Question about Edge blending. masking etc please help.

[m83nyc]

So I am kind of new to all this but learning fast. I understand most of VJing and resolume the part that gets me and prob most is the slicing, masking etc. I understand how it works and have tested some things. I do have a question about a setup I want to do.

So my gear is a MBP with a triplehead2go Digital/ I run two projectors out of this that convert to VGA setup works fine I am able to extend the displays. Here is my goal with my stage setup ....... I basically want to have a big back drop then objects in front that I can map/mask too. So in more simple terms I guess i am looking to edge blend to projectors for a bigger res and presence but also be able to map or mask to objects while edging.

I want to be able to edge bland project to a big screen then put objects in front of that projections I can mask,slice whatever you call this. Please advise me what to do ? I attached a crappy pic of what I am trying to do this will help me a lot. PS I have madmapper if I need it.

[Joris]

I would look into virtual screens for this one.

Assuming your projectors are each doing 1024x768, you can make both your composition and one big virtual screen at 2048x768. Rename the screen to "Masking" or something similar. You can use the comp to do your positioning, and the virtual screen to do the masking.

Then you make another screen at 2048x768 which is your dual output from your Trips. In this screen you create two slices, each setting their input to the "Masking" screen. Then you can blend them how you would normally do an edge blend.

phpBB [video]

[Oaktown]

@ m83nyc, the setup you're describing can be challenging because it sounds like you will have different focal planes so you'll have to decide which of the planes you want to be in focus. Also, if any of the 3D objects you want to map are in line with the blend for the background, it will need to be blended separately so you'll need to create two slices for that one object.

Take a look at this little simulation where both background and object 4 are different blends:

phpBB [video]

[Joris]

if any of the 3D objects you want to map are in line with the blend for the background, it will need to be blended separately so you'll need to create two slices for that one object.

That's why I'm recommending to use a Virtual Screen for the mapping/masking, and create the blend using this Virtual Screen as an input (instead of the comp).

[m83nyc]

Wow thanks for the input !!! this makes sense but dose not but I do understand. You guys seem to know exactly what I am looking to do here. I guess in a more scaled out way you are saying that ....

Virtual screen 1 = my canvas for mapping and masking. ? This makes total sense very straight forward.
Questions for this one is = how can assign one of my layers of videos in resolume to output to this Virtual screen objects as a whole? like in the video oaktown posted the same video layer is going on the four 3d ball objects ?
or how would I assign each mask/map 3d ball thing its own layer video ?


Virtual screen 2 = would be for the edge blend of both projectors. Get this and this video is key thank you. How would a assign a video layer to this well main the background behind the 3d objects ?



Is the way I am doing this correct for this type of stage setup i want to do or is there a better way ?

Will Madmapper make any of this easier ?




Guys I seriously really appreciate your help ! Very thankful

[Oaktown]

@ Joris, I'm sure there is a good reason behind the use of the Virtual Device the way you describe it but I'm not sure I see it so can you help me understand?

@ m83nyc, regardless of your approach (direct or using virtual device(s)), Since your objects will be in front of your background in real life, the projectors won't line up in both focal planes so I believe you will need to blend at least two pairs of slices. One pair to make your background and one pair to make your object map that's in line with the projectors' blend area.

[Joris]

The reason I'd use a virtual screen is that you can do all the mapping/warping/positioning in the virtual screen, using as many slices as you like. The actual output only uses two slices to do the edge blend. It doesn't matter to the blend screen what you do in the virtual screen, so you can easily move slices across the blend.

The caveat of this is that it assumes your objects are all on the same focal plane, like painted or thin object attached to the wall. The moment your objects extrude significantly from your background, you're up the creek, and there's no other way then to treat the object crossing the blend as a separate blend. This is just physics, as the projector overlap will be different on each depth

[Joris]

Virtual screen 1 = my canvas for mapping and masking. ? This makes total sense very straight forward.
Questions for this one is = how can assign one of my layers of videos in resolume to output to this Virtual screen objects as a whole? like in the video oaktown posted the same video layer is going on the four 3d ball objects ?
or how would I assign each mask/map 3d ball thing its own layer video ?

There's different ways to do this.

Oaktown will route the layer with the ball to four slices, hide it from the comp and then position and warp the slices in the output. This tends to be generally easier and quicker to set up.

I will use an effect like Tile, Iterate, Slide or Hive8's 4 Screen Duplicator effect to position the ball in the composition four times, and then uses slices to warp that exactly to fit. This tends to give you more freedom during performance, for instance when you realise half way through you want to add more layers.

Each technique can easily be adjusted to use separate balls for each object, just add a layer for each ball.

How would a assign a video layer to this well main the background behind the 3d objects ?

The idea is that you use a single input containing both background and balls, and blend it as a whole. As mentioned above, whether or not this will work depends on how far your objects are from your background.

[Zoltán]

The idea is that you use a single input containing both background and balls, and blend it as a whole. As mentioned above, whether or not this will work depends on how far your objects are from your background.

if you happen to be projecting from the same center position with the two projectors and use lens shift to left and right you can get away with only one blend, this is a special case.

also think about the shadows you'll have on the background from the 3d balls. depending on their distance from the background, and their distance from the others, you may be able to do a cross projection -left side from the right and right side from the left- to eliminate the shadows.
or maybe you would be able to find an angle from above(or below) with the projectors, where you can project to the background without the balls getting in the way while covering the balls also.
of course you would lose brightness with these last two options.

[Oaktown]

One additional note re. blending strategy. To economize resources, I think people should really get used to sizing their comps to match their actual projection area rather that creating comps that are the sum of the projector outputs.

@ m83nyc, since you're pretty new to blending, let me explain what I mean. If you have 2 XGA projectors (1024x768) that you're blending on a screen that's 9.5M x 4M there would be two main approaches:

• Option 1: You create a 2048x768 comp and adjust the slices' input area on the fly to match your projection area knowing that you'll have a 224x768 pixel area that won't be used so if you center your output, you'll end up with 112x768 unused areas on L & R.
  
  Option 2: You calculate the pixel size of your projection area which in this case would be (768/4*9.5) x 768 or and create a 1824x768 comp knowing that your blend area will be 224 pixels wide.

As far as I'm concerned option 2 has a lot of advantage such as:

• - comp size is smaller which means it'll run faster and the difference can be tremendous when you're working with large blends with lots of HD projectors
  - it gives you the ability to create artwork with the right width/height relationship which is very important if you work with a team
  - you can see the actual output in the preview window when you're in show mode