Resolume Forum

I have a gig coming up where I need to comp 2 FHD feeds into one 4k renderer to be sent to a screen processor.
My source is 1 FHD camera feed, one group set to FHD will be clean, minimal processing, and the other also set to FHD an effected camera.
My plan was to use a video router set to the cameraIn group, and put that in my proccessing group. But the router defaults to the comp resolution, and when set to FHD I have to scale and position the texture. And then my plan was to use clip effects above that in the group, but when I do that the effect is also fullscreen QHD res
This isn't expected behaviour to me. I'd expect setting a video router to FHD and the group to FHD would mean I am processing at FDH to. The comp size should be immaterial until the the groups are rendered into the comp and any FX would then be QHD.
I can of course just set the comp at FHD and advanced output the slices to the quardrants in the output window, but then the composition is totally irrelevant, except for lumiverses, and even then I would really want the lumiverse to be a group at a probably much lower resolution and that just be previewed in the comp.
I can see the reason for the comp and comp fx, I cant see the reason that groups arent determinded by their content or their set resolutions.
It seems to be bad for performance (rendering unused pixels) and usability, you can't setup a good over view using the comp window.
Can someone explain it to me, what is the render pipeline here?

I can't explain the rationale for it but the same issue made life difficult for me recently too.
Using camera input as a layer to be routed in a very large composition meant that the routed input was the size of the comp.
I was using effects and having such a large texture was expoensive on the GPU. I ended up having to use individual inputs instead of routing and it was a camera heavy show so re-adding all the inputs in changeover time at a festival was a ball ache.

Cropping and positoning the routed inputs was a PITA and resulted in images with incorrect aspect ratios which I didn't have time to investigate.

I've been thinking about feature requesting some kind of Advanced Input set-up. Something along the lines of the way Qtake software works with multiple inputs could be good, might even open up Resolume to another market for film and TV production playback if you could do frame accurate recording of multiple inputs and have the clips automatically fill in a comp etc

I'm glad its not just something I have misunderstood, I hope the devs might come in and give us some more info on how to deal with this.
I've gone with comp size = camera input size for now, and using an advanced output to render to the screen processor, but really the comp becomes a useless render target now, but at least all your fx aren't all rendering useless pixels.
Thankfully I'm just using one input and getting a premixed signal Although I was thinking about getting a quad multiplexed 4k input and cropping to quadrants, but I think I might fall foul of resolution again then. Seems odd to set resolutions and then them not be applied. ( for this I probably would have used wire so at least I could set the output res and the crop manually, I hope!)

viewtopic.php?t=19951 Seems to be related to this. I haven't had to check it in my scenario yet though

Scratchpole wrote: Thu Jun 13, 2024 08:46 have the clips automatically fill in a comp etc

Layer auto size options help.

Try duplicating camera clips instead of layer routers, they should not add more load than a single camera clip instance.
We have a ticket to improve Video router resolutions.

Glad you are planning to improve Video Router resolutions, I hope it gets a full rework.

Currently when using auto size on a router it only changes the scale to two different options for me (full comp size and input size, after a few clicks it stops changing it at all when it should be cycling through the options. I just added another router to my comp to test this again now and it's refusing to auto scale it at all. Layer routers are not working consistently.

The reason I use routers instead of camera clips is for programming efficiency not load. For example I was using 50 routers from one camera input on my last show, re-adding them all and scalling them all correctly within a multi screen/multi output composition would be nigh on impossible in a changeover at a festival.