Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[bob]

http://files.jriver.com/mediacenter/channels/v20/latest/MediaCenter-20.0.20.deb

20.0.20 (10/2/2014)

1. Fixed: https protocol broken in 20.0.18
2. Changed: Switched to gnutls from openssl for https support because of a bug in the openssl static libs that prevented some websites certificates from being verified (i.e. twitter)

20.0.18 (9/26/2014)

1. Fixed: JRWorker segfaulting occasionally prevented some file types from being recognized.
2. Fixed: MediaCenter also occasionally segfaulted on exit.

20.0.17 (9/19/2014)

1. NEW: Implemented the file "locate" function.
2. Fixed: Menu items that open a URL in a browser.
3. Fixed: Disabled menu items unsupported on linux.
4. Fixed: Default for managing links changed to external browser.
5. Changed: possibly more ability to play Internet streams. Needs testing.

20.0.16 (9/16/2014)

1. Fixed: Series and Season cover art downloaded to the wrong folders.
2. NEW: Video playback supports playing from a library server.
3. Fixed: Last.fm scrobbling.

20.0.15 (9/12/2014)

1. Fixed: Fullscreen video playback is functional.
2. Changed: The video renderer uses VSYNC if available in the graphics driver.

20.0.13 (9/8/2014)

1. Fixed: Corrected the latency calculation for the ALSA output.
2. NEW: Video support.

20.0.7 (8/14/2014)

1. Changed: Reverted to previous device detection scheme. Because of a bug in ALSA this can occasionally crash when enumerating devices. It fixes detection on some platforms however.

20.0.6 (8/12/2014)

1. NEW: File path dialogs for Options (requires gtk 2.0)
2. Changed: Retry audio device enumeration on failure. Fixes troubles finding devices on ubuntu x64.
3. Changed: More tweaks to eventing.

19.0.160 (8/5/2014)

1. Changed: (again) the root window updating. Reverted somewhat to the previous behavior. Only the right side, bottom and bottom right corner work properly for resize.

19.0.156 (7/24/2014)

1. Fixed: One of the types of HTTP redirects was not being handled properly. This fixes troubles with some internet streams and one particular DLNA controller.
2. Fixed: The ALSA Buffer time and Period time parameters were not being used.
3. NEW: The ALSA output format can be fixed to a specific type (i.e. S32_LE) as well as auto detected (previous method).
4. Fixed: Window Size/Position issues in some dialogs (i.e. dsp studio sample rate selection).
5. NEW: Rudimentary portable drive (handheld) support (no autodetect).
6. Changed: Root window move/resize. Position/Size isn't updated until left mouse button is released. Prevents jumping, etc. This will likely change again...

19.0.152 (7/15/2014)

1. Fixed: Detection of mountpoints under /media and /mnt, treated as "drives"
2. NEW: Implemented the functions which allow "Rename, move and copy files" to work.
3. NEW: For deleting files from the library, moving to the recycle (Trash) bin implemented. Since this is desktop dependent you'll need to symlink ~/Trash to your desktop trash directory.

19.0.148 (7/3/2014)

1. Fixed: Multichannel mapping.

19.0.146 (6/26/2014)

1. Fixed: Watched folders now see subfolders properly.

19.0.144 (6/24/2014)

1. NEW: Implemented DSD over DoP. Tested with Mytek Stereo192 DSD DAC.
2. NEW: Added S24LE audio format.
3. Fixed: Registration issues (losing previous registration, etc).
4. Fixed: Socket handle leak.

19.0.143 (6/19/2014)

1. NEW: Added DPMS control to the HDMI screen option under audio.
2. Fixed: Registration issues (losing previous registration, etc).
3. Fixed: WOL Mac address were not being supplied to the access key server properly.

19.0.141 (6/13/2014)

1. Changed: Default volume mode is now "Internal" rather than "System".
2. Changed: Added command line switch for restoring license from an MJR file (/RestoreFromFile).
3. Changed: Tooltips that obscure underlying clickable fields removed.
4. NEW: Screensaver control enabled (turn on hdmi screen option under Audio).
5. NEW: Added command line alsacap to the debian package (in MC's shared library location).
6. Changed: Audio driver configuration changed to be more like windows/mac.
7. NEW: The ALSA driver selection can be made from the WebRemote control panel.
8. NEW: DSD Encoder working.

19.0.138 (6/6/14)

1. Fixed: Asynchronous crashing triggered by Media Network being enabled.
2. Fixed: Ratings.
3. Fixed: Bug in eventing.
4. NEW: Licensing is now enabled. The trial period starts.

19.0.135 (5/9/2014)

1. Fixed: The watch folders function.
2. Fixed: A memory leak from above.

19.0.134 (5/15/2014)

1. New: The font dialog is enabled (Options->Tree and View->Select Font). You can choose from the default Lucida, GNU Unifont and a user defined font string. The User Defined font string has to be acceptable to XLoadQueryFont. If the font isn't found, MC will fall back to Lucida and then Mutt-Clearlyu. The font height can also be trimmed manually. Height in pixels of trim.

19.0.133-2 (5/12/2014)

1. Change: Reverted from requirement for dynamic ssl and crypto libs added inadvertently in 19.0.133

19.0.133 (5/9/2014)

1. Changed: The font loads orders of magnitude faster.
2. Fixed. The maximum size calculated for the font characters were not being calculated properly at load time resulting in imprecise placement of elements in the MC window.

19.0.129-3 (4/22/2014)

1. Changed: libcurl system shared lib dependencies reduced.
2. Fixed: Internet reader timeout.

19.0.129 (4/21/2014)

1. Changed: libcurl integrated into the JRReader. Greatly reduces dependencies and uses a newer version of libcurl. Needs testing to see if anything is missing and to check compatibility with distro's (unsupported) other than debian wheezy.
2. Changed: xz compression on the .deb greatly reduces download size.

19.0.128 (4/15/2014)

1. NEW: System volume control and mute control if supported by the device (controls the PCM output). The mixer is discovered during audio device config (reselect your audio device to get it).

19.0.124-5 (3/27/2014)

1. Changed: More maximize/not maximized work.


19.0.124 (3/25/2014)

1. Fixed: Use 64 bit stat functions. Fixes problems using mounted windows (and CIFS NAS) filesystems. Probably also fixes other random file problems.
2. Fixed: Moving the main window works properly now.
3. Changed: Method of screen position saving restoring (again). Needs testing.
4. Fixed: Resizing works (as long as nothing is playing).
5. Fixed: Restores maximized/not maximized state upon start.

19.0.122 (3/14/2014)

1. Fixed: Remembers size and position on re-open. Needs testing on different platforms / window managers.
2. Changed: The method used to allow moving/resizing. Needs testing on different platforms / window managers.

19.0.121 (3/11/2014)

1. Fixed: Memory leak introduced in 19.0.119
2. NEW: WMA playback support.
3. Fixed: HTTP authorization tokens weren't being done properly, so that only one server authorization was stored at a time.
4. Changed: Timeout advanced to July 10th 2014.
5. Fixed: Internet reader, HTTP redirects were not handled properly.
6. NEW: Use backing store for windows if the server supports it (UI speed increase).
7. NEW: Mouse cursor changes (on click) for main window move/resize. Intend this to work on hover in the future.
8. Changed: More tweaking of the tooltip and transient window hints (need feedback).

19.0.119 (2/28/2014)

1. Fixed: Movement and resizing of MediaCenter works properly now.
2. Fixed: HTTP authentication was broken in build 19.0.118.

19.0.118 (2/21/2014)

1. Fixed: The linux network reader fixed to handle custom methods (needed for DLNA events).
2. Fixed: When acting as a DLNA renderer, wouldn't move on to the next track in a playlist properly.
3. Fixed: Tooltip windows to work better with more window managers. Still not transparent however.

