Video stage & output resolution, scaling best pratices?

[Will]

Hi all,

I was wondering if there is recommended best practice for setting up your video stage and output routing when working with a display with a higher resolution than all your videos.

I searched the QLab documentation and this group for recommendations and didn't find any. Sorry if this has already been covered and I just missed it.

For example: I am programming QLab to run videos for a show, and the theater has a Barco F80-4K12 projector, which has a resolution of 3840x2400. However all of the videos I was given are 1920x1080. We will only be using 1 projector.

As far as I know there are 4 ways to set up QLab's video settings in this scenerio (the questions I'm specifically interested in getting answered are at the end):

I usually do option 1 or 2, but I'm curious if there is a better way.

Option 1

Set your stage and your output routing to be the same resolution as the display's native resolution.

Let QLab do all scaling and when you run a video cue in full screen, let QLab decide the best way to fill the display.

Option 2

Output routing: 1920x1080, Scaling mode: fit

Stage resolution: 1920x1080

Option 3

Output routing: Native projector resolution (in this case 3840x2400)

Stage resolution: 1920x1080

Warp your region to be the correct 16:9 aspect ratio

This is the most annoying option when dealing with a display with a different aspect ratio. As a test I tried to do this on my macbook display (3024x1964, a 1.54:1 aspect ratio), I ended up using an aspect ratio calculator to get the correct coordinates to set the warp corners.

Option 4

Change the display resolution in the mac display settings, to trick QLab into thinking the projector is a different resolution then it really is.

Let the projector do all the scaling calculations.

Questions I'm specifically interested in getting answered:

Note: I understand the answers to all of these questions might be different for a show with mixed resolution videos, so I'm just curious about a case where all your videos are 1 resolution and the display is a different resolution.

A) Is one of those options recommended over the others?

B) For option 4, would the projector actually be doing the scaling?

Or would the mac would be doing all the scaling, just outside of QLab (as long as the video cue is set to fill the screen)?

I'm not sure if the mac just scales it's output or if it renegotiates it's HDMI handshake and tells the projector that the mac is 1080p device.

I've never tried this option so I'm not sure if this would even work as I'm imaging it would though.

C) Would any of the options result in a better quality final image?

I'm defining "better quality" as the output is the most accurate to the source media after scaling. Also, smooth playback and least amount of artifacts.

D) Are some options more efficient for the computer, or easier for QLab to handle?

This is especially relevant for older macs.

E) Or is there no difference between all those options?

My dad is a sound engineer, and many years ago when he first taught me how to use a digital sound board he said, "The great thing about digital boards is that there is more than one way to do the same thing. The worst thing about about digital boards is that there is more than one way to do the same thing!"

I'm wondering if that applies here as well, there is more than way way to do video scaling in QLab, but they are all achieve the same result and it makes no difference which one you chose?

Thanks in advance!

[Chris Ashworth]

Hiya Will,

I recommend Option 2.

Option 1 is okay, Option 3 is a lot of extra work with no benefit, Option 4 might be okay but I believe the projector itself would not be doing the scaling and I am not 100% sure that it wouldn’t have unexpected consequences.

In terms of image quality, I think all of these theoretically will have the same result.

In terms of ease of setup, Option 2 is by far the best because it was designed for this purpose, to make it easy.

In terms of efficiency, Option 2 is also the best for the following reason:

The video rendering engine operates in two “passes”. The first pass is to render each video cue to a rectangular stage raster (set of pixels). The second pass is to render regions of that stage into various outputs. So if you are rendering your smaller video files to a stage that matches them exactly, each video cue will do the minimal amount of work to be rendered onto the stage, and then the scaling will happen exactly one time from the stage to the final output.

In contrast, if you render all the video cues to a larger stage, the scaling has to happen each time, for as many video cues as you run.

Best,

Chris

[Will]

Thanks! This was exactly the information I was looking for. That all makes a lot of sense.

The other reason I often use option 2 is because my stage grid matches the resolution of my videos, which is super helpful for lens shifting and zooming the projector. Glad to know it was also the best option for all the reasons you mentioned.

Can I recommend you put this in video documentation on the Qlab website?

Thanks again!

Will

[jimsta...@zoho.com]

Thanks, Chris.

  Another way it makes sense to me...

Running the main and backup systems through switchers (we normally carry two Roland V1HD) which are 1080x1960.

  A recent show required 4 HDMI outs through 4 V1HDs, feeding a 2160x12600 wall and the usual 1080x1960 for risers on stage.

3 of the 1080s made up the wall and were scaled through the in-house Barco E2 to 2160x4200. Looked fantastic!

The only quirk was the V1-HD doesn't have a sync in. With moving content stretched across three outputs, some transitions between the wall sections were imperfect.

Doubt anyone but our crew noticed.

Jim