Hi, I think this scenario has passed by on the mailinglist before, and I might because of how I work, be more attracted to these kinds of problems than others that mainly do non-linear work. After the recording, and after getting the main production idea, my projects are timewise mostly influenced by reconstruction of a multi track linear timeline. I am using multiple camera's that don't use any (world) timecodes, and have 2GB limitations. So the best thing they can do is giving me a relative timestamp by means the filename or creation_time metadata. What I would be interested in is the ability to: 0. having a forever scrolling canvas, canvas may do tricks like hiding places where no content is available 1. "automatically" construct such timeline each track(group) being from a single device, which typically atomically can only produce a single stream of content by means of metadata 2. having some tooling that can do macro alignment, independent of metadata (for example by audio fingerprinting) 3. having some tooling that can do micro alignment 4. export this in either some EDL format or multitrack format being NLE independent I wonder what other people are using for the above, other than pen and paper.
From my own experience within Cinerella (I know about mixers) I don't end up with the workflow that I want for non-continious recordings.
A second email follows on a different workflow issue. -- Stefan
сб, 28 янв. 2023 г., 21:49 Stefan de Konink via Cin < [email protected]>:
Hi,
I think this scenario has passed by on the mailinglist before, and I might because of how I work, be more attracted to these kinds of problems than others that mainly do non-linear work.
While I think I can't even start on implementing those ideas I have question: did you tried to use 'split timeline' feature, so each cam put on same teack but at wildly different time and you have 2-4 zones for working on each one? (mostly by cutting things to another track?) Or I completely misuderstand how split timeline works in cin ...
After the recording, and after getting the main production idea, my projects are timewise mostly influenced by reconstruction of a multi track linear timeline. I am using multiple camera's that don't use any (world) timecodes, and have 2GB limitations. So the best thing they can do is giving me a relative timestamp by means the filename or creation_time metadata.
What I would be interested in is the ability to:
0. having a forever scrolling canvas, canvas may do tricks like hiding places where no content is available
-- but how you will tell users there is like 1, min of content, and there is 5m? By some markets? Should those 'edits' expand individually when needed?
1. "automatically" construct such timeline each track(group) being from a single device, which typically atomically can only produce a single stream of content by means of metadata
---
how you think should work, is there something like cameraID or you need to associate some user-know metadata key to specific camera, so you can say send all medias with this metadata to specific track or media bin?
2. having some tooling that can do macro alignment, independent of metadata (for example by audio fingerprinting)
interesting idea, does any libre NLE currently have such functionality?
3. having some tooling that can do micro alignment
--
Like 'move this segment few frames left/right' as context menu option?
4. export this in either some EDL format or multitrack format being NLE independent
--- I'll try to look into it a bit more, but no promices (cmx3600 only had few audio tracks, but may be there were unofficial extensions ...)
I wonder what other people are using for the above, other than pen and paper.
From my own experience within Cinerella (I know about mixers) I don't end up with the workflow that I want for non-continious recordings.
A second email follows on a different workflow issue.
-- Stefan -- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin
On Sunday, January 29, 2023 2:21:13 AM CET, Andrew Randrianasulu wrote:
While I think I can't even start on implementing those ideas I have question: did you tried to use 'split timeline' feature, so each cam put on same teack but at wildly different time and you have 2-4 zones for working on each one? (mostly by cutting things to another track?) Or I completely misuderstand how split timeline works in cin ...
I cannot find the term 'split timeline'. But started looking; do you mean creating a clip from the main timeline. Open it with "Open EDL" in resources, and start editing that? <https://cinelerra-gg.org/download/CinelerraGG_Manual/OpenEDL.html#6784> I was not aware of this feature, but it would certainly make it easier to edit clips. Maybe this is more related to my second email "Workflow issues for multitrack editing". What I would still dislike is that my "0" EDL would be the initial place to align the material, and not the 'final' timeline. There is still no separation of concerns.
but how you will tell users there is like 1, min of content, and there is 5m? By some markets? Should those 'edits' expand individually when needed?
I would envision an entire project to be available, like it would be now when all content is aligned. The timeline would be something where the X-axis is creation time and duration. Resources are just placed on this timeline. This functionality would be a different presentation of the "resources file list". Before we get into the academic discussion of Cinelerra has frames and not seconds. You may express this in relative frames towards each other as well.
1. "automatically" construct such timeline each track(group) being from a single device, which typically atomically can only produce a single stream of content by means of metadata
how you think should work, is there something like cameraID or you need to associate some user-know metadata key to specific camera, so you can say send all medias with this metadata to specific track or media bin?
