[Draconix]

@Draconix do you use that file alongside the one for video cache? I mean can both files be used at the same time or only one, since the video cache buffer one is also called advancedsettings.

As stated by others here, you can only have one advancedsettings.xml file. If you want to try out my settings, you can add the code to your existing file. SoulReaver posted an excellent example of how to do this!

Just a little update: I've been using the latest nightly builds for a while now and everything is running smoothly. I still haven't encountered any audio dropouts since that one time, which leads me to believe that it was either a series of bad files or the issue was somehow solved in the more recent builds.

[Nuklear92]

@Draconix do you use that file alongside the one for video cache? I mean can both files be used at the same time or only one, since the video cache buffer one is also called advancedsettings.

As stated by others here, you can only have one advancedsettings.xml file. If you want to try out my settings, you can add the code to your existing file. SoulReaver posted an excellent example of how to do this!

Just a little update: I've been using the latest nightly builds for a while now and everything is running smoothly. I still haven't encountered any audio dropouts since that one time, which leads me to believe that it was either a series of bad files or the issue was somehow solved in the more recent builds.

Yeah, I saw his comment and actually did that and it worked. However, I think what messed up was that you have a video delay and when I tried it out, the audio was way behind than the video.

So, what should be the most default way. Just setting the max Passthrough only or you always need to set a video delay, etc?

Just to point out, I don’t have any audio dropouts anymore either running the Alpha 3 nightly build. However, I do seem to get micro stutters sometimes if I playback some big 4K DV/HDR10 files 40+ GB and above. So, I just wanted to tryout setting up that static maxpassthrough setting within my advancedsettings file and see if that prevents any micro stutters compared to using the a/v self learning algorithm solution introduced by Jogal when someone isn’t using any maxpassthrough manual settings.

Is there any downside by using the option compared to letting the self learning algorithm do the work instead?

Sorry, for all the questions in between. I just wanted to write everything in one comment instead of multiple. 😅

[MrMagic]

@Nuklear92 I only have this in my advancedsettings.xml:

Code:

  <audio>
    <maxpassthroughoffsyncduration>96</maxpassthroughoffsyncduration>
  </audio>

I have removed the video delay, because all audio is perfectly in sync without it.

[Nuklear92]

@Nuklear92 I only have this in my advancedsettings.xml:
 

Code:

  <audio>
    <maxpassthroughoffsyncduration>96</maxpassthroughoffsyncduration>
  </audio>

I have removed the video delay, because all audio is perfectly in sync without it.

Okay, so had you tested with and without the setting and noticed a difference in which works better to combat micro stutters. I'm currently using the default implementation of the self learning a/v sync algorithm as I said previously. However, I do encounter some micro stutters once in a while watching some large 4k remux files. Will this setting help make micro stutters happen even less?

[MrMagic]

Will this setting help make micro stutters happen even less?

In my experience, yes. Because the self-learning algorithm only uses the first 5 minutes (I believe). So if a higher value is needed later on in the movie, it will not be set.

96 is the maximum and if I remember correctly, @fritsch has discussed this with someone from Nvidia and 96 is also the recommended value for DD+ 5.1 Atmos audio tracks.

[Nuklear92]

So, 96 will work good for pretty much all codecs, including TrueHD, DTS-HD, etc. Aside from DD+/ATMOS?

I’ll try that option with some files and see how it behaves. 👍 what about audio dropouts, none?

[Hitcher]

Because the self-learning algorithm only uses the first 5 minutes

I think it's the first 30 seconds.

[b1ggles]

I understand, but my point is does adding multiple advancedsettings that does different tasks, say I already have the advancedsettings for video cache. And I decide to add the one for max video Passthrough. Does the file need to have the name “advancedsettings” to work or even if you add “advancedsettings(2).xml” does that still make the file work inside kodi.

In other words does it matter more what’s inside the file for it to activate and work in Kodi or does the “advancedsettings” name is also important and if it has any additional wording like “(2)” then it won’t activate. 🤔

You obviously didn't understand You can only have ONE file called advancesettings.xml that Kodi reads when it launchs and overrides the standard settings with whatever you've added, you can add as many overrides as you want. If you look at @faust1103 's example you'll see they have four different groups of settings... <cache>, <audio>, <video> and their log level.

[MrMagic]

I think it's the first 30 seconds.

You're right. It was too long ago and I wasn't sure about the exact duration

[faust1103]

Do you think his make a difference?
Thanks.

