Hi,
Just a bit more info on newly released Raspbian Stretch and Stretch-Lite.
Raspberry Pi gets a major OS update -- download Raspbian Stretch now!
One thing that caught my eye was that the BT stack apparently supports ALSA again, and if I recall from looking through the Stretch repo a while back there are new versions of ALSA and other audio components.
I'll post updates and screen shots after I complete a new build of moodeOS using Stretch-Lite 🙂
-Tim
It was such a glaring mismatch in the original Jessie distro.
Although it seems that the rev matters less than how you've configured it.
Tim, I've never felt the underlying infrastructure has dectracted from the presentation layer. You've done a wonderful job and I look forward to the next round of inteconnectivity.
We're so lucky with having such a cooperative group of keen diy audio aficionados.
It's encouraging to see efforts made in one area being validated in a completely different area, such as the efforts made with the Debian crew on making things better and more usable for us hackers.
I wish I had paid more attention in programming class 🙂
Hi Tim,
We did some listening to the new 3.8.2. comparing LL Kernel at 2ms to the RT and find the RT set to TS to be the best overall sounding in my system. The RT FIFO and RR still ticks frequently with the USB Belcanto dac.
We also think the LL 2ms sounds slightly better than the old 3.7 using RT.
Thanks for the hard work 🙂
botte
Hi Botte,
The TS scheduler policy is not really meant to be used with the Real-time kernel since it does not use the fixed priority setting that would be assigned to MPD by the RR and FIFO policies.
I'm leaning toward only allowing RR and FIFO for the Real-time kernel.
-Tim
Hi Botte,
The TS scheduler policy is not really meant to be used with the Real-time kernel since it does not use the fixed priority setting that would be assigned to MPD by the RR and FIFO policies.
I'm leaning toward only allowing RR and FIFO for the Real-time kernel.
-Tim
Hi Tim,
Well then maybe that is where the clicking is coming from in the RR and FIFO policies?
I listened for a couple of hours last night with the TS policy and only heard the click 6 or 7 times.
In 3.7 with the RT Kernel I had no clicks on any scheduler policy.
Thanks
Botte
It was such a glaring mismatch in the original Jessie distro.
Although it seems that the rev matters less than how you've configured it.
Tim, I've never felt the underlying infrastructure has dectracted from the presentation layer. You've done a wonderful job and I look forward to the next round of inteconnectivity.
We're so lucky with having such a cooperative group of keen diy audio aficionados.
It's encouraging to see efforts made in one area being validated in a completely different area, such as the efforts made with the Debian crew on making things better and more usable for us hackers.
I wish I had paid more attention in programming class 🙂
Hi Mike,
Stretch is looking pretty good so far. Thankfully it uses the same partition layout as Jessie but it also has the same annoying auto-expand-at-first-boot script that has to be pruned out of the root partition, and ssh is disabled by default and thus has to be re-enabled!
Probably won't see Stretch running under moOde until release 4.0 though (end of year).
-Tim
Hi Tim,
Well then maybe that is where the clicking is coming from in the RR and FIFO policies?
I listened for a couple of hours last night with the TS policy and only heard the click 6 or 7 times.
In 3.7 with the RT Kernel I had no clicks on any scheduler policy.
Thanks
Botte
Hi Botte,
Send me an email. I'd like to dig a bit more into your particular configuration and see what might be going on.
-Tim
Hi Tim,
I have the similar problem.
My system: Pi3, USB DAC SMSL M9, music on NAS Synology DS216J.
1. Standart kernel + TS + Performance: music is played without problem.
2. LL kernel + TS, or 10ms, or 5ms + Performance: without problem.
3. LL kenel + 2ms + Performance: music with click.
4. Real Time with any policy (TS, RR, FIFO) + Performance:
continuously glitches, click and others dirty
I have the similar problem.
My system: Pi3, USB DAC SMSL M9, music on NAS Synology DS216J.
