Black frame after creating PSA in DCP-O-Matic

[Mitchel Wagner]

Hello all,

My pc was recently reinstalles. Therefore I had to reinstall DCP-O-Matic. Since the new install (version 2.16.32) there seems to be an issue when I create a PSA.
The PSA's I create have a very short black frame at the end. The duration of the clip is 8 seconds. When I play the content in DCP-O-Matic it stops at 07.23
I exported this PSA to a MP4 video file and there doesn't seem to be a black frame. However when I view PSA via our Doremi in our theater there is a flash of black at the end of the PSA.
Could this be a setting in DCP-O-Matic ? I looked for it , but haven't found one yet.

UPDATE : I've tested this with a PSA that my co-worker made on his MacBook. Can't check wich version of DCP-O-MATIC he has, but it happends with his PSA to. I also checked both Theathers. Happends in the both of them.

[Carsten Kurz]

Maybe you can upload the video mxf of this DCP somewhere for us to take a look? A project with just a single PNG in it should not have any black added to it, or, only at the beginning if you moved it away from 00:00:00:00

[Mitchel Wagner]

https://dedomijnen-my.sharepoint.com...CU2hw?e=NOtq4h

Here is a link to the DCP in zip file.

[Marcel Birgelen]

https://dedomijnen-my.sharepoint.com...CU2hw?e=NOtq4h

Here is a link to the DCP in zip file.

The MXF looks clean, I don't see a black frame at the end. The CPL also seems OK to me. You can view the MXF directly using a player like VideoLAN.

You mention a black flash at the end while playing it back as part of your show. What exactly do you mean with a black flash? Does the image disappear and re-appear or is there simply a "black pause" of about one frame between this DCP and the next in the playlist?

[Mitchel Wagner]

The MXF looks clean, I don't see a black frame at the end. The CPL also seems OK to me. You can view the MXF directly using a player like VideoLAN.

You mention a black flash at the end while playing it back as part of your show. What exactly do you mean with a black flash? Does the image disappear and re-appear or is there simply a "black pause" of about one frame between this DCP and the next in the playlist?

It is a "Black Pause" of about a few frames between this DCP and the next DCP in the playlist. I will watch the MXF on VideoLAN, see if I can see the black frames there.

[Mitchel Wagner]

I've watched this in VLC. The image seems to be running ok. But it stops at 7 sec, while the file is 8 seconds long. I've attached an image with this, while it runs in VLC

[Carsten Kurz]

That DCP is okay. If you load it into DCP-o-matic or DCP-o-matic player, you can step through it frame by frame, and the last frame at 7s:23f is the image. I can only assume that your playback system causes the issue. Maybe there is a 'very different' DCP right ahead of behind it. Some playback servers manage more or less seamless transitions between clips with similar parameters. Under certain conditions, though, you may experience a short delay, or visible artefacts like flashes. You may rearrange your show playlist to analyse that issue.


VLC stopping at '7' is probably according to 7s23f

Remember, that display is time, not frame numbers. So, the first frame is a '0', and 7s23f indicates 7s + 24 frames.

[Marcel Birgelen]

I agree with Carsten. Your DCP seems to be fine.

VLC's progress indicator often is a bit wonky, especially with some more esoteric file types.

Unfortunately, cinema servers don't guarantee a seamless playback between playlist items, even though this used to be one of the strengths of Doremi's other, since discontinued, professional playback solutions.

As a rule of thumb: Add a short fade to black to the end of your DCPs if there is still a high-contrast image on screen at the end.

There might be something different between the to DCPs that requires a re-initialisation of the media block for example and in those cases, you're lucky this happens with just black on-screen, in the past, we've seen some servers just throwing flashes of digital garbage on-screen while this happened...

[Randy Stankey]

Don't forget that your editing timeline is zero-significant.

Your first frame will be 00:00:00 (MM:SS:FF) To make a video that is exactly eight seconds long, your video would have to end at frame 00:07:23.

Try going back to make sure that your project's timeline has the In-Point set at 00:00:00 and the Out-Point set at 00:07:23, not 00:00:01 >> 00:08:00, then re-export for processing by DCP-O-Matic. After ingesting and building the show, let the computer place the file, as-is, without trying to tell it to start the clip at any particular frame/time code.

If you try to force playback to start at 00:00:01 and end at 00:08:00, you'll be cutting off one frame at the beginning and there'll be one extra, blank, frame at the end which will likely be black because there's nothing to display.

You might also try editing in a 12-frame fade-in/fade-out at the beginning and end of the video. It will take half-second to come up from black and another half-second to come back down to black. That way, when you string clips together in your playlist, there will be a full second of transition between them and, if there are any one-frame glitches, they'll be hidden.

Also, be sure to check to be sure that you're NOT editing in drop-frame mode. If you are in drop-frame, your frame counter will have a colon between the seconds and the frame number. (00:07;23 instead of 00:07:23) Drop-frame is for conforming video to broadcast/television standard where the actual frame rate is 29.97 instead of 30.00. Two frames out of every minute (?) are dropped in order to make the math come out even. If you are in drop-frame mode when you don't need to be, you'll get wonky playback that might look like the problem you report.

[Ioannis Syrogiannis]

+1 to what people said before, about screening different types of DCPs and what happens in between.
One way to check if it has to do with the server/media block, is to have this CPL play twice in a playlist (same type, frame rate etc.) and check whether that effect appears both between the two identical CPLs and after those, when something else follows, as in your usual screenings.