[larsete]

Hi guys!

The first thing I have to say is that Kodi 17 works great even with 4k@60p videos that I was struggling to play before, so congratulations! That said, I will like to comment something about removing the possibility of having "Sync Playback to Display" and "Audio Passthrough" enabled at the same time.

I have one Kodi 15.2 in Windows and one Kodi 16.1 in Linux, and testing now with Kodi 17 in Linux too. For both Kodi 15.2 and 16.1 I have enabled both options at the same time ("Sync Playback to Display" and "Audio Passthrough"), because if I leave "Sync Playback to Display" off, I get the repeated frame about 40", and is something I can't stand

I know the best solution would be to turn off "Audio Passthrough", but I would prefer the AV doing that work, and having a TV and PJ able to use 24Hz, 50Hz and 60Hz refresh rates, I don't even notice any audio delay when "Sync Playback to Display" is on. I have done extensive tests, and even after 120 minutes of continue playing a 23.976fps movie, I can't notice any audio delay at the end. In fact, the "CodecInfo" is always reporting something like:

For 24Hz content in Linux:

Code:

refresh: 23.976 missed: 0 speed: 0.000%

Or the same screen + same movie in Windows:

Code:

refresh: 24 missed: 0 speed: 0.100%

So even if it does not seem to be doing anything/much, it removes this duplicate frame every 40" and everything plays smooth and without noticeable delays, at least for me

So the question is... is there any way or hack to be able to have both options enabled at the same time in Kodi 17?

Thank you very much in advance.
best.

[FernetMenta]

You have given no reason why it should be preferable to use passthrough audio.

[larsete]

According to my calculations, if the video speed is rose by 0.1 %, at the end of a 2h movie audio would be 7,5 secs off! but I swear is still in Sync and passing trough the audio to the AV, maybe the AV has something to do, but had 2 different ones during this time (Onkyo and Denon) with the same good results.

[larsete]

You have given no reason why it should be preferable to use passthrough audio.

Hi! thank you for your reply.

I'm not saying is better, but just trying to avoid any resampling

[FernetMenta]

According to my calculations, if the video speed is rose by 0.1 %, at the end of a 2h movie audio would be 7,5 secs off! but I swear is still in Sync and passing trough the audio to the AV, maybe the AV has something to do, but had 2 different ones during this time (Onkyo and Denon) with the same good results.

Not sure what you are trying to say. Why should audio not be in sync. The old method dropped or duplicated audio packets which was nonsense. If you experience a video glitch every 40 s without "sync playback to dispaly" means that audio speed has to be adjusted by 41ms every 40 secs too.

I will never understand why people prefer currupted audio (this is what you get when duplicating or dropping packets) just to see a useless codec indicator on the AVR. In most cases (like yours) quality is better if you disable passthrough

[larsete]

Thank you for your reply again FernetMenta.

Ok, that is what I thought (The dropped packets), but wasn't completely sure because some things I have read before about this were a bit confusing.

Before I could use WASAPI on Windows, any re-sampling had a noticeable effect on audio quality, so maybe that's why I always try to avoid this (Old habits die hard...), and since a don't notice any audio glitch with the old method I prefer this. Maybe also because I'm not sure how this works in Linux, so I'm afraid to have the same problem I had before on Windows without WASAPI. Sorry about my ignorance on this matter, but couldn't find before a good resource explaining everything without being confusing?

In any case, I guess the old method won't be available anymore? not even trough advancedconfig.xml right?

[FernetMenta]

Kodi does not resample if you don't want it to do so. Kodi can output bit perfect audio. Bottom line here is that things have improved for you and many others. You claim audio quality is important for you. As a result currupting audio by dropping or duplicating packts is not what you want.

So far I don't see any reason why I should wast time on makeing things worse again.

[larsete]

Ok now is more clear, will stick to this then. Thank you so much for the explanations.

I will bother you with a couple of additional questions though, just to be sure I get what I expect this way:

- In Windows I'm able to see it is doing resampling when using DirectSound vs WASAPI, so the latter is the one to go (Plus "Best Match" option in the Audio menu). In Linux, according to my tests, it is doing resampling when using ALSA, but seems to be delivering bit perfect audio when selecting the HDMI interface + selecting "Best Match". Is my assumption correct with this config? is something else I need to take into account to avoid Kodi or the SO resampling the audio?

- I have an ATMOS/DTS:X ready receiver, but not a correct speaker setup yet for it to work (will have in the future), so sorry if my question is stupid because I couldn't perform any test with this... will Kodi also deliver the LPCM audio with the metadata needed for ATMOS/DTS:X to the AV receiver this way?

Thank you again!!