19.0.117 (2/14/2014)

1. Fixed: Enabled filesystem statistics gathering function.
2. Fixed: Allow : in filename (was filtered out by windows code).
3. Fixed: Remove special treatment of UNC style paths (windows code).
4. Fixed: Remove special treatment of . in pathnames, i.e. foo./bar (windows code).
5. Fixed: The fixing of the documents path causes the automatic database backup to occur. (~/Documents/JRiver ....)

19.0.116 (2/10/2014)

1. Fixed: Prevent tooltips and transparent windows from stealing input focus. Fixes upper right corner search box.

19.0.115 (2/5/2014)

1. Changed: Completely reworked the linux ALSA audio output device control to be more efficient and reduce latency.
2: NEW: Allow advanced user control of the latency.

19.0.112 (1/30/2014)

1. Fixed: Stability improvements.
2. Changed: Moved timeout to April 9th 2014

19.0.109 (1/23/2014)

1. Fixed: Workaround a curl bug that could cause crashing on slow dns lookups.
2. Fixed: Window focus issues. Need feedback.

19.0.108 (1/21/2014)

1. Fixed: Stability improvements.

19.0.103-3 (1/14/2014)

1. Fixed: Stability improvements.

19.0.103-2 (1/13/2014)

1. Changed: Removed system requirement for libcrypto++ and libboost-regex.

19.0.103 (1/9/2014)

1. NEW: Implemented libboost for regular expressions to match the windows and mac versions.
2. Changed: More dependencies tuning.

19.0.102 (1/3/2014)

1. Changed: Bumped timeout to February 9th 2014
2. Changed: Package no longer depends on the various linux audio encoders to install, they are recommended but not required leaving it up to the user to install. These are Lame, Vorbis-tools, Musepack-tools. They can be 32 or 64 bit.

19.0.94 (12/16/2013)

1. Fixed: Bug introduced from the column header crash fix.

19.0.92 (12/11/2013)

1. New: Added Musepack (mpc) encoder (requires system musepack-tools).
2. Fixed: Pathname length for files increased from 260 to 4096.

19.0.91 (12/10/2013)

1. Changed: More shared library work (stability).
2. Changed: Now requires system libcrypto++9 (>= 5.6.1)
3. Fixed: MPC (Musepack) input plugin.

19.0.90 (12/9/2013)

1. Fixed: .deb pacakge dependencies will allow either i386 or amd64 versions of lame and vorbis-tool and un-arched xfonts.
2. Changed: Reliability in loading shared libraries, was causing problems when switching between audio formats during playback.

19.0.89 (12/6/2013)

1. New: Added OGG encoder.

19.0.88 (12/5/2013)

1. Fixed: OGG input plugin.

19.0.85-1 (12/3/2013)

1. Faster: Drawing from an internal image to the screen is much faster (requires 0 memory manipulations before draw; previously required two copies before draw).
2. Changed: More resolutions avaiable for the dynamic MC Icon.
3. Fixed: DLNA fixes from the main code.

[InflatableMouse]

2. Fixed: MediaCenter also occasionally segfaulted on exit.

Seems to be fixed for me too now. I (almost) always got a segfault on exit.

[mwillems]

Seems to be fixed for me too now. I (almost) always got a segfault on exit.

Same here, good work!

[hifi25nl]

I think you forgot the gnutls dependency. In 64 bit Archlinux for example without lib32-gnutls mediacenter20 will segfault.

[bob]

I think you forgot the gnutls dependency. In 64 bit Archlinux for example without lib32-gnutls mediacenter20 will segfault.

It's not using shared gnutls libs.
It is using libp11-kit0 (>= 0.12) and requires that in the debian dependency.

[InflatableMouse]

The option to remember last position remembers all tabs except Playing Now. That reverts to the default page.

