Exchanging a Video part in an existing DCP

[IoannisSyrogiannis]

Robert, I think that you take for granted that DCP-o-matic is an editor, while it is not. I wish it could have been incorporated in editors as part of an editor suite, but that is me dreaming. Using since long the program, I suppose you are already aware. It is not "super not obvious" that you will need to unmap the audio channels of the DCP, since there is no other way to remove the audio altogether. It seems like the only simple solution. When you have more than one video parts, I would also advise for using the "show graph of audio levels" on the "DCP>Audio" tab of the editor. That is the only way to get loudness readings for the whole DCP.
The list of the inputs works for timing, because mixing is not the goal here. Otherwise, we would have had discussions about crossfading etc. already. Fix timings with snipping out of the original DCP and you will get the video in one line instead of two.

[freezer]

Robert, I think that you take for granted that DCP-o-matic is an editor, while it is not.

I don't take that for granted, but it is offering the capability to do some simple manipulations for existing DCPs and those features could need some polish to shine. I will try to make some suggestions to Carl, but I need to think this through first.
Decades ago I had been working as an user interface designer for several companies. One important lesson in UI experience is to never let the user in the dark what is going to happen and never force a paradigm completely contrary to real world expectations - as that will likely result in a disaster.
Those principles just make sense. UI design is not an easy task - even if it is just about where to put this and that button/option/menu entry.

It is not "super not obvious" that you will need to unmap the audio channels of the DCP, since there is no other way to remove the audio altogether. It seems like the only simple solution.

One way to improve this would be an option to right click the DCP audio track in the timeline and select "Disable audio track" which would just unmap the audio channels and "Enable audio track" would map them back again. Of course it should then get displayed in the timeline as disabled and not just as invisible like now.

When you have more than one video parts, I would also advise for using the "show graph of audio levels" on the "DCP>Audio" tab of the editor. That is the only way to get loudness readings for the whole DCP.

Yes, I should have done that again, but I didn't because I had absolutely no idea that it would sum up the tracks - and I had checked audio before.

The list of the inputs works for timing, because mixing is not the goal here. Otherwise, we would have had discussions about crossfading etc. already. Fix timings with snipping out of the original DCP and you will get the video in one line instead of two.

Yes I did that - trimming out sections. But that started the whole misery because it also trimmed the audio so I added the original audio track too.
And it is not about mixing video tracks - it is about the correct order of the stack. If something sits on top it will hide what is below - but DOM does it show the other way round in the timeline. That is my issue here - it goes against a paradigm established everywhere else and also in the real world.

[IoannisSyrogiannis]

[...]it is offering the capability to do some simple manipulations for existing DCPs[...]
[...]- and I had checked audio before.[...]

It makes sense, that it would, but it doesn't offer such a capability in reality. What it does, is that it offers a schematic view of what one does manually, from the timing settings, etc.
Try to find who works with the timeline (do a search on older posts here), and you will find only a few disappointed users, because it was never in focus. That's why there is no practical way to be precise/productive when moving assets that way.

As for the audio check. I know many that have checked audio on a singular video/audio file (from the „Content“ tab), but not that many that have done that from the „DCP“ tab. We don't know each other, so I don't know if you have checked audio that way before. Regardless of what you have done before, I am pointing where the clue could have been. This exchange is between us two, but it might be of help for someone in a month or a year from now. That's what is super useful with forums.
I only checked on the beta version, but I saw that if you right click an input channel, you may set gain for that. So, you may have used that feature to lower each of the two mixed channels by 3dBs (?). But you/one would need to check from the „DCP“ tab and its audio segment there, if you are/one is using more than a single audio input.

[carl]

The timeline is certainly a bit simplistic, but I don't see a reason not to improve it - especially when there are obviously bad things like this example with the audio. I am no skilled UX/UI designer so any help is welcome.

If things were different I would be happy to say "DCP-o-matic is not a video editor" and remove all such features. But there are significant advantages in being able to manipulate DCPs in this way, so I think the timeline is worth keeping and improving where possible.

It's interesting that you say "things on top should play" - I haven't done thorough research but I do seem to remember at least one DAW that does "thing at bottom plays". Maybe I mis-remembered.

Then again, I have very little video editor experience and if that is the standard approach in that area it would probably make most sense also for DCP-o-matic.

[freezer]

It makes sense, that it would, but it doesn't offer such a capability in reality. What it does, is that it offers a schematic view of what one does manually, from the timing settings, etc.
Try to find who works with the timeline (do a search on older posts here), and you will find only a few disappointed users, because it was never in focus. That's why there is no practical way to be precise/productive when moving assets that way.