[centralseven]

Is it possible that there are sound "Improvements" done by the AVR's when they recognise a DD or DTS stream?
For example. With my Denon AVR the same movie sounds better when I play it native DD/DTS.
With multi channel input there is less bass and spatiality.
I did no manual sound ajustments, only audyssey measuring by the AVR.

Maybe I'm wrong, but aren't there audio format specific settings in the AVR? For center an souround things.
Does that still work with multi channel input?

[larsete]

Is it possible that there are sound "Improvements" done by the AVR's when they recognise a DD or DTS stream?
For example. With my Denon AVR the same movie sounds better when I play it native DD/DTS.
With multi channel input there is less bass and spatiality.
I did no manual sound ajustments, only audyssey measuring by the AVR.

Maybe I'm wrong, but aren't there audio format specific settings in the AVR? For center an souround things.
Does that still work with multi channel input?

Hi centralseven! in your tests, which system and audio settings are you using? (Linux/Windows, WASAPI/DirectSound/ALSA, Best Match/Optimized/Fixed), just to know if those are the ones I think are correct to play it in the same way when delivering a multi-channel input.

[FernetMenta]

Ok now is more clear, will stick to this then. Thank you so much for the explanations.

I will bother you with a couple of additional questions though, just to be sure I get what I expect this way:

- In Windows I'm able to see it is doing resampling when using DirectSound vs WASAPI, so the latter is the one to go (Plus "Best Match" option in the Audio menu). In Linux, according to my tests, it is doing resampling when using ALSA, but seems to be delivering bit perfect audio when selecting the HDMI interface + selecting "Best Match". Is my assumption correct with this config? is something else I need to take into account to avoid Kodi or the SO resampling the audio?

- I have an ATMOS/DTS:X ready receiver, but not a correct speaker setup yet for it to work (will have in the future), so sorry if my question is stupid because I couldn't perform any test with this... will Kodi also deliver the LPCM audio with the metadata needed for ATMOS/DTS:X to the AV receiver this way?

Thank you again!!

- DirectSound is not exclusivley openend, means it is shared with other applications. In order to do so it has to do internal manipulations to the sound. That is out of control of Kodi.
- Waspi is opened exclusively
- Kodi does not resample if the sink can be opened with the requested sampling rate.
- "Best Match" is for playback of music. If selected the audio sink gets reconfigured on every song instead of trying to play gapless.
- I don't think ATMOS metadata is decoded by ffmpeg yet. This is a format that benefits from passthrough audio (as long as ffmpeg has not catched up)

[larsete]

Thanks! And I guess using ALSA in Linux (Suspending PulseAudio), would have a similar effect as using WASAPI in Windows right? (playing directly to the hardware without mixing it up first).

In any case, taking into account ATMOS is not yet supported this way, and that I know more people feels more confident/relieved knowing everything is working fine when passing the stream to the AVR, would you (Or the Kodi team in general) consider adding the possibility of having this old syncing/passtrough method added again at some point in the future? (Maybe only for power users). Should I create a new thread requesting this feature or is something that is already settled and will never ever coming back?

I know that even if the quality is exactly the same (or better without dropping packets), now I find myself checking what the AVR is receiving everytime I play a video. Before, not all videos required syncing, but allowing both options at the same time, I could forget about it and just enjoy what for me was a perfect playback; while if now I want to bitstream a movie that does not require syncing I have to manually change the option everytime. Again, I know is better to leave it as you have said, but is something buried deep inside my head I think

Best!

[FernetMenta]

If the occasional video glitches bother you when passthrough is active, then what you want is a better method to sync video. Implementing this is much better than trying to add that old crap back at a different location.

[larsete]

Thank you for your advice FernetMenta. Just created a thread in here.

[centralseven]

Is it possible that there are sound "Improvements" done by the AVR's when they recognise a DD or DTS stream?
For example. With my Denon AVR the same movie sounds better when I play it native DD/DTS.
With multi channel input there is less bass and spatiality.
I did no manual sound ajustments, only audyssey measuring by the AVR.

Maybe I'm wrong, but aren't there audio format specific settings in the AVR? For center an souround things.
Does that still work with multi channel input?

Hi centralseven! in your tests, which system and audio settings are you using? (Linux/Windows, WASAPI/DirectSound/ALSA, Best Match/Optimized/Fixed), just to know if those are the ones I think are correct to play it in the same way when delivering a multi-channel input.

Hi!

Testsettings: Windows 10, WASAPI, 2.0 optimized, pass through, sync to viedeo off.

I sorted out some things last weekend.
Some audyssey features (dynamic eq/loudness) not working with multichannel input.
That is the reason for the big sound differences.

Due to the limitation of the LSPCON chip/firmware of my motherboard I have no other option to use multichannel output at the moment.
My settings now WASAPI, 7.1 best match, no pass through, sync on.
Setting a different fps as the video, to keep pt, is no option.for me.