Best practice for aut0-generating variations on video as elements swapped out


#1

GOAL:
Create template allowing a new batch of videos to be auto-created each time I switch out text and a few images. The image/text swap bit is already worked out by swapping .txt and .png image with same generic name in project folder.

ISSUES:
I’m under the impression this is trivial if the only difference between them is the render settings since I can use separate “take” per output setting, and tell it to render all takes. Will be digging into that when I’ve got a break tomorrow, but realizing now some of the templates are more complex.

In most of the templates, it’s not just the render settings that change, but rather everything is resized and moved around to best fit that aspect ratio. The ELEMENTS are always the same, but they are configured differently.

I was doing this in AE by setting up a different precomp within the same composition for each version where all the precomps draw on the same elements, then creating a render queue for each of the precomps.

Need to do same thing here where I swap elements once, and it renders out different versions where each uses those elements in different ways. I could just make different projects for each, but then I’d lose the ability to hit render once and have it do batch… plus I’d lose the ability to make a change once for something else like a material color and have it propagate across the various output formats. That last bit is important.

Is this still doable via “takes”, or is there something additional I should be looking into?


#2

yes this is doable with Takes. You would have a main parent take, then subsequent child takes would contain the variation.


#3

OK, Thx. Will dig in next time I’ve got a few hours to work out all the details. I see myself creating a lot more of these templates in the future, so it will definitely be worth the time to learn how to automate efficiently.


#4

Finishing up tuts on takes now, and looks super useful.

Wondering now what the best practice is for file management for the renders and text/image swapping of the takes.

Basically, I need to render out a bunch of videos where each batch contains mp4s rendered at various aspect ratios. The difference between each batch is that the text and images have been swapped out.

Ultimately, the output should be a folder named for the title of that batch, and each video should be output within that folder with a name that reflects the batch title and aspect ratio (so: CatVideo_1920x1080.mp4 within Desktop/Catvideo folder, for instance.

I know how to work out the scripting to pull text and image from png and txt files in a generic folder so it looks for Generic.png inside Desktop/Generic folder.

What I’m not clear on is how to use wildcards or whatever they would be called here such that I can use proper distinct batch titles for folder and asset names, and video take outputs. So, for instance, if the project is named “CatVideo” and resides in Desktop/CatVideo folder, then it uses that “CatVideo” string to look within it’s folder for “CatVideo.png” and txt file, and outputs videos with names per take like “CatVideo_1920x1080.mp4” (Last example would need one wildcard for video title, and another taken from the name of that render setting assigned to that take.)


#5

The feature you’re looking for is called tokens. Not sure about using them for referencing project assets, but they’re definitely used in c4d to create renders. One I use often (and which I enter in the output path for my render settings) is “./renders/$take/$prj_$take”. I’m sure you can get more complex than that but I haven’t had the need to so far.


#6

I think I’ve got most of the take process worked out now, and tokens are working beautifully. One question, though.

Currently, I have the default MAIN take which is just the generic unformatted version, then the various formatted versions I need are the child takes.

When I “render all takes”, it does all the children, but also the MAIN which isn’t something I need to be rendered.

Should I just delete one of the formatted children and instead make it the MAIN, or is there a way to render all CHILD takes without the MAIN parent?


#7

Yeah, in the Take Manager, there’s a column to select takes. “Render selected Takes”, well, renders only those.
Wording could be different, I use a non-english version…


#8

Yeah, sorry. Just caught that. It was just grayed out before since none selected.


#9

No problem :slight_smile:


#10

I also need a jpg thumbnail for each mp4 output.

Each take already has a render setting applied to it, and I’m not seeing how to also add a second.

I suppose I could create a new cilld take set to jpg for each of the existing child takes already set to mp4.

Will see if that works, but wanted to check if maybe there’s some other way to add a second render to each take without creating additional children.


#11

Not possible to add two render settings to a single take - you’ll need another take in order to assign the second tender setting.