Merging eight ProRes audio streams into 5.1 for DCP conversion

[mrbliss]

I hope somebody can help me.

I'm trying to make a DCP from a film that has been supplied to me as a ProRes file. I'm using DCP O Matic, which I have successfully used to produce feature length DCP's in the past. However, this time it seems to only include one track in the mix. This is my first experience of ProRes and I'm used to using single audio streams containing the left, right, centre etc tracks, but Pro Res seems have have eight separate mono tracks. Can someone tell me the easiest way to merge or remux these into one separate track, pref 5.1, but stereo or 2.1 should be ok for my purposes. Hopefully this can be one without converting the whole file. I'm using a PC, I don't have anything like Final Cut Pro, but I hope there might be some freeware tools that can help. I'm really out of my depth here!

[poisondeathray]

8 mono to 5.1 ?

You can use avisynth with QTInput() to do the mixing

eg.

orig=QTInput("input.mov", audio=2)

fr  = GetChannel(orig, 1, 2, 3, 4)
blr = GetChannel(orig, 5, 6)
slr = GetChannel(orig, 7, 8)
bs  = MixAudio(blr, slr, 0.5, 0.5)
final   = MergeChannels(fr, bs)

final

Did you want separate audio track 5.1 (e.g. WAV or AIFF), or muxed into MOV alone, or with muxed with video into MOV ? or something else like MXF ?

[El Heggunte]

Please post a MediaInfo report about the "problematic" source.

I can be wrong, but it seems you have an MXF file.

And no, there is no such thing as "ProRes audio" --- precise terminology does matter

[mrbliss]

Thanks both for you responses. Sorry, as i said, I'm way out of my depth here so I don't really know what the correct terminology is, but here's the MediaInfo:

General
Complete name : G:\PKG - KNOCKED_FOR_SIX_FEATURE_ONLY\KnockedForSix-Feature-01320203.mov
Format : MPEG-4
Format profile : QuickTime
Codec ID : qt
File size : 115 GiB
Duration : 1h 32mn
Overall bit rate mode : Variable
Overall bit rate : 178 Mbps
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:29
Writing library : Apple QuickTime 7.7.1
Media/UUID : 57DC9865-938A-470C-ADB3-91EB10291CC9
com.apple.quicktime.player.movie.audio.m : (Binary)

Video
ID : 1
Format : ProRes
Format version : Version 0
Format profile : High
Codec ID : apch
Duration : 1h 32mn
Bit rate mode : Variable
Bit rate : 169 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate mode : Constant
Frame rate : 23.976 fps
Color space : YUV
Chroma subsampling : 4:2:2
Scan type : Progressive
Bits/(Pixel*Frame) : 3.396
Stream size : 109 GiB (95%)
Writing library : Apple
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:29
Color primaries : BT.709
Transfer characteristics : BT.709
Matrix coefficients : BT.709

Audio #1
ID : 2
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : in24
Duration : 1h 32mn
Bit rate mode : Constant
Bit rate : 1 152 Kbps
Channel(s) : 1 channel
Channel positions : Front: L
Sampling rate : 48.0 KHz
Bit depth : 24 bits
Stream size : 759 MiB (1%)
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:10

Audio #2
ID : 3
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : in24
Duration : 1h 32mn
Bit rate mode : Constant
Bit rate : 1 152 Kbps
Channel(s) : 1 channel
Channel positions : Front: R
Sampling rate : 48.0 KHz
Bit depth : 24 bits
Stream size : 759 MiB (1%)
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:10

Audio #3
ID : 4
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : in24
Duration : 1h 32mn
Bit rate mode : Constant
Bit rate : 1 152 Kbps
Channel(s) : 1 channel
Channel positions : Front: C
Sampling rate : 48.0 KHz
Bit depth : 24 bits
Stream size : 759 MiB (1%)
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:10

Audio #4
ID : 5
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : in24
Duration : 1h 32mn
Bit rate mode : Constant
Bit rate : 1 152 Kbps
Channel(s) : 1 channel
Channel positions : LFE
Sampling rate : 48.0 KHz
Bit depth : 24 bits
Stream size : 759 MiB (1%)
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:10

Audio #5
ID : 6
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : in24
Duration : 1h 32mn
Bit rate mode : Constant
Bit rate : 1 152 Kbps
Channel(s) : 1 channel
Channel positions : Back: L
Sampling rate : 48.0 KHz
Bit depth : 24 bits
Stream size : 759 MiB (1%)
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:10

Audio #6
ID : 7
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : in24
Duration : 1h 32mn
Bit rate mode : Constant
Bit rate : 1 152 Kbps
Channel(s) : 1 channel
Channel positions : Back: R
Sampling rate : 48.0 KHz
Bit depth : 24 bits
Stream size : 759 MiB (1%)
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:10

Audio #7
ID : 8
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : in24
Duration : 1h 32mn
Bit rate mode : Constant
Bit rate : 1 152 Kbps
Channel(s) : 1 channel
Channel positions : Front: L (Matrix)
Sampling rate : 48.0 KHz
Bit depth : 24 bits
Stream size : 759 MiB (1%)
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:10

Audio #8
ID : 9
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : in24
Duration : 1h 32mn
Bit rate mode : Constant
Bit rate : 1 152 Kbps
Channel(s) : 1 channel
Channel positions : Front: R (Matrix)
Sampling rate : 48.0 KHz
Bit depth : 24 bits
Stream size : 759 MiB (1%)
Language : English
Encoded date : UTC 2014-04-24 00:04:26
Tagged date : UTC 2014-04-24 00:38:10

Other
ID : 10
Type : Time code
Format : QuickTime TC
Duration : 1h 32mn
Time code of first frame : 00:00:00:00
Time code settings : Striped
Language : English
Encoded date : UTC 2014-04-24 00:24:52
Tagged date : UTC 2014-04-24 00:24:52

[poisondeathray]

Got it figured out ? or more confused?

Can you answer the other question - did you want a separate 5.1 audio file like a WAV or AIFF ? or MOV with the audio only, or MOV with audio & video, or some other container?

Another option is to combine them in ffmpeg, then use eac3to or sox to downmix

[mrbliss]

Got it figured out ? or more confused?

Can you answer the other question - did you want a separate 5.1 audio file like a WAV or AIFF ? or MOV with the audio only, or MOV with audio & video, or some other container?

Another option is to combine them in ffmpeg, then use eac3to or sox to downmix

To be honest I'm not sure what I want! I think what I need is one file, containing video and audio, the audio bit being one stream containing several channels.

This is the software I'm using:

http://dcpomatic.com/

.mov is fine as a container, it doesn't seem to have a problem with the pro res video, that worked fine, but it just seems to pick the first of the 8 streams so there's no dialogue, only music and background sound. I'll have a look at the software you recomend, but any others pointers at this stage would be welcome.

[poisondeathray]

It's not a problem to remux (stream copy) the prores video, with downmixed 5.1 audio into a single MOV container

But there is a bit of a learning curve for avisynth, or to use ffmpeg / or sox / or eac3to . avisynth is run using scripts, and ffmpeg / sox / eac3to are all commandline programs

I'm not aware of any GUI's that will do what you want to do nicely with a push of a button


To use the avs script method for the audio, stream copy the video, into mov

ffmpeg -i input.mov -i downmix.avs -map 0:0 -map 1:1 -c:v copy -c:a pcm_s24le output.mov

The 2nd method without avisynth - To do it with ffmpeg and eac3to . The 1st step mixes to single 8ch wav, then eac3to downmixes to 6ch, then ffmpeg copies the video and muxes in the 6ch audio only into MOV

ffmpeg -i input.mov -vn -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6][0:7][0:8]amerge=inputs=8[aout]" -map "[aout]" -c:a pcm_s24le 8ch.wav

eac3to 8ch.wav 6ch.wav -downmix

ffmpeg -i input.mov -i 6ch.wav -map 0:0 -map 1:0 -c:v copy -c:a copy output.mov

[poisondeathray]

I can give you step by step directions if you are willing to download those programs and run a batch file

I'm going ask you to put everything into the eac3to folder , so you don't have to worry about file paths

1) Download eac3to, unzip the contents it to a folder

2) Download ffmpeg, put ffmpeg.exe into the same folder as the eac3to folder

3) Make a copy of your original MOV. Let's call it (and rename it) "input.mov" . Place "input.mov" into the same eac3to folder. Another reason I ask you do to this in case something goes wrong (but it shouldn't)

4) Download "batch.zip". Unzip it and inside is contained "batch.bat", put it into the same eac3to folder. If you open "batch.bat" in a text editor like notepad, you will notice it's the same commandline as the one posted in the earlier reply above

5) Double click batch.bat, and let it run. The file "output.mov" is the one you want.

6) Verify that the channels are correct (e.g you can open it up in audacity directly, if you have the audacity ffmpeg plugin.) I didn't specify deleting the temp files in the batch command, but you can delete them if you want (8ch.wav, 6ch.wav, and the eac3to log file)

[mrbliss]

Ok, no idea if what I'm doing is anywhere near correct, but this is where I'm up to:

Installed (I think!) Avisynth
Opend CMD window
pasted this:
ffmpeg –I D:PKG – KNOCKED_FOR_SEX_FEATURE_ONLY/KnockedForSix-Feature-01320203.mov -i downmix.avs -map 0:0 -map 1:1 -c:v copy -c:a pcm_s24le output.mov
Got this:
'ffmpeg' is not recognized as an internal or external command,
operable program or batch file.

[poisondeathray]

Try the step by step instructions with the batch file in post #8

[mrbliss]

Try the step by step instructions with the batch file in post #8

Thanks. Our posts crossed, will read post 8 now!

[mrbliss]

Thanks for much for you help with this. It seems quite clear, but I have a problem in that because of the size of the file I'm not going to be able to put the input & output in the same folder. I'm just going through a couple of portable drives to try and free enough space for a copy of the original and I can put the output in the eac3to folder. The drive the copy of the original will be in is I, so i just need to add "I:" before input.mov?

[smrpix]

A brief perusal of the manual shows DCP-o-matic has an audio channel assigning function. Since PCM is legal pretty much everywhere, you probably don't have to do any pre-conversions.

[poisondeathray]

Thanks for much for you help with this. It seems quite clear, but I have a problem in that because of the size of the file I'm not going to be able to put the input & output in the same folder. I'm just going through a couple of portable drives to try and free enough space for a copy of the original and I can put the output in the eac3to folder. The drive the copy of the original will be in is I, so i just need to add "I:" before input.mov?

Yes.


And really, there should be no chance of overwriting the original (but those are famous last words...so better safe than sorry)

Open up the batch file and specify the path with quotes where you want to put both the input and output file. If you want you can use the real original file as the input file

So you can edit the batch file, or just rewrite it . You don't even need to put everything in the eac3to folder, you can specify the paths of the executables (ffmpeg and exe)

For example , if "input.mov" was really in the root directory of drive I : (the entries I've changed are highlighted in red)

ffmpeg -i "I:\input.mov" -vn -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6][0:7][0:8]amerge=inputs=8[aout]" -map "[aout]" -c:a pcm_s24le 8ch.wav

eac3to 8ch.wav 6ch.wav -downmix

ffmpeg -i "I:\input.mov" -i 6ch.wav -map 0:0 -map 1:0 -c:v copy -c:a copy output.mov

pause

To edit batch files, you can just do them in notepad, and rename the extension from .txt to .bat





If you wanted to use the real path of the original file , you could have used

"G:\PKG - KNOCKED_FOR_SIX_FEATURE_ONLY\KnockedForSix-Feature-01320203.mov"

Same for the output files, same for the temp files, same for the executables - you can specify precise paths for everything

[poisondeathray]

A brief perusal of the manual shows DCP-o-matic has an audio channel assigning function. Since PCM is legal pretty much everywhere, you probably don't have to do any pre-conversions.

I think he said the problem was it only "reads" the 1st mono audio track

Unless it can accept multiple input files ? Is there anything about that ? Because then you could just demux it

[mrbliss]

A brief perusal of the manual shows DCP-o-matic has an audio channel assigning function. Since PCM is legal pretty much everywhere, you probably don't have to do any pre-conversions.

