A different strategy + notes

So on this film what I'm going to do is edit the film without sound... I mean the shots are already timed for the dialogue and the like as I animated to the dialogue but I gave them all a little extra head and tail... Why r I doing this?

I like building things from the bottom up... like making sure something can stand on its own before I add another level on top of it... So I'm going to make sure the film works visually before adding in music, dialogue and sound design... and well also before the skies and other extra stuff is added in the compositing phase... right now I'm cutting the raw shots that don't have skies, atmospheric fx, post processing etc...

Running my 4 quad render boxes + 8 core render server 24 hours/day for a month adds about $125 to the power bill... So budget that in before you start your cg film...

I started work at 2pm today and finished just now at 2am...

I'm going to posting a like casting call thang here for the voice acting soon... So if you want to do a voice in HSM watch out for that...

Also make sure all your monitors are calibrated and sync'd to the same color space... easy way to do this... set them all to Srgb and download some pluge bars and make sure they all look the same... I'm using crappy LCD's but will borrow an old NTSC CRT monitor to do the grade when its all ready... but as long as they are in the same color space and the pluge look similar you'll be ok for Color correction...

Comments

  1. Thanks for posting the electric bill. I was curious if you would gain anything from buying time on one of the render farms advertised online, but it sounds like you are coming out way ahead by doing it yourself...

    Judging from the $0.25/Ghz-hr rate seen on the RenderTitan site, $125 would buy you about 160 hours (about a week) on a 3Ghz machine, and it sounds like for the same amount of money you are getting about 3600 hours each month. If your machines cost you $2000 apiece, they pay for themselves after about four months... Good to know.

    I guess the only real advantage of those services is if you are on a tight schedule.

    Things will really start to get interesting when more renderers get rewritten to take advantage of GPUs; a recent paper showed a 10x speedup for a GPU-accelerated Renderman implementation. Just a few decent gaming graphics cards would make a decent render farm, and use less electricity, too.

    ReplyDelete
  2. My machines actually cost about $750 each so its worked out pretty well... as far as using online render farms... my project files + textures are huge... average about a gigabyte per shot...So I think any render speed gain from the online farm would be massively bogged down by transfer times... the render files average about a gig per shot as well.

    ReplyDelete
  3. Wow, I never thought of approaching sound design with dialogue that way, but it´s kinda like coating a painting or a car, which is a really cool. How would the lip sync work out? what would you use as reference?

    ReplyDelete
  4. @KIDFRAME

    So for the animatic and lip sync purposes I did all the temp voices... men, women, monsters and children...its pretty hilarious... If I have the guts I'll make it a selectable audio track on the DVD ^_^

    ReplyDelete
  5. lol

    now u're gonna have to put it on there mdot, sounds awesome!

    ReplyDelete

Post a Comment