[mwillems]

I'm getting loud static and distortion whenever I attempt to playback more than two channels with an Asus U7 (it's a 7.1 channel USB soundcard).

Specifics:
1) Speaker-test and other applications have no problem playing to all eight channels on the same box.  
2) When output format is set to "stereo," JRiver works fine when addressing any of the hardware devices (front, surround51, or surround71).
3) When output format is set to "5.1" or "7.1", JRiver outputs loud distortion that is modulated by the music that's playing (i.e. I can hear the source but it's heavily garbled).  I tried disabling all other DSP, and turning off JRSS mixing, but no effect.  I also tried resampling to 48KHz or 44.1KHz with no change. No combination of settings seems to allow ungarbled playback at >2 channels in output format.

As noted all channels work fine with speaker-test (targeting the ALSA hardware devices), so this seems to be JRiver specific.  This is on Arch Linux running Gnome.  EDIT: I tested on XFCE to rule out pulse and had the same results.   Let me know other testing would be helpful.  

Here's the relevant alsacap output:

Code: [Select]

 Device 0, ID `USB Audio', name `USB Audio', 1 subdevices (1 available)
    2..8 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE, S24_3LE
      Subdevice 0, name `subdevice #0'
  Device 1, ID `USB Audio', name `USB Audio #1', 1 subdevices (1 available)
    2 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE, S24_3LE
      Subdevice 0, name `subdevice #0'
  Device 2, ID `USB Audio', name `USB Audio #2', 1 subdevices (1 available)
    2 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE
      Subdevice 0, name `subdevice #0'

And here's a log with one successful playback and some garbled ones for comparison:

[InflatableMouse]

On what kind of pc is that happening? We may want to compare the USB controller because I had trouble getting sound at all via USB.

[mwillems]

It's a two year-old Asus laptop (K46CA).  It's an i5 with, I think, an intel USB controller.  I don't think it's a hardware issue as literally all other software appears to have no problem (vlc is fine, speaker-test is fine), and JRiver is perfectly fine for stereo.  I get no errors in dmesg.  I do get a few journal errors from the U7 on boot "kernel: usb 1-1.1: cannot get ctl value: req = 0x81, wValue = 0x0, wIndex = 0x1400, type = 3", but those don't recur when playback is attempted.

It kind of sounds like clipped output (i.e. it's much louder than it should be, and very distorted).

Based on the fact that it seems to work with other software whether or not pulse is loaded, it's seems like it's either some specific issue in my ALSA settings that nothing else cares about, or a bug in JRiver.  Mouse, you've got an Asus ST, right?  Do you or other folks have multi-channel working in JRiver for Linux, with any Asus/Cmedia based cards in particular?

[InflatableMouse]

I suppose you did try other USB ports?

[mwillems]

I suppose you did try other USB ports?

Yeah, I tried two different ports on the laptop, and then an attached USB hub for good measure, no change.  It sounds great in two channel, but I need the multi-channel capability for some active speakers I'm designing.  I guess one thing I haven't tried is testing the device on a windows machine with MC; if it works there, I may need to reintroduce another windows machine into the house for now :-(  Update: It works fine with JRiver on Windows (but there's a proprietary driver for windows so it may be a different ballgame).  At least I know the hardware's not defective.

[bob]

Hard to say what the issue is here.
We do channel mapping to map the state of MC's channels into the ALSA channel order. I don't think that would be the issue, you'd just get stuff on the wrong channel.
It's possible when we open the device in the hinted mode (the string you see for the device in our Audio Configuration), the device driver does something weird.
if it's possible to specify the hinted form of the device to the speaker test you should try that and see if the results are the same.

[mwillems]

Hard to say what the issue is here.
We do channel mapping to map the state of MC's channels into the ALSA channel order. I don't think that would be the issue, you'd just get stuff on the wrong channel.  It's possible when we open the device in the hinted mode (the string you see for the device in our Audio Configuration), the device driver does something weird.  if it's possible to specify the hinted form of the device to the speaker test you should try that and see if the results are the same.