When I compare the photogrammetry software Meshroom it has two variants. It can bundle assets based on camera serial number or a folder on disk. My current cinelerra projects are also organised by the same principles. I have nikon-50mm, nikon-35mm, nokia and tascam folders. Within resources we are just lacking this group information (the asset producing device) and the ability to pair between them (between devices).
2. having some tooling that can do macro alignment, independent of metadata (for example by audio fingerprinting)
interesting idea, does any libre NLE currently have such functionality?
In our previous discussion; - cinelerra can do aligning - kdenlive - for kdenlive someone also made an external tool <https://www.youtube.com/watch?v=mypKKK-Se7A>
3. having some tooling that can do micro alignment
Like 'move this segment few frames left/right' as context menu option?
Exactly, having multiple viewers available so you could ultimately also use it to view a clapperboard.
4. export this in either some EDL format or multitrack format being NLE independent
---
I'll try to look into it a bit more, but no promices (cmx3600 only had few audio tracks, but may be there were unofficial extensions ...)
MOV files certainly support it. It can obviously be used as an exchange format, but from performance perspective, playing a file that is 300MB/s is sub optimal. So even then you would like to split it in chunks that can be easily used. A (format)proxy may obviously solve this issue. -- Stefan
вс, 29 янв. 2023 г., 09:55 Stefan de Konink <[email protected]>:
On Sunday, January 29, 2023 2:21:13 AM CET, Andrew Randrianasulu wrote:
While I think I can't even start on implementing those ideas I have question: did you tried to use 'split timeline' feature, so each cam put on same teack but at wildly different time and you have 2-4 zones for working on each one? (mostly by cutting things to another track?) Or I completely misuderstand how split timeline works in cin ...
I cannot find the term 'split timeline'.
sorry, it was 'split pane' https://cinelerra-gg.org/download/CinelerraGG_Manual/Multi_Pane_Support.html But started looking; do you mean
creating a clip from the main timeline. Open it with "Open EDL" in resources, and start editing that?
<https://cinelerra-gg.org/download/CinelerraGG_Manual/OpenEDL.html#6784>
I was not aware of this feature, but it would certainly make it easier to edit clips. Maybe this is more related to my second email "Workflow issues for multitrack editing". What I would still dislike is that my "0" EDL would be the initial place to align the material, and not the 'final' timeline. There is still no separation of concerns.
but how you will tell users there is like 1, min of content, and there is 5m? By some markets? Should those 'edits' expand individually when needed?
I would envision an entire project to be available, like it would be now when all content is aligned. The timeline would be something where the X-axis is creation time and duration. Resources are just placed on this timeline. This functionality would be a different presentation of the "resources file list".
Before we get into the academic discussion of Cinelerra has frames and not seconds. You may express this in relative frames towards each other as well.
1. "automatically" construct such timeline each track(group) being from a single device, which typically atomically can only produce a single stream of content by means of metadata
how you think should work, is there something like cameraID or you need to associate some user-know metadata key to specific camera, so you can say send all medias with this metadata to specific track or media bin?
When I compare the photogrammetry software Meshroom it has two variants. It can bundle assets based on camera serial number or a folder on disk. My current cinelerra projects are also organised by the same principles. I have nikon-50mm, nikon-35mm, nokia and tascam folders.
Within resources we are just lacking this group information (the asset producing device) and the ability to pair between them (between devices).
2. having some tooling that can do macro alignment, independent of metadata (for example by audio fingerprinting)
interesting idea, does any libre NLE currently have such functionality?
In our previous discussion;
- cinelerra can do aligning - kdenlive - for kdenlive someone also made an external tool <https://www.youtube.com/watch?v=mypKKK-Se7A>
3. having some tooling that can do micro alignment
Like 'move this segment few frames left/right' as context menu option?
Exactly, having multiple viewers available so you could ultimately also use it to view a clapperboard.
4. export this in either some EDL format or multitrack format being NLE independent
---
I'll try to look into it a bit more, but no promices (cmx3600 only had few audio tracks, but may be there were unofficial extensions ...)
MOV files certainly support it. It can obviously be used as an exchange format, but from performance perspective, playing a file that is 300MB/s is sub optimal. So even then you would like to split it in chunks that can be easily used. A (format)proxy may obviously solve this issue.
-- Stefan
On Sunday, January 29, 2023 9:03:35 AM CET, Andrew Randrianasulu wrote:
sorry, it was 'split pane'
https://cinelerra-gg.org/download/CinelerraGG_Manual/Multi_Pane_Support.html
Concept is interesting to get some scratch space. At this moment I wonder if the main window could be made available for each "OpenEDL". The split X would be a workaround to make some area available at the end of the project to do the multitrack editing on a single screen. -- Stefan
2. having some tooling that can do macro alignment, independent of metadata (for example by audio fingerprinting) for kdenlive someone also made an external tool <https://www.youtube.com/watch?v=mypKKK-Se7A>
Alignment using timecodes for Professional productions was addressed in MantisBT 448 (https://www.cinelerra-gg.org/bugtracker/view.php?id=448). The "Align timecodes" tool came into being there. But the pro who created the issue concluded: "I have tested the 'align timecodes' function in the new version and am very happy to say that is working wonderful! Along with the 'To clip' function I can now make synced up subclips. I believe that to my knowledge Cinelerra GG is now the first FLOSS video editing program to support timecode syncing with subclip functionality: that's a big premiere! Thanks to everyone involved and especially PhyllisSmith. Few notes for further improvements: - A batch function would be great, especially to already sync up the clips from the 'Resources' window. - When I use the 'align timecodes' function the timecode of the whole sequence changes to the timecode of the clip I'm syncing. I have not seen this in other editing programs and don't think this is a useful feature, I think it would be better if the timecode of the sequence stays untouched and the clips stays in place." It would be nice to be able to do alignment of various audio/video tracks with an external audio track in batch, all at once, in the Resources window, before getting to the timeline. Question: is it possible to modify and use the script linked by Stefan in CinGG?
4. export this in either some EDL format or multitrack format being NLE independent.
As for interchange between various programs, I don't think CMX3600 is viable; it is too limited and not well supported by modern programs. Nowadays AAF (https://en.wikipedia.org/wiki/Advanced_Authoring_Format) is used as an interchange format; I believe it is based on Avid MXF.
вс, 29 янв. 2023 г., 13:35 Andrea paz <[email protected]>:
2. having some tooling that can do macro alignment, independent of metadata (for example by audio fingerprinting) for kdenlive someone also made an external tool < https://www.youtube.com/watch?v=mypKKK-Se7A>
Alignment using timecodes for Professional productions was addressed in MantisBT 448 (https://www.cinelerra-gg.org/bugtracker/view.php?id=448). The "Align timecodes" tool came into being there. But the pro who created the issue concluded: "I have tested the 'align timecodes' function in the new version and am very happy to say that is working wonderful! Along with the 'To clip' function I can now make synced up subclips. I believe that to my knowledge Cinelerra GG is now the first FLOSS video editing program to support timecode syncing with subclip functionality: that's a big premiere! Thanks to everyone involved and especially PhyllisSmith.
Few notes for further improvements: - A batch function would be great, especially to already sync up the clips from the 'Resources' window. - When I use the 'align timecodes' function the timecode of the whole sequence changes to the timecode of the clip I'm syncing. I have not seen this in other editing programs and don't think this is a useful feature, I think it would be better if the timecode of the sequence stays untouched and the clips stays in place."
It would be nice to be able to do alignment of various audio/video tracks with an external audio track in batch, all at once, in the Resources window, before getting to the timeline. Question: is it possible to modify and use the script linked by Stefan in CinGG?
I think it uses kdenlive specific editing functions? And I do not know enough about xml generated by cingg for making it work for us or guide its author to do it.. but there are some python projects on github https://github.com/align-videos-by-sound/align-videos-by-sound https://github.com/benfmiller/audalign in java https://github.com/JorenSix/SyncSink none of them (?) works as is for providing timecode offset for embedding in media files so align via timecode will work, for example. I'll keep searching! (there is related problem of aligning text with its audio, but this is another area of research, it just showed up initially via github search,)
4. export this in either some EDL format or multitrack format being NLE independent.
As for interchange between various programs, I don't think CMX3600 is viable; it is too limited and not well supported by modern programs. Nowadays AAF (https://en.wikipedia.org/wiki/Advanced_Authoring_Format) is used as an interchange format; I believe it is based on Avid MXF.
вс, 29 янв. 2023 г., 14:31 Andrew Randrianasulu <[email protected]>:
вс, 29 янв. 2023 г., 13:35 Andrea paz <[email protected]>:
2. having some tooling that can do macro alignment, independent of metadata (for example by audio fingerprinting) for kdenlive someone also made an external tool < https://www.youtube.com/watch?v=mypKKK-Se7A>
Alignment using timecodes for Professional productions was addressed in MantisBT 448 (https://www.cinelerra-gg.org/bugtracker/view.php?id=448). The "Align timecodes" tool came into being there. But the pro who created the issue concluded: "I have tested the 'align timecodes' function in the new version and am very happy to say that is working wonderful! Along with the 'To clip' function I can now make synced up subclips. I believe that to my knowledge Cinelerra GG is now the first FLOSS video editing program to support timecode syncing with subclip functionality: that's a big premiere! Thanks to everyone involved and especially PhyllisSmith.
Few notes for further improvements: - A batch function would be great, especially to already sync up the clips from the 'Resources' window. - When I use the 'align timecodes' function the timecode of the whole sequence changes to the timecode of the clip I'm syncing. I have not seen this in other editing programs and don't think this is a useful feature, I think it would be better if the timecode of the sequence stays untouched and the clips stays in place."
It would be nice to be able to do alignment of various audio/video tracks with an external audio track in batch, all at once, in the Resources window, before getting to the timeline. Question: is it possible to modify and use the script linked by Stefan in CinGG?
I think it uses kdenlive specific editing functions? And I do not know enough about xml generated by cingg for making it work for us or guide its author to do it..
but there are some python projects on github
https://github.com/align-videos-by-sound/align-videos-by-sound
https://github.com/benfmiller/audalign
in java
https://github.com/alopatindev/sync-audio-tracks also this. Uses same process as kdenlive related program , but compiles into standalone program and including scripts use it + ffmpeg (downsample audio for faster processing) for calculating delay, but between two files only ....
none of them (?) works as is for providing timecode offset for embedding in media files so align via timecode will work, for example.
I'll keep searching! (there is related problem of aligning text with its audio, but this is another area of research, it just showed up initially via github search,)
4. export this in either some EDL format or multitrack format being NLE independent.
As for interchange between various programs, I don't think CMX3600 is viable; it is too limited and not well supported by modern programs. Nowadays AAF (https://en.wikipedia.org/wiki/Advanced_Authoring_Format) is used as an interchange format; I believe it is based on Avid MXF.
вс, 29 янв. 2023 г., 09:55 Stefan de Konink <[email protected]>:
On Sunday, January 29, 2023 2:21:13 AM CET, Andrew Randrianasulu wrote:
While I think I can't even start on implementing those ideas I have question: did you tried to use 'split timeline' feature, so each cam put on same teack but at wildly different time and you have 2-4 zones for working on each one? (mostly by cutting things to another track?) Or I completely misuderstand how split timeline works in cin ...
I cannot find the term 'split timeline'. But started looking; do you mean creating a clip from the main timeline. Open it with "Open EDL" in resources, and start editing that?
<https://cinelerra-gg.org/download/CinelerraGG_Manual/OpenEDL.html#6784>
I was not aware of this feature, but it would certainly make it easier to edit clips. Maybe this is more related to my second email "Workflow issues for multitrack editing". What I would still dislike is that my "0" EDL would be the initial place to align the material, and not the 'final' timeline. There is still no separation of concerns.
but how you will tell users there is like 1, min of content, and there is 5m? By some markets? Should those 'edits' expand individually when needed?
I would envision an entire project to be available, like it would be now when all content is aligned. The timeline would be something where the X-axis is creation time and duration. Resources are just placed on this timeline. This functionality would be a different presentation of the "resources file list".
Before we get into the academic discussion of Cinelerra has frames and not seconds. You may express this in relative frames towards each other as well.
1. "automatically" construct such timeline each track(group) being from a single device, which typically atomically can only produce a single stream of content by means of metadata
how you think should work, is there something like cameraID or you need to associate some user-know metadata key to specific camera, so you can say send all medias with this metadata to specific track or media bin?
When I compare the photogrammetry software Meshroom it has two variants. It can bundle assets based on camera serial number or a folder on disk. My current cinelerra projects are also organised by the same principles. I have nikon-50mm, nikon-35mm, nokia and tascam folders.
Within resources we are just lacking this group information (the asset producing device) and the ability to pair between them (between devices).
2. having some tooling that can do macro alignment, independent of metadata (for example by audio fingerprinting)
interesting idea, does any libre NLE currently have such functionality?
In our previous discussion;
- cinelerra can do aligning - kdenlive - for kdenlive someone also made an external tool <https://www.youtube.com/watch?v=mypKKK-Se7A>
---
I think shotcut also should have something since 2022 https://forum.shotcut.org/t/audio-alignment-implementation/30420/44
3. having some tooling that can do micro alignment
Like 'move this segment few frames left/right' as context menu option?
Exactly, having multiple viewers available so you could ultimately also use it to view a clapperboard.
4. export this in either some EDL format or multitrack format being NLE independent
---
I'll try to look into it a bit more, but no promices (cmx3600 only had few audio tracks, but may be there were unofficial extensions ...)
MOV files certainly support it. It can obviously be used as an exchange format, but from performance perspective, playing a file that is 300MB/s is sub optimal. So even then you would like to split it in chunks that can be easily used. A (format)proxy may obviously solve this issue.
-- Stefan
participants (3)
-
Andrea paz -
Andrew Randrianasulu -
Stefan de Konink