Projection Mapping: Best Workflow for Sending Sections of Video File to Different Screens

[Starbuckle]

I'd just like to check that I've understood the best way to isolate portions of a video file to send to different screens. 

Imagine I had some video content, laid out like the above - with UVs or sections of unique content arranged inside a single video file. Assuming I want to map the three example content areas to three different screens, am I right in thinking that I need to do the following:

- Create three surfaces, each with the pixel dimensions of their respective content areas

- In the Display & Geometry tab, assign their corresponding surface in 'Video Surface'

- In Mode: choose Custom Geometry and enter the appropriate translation values for the surface to match up to the content area. 

Is this the best workflow or have I gone about things the wrong way?

 

And what if your content area is not rectangular? Would you make a surface that contains the content area as well as possible inside a rectangle and then do the rest with masking? 

Any help much appreciated :) Thanks 

[micpool]

In theory what you suggest would work, but it is overcomplicated and inefficient.

QLab requires a video cue for each surface, it's not like a true mapping program where a single video or image can be mapped directly to several  different output geometries.

So it is probably more sensible just to create 3 smaller image files 1 for each surface with just the content for that surface.  For stills your approach would probably give adequate performance, but for video, if you had a HD projector, you would be running 3 cues each playing back a full frame video, when all you need is 3 videos of much smaller pixel dimensions, and when crossfading you would need to run 6. If you have separate content for each surface you have far more control, and it's much simpler, and places far less strain on your computer.

Surfaces can be masked by dropping grey scale images into the mask well in the surface editor  e.g this would be a mask for a soft edged triangle

You may find this tutorial helpful

https://qlab.app/cookbook/masks-for-sail

Mic

[Starbuckle]

Ok, I get it. I was approaching things the way I would with mapping software like MadMapper and I see now that things work differently. 

And this would still be your recommended approach even if I had one video file, portions of which need to be sent to different screens, that contains animation that is continuous across the different screen portions? (like below)

I think you're saying that it's more efficient to invest time in prepping the content and splitting it up into its screens before you bring it into Qlab. This avoids asking Qlab to play back video content that doesn't even get seen. 

[micpool]

Ah, I may have misunderstood your original question. Do you have 6 display devices as opposed to a single projector and 6 screens within it's beam?

If so, if you have a 1 to 1 pixel relationship between  your content and the devices it is displayed on, then you can just put your multiple displays as screens on a single surface.

I don't have 6 displays available but in this screenshot  I am showing 4 partial screens placed on a 2160p sized surface.

if the relationship between your displays and your content is more complex, then the best solution is to Syphon out of QLab to MadMapper or similar and do the mapping in that.

This tutorial may be helpful 

https://qlab.app/cookbook/syphon-to-madmapper

Mic

[Starbuckle]

I'm talking about 6 screens from one projector. I was using 'screens' in the abstract as in a discrete physical surface onto which I want to project some unique content. 

Does your first piece of advice apply in this case?

On Tuesday, March 10, 2020 at 9:50:08 PM UTC, Starbuckle wrote:

[Richard Williamson]

If your content is always in sync (even if it doesn’t look like it’s from the same source) then I would render out a single video file with the content for all three screens on it, and with black everywhere else - this would be the least load on the machine and so the most reliable.

If you are after three areas which fire at different points then you would probably want three surfaces set up separately with separate files for each - although you could do a combination depending what is happening at different points of the show

Richard

--
Contact support anytime: sup...@figure53.com
Follow QLab on Twitter: https://twitter.com/QLabApp
User Group Code of Conduct: https://qlab.app/code-of-conduct/
---
You received this message because you are subscribed to the Google Groups "QLab" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qlab+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/qlab/8398a34d-d686-4e42-8ffb-109fe9e31576%40googlegroups.com.

[micpool]

On Wednesday, March 11, 2020 at 11:02:54 PM UTC, Starbuckle wrote:
> I'm talking about 6 screens from one projector. I was using 'screens' in the abstract as in a discrete physical surface onto which I want to project some unique content. 
>
>
> Does your first piece of advice apply in this case?

In that case yes, make separate videos for each surface, or use a single surface in QLab and Syphon to MadMapper or similar for mapping.

Mic.

[Starbuckle]

Thank you for all the brilliant advice :) 

On Tuesday, March 10, 2020 at 9:50:08 PM UTC, Starbuckle wrote: