Resolume Forum

The reason why this works in v5 is that in v5 the transition's phase actually affects the layer mixer's opacity. We had to change this and turn the transition into a mixer on it's own because we have different 'kinds' of mixers that were not unitable otherwise and would've never resulted into a correctly composited image. I don't know if you remember, but we've had some 'pops' in v5 when alpha transitions were involved. We had to change this to fix that.

Having said that, i do think that the multiply mixer is currently broken.

I've been working a lot with premultiplied colors lately because i discovered that a lot of our mixers/effects/generators weren't handling premultiplied colors properly. An issue that pops up every now and then is where a color's transparency is considered as part of the color itself, while in fact it's not related to color at all. In terms of the multiply mixer, it's currently doing a naive component wise multiplication (f(a,b) = ab if you will). The problem here is that it's not respecting b's opacity to be a factor of the function.
I think what you guys instead want is for transparency in the destination texture to be treated as 'there's nothing here to multiply with' rather than 'there's nothing here, so multiply with the premultiplication color (which happens to be black)'

When Joris gets back i'll discuss the options with him. We'll need to make a choice if we want to modify the current multiply behaviour or if we should create a new one.

Menno wrote:I think what you guys instead want is for transparency in the destination texture to be treated as 'there's nothing here to multiply with' rather than 'there's nothing here, so multiply with the premultiplication color (which happens to be black)'

Correct!

That would mean that using the same two files, if I adjust the opacity of the mutiplied clip to 50%, I would get the following result (which is what I get in both Adobe Photoshop and Dataton Watchout): instead of:

You dont want alpha to be multiplied at all right? Output should be b and not a?

Correct, b not a!

And at 5% clip opacity I would get this: All that being said, the ability to multiply with just alpha where a=1 create full transparency on top of any color and a=0.5 create 50% transparency would be a great effect on its own especially with groups!

I'm so glad to read these new post here!!!!!!
I agree with Oaktown in everrything he said that's exactly my point.

Best,

Davide

If you haven't yet, take a look at this thread and download 6.1.0 which has the corrected multiply blend mode.

THANK YOU!!!
I was aware and I'm so glad it has been fixed!!!!!

Thanks again to you and to all the resolume team!!!

Davide

Hi there,
I've just seen the problem has been fixed when you use the general blend mode but not when you use the clip blend mode.
Let's say I want to use on layer 3 screen blend mode and I'm forced to have a clip on layer 3 that requires a different blend mode, let's say multiply.
If I tweak the blend mode of the clip changing it from layer determined to multiply when I play that clip
I still see the jump.
Please fix that part as well!!!
Having the chance to bound the blend mode to the clip is a very USEFUL feature that prevents you to mess the projection when you are doing things very fast during a performance.

Please let me know what and when you'll be able to fix it.

Best,

Davide

davide93 wrote: Wed Oct 10, 2018 23:05 If I tweak the blend mode of the clip changing it from layer determined to multiply when I play that clip
I still see the jump.

This is a different issue we're working on, and affects all blend modes not only the Multiply. (#11504)
What happens is that the layer blend switches to the Clip blend around 50% in the transition.

Yes, indeed after I reported the issue I've seen it happens with every blend mode!!!
Please can you tell me when do you think the fix will come?
Thank you so much in advance Zoltán!!!