I'm not 100% sure what you mean by "hinted form";  if you mean the PCM devices, like

Code: [Select]

surround71:CARD=U7,DEV=0,that's the device string I used with speaker-test, and it worked for all channels.  It looks like that's the format MC uses in setting the output device (although there's some whitespace in the MC device name), but if you mean a different type of device addressing format (hw devices, etc.), I'm happy to test but an example would be helpful.

MC successfully plays stereo to that same device (surround71:CARD=U7,DEV=0), but it just produces garbling/clipping when output format is set to >2 channels.

Edit: One observation that I have is that the alsacap (and aplay -l) output attached below shows three devices for the U7; only one of those three devices (device 0) has 8 channels, the others are 2 channel devices (presumably the separate headphone output, and possibly the spdif out?).  The device switches between these modes with a switch on the device; I notice that I get the same results in MC when trying to play 7.1 to the headphone out (which only really supports 2 channel), but that's probably appropriate: I would expect there to be an error or bad behavior when trying to play 7.1 to a 2 channel interface.  

Is it possible that MC is addressing itself to the wrong one of the three U7 devices?  The PCM devices are unitary (there aren't separate PCM handles for each of those three devices thrown by alsacap, just one set), so maybe I can resolve it by creating a custom handle that points only to the eight channel device?  I don't know much about the innards of ALSA, so any thoughts as to whether I'm barking up the wrong tree would be appreciated.

aplay -L output

Code: [Select]

sysdefault:CARD=U7
    Xonar U7, USB Audio
    Default Audio Device
front:CARD=U7,DEV=0
    Xonar U7, USB Audio
    Front speakers
surround21:CARD=U7,DEV=0
    Xonar U7, USB Audio
    2.1 Surround output to Front and Subwoofer speakers
surround40:CARD=U7,DEV=0
    Xonar U7, USB Audio
    4.0 Surround output to Front and Rear speakers
surround41:CARD=U7,DEV=0
    Xonar U7, USB Audio
    4.1 Surround output to Front, Rear and Subwoofer speakers
surround50:CARD=U7,DEV=0
    Xonar U7, USB Audio
    5.0 Surround output to Front, Center and Rear speakers
surround51:CARD=U7,DEV=0
    Xonar U7, USB Audio
    5.1 Surround output to Front, Center, Rear and Subwoofer speakers
surround71:CARD=U7,DEV=0
    Xonar U7, USB Audio
    7.1 Surround output to Front, Center, Side, Rear and Woofer speakers
iec958:CARD=U7,DEV=0
    Xonar U7, USB Audio
    IEC958 (S/PDIF) Digital Audio Output

compare with aplay -l output

Code: [Select]

card 1: U7 [Xonar U7], device 0: USB Audio [USB Audio]
  Subdevices: 0/1
  Subdevice #0: subdevice #0
card 1: U7 [Xonar U7], device 1: USB Audio [USB Audio #1]
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 1: U7 [Xonar U7], device 2: USB Audio [USB Audio #2]
  Subdevices: 1/1
  Subdevice #0: subdevice #0

and alsacap

Code: [Select]

 Device 0, ID `USB Audio', name `USB Audio', 1 subdevices (1 available)
    2..8 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE, S24_3LE
      Subdevice 0, name `subdevice #0'
  Device 1, ID `USB Audio', name `USB Audio #1', 1 subdevices (1 available)
    2 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE, S24_3LE
      Subdevice 0, name `subdevice #0'
  Device 2, ID `USB Audio', name `USB Audio #2', 1 subdevices (1 available)
    2 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE
      Subdevice 0, name `subdevice #0'


[bob]

I'm not 100% sure what you mean by "hinted form";  if you mean the PCM devices, like

Code: [Select]

surround71:CARD=U7,DEV=0,

Yes, that is what I mean by the hinted device. We are asking ALSA to enumerate all devices. That is what you see in MC's list of ALSA devices.
We open that exact device string that you select in MC. (minus the " [ALSA]" at the end)
I don't see any whitespace in any of my devices in MC, the strings are the same as with the aplay -L command.
The point of asking the device to open with the hinted string is for it to do any subdevice choosing it needs to do based on the hinting.
That's why I asked you to try it with the speaker test and the same string.

I don't know why there should be any difference
that's the device string I used with speaker-test, and it worked for all channels.  It looks like that's the format MC uses in setting the output device (although there's some whitespace in the MC device name), but if you mean a different type of device addressing format (hw devices, etc.), I'm happy to test but an example would be helpful.

MC successfully plays stereo to that same device (surround71:CARD=U7,DEV=0), but it just produces garbling/clipping when output format is set to >2 channels.

Edit: One observation that I have is that the alsacap (and aplay -l) output attached below shows three devices for the U7; only one of those three devices (device 0) has 8 channels, the others are 2 channel devices (presumably the separate headphone output, and possibly the spdif out?).  The device switches between these modes with a switch on the device; I notice that I get the same results in MC when trying to play 7.1 to the headphone out (which only really supports 2 channel), but that's probably appropriate: I would expect there to be an error or bad behavior when trying to play 7.1 to a 2 channel interface.  

Is it possible that MC is addressing itself to the wrong one of the three U7 devices?  The PCM devices are unitary (there aren't separate PCM handles for each of those three devices thrown by alsacap, just one set), so maybe I can resolve it by creating a custom handle that points only to the eight channel device?  I don't know much about the innards of ALSA, so any thoughts as to whether I'm barking up the wrong tree would be appreciated.

aplay -L output

Code: [Select]

sysdefault:CARD=U7
    Xonar U7, USB Audio
    Default Audio Device
front:CARD=U7,DEV=0
    Xonar U7, USB Audio
    Front speakers
surround21:CARD=U7,DEV=0
    Xonar U7, USB Audio
    2.1 Surround output to Front and Subwoofer speakers
surround40:CARD=U7,DEV=0
    Xonar U7, USB Audio
    4.0 Surround output to Front and Rear speakers
surround41:CARD=U7,DEV=0
    Xonar U7, USB Audio
    4.1 Surround output to Front, Rear and Subwoofer speakers
surround50:CARD=U7,DEV=0
    Xonar U7, USB Audio
    5.0 Surround output to Front, Center and Rear speakers
surround51:CARD=U7,DEV=0
    Xonar U7, USB Audio
    5.1 Surround output to Front, Center, Rear and Subwoofer speakers
surround71:CARD=U7,DEV=0
    Xonar U7, USB Audio
    7.1 Surround output to Front, Center, Side, Rear and Woofer speakers
iec958:CARD=U7,DEV=0
    Xonar U7, USB Audio
    IEC958 (S/PDIF) Digital Audio Output

compare with aplay -l output

Code: [Select]

card 1: U7 [Xonar U7], device 0: USB Audio [USB Audio]
  Subdevices: 0/1
  Subdevice #0: subdevice #0
card 1: U7 [Xonar U7], device 1: USB Audio [USB Audio #1]
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 1: U7 [Xonar U7], device 2: USB Audio [USB Audio #2]
  Subdevices: 1/1
  Subdevice #0: subdevice #0

and alsacap

Code: [Select]

 Device 0, ID `USB Audio', name `USB Audio', 1 subdevices (1 available)
    2..8 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE, S24_3LE
      Subdevice 0, name `subdevice #0'
  Device 1, ID `USB Audio', name `USB Audio #1', 1 subdevices (1 available)
    2 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE, S24_3LE
      Subdevice 0, name `subdevice #0'
  Device 2, ID `USB Audio', name `USB Audio #2', 1 subdevices (1 available)
    2 channels, sampling rate 44100..192000 Hz
    Sample formats: S16_LE
      Subdevice 0, name `subdevice #0'


If you turn in logging in MC, the logfile will contain detailed messages about the device opening calls. You could post that here or look for any ALSA "failed..." messages.

