ProRes is a very good codec. You can re-render it six generations deep before you see any kind of image degredation. I don't know ANYBODY who would re-render six generations deep -- they would instead prerender to a lossless intermediate codec for further work in AE, then either render to the specified delivery codec or again to a lossless codec.
You can set the codec used for RAM Previews in the Templates section of AE. It should default to Losless: if you haven't messed with it, you're seeing the best possible quality
Pierre Devereux wrote:
When I feed this footage into After Effects, where can I go to read up what it is that After Effects does with the file? Does it uncompress it? Does it work on the file within its compressed state?
Yes. After Effects decompresses everything in order to process it, even if you don't adjust it or apply any effects.
When i take a shot of a background, and I add another shot of a character (keyed out and ready) and I place the character in the background and preview 5 seconds, what am I looking at? Final quality, or an intermediate step.
If AE is set at full resolution in the preview window, the resulting previews will be full quality. Be warned that AE previews are not interlaced. Only renders from the render queue process fields correctly.
Ill spend some more time with the camera and Samurai then to make sure I am getting the best possible quality to the device. Then Ill start reseraching better workflows. Almost all of our shots will have characters composited in, and there promises to be many action sequences with special effects that ill have to add, so there should be no end of research happening from this side! :-)
Thanks once again, and ill send an update when I figure out if it was just me that did not like the quality, or if I had some settings fiddled with ( being an ex Microsoft Server engineer, I make no excuses for my fiddly habits - I must remind myself not to change settings jsut to "see what it does!" :-)
I do have one more question around this though:
When i record a scene, the largest file size i can generate is 27.6MB per second (megabytes), Yet some of the tutorials i have seen, specifically around large file sizes, puts there 17 second footage at 14GB - which, if my math is correct (and I dont promise it is!) puts their data at around 80MB per second. How did they manage this, or did they take the footage into a NLR already, and render the larger file out? As I have noticed, that i can capture at a certain size, and with the lossless settings in AE, generate a bigger file.
I promised I would post again if i figured this out - I wish I had given that more thought, as the issue, turned out to be me, being completely silly.
BUT, in the hopes that I am not the only crazy beginner out there, here is what I found ( I hope someone else can benefit from this hard lesson)
CHECK YOUR MONITORS RESOLUTION SETTINGS!
Yes, that was my issue. when I first purchased the HD LED TV, I used an old interim HDMI cable that was far from ideal. It caused many issues with my display setup, and I had to eventually put it on a very low resolution to get it to work. As a result, when I got the proper HDMI cable, and have been using that one ever since, for some reason , and I wont even attempt to make excuses, it completely slipped my mind to reconfigure the resolution of the T=HD LED TV back to 1080p.
Sigh, anyway, the quality coming form the AF101 and the Atomos Samurai through to AE now, is perfect!
Next challange - figuring out the AF101 shallow depth of field - I need to find out what settings onthe camera will allow me to get more of the scene into focus.
Thanks for the help guys, I hope that by reading about this silly mistake, others will be able to avoid it! :-)