@ray said:

have an rtx 4090 and an rtx 3070 (externally mounted) and would like to use these cards also

 Unfortunately this type of multi card config isn't officially supported. Primarily because we can't possible test every configuration, so we don't have 'best practices' outlined.
With this type of config you will undoubtedly run into a bottle neck somewhere, although it may or may not impact your project. Generally 1 primary dedicated gpu will be assigned to Isadora by Windows, and this is where vide textures etc are loaded to for processing (fx, mixing, mapping etc) then the finished texture will be sent to the GPU physically connected to the output device (projector etc).
So, transfer speed to these output cards is often where performance hits are seen.
As far as the stage IDs/Numbers. Isadora uses the connected displays as Windows reports them, so if windows moves anything, Isadora will see them differently.
I know others have used edid hdmi emulators, to lock device to specific positions.
I think this post has a fair amount of info on this: https://community.troikatronix...

@mclefferson

We have Art Net support via the Art Net actors, but do not have sACN support at this time.
If sACN is an absolute requirement, it could be handled via Pythoner, but would take some custom coding.

As far as VLANs go, there isn't an Isadora-specific requirement to separate Art-Net and sACN, or to put all DMX-over-IP traffic on a single VLAN. They can coexist on the same lighting/show-control VLAN, and for a relatively small network that's probably the simplest setup.

@ray said:

I have an rtx 4090 and an rtx 3070 (externally mounted) and would like to use these cards also.

 @DusX I'm going to leave the multi-card explanation to you

@dbini Thanks! I'll give it a whirl 

@michel This is perfect thank you! 

@ray

Here is the Isadora Duncan game using 3D particles - use the 's' and the '+', '-', keyboard characters to interact with it. Download the Isadora files here.

@ray I always guessed it was for Isadora Duncan.

@as1ama1and looping the video through a set of 3 shader actors does quite a good job. 

Freeze shader: grabs a frame at the start of each cycle and holds it. It does this through a feedback loop (output → pass-through shader → back into video in 2), because Isadora's GLSL actor has no persistent buffer.
Pass shader: an empty shader that just closes the loop.
Mix shader: lays the frozen frame over the live image and fades it out over time. blend mode decides whether the clone shows only where it's darker or lighter than the live image, so the live video is never covered.
A Wave Generator (sawtooth, 0–100) drives the clock input. One cycle equals one snapshot, so freq sets the cycle time.

Here's the patch: echo trails

Best Michel

@as1ama1and

try feeding the video into an Effects Mixer on channel 1 and running the output through a Video Delay into channel 2. By adding a Video Mixer into the loop you can control how much of the output feeds back into the input. Play with different blend modes until you get the effect that you like.

Howdy! 

I am trying to recreate a live version of the picture below. I made that picture by blending a few layers using a darken blend. For the live version, I want live feed with the stage being a tv. 

The idea is a dancer is in front of the camera standing still and then their small movements create the same effect which accumulates. Though at a certain point I'd want older layers to dissolve away completely. I know there are blend modes in the effect mixer actor but I don't know how to go about setting the rest up. Any help would be greatly appreciated!!