It works - if you are aware of the pitfalls to avoid. But I had to that under time constraints and pressure - never a good starting point for doing something you have never done before.

As for the audio check. I know many that have checked audio on a singular video/audio file (from the „Content“ tab), but not that many that have done that from the „DCP“ tab. We don't know each other, so I don't know if you have checked audio that way before.

I always check the audio. But because the audio had not been altered in any way, had been checked in Resolve Fairlight, then in iZotope RX, then in DOM in the content tab. But nothing in the interface suggests, that there is a difference between the Audio Levels Graph in the content tab vs. the DCP tab. Also - when I let the Audio check in the DCP tab - it takes a very long time to start, hovering at 0% and a green bar is moving across, then after a while it starts to analyzing audio and counts up to 100%. Then says OK and the graph isn't shown.
Anyway - I had no reason to think that copying audio through would have altered anything.

I only checked on the beta version, but I saw that if you right click an input channel, you may set gain for that. So, you may have used that feature to lower each of the two mixed channels by 3dBs (?).

No - I did not want any mix at all and lowering all channels by 3dB would have also altered the endcredits tracks - which had no summed mix because of the trimmed DCP.

[freezer]

The timeline is certainly a bit simplistic, but I don't see a reason not to improve it - especially when there are obviously bad things like this example with the audio. I am no skilled UX/UI designer so any help is welcome.

When work slows down here, I will have a look and do some UI design suggestions. I know that it can be hard for programmers to design interfaces - because they are used to think in a completely different way and program logic. That's why normally the interface designer isn't the programmer and vice versa

If things were different I would be happy to say "DCP-o-matic is not a video editor" and remove all such features. But there are significant advantages in being able to manipulate DCPs in this way, so I think the timeline is worth keeping and improving where possible.

Yes - the feature is very cool and helpful.

It's interesting that you say "things on top should play" - I haven't done thorough research but I do seem to remember at least one DAW that does "thing at bottom plays". Maybe I mis-remembered.

DAWs are not video editors and audio is different from video.
For audio it does not make any difference where you place the track in the stack. Waveforms will always mix up together.
I guess what you are referring to is the master track, which is usually at the bottom of the stack.
You can route audio tracks any way you want due to audio's nature. That does not apply to video in the same way.

I've done the audio editing and 5.1 mixing for our feature film. All in all I had 48 tracks for dialog, ambiance sound, sound effects, ambiance sound effects, foleys, reverb, music. I can group for example all music tracks together into a new track called a sub-mix bus. So any audio mastering like EQs, dynamics I do can be done to that mix bus where all music comes together.
And all mix buses are then routed into the master busses like a 5.1 main mix and an additional 2.0 stereo mix.

On the mixing console (where you add plugins, adjust levels and more) the tracks and buses are shown from left to right, but the tracks itself are shown from top to bottom.

For video it is different. Just think of a stack of cards. The upmost card is always the only one you can see and if you pull that away you will just see the next card in the stack.
You can work with different mathematical mix modes or transparent parts of a video by using alpha channels. But the top most video will still sit on top.

[IoannisSyrogiannis]

[...]
But nothing in the interface suggests, that there is a difference between the Audio Levels Graph in the content tab vs. the DCP tab. Also - when I let the Audio check in the DCP tab - it takes a very long time to start, hovering at 0% and a green bar is moving across, then after a while it starts to analyzing audio and counts up to 100%. Then says OK and the graph isn't shown.
[...]

That seems like a bug. I wonder, though, what happens if you press the button again. It would either start from zero, or it would draw the graph from the info collected.

What might give away the fact that the content and DCP tabs are different on audio levels' graph, is the fact that all different audio on the content has its individual graphical representation, when chosen and run. So -especially if one is muxing separate audio channel files- one can't get total loudness levels from the individual graph. Happened to me so many times, to need to go with making a DCP with audio out of discrete audio channels that I learned my lesson.

If this post leads to a GUI amelioration, „hurrah“!

[carl]

DAWs are not video editors and audio is different from video.
For audio it does not make any difference where you place the track in the stack. Waveforms will always mix up together.
I guess what you are referring to is the master track, which is usually at the bottom of the stack.

I was referring to DAWs that have some comping/"take lane" workflow whereby there can be overlapping clips/regions on a single track, so you can comp takes together. I'll check my memory on this though!