(Since this is a bit of a complicated question, I recorded a video to show what I'm doing that may be more clear: https://youtu.be/VyaUT6KMYZA)
I am trying to understand how MadMapper outputs projects to the MiniMad, specifically regarding a quad's input position relative to media content. I'm trying to have multiple cues that all use the same quads, and each cue has a different video assigned to all quads (since MiniMad only supports a single video at a time). The input position for those quads can be different in each cue, as I'm trying to have different parts of the video show up in some inverted masks. But each time I export the project to my MiniMad, some of the cues have the correct quad input position, and others don't. If I export again, sometimes some cues get fixed while others get broken which is super confusing and annoying.
In order to better understand the behavior, I created a couple of simple experiment projects and got some confusing (to me at least!) results. Based on these experiments, I think I have discovered two limitations of the MiniMad:
- When you export to MiniMad, you are exporting media, not cues that use the media. If you have multiple cues with the same media used, the settings for only one cue are used. I'm not sure how MadMapper decides which cue's settings to use.
- Each quad can have a single input position for all media that is exported to MiniMad. You can't have one cue where the quad's input is in one position relative to its media, and another cue where the quad's input is in another position relative to its media. I'm not sure how MadMapper decides which input position to use for the quad if there are multiple cues with different input positions in the project.
Is this correct?
Here's what I did that leads me to this conclusion:
Setup: One 10-second video applied to a single quad. The quad has an inverted mask attached to the output. The quad was then dragged into three cues, and each cue had the quad's input positioned differently so for each cue, a different part of the video is visible in the mask. The project was then exported to a MiniMad.
Expected result: I expected the MiniMad to play three different cues for 10 seconds each. Each cue would then show a different part of the video in the (inverted) masked area.
Actual result: I seem to have gotten only a single cue. Basically, I see the same content restarting every 10 seconds (the length of the video) with the media in the same position.
So I thought that maybe the MiniMad only supported each video in a single cue. So ran another experiment:
Setup: I took my test video from the first experiment and made three versions of it, and then applied those to the three cues. So each cue has the same quad but with a different video, and in each cue the quad's input position is shifted to different parts of the input video for the cue. The project was then exported to a MiniMad.
Expected result: I expected the MiniMad to play three different cues for 10 seconds each. Each cue would then show a different part of the video.
Actual result: I got three cues, but it seems like the quad's input position is the same for all 3 cues. In my case, cue 3 looks correct. Cues 1 and 2 show the content as if their quad's input position was the same as it was set for the third cue, not the content that should have been visible based on their quad's input position.