I'm just trying DCP O Matic again (it takes an hour just to scan the file on an i7 pc!) to see if I can work this out, but I think it only showed one channel when I did it last time (the sample from the manual shows 6).

[smrpix]

Looks like I too briefly perused the original post.

I wonder if the files show up as discrete tracks in Quicktime or Mpegstreamclip. Might be another way to extract them.

[mrbliss]

Ok, its finished scanning, and under audio / content it still just shows one channel, it doesn't seem to be able to recognise the other 7.

[mrbliss]

Thanks both for your help. I now have the bat file running and 8ch.wav is being built. There was one error on the way:

[Parsed_ameged @000000000492ca001] No channel layout for input 1

Not sure how much of a problem that is going to be?

[poisondeathray]

Thanks both for your help. I now have the bat file running and 8ch.wav is being built. There was one error on the way:

[Parsed_ameged @000000000492ca001] No channel layout for input 1

Not sure how much of a problem that is going to be?

It won't matter, because you're manually mapping the channels in the order that they come in

7.1 is a standard configuration : FL+FR+FC+LFE+BL+BR+SL+SR. That's the way they are usually mapped

If you look at the mediainfo report you posted above, you will notice that the metadata for each audio channel says the same thing. That's the channel ordering that all programs expect

However, if the file you got was mapped incorrectly, and mislabelled, in the first place - then that might be a problem. That's why I asked you to double check the output file channels

[poisondeathray]

Also double check with ffmpeg -i input.mov and see if the numbering matches what mediainfo says

Verify that it says something like

0:0 prores
0:1 audio
0:2 audio
0:3 audio
.
.
.
0:9 audio

Some quicktime files insert a timecode track as track 1. (In your mediainfo report it says 10 for the timecode track, but mediainfo numbering doesn't always correlate with ffmpeg track numbering). If that's the case , you need to adust the channel mapping, because 0:1 won't correspond to the 1st mono audio track

[mrbliss]

I think I'll wait until this has finished running before examining it. It's looking pretty good so far. The wav file is still being muxed (3gb now). When this bits done, will the next stage be quicker? Getting late here in the UK and I next to get some shut eye, but if its not going to be too much longer then I'd like to start the DCP rendering asap.

[poisondeathray]

Last step should be fast, it's limited mostly by your HDD speed. It's just stream copying the video and new audio into a new MOV container. However, if your HDD's are slow and fragmented...

[mrbliss]

Its now just started creating 6ch.wav. About the same time for this as 8ch.wav?

[mrbliss]

Ah, already building output.mov!

[mrbliss]

Quick question while I'm waiting for this to finish. I can put the source and destination on two different drives, one in my PC and the other a portable USB 2. Which would be the best way around for speed purposes?

[mrbliss]

Also, DCP O Matic will accept 8 channels, but the last two are labled HI and VI. The Mediainfo report calls the last two channels Front L (Matrix) and Front R (Matrix). Should I just mix them into the L & R channels?

[mrbliss]

Well, its just finished and it all works perfectly... for the first eight minutes, then the sound just stops. Why would that be?

The log reads:

eac3to v3.27
command line: eac3to 8ch.wav 6ch.wav -downmix
------------------------------------------------------------------------------
WAV, 7.1 channels, 1:32:08, 24 bits, 9216kbps, 48kHz
Reading WAV...
Mixing surround channels...
Writing WAV...
Creating file "6ch.wav"...
The original audio track has a constant bit depth of 24 bits.
The processed audio track has a constant bit depth of 24 bits.
Caution: The WAV file is bigger than 4GB. <WARNING>
Some WAV readers might not be able to handle this file correctly. <WARNING>
eac3to processing took 2 minutes, 56 seconds.
Done.

[poisondeathray]

Splitting source & dest HDD's is a good idea, but not sure which configuration would be faster

No idea why it stops - was that the mov that you tested, the 8ch wav, or the 6ch wav ?

Do any/all of them play ok in a media player ? e.g. vlc, mpchc, potplayer etc.... ?

[El Heggunte]

Proper .WAV files cannot be bigger than 4GB, because the header field that indicates the stream size is a 32-bit integer value

(yes, Micro$oft is stupid, as usual).

eac3to can output to .W64 or .RF64 though.