AVR chops off beginning of ATMOS Tracks (SEE MY WORKAROUND IN POST #41)

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I think that there will always be a small time, either for hdmi handshake or for AVR change format, or both, for which the sound will be muted at the beginning of a file.

The Solution, valid for all cases, would be to make files that have enough silence time at the beginning to deal with this ‘sync’ process.

The files from CDs or Blu-rays or whatever mastering should be adapted in this way by ‘adding’ some amount of silence at the beginning, around 5 to 10 seconds.

One very good example of this can be found in the track samples (snippets) that can be downloaded from the album “Silent Work – Sonic Leaks” from the GrobiTV site:

Pure Audio SONIC LEAKS von Silent Work in Dolby Atmos und Auro-3D

Those MKV files have 10 seconds silence at the beginning that is explicitly shown by the video animation displaying a countdown “Handshaking: xxs”.


Proposal to IAA

@sjcorne , I want to suggest that a site like IAA could implement this solution for its download files.

I have done a successful test with the track “Dear life.mkv” (Atmos TrueHD) from Bruce Soord’s Luminiscence album, downloaded from IAA. In my system I always lose the first wonderful three notes of the guitar, as they start just at the beginning of the file. I also lose those same notes when playing stereo CD ripped files and the DVD DTS, as well. Not in all cases I can restart the song by pressing ‘rewind’ button, which is something annoying for exquisite people like us.


How I did it:

The idea is to ‘append’ via MKVToolNix, a silence file first before the original MKV. As the characteristics of both files have to be similar, I had to elaborate ‘a little’ the silence track, but then, it can be used later for all tracks.

Make the silence track:

- Create a mono WAV with 10 seconds silence.
- Using Audacity, copy 12 times that mono silence file to get a 12 channel. Then Export to a 12 Channel WAV file.
- Using MMH, with that WAV source 7.1.4 encode an Atmos TrueHD file MKV

Make the compatible silence file:

I had errors when trying to append the original MKV to the silence file, that was a MKV without video track. So I prepare the silence file with ‘same’ tracks as the original MKV:

- MKVToolNix, merge original MKV and silence MKV, and select only the video track and the audio track from silence.
- MKVToolnix, Output tab, select “Split mode: After output duration / Duration: 10s”
(I tried first with 5s but it generated 10s files, that I guess is the minimum??)
- MKVToolNix, Start Multiplexing and then get only the first of the ‘splited’ files generated.

That will be the silence Atmos compatible file to append for all tracks. Of course it has the video part of the first track, but with a more generic video track, this file would be universal for all subsequents append for all kind of MKV (video+audio) Atmos TrueHD.

Make the final Customer file:

- MKVToolNix, append the compatible silence file MKV with the original MKV

Make Full Albums:

This can be done for each individual song, when needed, if there is important sound during about the first 5 to 10 seconds or so.

But when a complete gapeless is needed for the album, the solution is to append all tracks with MKVToolNix, and aply the silence Atmos track process only for the first song, as @ar surround explain in the post 41.


CONCLUSION:

This has corrected the problem for me for my players and the PC (with MPC-HC) and I guess with any players.

I can now fully enjoy the Atmos TrueHD from IAA of Bruce Soord's Luminiscence listening properly the good first guitar notes of the first song. :)
This isn't a perfect solution, but I've found that if I rip the MKV to separate MKA files, and then play it in Windows Media Player (I guess it's just called "Media Player" now), if I press play it does the handshake, but if I then skip back to the beginning of the song, it does not do the handshake again and will play immediately from the beginning of the song. I haven't been able to get this functionality in any other media player. I agree that a bit of silence (maybe more like 5 seconds than 10) would be best though.
 
This isn't a perfect solution, but I've found that if I rip the MKV to separate MKA files, and then play it in Windows Media Player (I guess it's just called "Media Player" now), if I press play it does the handshake, but if I then skip back to the beginning of the song, it does not do the handshake again and will play immediately from the beginning of the song. I haven't been able to get this functionality in any other media player. I agree that a bit of silence (maybe more like 5 seconds than 10) would be best though.
In my system I think 5 sec would be enough, using TV display for music listening. If I use the projector, the handshake time increases.
I talk about 10 sec, because with the tests I've done with MKVToolNix, splitting "after output duration", it does not work for 5 seconds and still generate a minimum of 10 seconds file. (I didn't investigate the matter much nor did I try to do it again either). Also the Silent Work - SONIC LEAKS that I mention above, do use 10 seconds.

Maybe it's only my system, but even with a normal stereo CD, ripped to FLAC image/CUE, I lose the initial sound if the first song starts immediately. The video handshake may impact sound that would be muted for a moment. The audio/video handshake can behave very different from one system to another, and can be fired (or not) when changing from one playback file to the next. Gapless implementation is not done for every systems and file types, and sometimes does not work perfectly.

IMO, it should be a mandatory QA for the producers, from the studio master, to generate files with a silence enough at the beginning. Mandatory to have the entire album in a single file. Either a M2TS file on Bluray or a single MKV file with chapters. This way we would eventually have a problem only with the first song. 10 seconds is not too much for the beginning of an album. Just prepare it for the "audition". Put a fade in video for the start. We could lose part of the fade in video, but with silence.
 
In my system I think 5 sec would be enough, using TV display for music listening. If I use the projector, the handshake time increases.
I talk about 10 sec, because with the tests I've done with MKVToolNix, splitting "after output duration", it does not work for 5 seconds and still generate a minimum of 10 seconds file. (I didn't investigate the matter much nor did I try to do it again either). Also the Silent Work - SONIC LEAKS that I mention above, do use 10 seconds.