1. Standart kernel + TS + Performance: music is played without problem.
2. LL kernel + TS, or 10ms, or 5ms + Performance: without problem.
3. LL kenel + 2ms + Performance: music with click.
4. Real Time with any policy (TS, RR, FIFO) + Performance:
continuously glitches, click and others dirty
Hi @Valulik,
Good info.
Unlike I2S audio devices which use kernel drivers specifically written by the manufacturer for their devices, USB audio devices all use a single, universal snd-usb-audio kernel driver. The only exception to this that I'm aware of is a custom USB audio kernel driver for certain M2Tech Hiface devices. This driver is in mainline kernel source tree but not part of the config. It has to be enabled and compiled into custom kernel as was done for the Advanced kernels in moOde 3.8 series.
There continue to be audio glitch issues that occur with snd-usb-audio driver and certain high speed USB (UAC2) chipsets. Sometimes these can be eliminated by using the USB (UAC2) fix in System config but no guarantee.
Its also possible that certain Advanced kernel settings + USB interface + certain USB audio devices yields glitches. These scenarios are where I'd like to understand more about specific configs.
Email me output from System info screen when you get a moment and I'll take a look.
-Tim
Good info.
Unlike I2S audio devices which use kernel drivers specifically written by the manufacturer for their devices, USB audio devices all use a single, universal snd-usb-audio kernel driver. The only exception to this that I'm aware of is a custom USB audio kernel driver for certain M2Tech Hiface devices. This driver is in mainline kernel source tree but not part of the config. It has to be enabled and compiled into custom kernel as was done for the Advanced kernels in moOde 3.8 series.
There continue to be audio glitch issues that occur with snd-usb-audio driver and certain high speed USB (UAC2) chipsets. Sometimes these can be eliminated by using the USB (UAC2) fix in System config but no guarantee.
Its also possible that certain Advanced kernel settings + USB interface + certain USB audio devices yields glitches. These scenarios are where I'd like to understand more about specific configs.
Email me output from System info screen when you get a moment and I'll take a look.
-Tim
Hi,
For those having issues with AIFF tags not being read by MPD, please follow issue #98 in MPD Git repo.
AIFF tags not being read * Issue #98 * MusicPlayerDaemon/MPD * GitHub
-Tim
For those having issues with AIFF tags not being read by MPD, please follow issue #98 in MPD Git repo.
AIFF tags not being read * Issue #98 * MusicPlayerDaemon/MPD * GitHub
-Tim
Tim - I just looked at your link, it looks like MaxKellermann closed Issue #98 because it was a duplicate of Issue #87 and he also closed Issue #87 stating something to the effect that it kinda works...?
Dunno for sure if that is right, I'm not one of those that maintains the software in question.
What I can tell you for sure is that AIFF files are still not displayed in the Moode Audio library, at least with 3.7 and 3.8.X
Dunno for sure if that is right, I'm not one of those that maintains the software in question.
What I can tell you for sure is that AIFF files are still not displayed in the Moode Audio library, at least with 3.7 and 3.8.X
Hi,
For those having issues with AIFF tags not being read by MPD, please follow issue #98 in MPD Git repo.
AIFF tags not being read * Issue #98 * MusicPlayerDaemon/MPD * GitHub
-Tim
Does the fix in MPD => 0.20.6 unable to parse .aif metadata * Issue #87 * MusicPlayerDaemon/MPD * GitHub work in MoOde?
Cheers,
Phil
Hi Phil,
It wasn't stated as a universal fix and it didn't work on any of my test files, including the ones referenced in the issue I posted, but it looks like MaxK feels the issue is resolved and thus closed.
I'm not planning to add the params to mpd.conf unless I get confirmations that they actually resolve aiff tag issue.
The issue I posted can always be reopened and if someone wants to do this and add new information that Max can work with then by all means go for it!
-Tim
-Tim
It wasn't stated as a universal fix and it didn't work on any of my test files, including the ones referenced in the issue I posted, but it looks like MaxK feels the issue is resolved and thus closed.
I'm not planning to add the params to mpd.conf unless I get confirmations that they actually resolve aiff tag issue.
The issue I posted can always be reopened and if someone wants to do this and add new information that Max can work with then by all means go for it!
-Tim
-Tim
To me the answer lies in why did it work before and it doesn't work now? If nothing changed, then it should still work. But, something must have changed because it no longer works. Irrespective of what MaxKellermann did or thinks...
Tim, I can appreciate that many users of Moode probably just stream radio or use some other file type. However, I personally know of at least a dozen people off the top of my head who use Moode and are disappointed that they have to browse across their attached drive to get to the AIFF music file they want - they were all former users of iTunes or Audirvana and, in those programs AIFF files sound best.
Maybe the solution is change all the AIFF files to some other fill type and avoid the problem all together???
Anybody have any experience with what file type sounds best in Moode?
Tim, I can appreciate that many users of Moode probably just stream radio or use some other file type. However, I personally know of at least a dozen people off the top of my head who use Moode and are disappointed that they have to browse across their attached drive to get to the AIFF music file they want - they were all former users of iTunes or Audirvana and, in those programs AIFF files sound best.
Maybe the solution is change all the AIFF files to some other fill type and avoid the problem all together???
Anybody have any experience with what file type sounds best in Moode?
To me the answer lies in why did it work before and it doesn't work now? If nothing changed, then it should still work. But, something must have changed because it no longer works. Irrespective of what MaxKellermann did or thinks...
Tim, I can appreciate that many users of Moode probably just stream radio or use some other file type. However, I personally know of at least a dozen people off the top of my head who use Moode and are disappointed that they have to browse across their attached drive to get to the AIFF music file they want - they were all former users of iTunes or Audirvana and, in those programs AIFF files sound best.
Maybe the solution is change all the AIFF files to some other fill type and avoid the problem all together???
Anybody have any experience with what file type sounds best in Moode?
Someone with a lot of time and patience would have to do several bisections of the mpd git tree (plus compile and test) to find which commit broke things.
Cheers,
Phil
To me the answer lies in why did it work before and it doesn't work now? If nothing changed, then it should still work. But, something must have changed because it no longer works. Irrespective of what MaxKellermann did or thinks...
Tim, I can appreciate that many users of Moode probably just stream radio or use some other file type. However, I personally know of at least a dozen people off the top of my head who use Moode and are disappointed that they have to browse across their attached drive to get to the AIFF music file they want - they were all former users of iTunes or Audirvana and, in those programs AIFF files sound best.
Maybe the solution is change all the AIFF files to some other fill type and avoid the problem all together???
Anybody have any experience with what file type sounds best in Moode?
Hi,
I don't disagree. It seems like a regression bug which I suggested as such in the issue but MaxK doesn't feel the need to investigate that avenue based on the partial fix in the other referenced issue.
The issue can always be updated or reopened if there is new information or a better way to repro the issue.
-Tim
Someone with a lot of time and patience would have to do several bisections of the mpd git tree (plus compile and test) to find which commit broke things.
Cheers,
Phil
Hi Phil,
I've updated the issue with an A/B repro.
AIFF tags not being read * Issue #98 * MusicPlayerDaemon/MPD * GitHub
-Tim
Is the volume of moode compatible with Allo Ralay Attenuator?
Thanks
Hi,
Can you provide a link that explains the Allo Relay Attenuator?
-Tim
Hi Phil,
I've updated the issue with an A/B repro.
AIFF tags not being read * Issue #98 * MusicPlayerDaemon/MPD * GitHub
-Tim
Does this patch fix it?
tag/Aiff: the FORM chunk size is big-endian * MusicPlayerDaemon/MPD@dd9fd3d * GitHub
Cheers,
Phil
Hi Phil,
Maybe. I guess we will find out when 0.20.10 is released or Max responds to the update I posted to issue #98 and confirms that the patch addresses the issue.
-Tim
Maybe. I guess we will find out when 0.20.10 is released or Max responds to the update I posted to issue #98 and confirms that the patch addresses the issue.
-Tim
- Home
- Source & Line
- PC Based
- Moode Audio Player for Raspberry Pi