[mwillems]

Yes, that is what I mean by the hinted device. We are asking ALSA to enumerate all devices. That is what you see in MC's list of ALSA devices.
We open that exact device string that you select in MC. (minus the " [ALSA]" at the end)
I don't see any whitespace in any of my devices in MC, the strings are the same as with the aplay -L command.
The point of asking the device to open with the hinted string is for it to do any subdevice choosing it needs to do based on the hinting.
That's why I asked you to try it with the speaker test and the same string.

Ok, looking again, you're right, there isn't actually any added whitespace in the MC device list, it just looks like there's an extra space or two in there on one of my screens (possibly due to DPI scaling or font size issues).    

Regardless, I can confirm that I used the "aplay -L" device names in speaker-test and got correct playback in all channels.

If you turn in logging in MC, the logfile will contain detailed messages about the device opening calls. You could post that here or look for any ALSA "failed..." messages.

I attached a log above showing one successful playback attempt and a few garbled ones (http://yabb.jriver.com/interact/index.php?topic=92309.msg636492#msg636492);  I just looked through again and I don't see any ALSA failure messages in there; the successful attempt seems identical to the garbled ones.  MC seems to think it's playing back normally, and playback appears to proceed normally from the UI (the progress bar moves forward, the stop button works, etc.).  

Any other suggestions?  Meanwhile, I'll keep chewing on it.

[bob]

Speaker-test uses the same method to set the channels as MC does.
Do you have DSP studio set to do "source number of channels" and the mixing set to "none"?

[mwillems]

Speaker-test uses the same method to set the channels as MC does.

Not exactly sure what the difference could be. This command produces nice clean-sounding sine waves in all channels

Code: [Select]

speaker-test -Dsurround71:CARD=U7,DEV=0 -c8 -tsine

Do you have DSP studio set to do "source number of channels" and the mixing set to "none"?

I had not previously tried "source number" because it would not work for my intended application, but I just tested with those settings.  Those settings exhibit the same basic problem: clean output with stereo sources, garbled output on all channels with multichannel sources.  

[mwillems]

Well I cracked it; it's kind of convoluted how I got there, but, long story short, the issue is either with MC's auto detection of what bitdepths/formats the U7 supports, or more likely the U7's reporting of what bitdepths/formats it supports.

When I use any of the hinted PCM outputs with a stereo output (front, surround71, etc), MC appears to detect that the device supports S24_3LE and plays that, which plays correctly.  When I used any of the hinted outputs with more than two channels selected in output format, MC's autodetection was likewise detecting that the device supported S24_3LE, and playing that, but the device does not appear to support that format in multichannel.  Attempting to manually set the output to S24_LE does not work and throws a "something went wrong with playback" type error.  Attempting to manually set the format to S24_3LE produces garbled output with any multichannel output format.  

So here's how I fixed it for now:  I created the following .asoundrc to create a hinted PCM alias:

Code: [Select]

pcm.u7hw {   type hw;
                card "U7";
                device 0;
                subdevice 0;    }

pcm.u78c {     type plug;
                slave.pcm "u7hw";
                hint {  show on;
                description "Asus U7 8 Channels";       }       }

Media Center correctly plays 8 channel outputs to "u78c" at S24_LE with no issues, and will actually even play S32 outputs (which my device doesn't actually support to my knowledge, but it's an ASIO device on the windows side, so maybe it takes a 32-bit word but only uses 24 bits).  So my alias works fine, but I never did find a way to get anything other than garbled/clipped output from MC with multichannel when playing to "surround71:CARD=U7,DEV=0," but that's because MC is detecting ot it's reporting to MC that it supports S24_3LE instead of S24_LE, which it doesn't with multichannel output.  

