Kevin Camp
Forum Replies Created
-
unless there is a ‘render lens flare only’ setting in the effect (and i don’t think there is), the lens flare will be rendered on top of the footage/layer that it is applied to (in your case a black solid). what you can do, is simply duplicate the lens flare layer, with all the lens flare settings and keyframes that you have currently, and sett the lower lensflare layer to use the upper as a lumanence track matte (its a pulldown in the modes section of the timeline).
the flare will look a little muddy, since it was over black. so, you will need to render that with alpha and import back into ae, make sure the interpret footage setting (select footage in project window and right-click or file>interpret footage>main…) is set to premultiplied (this works in ae7, not sure about 6), place in a new comp and rerender with alpha and ‘straight’ selected. if the interpret footage options are different in ae6, you can bring your flare in to a comp and apply unmult effect (freebie from red giant) or xmult (another freebie i found somewhere) or other similar effect taht can remove the black background, then render straight with alpha.
avid may have a setting to pull the black out of a keyed image too…. it’s a common problem that’s been around for as long as people have used alpha channels.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
the only thing i can think of is that divx is using temporal compression and needs several frames to create that type of compression, where the other codecs are just rendering and writing one frame at a time.
if there is a setting for the divx codec to set the frequency of frames (often referred to as key frames) that it uses for temporal compression, you might try lowering that number. i know you mentioned that you would like to render once, but it would not be a big workflow hit to render to a lossless codec, setting a post-render action to import. then re-rendering the imported render to the divx codec.
you may even be able to set up you’re divx comp with your other comp already nested in it, then set the lossless comp to render in the render queue with import and replace usage as the post-render action and set your divx comp up to render after that, all in the same queue.
additionally, since 1gb of ram is not that much, if the ram in your 2 pcs are the same, put it all in one pc and see how that works.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i can’t remember if ae6.5 can change the shadow map resoultion, but that may help… in ae7, choose comp settings, advanced tab, advanced 3d, options… select a smaller shadow map.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i assume 3000×3000 is pixels… if it is pixels then there is no dpi (or ppi) to define (if 3000 is inches its a different story).
if 3000×3000 is pixels, all you have to do is create a simple action in photoshop that uses image size with scale unchecked and enter a resolution in ppi (or ppc if you’re metric). then create a droplet through the automate menu (file>automate) that uses that action you just created. now you can drag and drop an entire folder full of ae rendered images on the droplet and photoshop will make them the correct ppi (and your files won’t get any bigger).
also, save out of ae as psd (or something without lossy compresssion) then when you set you ps batch process, have it create the jpg (no need to compress twice).
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
if you imported a file from ps you can check the interpret footage window (select the image in the project window, then right-click, or file>interpret footage>main) and check if ae is interpreting the image correctly.
but i’m not sure that importing the file will allow you to use the vector data as an animation path…
if you are copying and pasting the path as a mask, i think it may get pasted as square pixel. to compensate you might try one of two things. you could work in a square pixel comp for that part of the animation, then place it into a .9 comp to composite with your other elements. or you may be able to create a solid in your .9 pixel aspect comp that is square pixel size (720×540 or maybe 648×486), paste your path, then scale the layer to fit your .9 comp.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
you might look at the physical differences between the 2 machines along with the memory and cache settings. it may vary well be that the 3d render engine in ae7 has been improved from 6.5, but what you’re describing sounds a lot like ae6.5 running out of memory. you could try enabling disk cache on ae6.5 (or increasing it) and see how that effects the lines.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
it’s been rumored for a while that the next macpro will have 4 cores per cpu (2 x core quatros, i suppose). and the next mac pro generation would probably be due right around nab, maybe a bit later. so, it wouldn’t surprise me if they had one to show.
although it wouldn’t surprise me either if apple showed off the new osx at nab, get people hyped up to buy a new mac, then 3 months later release an 8 core mac pro.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
glad i could help… there’s just a few too many libraries in osx for my taste.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
also, are you seeing the difference on different computer monitors? pc and mac gamas are different unless they’ve been set otherwise, so that could account for the clips looking different. you’d want to compare the footage on a broadcast monitor, or at least take the clip back to the other machine to compare to original footage.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
when i work in 3d and composite in ae, i prefer to render the 3d at 60p… the progressive allows me a greater ability to manipulate the 3d footage in ae and 60 fps allows me to render interlace if needed.
there is a plugin called reel smart motion blur by re:vision that does a good job of creating motion blur on footage that needs more motion blur. it may help too.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW