today i learned i’m the 100th reply posted on this topic!
3 Likes
Watch out, in some different softwares it would be reply 101 
Edit: Was a bug, fixed after restart…
Today I learned, that Output settings are not saved within, when loading an action preset for the timeline. Not great when there are already 40 actions, that now need manually changed the Background from BG Clip to black.
only if it counts the initial post as a reply 
2 Likes
Just today I learned, that when you move/duplicate a Blur node (even with keyframes) from one resolution into another, the values of the keys are getting multiplied with the new size 
1 Like
Yeah I actually hate this behavior and wish there was a way to turn it off. Very unhelpful sometimes.
2 Likes
It’s the same for a variety of parameters. Text changes the scale if you pipe a different aspect ratio into it. The blur in action changes if you swap the input to a different rez/aspect.
Minor time saving - but regardless.
I’ve seen in a few of the tutorials when setting up extended bicubic to remap things like gaps, that you set the shot up by moving the UVs and then copy to the vertices to get the initial setup for the part of the shot you’re working on. In fact if you Shift+U (or select Move UV+Vertices) in the tool menu, it moves both of them together. So you can stay in results view and move them into the place, then switch to either view to make your adjustments, no copying needed.
4 Likes
I just wanted to say how much I love this topic 
7 Likes
Another one from recent work, speaking of extended bicubics. If you have a seam you want to get rid of entirely, not just shrink, cross over the control points, meaning drag them beyond their next neighbor. Requires a bit of finesse as extreme moves can tear the image and leave artifacts. But it will save some clone paint, and is nice and neat.
It’s like sliding one plate underneath the other.
1 Like
I would really like this changed! I made a improvement request for it a while ago. FI-02979 - Support for displays running >60hz
3 Likes
Just ran into this same issue, upvoted!
Today I tried (and succeeded) sharing a mask between multiple selectives in the image node to avoid copying it. Copies of masks are a headache once you start tweaking them.
7 Likes
You can do this with gmasks in action as well, and also with shadows.
1 Like
In the selectives menu you can invert gmask. So one mask connected to two selectives can let you affect the outside and the inside without duplicating the mask.
4 Likes
Right, also can use two Mastergrade nodes in the same selective and set one to outside.
In my use case I had one mask for an object, but needed different diamond keyers, as one was for the bright, the other for the shadow side of the object.
1 Like
@allklier Do you have a way of bringing in multiple mattes into effects/image
Do you have a but more detail on context?
Image can refer to multiple mattes, but only when used in BFX or Batch. In the TLFX image has only a single media front input with matching matte.
I’ve experimented with using BFX to assemble complex mattes and then passing them through TLFX for similar reasons. It almost works. You can pass on the matte via the back node in BFX and Image can see it. But there is a side effect of this matte being premulted on the fill channels. Haven’t found a toggle to leaving the fill alone. May exist, haven’t looked that hard.
In the image node under the selective, under the control menu there is a section for ‘Input Matte’ and you can select the media input index. In TLFX it’s fixed to 0, but in other environments you can change it.
I would like access to different mattes that were created into image on the timeline. I know you can do this in bfx with multiple inputs. I was playing around just now trying to generate a multichannel clip to use in effects/image. I almost got what I wanted.
Ideally, it would be great to be able to drag media into the image schematic. Feature request on its way 
1 Like
Right, the Mistika timeline maybe a pattern for that, where you can have multiple inputs by referencing tracks beneath you by relative index. So media 0 is the track right underneath, media 1 is two tracks below, etc.
I think dragging in media directly is architecturally a bit weird for a timeline environment. But using tracks to manage and supply multiple sources is consistent.
Maybe a concept like the matte container. Have a multichannel or a bunch of layers in the container. That way we dont have a deep number of layers in the main timeline.
1 Like