 parameters is not "nice"; maybe the useable range needs to be re-calibrated?

    All the best.

    Erik.

  12. #12
    Commodore con Forza
    Join Date
    Dec 2011
    Posts
    523
    Thank you for your testing. Multi channel output is currently at an very early stage - I'm currently trying to get the initial version in a mergeable state.

    Quote Originally Posted by erikds View Post
    - Is there any logical explanation why playing manual 1 and 3 works and manual 2 and pedal not?
    The midi channels used are: manual 1 = 1, manual 2 = 2, pedal = 3, manual 3 = 4.
    Channels 1 to 3 come from the midi output of the Focusrite Saffire, channel 4 comes from M-Audio 61 usb keyboard.
    The MIDI channels/manuals should not influence the audio output. The unit of work are windchests. Either one output channel and/or a number of windchest in some audio groups can be the source of sound errors.
    As Pitea features one windchest per manual and you likely put all stops of a windchest in one audio group, you can see this result.

    I need more information of your audio configuration (audio groups, pipe/register assignment to audio groups, audio routing configuration of the setting dialog) to make further assumtions.

    From your statements, I would conclude, that you added three "audio interfaces" a 2 channels in GO. The multi device synchronization is in an early state.
    Adding one audio interface with more than 2 channels (eg. my onboard sound card offers a front+rear connector as a audio device with 4 channels in GO) should result in a more stable sound.

    Quote Originally Posted by erikds View Post
    - Is it logical that Windows Direct Sound using PortAudio gives better latency than WASAPI?
    GO reuses existing audio libraries to avoid dealing with all the sound output details on three plattforms. Only two settings in GO influence the sound output via Portaudio:
    * the buffer size in samples
    * the latency setting for the audio interface (Properties button)

    Quote Originally Posted by erikds View Post
    Please explain what the concurrency parameters actually do.
    (Testing haphazardly "in the blind" does not seem very efficient.)
    Release concurrency factor specifies in how many parts the release sample processing is splitted.
    Concurrency specifies, how many helper threads are used to offload processing the various subtasks (eg. releases, windchests) from the main audio threads.

    Quote Originally Posted by erikds View Post
    Any idea why WDMKS latency does not follow buffer settings?
    The actual latency in the audio dialog is returned by PortAudio. You could check the latency setting for this device.

    Quote Originally Posted by erikds View Post
    Having to set the upper limits for all 3 parameters is not "nice"; maybe the useable range needs to be re-calibrated?
    The upper limit constant for the buffer size in the GO source is set to 1024 - raising it to n * 1024 should be easy. Propably it was set to 1024 to limit the maximum respone time (23 ms).

    Release concurrency should never ever be more than the number of cores.

    The concurrency limit is probably too low for a high number of audio groups with lots of windchests in the current implementation. A high number of threads will be causing other problems, so I intend to rework the threading schema.

    PS:
    Your statement about the order of setting the concurrency parameters sounds strange. The order, in which is anything done in the setting dialog should be irrelevant. At the moment, I can't imagine, what kind of side effect of your actions could result in the described behaviour.

  13. #13
    Ensign, Principal
    Join Date
    Oct 2009
    Posts
    58
    The MIDI channels/manuals should not influence the audio output. The unit of work are windchests. Either one output channel and/or a number of windchest in some audio groups can be the source of sound errors.
    As Pitea features one windchest per manual and you likely put all stops of a windchest in one audio group, you can see this result.

    I need more information of your audio configuration (audio groups, pipe/register assignment to audio groups, audio routing configuration of the setting dialog) to make further assumtions.

    From your statements, I would conclude, that you added three "audio interfaces" a 2 channels in GO. The multi device synchronization is in an early state.
    Adding one audio interface with more than 2 channels (eg. my onboard sound card offers a front+rear connector as a audio device with 4 channels in GO) should result in a more stable sound.
    See png's for above info

    more speakers channels for GrandOrgue ??-go-audio-groups-pngmore speakers channels for GrandOrgue ??-go-settings-pngmore speakers channels for GrandOrgue ??-go-audio-output-png

    Windchest 1 all pipes (Pedal) to Center-Bass
    Windchest 2 and 4 all pipes to Front
    Windchest 3 all pipes to Surround

    Under Sample Loading: all settings are Parent default

    Release concurrency should never ever be more than the number of cores.
    My Q6600 has 4 cores. If i set release concurrency to anything less than 7 i get stuttering sound from playing manual 2.

    Your statement about the order of setting the concurrency parameters sounds strange. The order, in which is anything done in the setting dialog should be irrelevant. At the moment, I can't imagine, what kind of side effect of your actions could result in the described behaviour.
    If i change parameters on the Settings window in any other way than described the penalty is stuttering sound.
    Maybe one of the idiosyncrasies of Windows 7 "overzeleaous" User protection "feature"?
    It looks as if after changing one or more parameters they do not "take" unless strictly following the procedure as described.

    As soon as you have something new to test, just let me know.

    All the best.

    Erik.
    Last edited by erikds; Mar-25-2012 at 21:18.

  14. #14
    Commodore con Forza
    Join Date
    Dec 2011
    Posts
    523
    Quote Originally Posted by erikds View Post
    See png's for above info

    Windchest 1 all pipes (Pedal) to Center-Bass
    Windchest 2 and 4 all pipes to Front
    Windchest 3 all pipes to Surround

    Under Sample Loading: all settings are Parent default
    Your configuration is as I expected. You completly put one windchest into an audio section, one audio section is only one audio group and an windchest corresponds to a manual => a problem with an audio output affects a complete manual.

    Additionally, I had yet spend most effort on one sound interface in GO with more than two channels case, not multiple sound interfaces.

    Quote Originally Posted by erikds View Post
    My Q6600 has 4 cores. If i set release concurrency to anything less than 7 i get stuttering sound from playing manual 2.
    I'm aware of this behaviour.

    Quote Originally Posted by erikds View Post
    If i change parameters on the Settings window in any other way than described the penalty is stuttering sound.
    Maybe one of the idiosyncrasies of Windows 7 "overzeleaous" User protection "feature"?
    It looks as if after changing one or more parameters they do not "take" unless strictly following the procedure as described.
    I don't think, that it is related to something windows specific. GO stores the complete configuration in an object and after that configures sound output from it => Regardless how you configure GO, the same configuration should be set up always in the same way.
    I can only speculate: In my option, it is most likely a strange side effect (eg. affecting the process memory layout).

  15. #15
    Commodore con Forza
    Join Date
    Dec 2011
    Posts
    523
    http://sourceforge.net/projects/ourorgan/files/Development releases/

    Support for one audio interfaces with more than 2 channels has been merge into GO trunk.

Similar Threads

  1. Hmmmm :rolleyes: Possible to use Hauptwerk via mounted instrument speakers?
    By Ntalikeris666 in forum Electronic/Digital Organs
    Replies: 12
    Last Post: Mar-31-2010, 16:51
  2. Need suggestions on Speakers........
    By tittualex in forum Electronic/Digital Organs
    Replies: 3
    Last Post: Aug-20-2009, 18:30
  3. Online Music channels in PC
    By Alex1985 in forum Classical Music Forum
    Replies: 4
    Last Post: Jun-20-2009, 12:50

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •