Try normalizing the layer. Now everything placed on that layer will scale to fit the size, position, rotation, opacity, blend mode, map, mask, effect and everything you picked for that layer (it preserves the original ratio of course and is restricted by the first boundary met).
Media that are smaller will be enlarged. Media that is bigger than your target will be shrunk. Despite the differences in sizes of media placed on a normalized layer it will all appear to play in the same size window (or target area).
This saves having to have layers for each media size you need to use.

Once again Philippe and the Millumin team have done a great job.

This is not scaling each media separately but it does take care of many situations you would want this feature for.

In Terms of less or more processor power needed, i wouldn't care and i doubt it is much more then how it is now, to use more layers for the stuff that needs adjusting from one show to another...

I see it from a theater environment, where a play is touring and you need to adjust small things in every new venue... for example, i us millumin for a fully projection mapped set in a theater piece, where the set canvas can vary from venue to venue. At the moment i adjust my small things layer based, but in terms of show mode, id rather have 3-4 layers on screen instead of 20+ and no, there is no point to render things like a clock, notice board, steam from a kettle and other small things in the set in full resolution or embed it in the background(where the background now is only a image)... hope that makes sense...

simple workflow would be: Layer scale/pos has priority over clip scale/pos within the layer... the very same as in resolume...

I would find this confusing and I see it as going against the Millumin paradigm of scaling by layer. If this was to be implemented, there would have to be a mode you could set for any layer with independent media scaling.

This idea could save a lot of layers.

A compromise for now is to encode the media at different sizes if you want to change scale on a given layer. A second option is to create a composition that does the scaling for each media.

I'm not sure this idea is a good use of 123 votes.

Even if you could click on the clip & give it an X & Y position & a pixel width & height, safe to assume though it would be asking the processor to work exceptionally hard with regards to the size of each image/video in the library that you want to change?

This would be really awesome as I find with multiple layers when I do scroll down to the bottom (main video layer) the output on screen content tear & you have to wait a period of time until it has rectified itself!