Brendan Coots
Forum Replies Created
-
There is an option called Multiprocessing in your After Effects preferences that must be turned on, first and foremost. Once you turn it on, the multiprocessing dialog window should tell you how many cores it’s using. If it says it will use both of your CPU cores, you’re golden and nothing else can really be done. If it says one core will be used, read on.
For AE CS3 to function at peak performance, it is recommended that you have at least 1GB RAM for general system resources, PLUS 2GB RAM per CPU core. In your case, this means 5GB RAM would be ideal (1GB for system + 2GBx2 cores = 5GB). Since you don’t have this much, you might need to tweak your after effects settings to convince it to use both cores. Go to the memory page of your preferences and very gradually dial down the 60% number, while going back to the multiprocessing page to see if the change convinced it to use both processors. Repeat until it does.
Beyond that, the only optimization a person can really do is to add more RAM, add a fast secondary drive (as in, NOT a single disk external drive), keep their Mac optimized with a program like Cocktail, etc.
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
There is no point in creating an uncompressed file out of Final Cut if the footage is Mini-DV, you aren’t gaining any quality and as you saw it just results in a huge file for no reason. 99% of the time you should export footage in its native format for use in After Effects. In your case, this would be the DV codec which After Effects should have no problem with.
When you render back out from After Effects, it’s a totally different story. The key and any other effects that you do in AE are done at 4:4:4 uncompressed quality, that’s just how After Effects works. Therefore, it DOES actually make sense to render to an uncompressed or lossless codec out of AE because there is new, higher quality elements there than the DV footage you imported. To save space, don’t use uncompressed – lossless codecs don’t lose data but save space wherever possible, resulting in a file visually identical to an uncompressed file, but at a much smaller size. The Animation Codec is probably a good option here.
I should also add that doing the type of work you are attempting to pull off requires that you have a relatively fast computer with at least 2GB RAM or more, and at least one fairly large secondary hard drive (320GB at least) dedicated solely to storing your media. Anything less than these specs and you will be limited to working at lower quality and taking much longer to do basic tasks, as you experienced.
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
It would help to know what part of this image you are unhappy with.
If you are referring to the banding going on in your gradient areas, that could be fixed by making the project 16-bit instead of 8-bit (go into Project Settings).
Outside of that you need to be more specific about what you are trying to fix, so that people can address the particular issue.
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
Oh yeah, I forgot that little chestnut even though it’s really the worst aspect of HDV.
I should add one more thing here – We’ve been testing the SheerVideo codec and the results are truly stunning. Way smaller file sizes and the same quality level as Animation codec. If you are looking for an intermediary codec Ian, this one would work great. We may even adopt it as our primary codec for everything.
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
You may not realize this, but HDV is the same exact quality as Mini-DV (25Mbps). You could convert it to cave painting and it would hold up, so I wouldn’t even worry about which codec you convert to, as long as it supports HD it will be better than HDV and will play better with your apps. ProRes is an option, but you might be better off with something less processor intensive since that is the core of your post. The way you are handling output from FCP right now is probably fine.
It sounds to me like your workflow is ok right up to the point that you render out of AE for re-import to FCP. At this stage, maybe you should consider rendering proxies out of AE (motion-jpeg, etc.) to be used in the edit until the cut is locked, THEN render your lossless files from AE and fold back in to FCP for final output. The only way around this is to render to a lossy format out of AE, which you probably don’t want in your final cut anyway.
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
Two points on bit-depth within AE:
1. 16-bit can offer noticeable quality improvement over the standard 8-bit, but mostly on things like gradients, edge detail etc.
2. 32-bit is ONLY useful if you are working with HDR plates or want certain effects to take on a more photographic look (mostly blurs and such).
Either way, these choices really only affect what is happening WITHIN After Effects as you work. 99% of the time you will be outputting to an 8-bit or 10-bit file format anyway, such as Animation codec, Prores422 etc. Outputting to higher bit-depth than 10-bit requires using frame sequences or specialized (read: expensive and tricky) codecs and most of your apps will not accept them anyway (including FCP). In the end, that’s not even the point of higher bit-depths in AE – the working bit-depth simply defines the precision at which effects take place, so that your work looks cleaner when output to 8 or 10-bit. Working in these formats is just a way of preserving certain information and detail while working in AE. When you output the file to 8-bit you really aren’t going to lose much (if any) of that added detail, but if you’re really concerned feel free to render to a 10-bit codec.
Bottom line, it couldn’t hurt for you to work in 16-bit given your source materials, but 32-bit would be total overkill with painfully long render times and minimal, if any, improvements over 16-bit.
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
I’ve personally given up on looking at people’s schooling (and really their resumes as well) when I’m hiring, because it doesn’t really matter to me. All that I want to see is a demo reel, because it tells me all I need to know – can the person shoot good video, do they understand lighting, staging etc.?
There are a lot of degreed people out there who aren’t qualified to WATCH movies, let alone make them. Your son would almost, but not quite, be better off spending the money on a good camera, lights etc. and taking a few months to really dive into production. He’d probably learn just as much if he’s self-directed. Plus, if he decides he doesn’t like it, you have the gear to sell off and recoup somewhat.
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
Really stunning work – sound design, 3d, graphics all working together extremely well. That is the best example of motion graphics work I have seen in years. Want a job? 😉
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
Lloyd Alvarez has a script that lets you distribute layers in 3d space randomly:
https://aescripts.com/2006/06/12/3d-layer-distributor/
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com -
The square pixel dimensions for widescreen DV is 864×480. Create an 864×480 comp, drop in your footage and stretch horizontally to fit.
Brendan Coots
Splitvision Digital
http://www.splitvisiondigital.com