So it's either a bug in MC's auto format detection or (more likely) the hardware is just misreporting what it can and can't do.  If you want to try and run it down, I'm happy to do more testing if that would be helpful, but I'm sorted out over here.

[bob]

Well I cracked it; it's kind of convoluted how I got there, but, long story short, the issue is either with MC's auto detection of what bitdepths/formats the U7 supports, or more likely the U7's reporting of what bitdepths/formats it supports.

When I use any of the hinted PCM outputs with a stereo output (front, surround71, etc), MC appears to detect that the device supports S24_3LE and plays that, which plays correctly.  When I used any of the hinted outputs with more than two channels selected in output format, MC's autodetection was likewise detecting that the device supported S24_3LE, and playing that, but the device does not appear to support that format in multichannel.  Attempting to manually set the output to S24_LE does not work and throws a "something went wrong with playback" type error.  Attempting to manually set the format to S24_3LE produces garbled output with any multichannel output format.  

So here's how I fixed it for now:  I created the following .asoundrc to create a hinted PCM alias:

Code: [Select]

pcm.u7hw {   type hw;
                card "U7";
                device 0;
                subdevice 0;    }

pcm.u78c {     type plug;
                slave.pcm "u7hw";
                hint {  show on;
                description "Asus U7 8 Channels";       }       }

Media Center correctly plays 8 channel outputs to "u78c" at S24_LE with no issues, and will actually even play S32 outputs (which my device doesn't actually support to my knowledge, but it's an ASIO device on the windows side, so maybe it takes a 32-bit word but only uses 24 bits).  So my alias works fine, but I never did find a way to get anything other than garbled/clipped output from MC with multichannel when playing to "surround71:CARD=U7,DEV=0," but that's because MC is detecting to it's reporting to MC that it supports S24_3LE instead of S24_LE, which it doesn't with multichannel output.  

So it's either a bug in MC's auto format detection or (more likely) the hardware is just misreporting what it can and can't do.  If you want to try and run it down, I'm happy to do more testing if that would be helpful, but I'm sorted out over here.

I'm thinking it's a shortcoming in ALSA. It seems to assume you can do all supported channels at all supported formats.
After initializing the device we ask it among other things how many channels it supports then what output formats it supports.
There doesn't seem to be a way to ask it how many channels it supports at whatever output format you can get.
MC's Auto (default) format method starts with the highest resolution output format the device supports then works it's way down.
You can force it to a particular format with the Device Options under the ALSA audio configuration in MC.

[mwillems]

I'm thinking it's a shortcoming in ALSA. It seems to assume you can do all supported channels at all supported formats.
After initializing the device we ask it among other things how many channels it supports then what output formats it supports.
There doesn't seem to be a way to ask it how many channels it supports at whatever output format you can get.
MC's Auto (default) format method starts with the highest resolution output format the device supports then works it's way down.
You can force it to a particular format with the Device Options under the ALSA audio configuration in MC.

That's the weird part. I tried manually setting it, andw hen I try to force output to S24_LE using one of the default hinted PCM devices, MC throws a "something went wrong with playback" error.

When I try to force the exact same format using my custom alias, MC works without a hitch.  
That's the part I haven't really been able to sort out.  It seems like the default PCM devices are providing inaccurate information about what formats the device does or doesn't support, so it may just be a device specific issue.

[bob]

That's the weird part: I tried manually setting it. When I try to force output to S24_LE using one of the default hinted PCM devices, MC throws a "something went wrong with playback" error.

When I try to force the exact same format using my custom alias, MC works without a hitch.  
That's the part I haven't really been able to sort out.  It seems like the default PCM devices are providing inaccurate information about what formats the device does or doesn't support.

That's a bit weird.
I think you could though use the Front: device in MC and set the format in in the Device options. I'd be pretty surprised if that didn't work for the supported format/channels.
You could check that vs the speaker-test too choosing the output format is something I think it's capable of...