Kevin Camp
Forum Replies Created
-
there are 2 little pulldowns at the bottom of your preview window, one has a percent value(probably 100%), the other a term for a fraction (probably full). you can change those settings to help render more of your comp and render it more quickly.
try setting the percent to a lower number, try 50%. this will decrease the preview size by half (don’t worry this is for preview only, your comp is not effected). and generally you’ll also set the other pulldown to a corresponding fraction.. ‘half,’ in this case.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
you just need to find a data rate that will work on the other pc… dv will keep the data rate within firewire 400 limits, although quicktime dv compression isn’t the greatest (not that dv compression is great anyway). photo jpeg would be another that you should be able to scale to a data rate that will work. you can create multiple output modules when you render, so you can, in effect render once, but create multiple media files with various formats/settings to try out.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
yeah… i was interested in the drive setup used in the barefeats test, after the previous posts by steve forde about the potential bottleneck, just to see if the ae results might get even better if using a 4 drive stata2 raid. rob has posted raid performance tests before, i’m hoping he will set one up and re-test ae cs3.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i take it the surfer is moving very fast… how many frames is the surfer moving across the screen?
you can crank up the shutter angle to 720, which will increase the motion blur effect. go to the composition settings for your comp, click the advanced tab, set the shutter angle to a higher value.
just to make sure… you are using both the motion blur switch for your surfer layer by clicking the motion blur checkbox in the switches for that layer AND enabling motion blur by clicking the motion blur button at the top of the comp window, correct?
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
1. transparency… to create transparent areas you will need 2 things and alpha channel and a codec/file format that supports alpha channels. first, click the checkerboard icon at the bottom of the preview window to make sure you can see the transparency (you should see the checkerboard pattern in the transparent areas, like in photoshop). now, send your comp to the render queue via make movie or add to render queue and click the underlined term next to output module (usually says lossless). the lossless animation codec supports alpha and works well with fcp, so keep that, but set the channels to rgb+alpha and set the color to straight (unmatted). now, when you import it into fcp, just make sure that fcp reads the alpha channel correctly and you should have a nicely keyed graphic in fcp (dvcpro50 does not support alphas so it is not an option)
2. shadow behavior… the drop shadow effect is kind of a fake shadow, it won’t behave as a true shadow because you don’t have a true light source casting the shadow. you could set one up using ae’s 3d abilities, but i think it is easier to fake it with the drop shadow effect in this case. you need to set the direction property of the drop shadow to rotate opposite the direction of your circle… so if your circle rotates clockwise 720 degrees, your shadow direction should rotate -720 degrees over the same time. if you want to get into expressions, you can link the two properties, but setting keyframes with opposite values will work fine.
3. pixel aspect… by your description, i believe things are set corrctly in ae. if you import your movie back in to ae and it looks correct, then it must be a problem importing into fcp. i haven’t used fcp much, only to capture, so i can’t be much help there.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i think your after effects settings are good for creating transparency… i’m not sure how flash interprets the alpha channel… you could try a png image sequence, i’ve had good results with pngs and software that doesn’t interpret alpha channels.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
if the entire layer is flickering (maybe more like strobing), and you are rendering with the intent to display on tv, try rendering with fields. to set ae to render fields, send your comp to the render queue (via make movie or add to render queue). click ‘best’ next to render settings to open the render settings window. set to render with fields (use appropriate dominance, lower for most ntsc applications).
if only parts of the surfer are flickering, like fine detail or patterned areas, then this is usually called field judder or jitter. this can be fixed by applying a slight vertical blur to the image in ae, or even better to do it in photoshop… there is an action set called defickerator (google should return the link) that is designed to soften the details that lead to judder, it works well.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
if it previews bad, but renders fine, it may be an opengl problem.
in ae the preview setting does not carry over to the render settings, so if opengl is used for previews, and giving you problems, the final render may not show the same preview issues.
check your previews preference (or click the lightning bolt icon under the preview window) and set to use adaptive resolution, rather than opengl.
fyi, if you ever wanted to use opengl for renders a well, from the render queue click the render settings and click the use opengl render.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i’ve seen this with an older version of cinema 4d (it may still be a problem with c4d, i don’t know). but since ae handles premultiplied images well if you tell it that the image/footage is premultiplied, i’ve stopped using the straight with alpha option in c4d if its going into ae.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
if you are rendering to send to fcp to be edited with other footage in fcp, then you will want to render the same frame size, frame rate and par. so, if they are editing in d1/pal widescreen, 25 fps interlaced, then you should give them the same.
they probably can convert square to non-square, but you can save them and yourself some time by working in with the non-sqaure pixel format.
also, make sure they are aware that you will be rendering in rgb and that they should import accordingly. i haven’t used fcp, but it is normally an easy adjusment… you could also make the adjustment at render, in ntsc land you add a levels adjustment layer over the whole comp, set the black output to 16 and the white output to 235. i don’t know if thats the same for pal, but if pal in yuv colorspace, then it probably is the same.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW