Kevin Camp
Forum Replies Created
-
well if your looking to add to your arguement for progressive, you will also have a more natural (film-like) motion blur if you shoot progressive than you will if you deinterlace interlaced footage, which will look a bit choppy unless you effect it further with frmae blending, directional blur, reelsmart motion blur, etc…
so, better resolution, kinder compression artifacts and film-like motin blur… who can argue with that.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
download the beta at adobe labs and try it out. many have been using ae7 with windows on intel macs due to the numerous issues with running 7 under rosetta and are very happy. most of them, i believe, are running windows via boot camp though.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i suppose you are right… i mean, you can’t argue with steve roberts 😉
but i’ve never worked with anything that used a field rate… it just seems that saying 30 (29.97) interlaced frames per second would be less confusing than bringign in another number for field rate…. but there are so many confusing things with video standards.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
if the destination is progressive, i would shoot progressive. deinterlacing does involve some interpolation to generate the missing field data. even though it does use the data from the field that it is dropping, it is not going to be quite as good as having it shot progressive from the start. the progressive image should be a little sharper and cleaner than deinterlaced footage. also, compression is generally a little more kind to progressive frames that interlaced frames.
if your destination is going to be interlaced, then you may want to stay interlaced… deinterlace to process, but render that back out to interlaced.
the effects you mentioned will be better on progressive footage, but we’ve all used them on interlaced footage with fields separated and gotten good results, so i think it comes down to destination and asthetic preferences (some love the smoothness of interlacing, others the sharpness of progressive).
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
with opengl off, the cpu is doing the work and the ram is storing the rendered frames. essentially, faster cpu = faster renders; more ram = more frames stored.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
you may need to give us your workflow, from the original footage format (dv, dvcpro, beta), to the capture setting, to the export setting and the specifics fo what ae thinks the footage settings are.
you say ‘shot 60i’… we assume you are talking about 30i, which would be an interlaced frame with a frame rate for 29.97. is this correct?
you mentioned you captured to dv… was it dv tape you captured from? what did you export out of premiere, dv, lossless…? high compression settings and compressing and recompressing can start to muddy up interlacing, and that may create problems.
in ae, if the footage is interpretted without separate fields and you step through it frame by frame, you are seeing interlacing wherever there is movement, right?
explain your shot a little more… you said the building looks clean when separate fields is off, but the guy walking is interlaced. is camera locked down for the shot, no pan or other movement? by your footage’s field description i would assume this to be correct. did you try to separtate fields (lower) with preserve edges checked? this can often help.
remember, deinterlacing footage that was shot as interlaced (not 24p witha pulldown added) will involve interpolation by the software, and may not be perfect. you may end up maksing out a deinterlaced guy walking over an interlaced building to get the best results.
if you can get the interlacing problems worked out, and you have 30i (29.97, interlaced)you can slow your footage down by deinterlacing, but interpretting the footage as half of what your comp’s frame rate is. so if you want to slow 30i footage by half, interpret footage as separate fields (possibly, with preserve edges), but set the frame rate to 14.94 then drop that into a comp that is 29.97… 30i to 24p (23.976): do the same, but set the footage to 11.988 and put that into a comp that is 23.976.
you can go slower, as steve mentioned, you will need to use time remap or timewarp or a third party effect (like twixtor) to generate the extra frames.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i don’t believe adaptive resolution will see any benefit from a better graphics card.
if you want to upgrade the pc, you’ll get the best return for your money on ram… max it out.
if you have muliple processors or cores, you may then want to purchase nucleo ($50-150) or upgrade to ae cs3 (out very soon, and if you don’t have the pro version of 7 you’ll get it with the upgrade to cs3 free) to take better advantage of the multiple processors.
if you don’t have multiple porcessors you may benefit from an opengl 2.0 card. search the cow for an opengl article written by mylenium about a year ago. in it, he explores using opengl in ae7, and using it for previews and renders. you can also search the ae’s help for opengl effects and find a list of the accelerated effects and features to help determine how well it might benefit you. opengl does render differently than the software render engine, notably the anti-aliasing. in my experience with opengl, i found that it does a poor job compensating for pixel aspect ratio mismatches (ex: placing a square px image into a non-square px comp). so you will want to convert images (or other elements) to the pixel aspect ratio you are working in, in another pieces of software.
upgrade the ram first, then look at your processor situation… if you have a single processor, you may benift from opengl acceleration enough to make it worth while, but you will want to develop a workflow around opengl. if you have multiple processors, you may get better render speeds with nucleo or upgrading to cs3.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
it could be a problem with a version 6.5 plugin getting loaded by 7.
if you can’t find the offending plugins to remove them, you may want to reinstall ae7. make sure that you only install the cc effects from the ae7 install disc. for any other third party effects/plugins you may have and want to install, you should download the latest versions (preferably they will say for ae7).
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
i don’t believe adaptive resolution will see any beneift from a faster graphics card….
if you are looking to spend money upgrading a pc for use with ae, you’d be much better off maxing out the ram.
however, if you have already maxed the ram and you have a single processor pc, you may see some benefit with a good opengl card… there is an article here at the cow that explores opengl in ae. note that opengl does render a little differently, particularly anti-aliasing. and in my experience it does a poor job compensating for pixel aspect ration mismatches… so if you have a non-square pixel comp, and bring square pixel images in, it won’t look so good. or the other way around. so just convert your images and other elements to the pixel aspect ratio that you are working in.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW -
if the duplicate frames are regular… ie always 3-in-a-row, then you could just drop the footage in to a comp that is the frame rate divided by the number of duplicate frames (frame rate/3, in my example). or use interpret footage to set the frame rate of the clip to the number of frames times the actual frame rate and put that into a comp that is the actual frame rate.
either way, you may be able to precomp that and apply time remap on the precomp, or render the new footage, import and apply time remap.
Kevin Camp
Designer – KCPQ, KMYQ & KRCW