I must admit I've made a few posts in the past on this subject and it's tough to get any answers to the question of slowed down fluids in maya.
Now I'm just looking for a little clarification here as its taken me a very long time to get my head around the whole idea mainly due to the amount of variables such as sim rate scale, substeps, cache steps and frame rate. not to mention the trax editor method. Now I'll be totally honest all of these things seem to do very similar things but not what i've been after. They haven't helped me slow doen a small scale sim one bit without a stepping effect or in some way altering the look of the sim.
Would I be accurate in saying that the only way to get a high quality slow fluid sim is to up the container scale very high and adjust your fields accordingly? The fact is I just cant seem to create the effect of a small scale smoke being filmed at a high frame rate.
Depending on how slow you need to go have you tried to export out at a higher frame rate and interpret it as a slower rate? I have read more than once that fx houses have rendered out at say, 120 fps and played it back at 24 to achieve the desired effect. Yes I understand the amount of time and processing power is insane, but what do you do? Can you afford twixtor? And would it even work?
The movie Guardians, (that owl movie from Snack Snyder) did that for their slo shots and there was a discussion/article on it and their rendering challenges. (of course I can't find it now, maybe it's in a Cinefx mag?) Yes, to simulate real world you sometimes have to shoot it like real world.
Thats a great solution. I have tried it and like you mention there are render time issues. Fluids are pretty expensive as they are without having to do a tonne of extra frames.
However I'm only a little concerned that by cacheing the fluid at 300 fps it is in some way altering the fluid's behaviour. Mad as it may sound the 24th frame at 24fps and the 300th frame at 300fps dont look quite the same. I'm probably splitting hairs and being picky because they are similar but not quite. Have you ever come across this?
Yes, actually I have with a cloth sim I did awhile ago, and I'm not sure what causes it or if it's a quirk of the increased frame rate or not. But since my client didn't notice anything I didn't say anything. I won't be able to do a test this week or research it a ton as i am behind schedule and loaded with projects.
But as you say they are similar but not exact? So you could live with it as is since it'll pass?
It would be good enough for me.