Hi John and Everyone,
I'm attaching the memories file I worked from and the the last movie test, ignore the lack of aesthetics please.
I waited until after 8 or 9 tries before writing; I'm at a loss to explain the black frames - what I've seen in this particular animation attempt is that it renders to canvas up until frame 8 and then I see a black frame for the duration, not knowing that it is in fact rendering the remainder of the timeframe.
On the first few attempts I had my system preferences settings set without realizing it to shut off the monitor and the screensaver on. I thought well that might certainly be a factor, it had been in the past with other rendering software, so I shut those off and the screen was on for the entire render. Same thing, hanging on frame 8 and you will see the results; intermittent black frames afterwards.
So I'm curious to know a couple of things that depending on your feedback could be suspect. Might it be that you have to develop enough of a depth of knowledge of processor chains to know what elements in a chain are going to cause this and you can't simply drag something out of the evolution editor and into the advanced msg editor memories?
Also as I hope you'll see, from doing "test renders" of the msg files from the evolution editor or the advanced msg editor to the canvas before the animation and assuming these would be more or less in the same "style" and color I'm unsure about why some frames that I would not have expected in terms of content or color appear in the movie.
In my non-programming "front-end" high tech mind my sense tells me that interpolating between frames in a PASeq timeline, theoretically, wouldn't take longer than if I were manually doing an evolution in the editor and dragging that evolution to the canvas? Is that way off? Because it's taking I'm guessing here 5-7 times longer when this particular animation is rendering to render a frame. Perhaps (if you folks have time,) you can set me straight as to what may actually be going on.
sorry for all of the questions at once.
I also have that "older" 2008 MacPro tower "2,1" which flew faster than I could for years until 10.6 or so and now according to the apple store employees I mentioned the OS and the programs have become so much larger and written for newer processors; What is the earliest OS X I can run Studio Artist 4 X because I'm thinking of going retrograde on at least one drive and trying it out.
Many thanks and best regards,
Mark
Replies
the MSG file will need to be zipped before positing if you want us to be able to use it.
You're keyframe interpolating between different processor chain configurations. As opposed to interpolating between different parameter settings associated with a fixed processor chain configuration.
You can get pretty strange results when you try and interpolate between different processor chain configurations. You might also get wildly incorrect interpolated parameter values (which could cause some weird behavior, or cause things to take a really long time to process). Because the parameter interpolation will be working between disparate processors.
You should really try to stick to keyframe interpolating between different parameter settings associated with the same fixed processor chain configuration.
You can always open up a rendered animation and delete individual frames in something like quicktime player 7 pro to get rid of frames that visually drop out or go to solid colors or otherwise seem out of place with what you want visually.
As far as what mac os versions can studio artist run on. You should be able to run the last update of osx 10.4, and anything latter than that.
Because you are interpolating between disparate processors, you are getting really huge spatial area parameter values for some of the processors that do area filtering. This is why the processing time goes up dramatically. It's trying to interpolate based on a random seed parameter associated with a different processor in the next keyframe. Again, because you have different processors at different index position in the processor chain for 2 adjacent keyframes.
So for a given processor index position in the processor chain editor, you have 2 different processor types for 2 adjacent keyframes. the interpolation tries to interpolate the parameter values for these 2 different processors. The first parameter for the one processor is a spatial area parameter. the first parameter for the other processor is an error diffusion processor, which has a huge random seed value for it's first parameter. So the interpolation tries to set the spatial area parameter for the filter processor to a huge value (not one you would ever use), which then ends up takeing a long time to run.
Yes, that explains an awful lot. I might wish it would do that sort of thing, but maybe in some huge render farm in version 12 :-) Thanks very much!