Yes. You can just add his settings to your own advancedsettings.xml so it looks like this in the end and test it again.

xml:
 <advancedsettings>

<cache>

<buffermode>1</buffermode>

<readfactor>20</readfactor>

<memorysize>278921216</memorysize>

</cache>

<audio><maxpassthroughoffsyncduration>92</maxpassthroughoffsyncduration>

  </audio>

  <video>

    <latency>

      <delay>-40</delay>

    </latency>

  </video>

  <loglevel>-1</loglevel>  

</advancedsettings>

Hi,

Thanks for the advice. I will be testing this weekend with these advance settings and report back:

<advancedsettings>

<cache>

<buffermode>1</buffermode>

<readfactor>20</readfactor>

<memorysize>278921216</memorysize>

</cache>

<audio>

<maxpassthroughoffsyncduration>96</maxpassthroughoffsyncduration>

</audio>

</advancedsettings>

[Howardhuang]

Dears,
I have installed the latest version of Kodi on my android TV.
But I can't find the option to turn on the Dolby TrueHD passthrough, there even no this option.
Does anyone have the same issue? Did you already solve it? or know the reason of it.
Kodi version: kodi-20230914-55ef84ad-master-armeabi-v7a  Kodi v21 (Omega) Pre-release Builds
I can only see below passthrough options:
Passthrough device: AudioTrack(RAW), Android IEC packer
Dolby Digital(AC3)
Dolby Digital Plus
DTS
DTS-HD

[Hitcher]

Then your device doesn't support it.

A debug log will provide more info.

The instructions are here... debug log (wiki)

If you are using the Basic Method, then ensure the following is applied...
1.Enable debugging in Settings>System Settings>Logging,
2.Restart Kodi
3.Replicate the problem.
4.Upload the log to Kodi Paste Site manually or use the Kodi Logfile Uploader. (wiki) With either method post the link to the log back here.

If you are using the Advanced Method ensure you have correctly created and applied the advancedsettings.xml file (wiki)

In both instances, you should see the word DEBUG throughout the log.

Note: Full logs only. No partial or redacted logs
Do NOT post your logs directly into the forum. Use the Kodi Paste Site. Post the link to your pasted log in the forum

[Nuklear92]

@jogal @fritsch wouldn’t be better for the self calibration feature to run for 1 whole minute, instead of 30 seconds only?

I mean, I say because I’ve noticed that a lot of content I play being with lossless audio or lossy Dolby d plus, etc in just 30 seconds may jump from 10-45ms and rarely end up ending the the sweet spot the Shield is supposed to be which as you guys said it’s between 88-100ms to have less chances of any micro stutters. So, maybe I’m guessing letting the feature run for a whole minute could have more chances of getting closer to those recommended numbers instead of just 30 seconds which I think sometimes is too small of a window time to dictate the best “ms” Kodi think it should be set.

Maybe I’m totally wrong, and there are other variables on why most of the content setttle between 10-45ms or 10-65ms instead of always staying in the higher numbers between 88-100ms. I just thought a whole minute of self calibration could be a better bet for Kodi to set to a higher number instead of staying on the lower ones as mentioned above.

[jogal]

<readfactor>20</readfactor>

These high values do not make any sense for Shield (or in any system). This saturates the network and can cause CPU spikes that in Shield produce stutters. The cache must (slowly) load data in the background to compensate for sudden variations in the connection and/or the stream, which itself has a variable bitrate. In fact I think the default value 4.0 is too high.

I would recommend using a memory size of 128 - 256 MB (256 x 1024 x 1024) and a readfactor of 1.5. Reading at 1.5x will cause data to accumulate in the cache and in this way when there is a peak in demand the cache will be emptied a little so that it is slowly recovered afterwards.

This is valid for both the local network and Internet (today there is not much difference).

The cache should be like a shock absorber for a suspension.

Read 80 Mbit/s stream at 20x is unnecessary and may cause additional disturbances and no benefit.

[jogal]

Maybe I’m totally wrong, and there are other variables on why most of the content setttle between 10-45ms or 10-65ms instead of always staying in the higher numbers between 88-100ms. I just thought a whole minute of self calibration could be a better bet for Kodi to set to a higher number instead of staying on the lower ones as mentioned above.

The value obtained is the one measured in real conditions for a specific system and stream (other users may obtain other values even for same stream).

If the calibration says A but you want to use B then you should not use auto-calibration, you should use advancedsettings (which is also fine).