Maybe it's only my system, but even with a normal stereo CD, ripped to FLAC image/CUE, I lose the initial sound if the first song starts immediately. The video handshake may impact sound that would be muted for a moment. The audio/video handshake can behave very different from one system to another, and can be fired (or not) when changing from one playback file to the next. Gapless implementation is not done for every systems and file types, and sometimes does not work perfectly.

IMO, it should be a mandatory QA for the producers, from the studio master, to generate files with a silence enough at the beginning. Mandatory to have the entire album in a single file. Either a M2TS file on Bluray or a single MKV file with chapters. This way we would eventually have a problem only with the first song. 10 seconds is not too much for the beginning of an album. Just prepare it for the "audition". Put a fade in video for the start. We could lose part of the fade in video, but with silence.
Interesting. For me the problem is actually just Dolby content, and realistically it's really only my current AVR (Onkyo RZ-50) that I'm having the issue with. Ironically, my cheap old Denon did not have this much of an issue with Dolby content. PCM audio plays immediately for me.
 
PCM audio plays immediately for me.
Atmos is all PCM.
But it is probably the combined delay of the HDMI handshake and the AVRs detection of the Atmos metadata before the codec activates the data stream. I get the same with varying degrees of delay with a small amount of material. I never bothered to document the "what does what"
to try and isolate things.
Garys time delay software should solve all but the most severe issues.
 
IMO, it should be a mandatory QA for the producers, from the studio master, to generate files with a silence enough at the beginning. Mandatory to have the entire album in a single file. Either a M2TS file on Bluray or a single MKV file with chapters.

I completely agree. Funny I only find this thread now after complaining in other threads and forums about the recent Gentle Giant Missing Piece disc which cuts off the beginnings of every song for me. Others responded that it only cuts off the beginning of the first track for them, which is seemingly ok by them. I agree with the above - this is simple mastering/authoring care. While they can't always account for every system, they can certainly do better to avoid these problems which obviously many people are having on various systems.
 
I would like to point out that having the album in separate files on the Blu-ray shouldn't be an issue: a lot of film Blu-ray Discs have Atmos and have the film stored in separate files on the disc and work fine. I think the issue is lack of proper "seamless-branching" authoring or whatever the thing is where each file has overlapping samples with the other files to prevent drop-outs.
 
I would like to point out that having the album in separate files on the Blu-ray shouldn't be an issue: a lot of film Blu-ray Discs have Atmos and have the film stored in separate files on the disc and work fine. I think the issue is lack of proper "seamless-branching" authoring or whatever the thing is where each file has overlapping samples with the other files to prevent drop-outs.
...shouldn't be an issue.... But the reality is that IT IS an issue.

I guess not all Authoring programs may support seamless branching, and that the concept may be better known in the world of complex movie authoring more than music authoring.

I think the problem could lie in the way the master files are delivered to the authoring technician who has to "place" them in the menus. Depending on whether the delivery is a single file for an album or different files for each song, the technician who does the Bluray authoring must worry about merging the files (or not) without knowing the impact it may have. The authoring guy would not care about inserting any additional silence at the beginning. That sould be done in the previous step when generating the masters?

I would be insterested in knowing how all this work for big companies with different technicians in the production chain. I would like to hear from anyone who knows about this.
 
Atmos is all PCM.
But it is probably the combined delay of the HDMI handshake and the AVRs detection of the Atmos metadata before the codec activates the data stream. I get the same with varying degrees of delay with a small amount of material. I never bothered to document the "what does what"
to try and isolate things.
Garys time delay software should solve all but the most severe issues.
Not just Atmos but anything Dolby has the issue for me.
 
Dolby only, not DTS?
Then that appears to be some issue with your AVR and it's being slow to recognize the codec.
What is the source of the file and how does it connect to your AVR?
I also think it's my AVR but I'm not the only one with the issue and there isn't an obvious resolution. I haven't noticed it with DTS but I rarely use DTS so I'd have to double-check. This is the case no matter what the source is, so long as it is using Dolby encoding - MKV/MKA files running on a PC, MKV files streaming on Plex, or streaming services on the SHIELD e.g. Netflix, Disney+, Amazon etc.
 
@LuvMyQuad where in Kodi can I find this setting?
Be sure the global settings level is set to expert.

Go to: system/audio/output configuration. Set it to "fixed". (I think it defaults to "best match"). Then set "number of channels" and °limit sampling rate" as needed.

For example, if you set it to 5.1 channels with a 48 kHz sampling rate, all playback, even something like a stereo 44.1 CD rip will be played back with the additional channels padded out (but silent), and upsampled to 48 kHz. Likewise, 7.1 96kHz tracks will be down mixed to 5.1 and resampled to 48kHz
 
I also think it's my AVR but I'm not the only one with the issue and there isn't an obvious resolution. I haven't noticed it with DTS but I rarely use DTS so I'd have to double-check. This is the case no matter what the source is, so long as it is using Dolby encoding - MKV/MKA files running on a PC, MKV files streaming on Plex, or streaming services on the SHIELD e.g. Netflix, Disney+, Amazon etc.
Well handshake timing issues are something that can effect us all.
I'm rarely effected by it and your widespread problem doesn't hint at any easy solution.
Sorry
 
Back
Top