mesut wrote:... sooo... if there's a bug... where is it? in c4d, in octane-c4d-plugin, in standalone... or in the abc file-format?
let me pick up the ball at this point again.
i would really like to solve this problem somehow and i still am not sure, what's the right answer to the above question.
it's not only the fact, that every 20 frames one receives this failed-mblur-frame.
when dialed up the count of particles in the scene, this particular fail-frames also produce a huge amount of rendertime. what is about 30secs for the normal render, these failframes take up to an hour to render their faulty result.
which of course renders batchrendering an animation job out of standalone rendertimes-wise ridiculously unpracticle.
and sometimes, on one of those frames, one of two cards crashes. system reboot needed afterwards.
workaround methods:
1. don't use motionblur:
- yeah, sure.
2. export abc-files out of realflow, stich them and import them into the exported orbx-scene, into standalone:
- not helpful, because all the modifications that need to be done with the particle simulation within in c4d are not being taken into account, which renders this method useless.
3. find out which regularity (in this case every 20 frames) is present and render accordingly, while rendering the "failframes" singularily - not as an animation range of frames - out of c4d.:
- that would lead me to splitting 1 c4d-orbx-export of an originally 100frames-animation into 5: frames 0-19, 21-39, 41-59, 61-79, 81-99. one can see the amount of handwork and supervision that'd be involved in this. even if all the "parts"-orbx exports were to be imported as rendertargets into one gigantic uber-orbx-scene, batch rendering doesn't allow (afaik) seperate frameranges to be rendered for the seperate rendertargets that are checked in the batchrender dialouge.
so, the workarounds i see are either unusable or extremely unpracticle.
is there any other trick to make the best out of this situation?
and again:
what is the core of the problem here? is it c4d, c4d-octane-plugin, standalone, the exported abc-range? and further: is this a known issue, is it acknowledged and something that is being worked on? has it any kind of priority or is it something along the way?
while i am indeed thankful, that i had received a response at all, i'd very much appreciate it, if there could be some kind of closure. may it be an answer like: "known issue, sucks since years, will take a while to fix/not sure if fixable at all" or "wow, that's new, never seen that, we'll see what we can do", or a "hey kiddo, can't reproduce, seems you suck at doing what you do, here everything's working just great, we use it every day (TM)". anything allong those lines please, so i can decide in which direction i'd need to move, 'cause i need to keep on moving.
right now, it's a bit like being left hanging.
thx.