Advanced festival LED mapping

[flexcube]

Hi Resolume

I'm slowly getting into the more advanced settings of resolume.
A friend of mine is going on a small tour with a DJ shortly and I have some questions
on how to set up Resolume and my graphics effectively on complicated LED screens.
I am wondering how professional VJ's are approaching these kind of setups.

As example for my questions I'll use this setup from the last edition of dancevalley. (Cause this setup has most of the challenges you will face on tour) Wonderfull stage, visuals by the guys from Eyesupply.

Dancevalley:

Stage in action:

phpBB [video]

The screen on the Dance Valley setup is something like this:
- 1. Screen middle high
- 2. Screen behind DJ
- 3. Screen DJ Booth
- 4. Screens left and right

and the 3 long strips that go all across the stage


So there are 2 situations as a VJ for a festival.


1. You VJ for a stage the whole day. Way in advance you get send the LED template by the company.

How would you approach this setup?

The easy part:
I would just put screen 1 till 4 each on a different layer and load the corresponding content in to that layer.
Route the layers to Layer output only. This method can be used to show any content you want with a bit of resizing. because all these screens have the average 16:9 shape and are just rectangle screens.

Only downside is that effects don't really work anymore due to layer routing bypassing the composition.

The hard part:

But how would you put your content on the large strips that go across the stage, while making use of the strip shapes. For example you want an arrow that follows the path of the outer strips. Going from low left up, then to the right and then going down at the right side.
Is this something that is always custom made in after effects with use of the LED template? It seems like the graphics follow the curves of the screens and that it is not just a graphic with masked out parts.


2. Touring with a DJ
The reason I am asking is because when you are touring and therefore already have the content that has to be put on screens like this and your content is not matched with the template. How would a professional approach this LED setup in resolume? Would you just throw your content on the screens, and not really worry too much about the shapes of the screen?


What it all comes down to, I am a bit lost about what has to be customly made with the LED template in advance and what can be done inside of resolume while on tour.
When you are on tour, you often don't have the time to create custom screen setups. So how would you approach the situation as a VJ for the whole day and as a VJ that tours with the DJ?

[Joris]

Very good question! You're close. Most VJ teams I've seen operate on this level approach both options the same way.

The trick is to create what's called an input map in your composition. Using a 1920x1080 comp, you recreate the layout of the stage. Using a combination of layer positioning and clever use of effects, you can have layers which show 1080p content scaled down and cropped only on certain screens, as well have content which spans the entire output. The advantage is that you can use all your existing content in combination which custom content for that show.

In the advanced output, there is no need to use Layer to Slice routing. Since everything is positioned in the composition, all you have to do is make the necessary slices to adjust the output of your comp to what the LED processor expects.

When you are operating the stage for the entire day, you can spend some extra time creating custom content. This content is created to fit exactly to the input map, so you can use it in combination with your existing library. This is why I keep hammering everyone not to use Layer to Slice routing unless you absolutely have to. Layer to Slice routing forces you to either use only custom content, or only existing. You can't mix and match anymore.

When you are touring, you usually have a generic input map comp, that takes into account most common stage setups ( main screen, booth, side screens, wings ) as basic blocks. You can then reroute these to fit the LED processor setup in the advanced output. This can get tricky with for instance lyrics that won't fit in a certain aspect ratio, but you can also prepare for this by using prerendered lyric blocks for different common ratios.

Although in my experience, most guest VJs don't know much about pixel mapping, or even care about it. They usually have their 1080p output scaled in to fit the pixel map via a capture device. Since the 1080p output is just a clip, the person that runs the stage can send it to the different screen on the stage by triggering it in the desired layer.

[Oaktown]

To add to Joris' very detailed response, I would use a larger than 1920x1080 comp since you have strips that go accross the stage. So for instance, I would make my composition 1920x1580 so that I can have:

• 5 @ 100x1920 strips at the top since it look like you have five independent strips (Note you could use just one if all your strips have the same content)
  1 @ 1920x1080 at the bottom split in 4 @ 960x540 which you'll scale back (to 200% if your screens are 1920x1080) to fit your screens on stage

Then you can:

• have presized layers with X & Y offsets to match your composition map
  have 9 slices (or whatever you need) in the advanced output to assign all the various content to the screens

[Joris]

To give you an idea of what I mean, here's an quick mock-up of a pixel map for this stage. Note it's 1920 wide, but takes into account the aspect ratio of the stage, so it's not fully covering 1080 pixels in height.

It wouldn't really work for this stage, you'll notice I have to smuggle a bit to get the wings that face inwards in there, but I think it gives the idea that I want to get across.

If you want to get pixel perfect, I think this actual stage was over 4K in outputs, and Eyesupply had to do some crazy sorcery and dark voodoo juju in their input maps. Their ability to make it happen after the rest has gone home to cry is why they get do those huge stages in the first place.

[Oaktown]

Hey Joris, I totally understand your approach but I'm always trying to minimize real estate to maximize fps in my comps so I tend to have as little unused space as I can which is why I was giving another approach.

(as a note I realized I wrote 2420x1080 when I meant to write 1920x1580 but I corrected the post already)

Here are two quick mock-up of the alternative:

1920x1580 as per my last post

and 2920x1580 if one wanted to use vertical mapping for the 5 bands as well

Then if I needed the visual aid I would create an additional screen (not assigned to any output) with slices to create the exact layout in advanced output to get a sense of the final output on the stage.

[cosmowe]

arrrg guys .......so I will now close photoshop now!...you where faster.

[chenthemagician]

Greetings, I'll try to add my two cents to this convo..usually it goes a bit more technical than that...to be fully pixel accurate you have take into account the size led screen, the actually distances between the screens themselves ( so that the image doesn't appear disjointed) this can be calculated by the physical width and pixel width of the tiles used, and once done on paper, you can create your composition to full stage scale with pixel accuracy to the level of about 98%....and then you have your slices put in the appropriate positions to fit the physical location as it would appear on the stage...this is a regular practice which allows me to have full flexibility of using my layers to then put content how I see fit on the output...

I did a show that was two physical output from the machine to a LED screen split into 28 different stripes plus the screen in the back and each was set up as a slice and spaced accordingly and a DJ FACE LED screen...so any image looked like it looked on my composition output window... Check it out below...

Advice: Get your output pixel accurate...then have fun after...

much love from Jamaica

[chenthemagician]

So...re-reading the whole thing...I think we are saying all the same thing...lol...

[Joris]

Everyone has their own technique and each has benefits and drawbacks. The reason I tend to go with input map at the expense of comp real estate and processing power, is that I can then still use my entire library of content. This is not so important on bespoke theatre shows or corporate gigs. But it's a must on VJ gigs, where there is never enough budget to run the entire show on custom content.

Using an input map like Chen and I describe, you can still play content 'fullscreen' over the entire comp, and it will look correct on the stage. Elements with movement will leave one screen and appear on another, as if the stage is one big screen with a few gaps in it. This means I can 1. use any custom content, 2. place my entire library of existing content on specific screens using layer positioning, as well as 3. use my entire library 'fullscreen' without the need to re-render. And I'm able to mix and match in realtime between those 3 options.

In my experience, it's still very rare to get 6mm LED screens at very high definition resolutions. So most stages can be run using a laptop and a 1080p comp, with 5 to 6 active layers and effects playing at around 30 fps. Of course, when you get to run main stage at EDC Vegas or Ultra Japan, this becomes a different story, but by that time you probably got some other tricks up your sleeve anyway.

[Oaktown]

Using an input map like Chen and I describe, you can still play content 'fullscreen' over the entire comp, and it will look correct on the stage. Elements with movement will leave one screen and appear on another, as if the stage is one big screen with a few gaps in it. This means I can 1. use any custom content, 2. place my entire library of existing content on specific screens using layer positioning, as well as 3. use my entire library 'fullscreen' without the need to re-render. And I'm able to mix and match in realtime between those 3 options..

Point well taken! I don't VJ so this is a good thing for me to keep in mind. In my line of work, all the media is tailored to the spec of my composition so I end up with the same effect (full screen on multiple screens with negative space) but there is no way to change things in real time like you would need to in this type of gig.