It's certainly possible. What do you get when you play back? Nothing at all? Picture but no sound?
Please post the metadata.xml from your DCP-o-matic project as that will help with debugging.
Creating a 5.1 DCP
-
Eliza
- Posts: 3
- Joined: Tue Dec 11, 2018 2:44 pm
Re: Creating a 5.1 DCP
We get the image but the sound is only coming from the screen. Is there any chance I could put you in direct touch with the tech guy working on it this evening? We have a live "test" audience tomorrow for the film and are rather in despair of getting the surround sound up and working, which I also have to proof for a sound mix deadline this week. ~Eliza
-
carl
- Site Admin
- Posts: 2548
- Joined: Thu Nov 14, 2013 2:53 pm
Re: Creating a 5.1 DCP
Sure, email to carl@dcpomatic.com is fine. Though it's late evening here so I won't be reading email very much for the next 8 hours or so. If you can send the metadata.xml file from your DCP-o-matic project that might give an answer more quickly.
-
Eliza
- Posts: 3
- Joined: Tue Dec 11, 2018 2:44 pm
Re: Creating a 5.1 DCP
I did e-mail you directly & sent the metadata. We aren't even getting stereo. Just sound from one single central channel. Very frustrating. We have a screening at 12:30 EST tomorrow (Wednesday) to a test live audience.
-
carl
- Site Admin
- Posts: 2548
- Joined: Thu Nov 14, 2013 2:53 pm
Re: Creating a 5.1 DCP
OK I have replied via email.
-
Carsten
- Posts: 2804
- Joined: Tue Apr 15, 2014 9:11 pm
- Location: Germany
Re: Creating a 5.1 DCP
As this is a fairly generic topic, and people may drop by easily through Google - this is nothing that needs Carl to fix something, although he may have adjusted Elizas Project file to correctly assign the audio content to the proper channels.
But - making 5.1 DCPs is perfectly supported by DCP-o-matic in more than one way. Just as 7.1 or 7.1+HI/VI-N.
Basically - you either create your source footage/audio in a way that helps DCP-o-matic in assigning the audio channels/streams to the proper DCP channels automatically. Or, you need to manually assign each track/file in the audio matrix (and possibly timeline view) so that the content audio matches the proper DCP tracks for Left, Right, Center, LFE, Left Surround, Right Surround.
Premiere and other editing tools allow to output audio in very different formats, and there is no single implementation of multichannel audio. Especially Quicktime, MKV and MXF file containers can be complex and support a variety of track arrangements. Each one may need different settings in DCP-o-matic to correctly assign incoming content audio tracks/files to outgoing DCP audio channels.
Normally, using a single stream interleaved multichannel audio format is the easiest way - this can be a separate audio file, or contained in a video file like MP4. Note that a single stream audio file can still contain an unlimited number of individual audio tracks, thus a single stream also caters for a surround mix. 'Stream' means that, typically, all audio tracks within that stream are related to each other, whereas multiple streams may contain unrelated audio (as e.g. a stereo music track and a commentary, or multiple language versions of a feature - which quite obviously are not meant to play together). You may carry a 5.1 surround mix contained in 6 channels of a single interleaved stream (the best way), or as six single channel streams (the worst way to do it). Or in 6 separate mono files.
Your audio/editing application manual should tell how to output audio in the desired format.
The screencast Carl linked to earlier in this thread shows both interleaved and single channel/file examples: https://dcpomatic.com/videos/audio_mapping.mp4
For an interleaved audio file, if possible, the channel order should follow the digital cinema scheme:
1-Left, 2-Right, 3-Center, 4-LFE, 5-LeftSurround, 6-RightSurround. This will allow the easiest setup in the audio matrix and will also be routed 1:1 correctly and automatically on content import, so, no need for manual override.
Using single files for every audio channel, like left.wav, right.wav is NOT the best way to do it, as it makes adjustments to the composition cumbersome, and you need to assign every track individually in the audio matrix to send them to the proper DCP channels. Per default, single channel files are always assigned to the DCP center channel upon import (unless a special file naming scheme is used that helps DCP-o-matic to assign files automatically). And that is probably the issue Eliza had, they imported a bunch of audio files and left them all assigned to the default center channel, probably overseeing the meaning of the audio matrix routing function.
If you want to help DCP-o-matic with the channel assignment of single channel track files, follow this scheme naming the files:
<content-name>-L.wav
<content-name>-R.wav
<content-name>-C.wav
<content-name>-LFE.wav
<content-name>-LS.wav
<content-name>-RS.wav
The hyphen needs to be present in order for DCP-o-matic to recognize the channel label.
As an aid, use the track designators under the 'DCP' row in the audio matrix unter content->audio (case insensitive):
Again, if one-file-per-channel is everything you have, it is perfectly okay to use them in DCP-o-matic that way, you just need to take special care when you need to apply identical settings to every channel. Settings (e.g. gain or delay) assigned to an interleaved audio file will automatically have that setting applied to all channels (which usually is the correct intention). Settings applied to a single file/track will need to be set identically to all the other tracks manually. That can lead to errors, if e.g. you forget to set a delay or gain to e.g. LeftSurround only. Imagine when working with multiple reels and e.g. 7.1 audio could mean that you would need to make identical adjustments to nearly fifty audio content items manually!
P.S. - Carl, I think I wrote that elsewhere on the forum, and I will file a feature suggestion - so far the automatic channel/matrix assignment for single file import only works up until LS and RS e.g. 'reel1-LFE' - it doesn't work up from there, e.g. for HI, VI, Lc, Rc, BSL, BSL. It's probably easy to extend that scheme at least to these. 13/14/15/16 - don't know wether it's necessary. I think it's good to follow exactly the DCP track labels as seen in the audio matrix. Case insensitiv, of course. And maybe you can also add reel number detection, like <reelname>1-LFE,<reelname>2-LFE, etc.
And - it doesn't matter much, but maybe you could also follow the track order for parallel files in the timeline view, so people are not mislead by the vertical visual arrangement. Especially since up/down will result in sequential arrangement, not in a clean visual order in timeline view. Some people may misread vertical order for track order, and then try to adjust it visually, which results in a complete mess at first.
- Carsten
But - making 5.1 DCPs is perfectly supported by DCP-o-matic in more than one way. Just as 7.1 or 7.1+HI/VI-N.
Basically - you either create your source footage/audio in a way that helps DCP-o-matic in assigning the audio channels/streams to the proper DCP channels automatically. Or, you need to manually assign each track/file in the audio matrix (and possibly timeline view) so that the content audio matches the proper DCP tracks for Left, Right, Center, LFE, Left Surround, Right Surround.
Premiere and other editing tools allow to output audio in very different formats, and there is no single implementation of multichannel audio. Especially Quicktime, MKV and MXF file containers can be complex and support a variety of track arrangements. Each one may need different settings in DCP-o-matic to correctly assign incoming content audio tracks/files to outgoing DCP audio channels.
Normally, using a single stream interleaved multichannel audio format is the easiest way - this can be a separate audio file, or contained in a video file like MP4. Note that a single stream audio file can still contain an unlimited number of individual audio tracks, thus a single stream also caters for a surround mix. 'Stream' means that, typically, all audio tracks within that stream are related to each other, whereas multiple streams may contain unrelated audio (as e.g. a stereo music track and a commentary, or multiple language versions of a feature - which quite obviously are not meant to play together). You may carry a 5.1 surround mix contained in 6 channels of a single interleaved stream (the best way), or as six single channel streams (the worst way to do it). Or in 6 separate mono files.
Your audio/editing application manual should tell how to output audio in the desired format.
The screencast Carl linked to earlier in this thread shows both interleaved and single channel/file examples: https://dcpomatic.com/videos/audio_mapping.mp4
For an interleaved audio file, if possible, the channel order should follow the digital cinema scheme:
1-Left, 2-Right, 3-Center, 4-LFE, 5-LeftSurround, 6-RightSurround. This will allow the easiest setup in the audio matrix and will also be routed 1:1 correctly and automatically on content import, so, no need for manual override.
Using single files for every audio channel, like left.wav, right.wav is NOT the best way to do it, as it makes adjustments to the composition cumbersome, and you need to assign every track individually in the audio matrix to send them to the proper DCP channels. Per default, single channel files are always assigned to the DCP center channel upon import (unless a special file naming scheme is used that helps DCP-o-matic to assign files automatically). And that is probably the issue Eliza had, they imported a bunch of audio files and left them all assigned to the default center channel, probably overseeing the meaning of the audio matrix routing function.
If you want to help DCP-o-matic with the channel assignment of single channel track files, follow this scheme naming the files:
<content-name>-L.wav
<content-name>-R.wav
<content-name>-C.wav
<content-name>-LFE.wav
<content-name>-LS.wav
<content-name>-RS.wav
The hyphen needs to be present in order for DCP-o-matic to recognize the channel label.
As an aid, use the track designators under the 'DCP' row in the audio matrix unter content->audio (case insensitive):
Again, if one-file-per-channel is everything you have, it is perfectly okay to use them in DCP-o-matic that way, you just need to take special care when you need to apply identical settings to every channel. Settings (e.g. gain or delay) assigned to an interleaved audio file will automatically have that setting applied to all channels (which usually is the correct intention). Settings applied to a single file/track will need to be set identically to all the other tracks manually. That can lead to errors, if e.g. you forget to set a delay or gain to e.g. LeftSurround only. Imagine when working with multiple reels and e.g. 7.1 audio could mean that you would need to make identical adjustments to nearly fifty audio content items manually!
P.S. - Carl, I think I wrote that elsewhere on the forum, and I will file a feature suggestion - so far the automatic channel/matrix assignment for single file import only works up until LS and RS e.g. 'reel1-LFE' - it doesn't work up from there, e.g. for HI, VI, Lc, Rc, BSL, BSL. It's probably easy to extend that scheme at least to these. 13/14/15/16 - don't know wether it's necessary. I think it's good to follow exactly the DCP track labels as seen in the audio matrix. Case insensitiv, of course. And maybe you can also add reel number detection, like <reelname>1-LFE,<reelname>2-LFE, etc.
And - it doesn't matter much, but maybe you could also follow the track order for parallel files in the timeline view, so people are not mislead by the vertical visual arrangement. Especially since up/down will result in sequential arrangement, not in a clean visual order in timeline view. Some people may misread vertical order for track order, and then try to adjust it visually, which results in a complete mess at first.
- Carsten
You do not have the required permissions to view the files attached to this post.