I ran into a problem. For example, the Lower Thirds layer. I wanted to use push from left. But I can’t define the edges/margins. The push-in comes from the edges of the screen. Effectively the layer displays full width. Yes I know you could use images with transparency. But it doesn’t solve the boundaries issue for animations as it is implemented on the full width of the screen.
I would like to be able to set top, bottom, left and right off set for the layer so the contents only shows and animates within that area. So for example by setting a left offset of 100px the push-in would start from 100px in.
I think this feature should be on most layers that need push-in etc from inside the full screen area.
Thanks for your comments. Have you tried to use the Slideshow Source instead of the Slideshow layer? It allows you to use the slideshow as a source for any layer, so it’s easy to do 2-ups or any other layout.
And finally, every Source (and I know this isn’t the topic) should have a Keep-Alive-Flag, which overrides all other mimoLive performance settings. Especially for sources where no preview is necessary, this could help to avoid having an overdozed Multi-View which will be rendered for nothing. <3 <3
I wish you had asked this a few years ago when I was making custom layers for bonixTV/mimoLive!… I had a lot of ideas about where mimoLive was limiting production possibilities. I have a vague recollection that having multiple stacks would open possibilities (which I guess you do with Layer groups today) like CamTwist Studio had more than one vertical stacks of layers. Achim suggested multiple documents and syphon to glue them together at the time. I’ll try and think about what I wanted most over the next few days. Sometimes having the document dimensions dictate the layer dimensions is an issue but I forget why, again the workaround is multiple documents and syphon.
I’m not that good with QC to make everyrhing I need happen, so: no. The Syphon work around (which I call left of below method) causes lots of unnecessary never directly used generations/over head. It does work for some tiny things, by the cost of 4444-like temporary load over localhost. It’s never the thing of only one use case. Based on several action, I’d currently need up to 16 parallel documents, all based on the same video inputs… memory, GPU and CPU bottle necks are responsible, why this is impossible for me to embed all Syphons inside a main document. Maybe Apple Silicon makes this possible in about 10 years or so. And up to this, I just had combined something on screen, nothing else.
Yes, it’s easy and already available in some ways. This thread was about what every layer should support. Filters are causing generations. When my eyes do not trick me, also a loss in quality, due to prerendering and an assumed syphon action inside the filter chain/color conversion issue/or other phenomenas already and often discussed here… Best result was - up til now - inside a layer.
yea i realised after i wrote that that you were talking about adding flip horizontal to every layer. also realised Placer later already has it. filters do not use syphon. mimoLive uses a stack of Quartz Composer compositions and any given layer uses the output of the layer below it as the “background” image to its renderings. so will show through the background, while overs will cover it over entirely.
filters are just like layers, i that they are also QC compositions but applied instead to video “sources” the same way a layer applies it’s composition to the layer below but it inherently treats the “source” as its renderer input. there’s no Syphone involved in mimoLive native effects at all and in fact BonixTV, the predecessor of mimoLive predated Syphon’s original release i think. you could make an effect that used Syphon in some way in QC though.
i really like this idea of layers as functions and was pushing on something in the same direction a few years ago. Functional programming is based o. the idea of lamda calculus and the Quartz Composer basis of mimolive layers and effects are all functional programming (though no lambdas).
i used to think the ability to “patch” the output fro. one layer to the input of another would be excellent. when i was building a vision mixer for speaker and slides i needed to make it adaptable for 4:3 and 16:9 camera and slide aspect ratios. and really needed to crop the camera vision from landscape to portrait to best use screen realestate.
i was programming it in QC and it got really complicated and did my head in making it super adaptable to different use cases and being as dynamic as i wanted it. if i could have simply chained a couple of placer layers together i think it would have made it much easier to implement, i’d still have needed to code for the 3D aspects i wanted but not get so bogged down in the geometry of the incoming signals.
so yeah function layers that can be applied to existing layers (like adjustment layers in photoshop for example) or patching cables to send output to inputs between layers would be cool. adjustment layers wouldn’t need Syphon, patching would need Syphon but is more powerful i think than simple adjustment layers. functions that could be applied to inputs would be like effects already are i guess?
yeah i realise that. that’s where adjustment layers to add extra features to any layer could be a good tool. save all layers bloating with extra features.
flip and transform “layers below” is easy to write in QC for achim. it could be a new layer in layers list or it could be in a set of new adjustment layers that look different, so as to distinguish them from standard layers. and they could be thinner and maybe even collapse to sit four to a row, to save screen realestate.
Thinking about this some more, patching needn’t require use of Syphon (not that Syphon is so bad) if mimoLive can use the output of one QC composition and route it to the “Layers Below” (background) layer of any other composition.
I’m not sure the best way this would be indicated in the mimoLive UI. One possibility is to have a check box on all layers that can output the layer result to “Patches” and “Patches” would appear as a folder of sources o the “patch” type in the Sources column on left of screen (where Media and Cameras etc are). Another way would be to do what reason does and have actual visual cables (see skeuomorphism) joining any layers the have data flows between them when in “patching” mode.
