Flash animation -> FCP / DVDSP
I've made an animation sequence in Flash, that's supposed to end up as a motion menu in DVDSP, after some editing in FCP. I work woth SD PAL. Should I have a 720x576 or a 768x576 canvas in flash? And which settings should I have in FCP to get the best result?!
At the moment I have a 720x576 canvas in Flash, exporting 8 bit uncompressed, Framerate 25, 720x576. In FCP I have a sequence with CCIR 601/DV PAL, square pixel aspect ratio, upper field dominance, uncomp 8 bit. But the end result doesn't look good - graphic curves are very "hard", cut off...
I know this is an old issue, but I can't seem to find the exact answer anywhere at the moment...
Thank you very much in advance.
Your end product is going to be video, PAL, so it will be 720x756 pixels. For video output those pixels will never be square (though of course your sources can have square pixels.) In PAL, if your video is 4x3 aspect ratio (traditional), it will have a pixel aspect ratio of 1.066 (giving the notional 768x576) or if it's widescreen the pixel aspect ratio will be 1.42.
Flash only deals in square pixels, so your neatest route might be to work at 768x 576 in Flash, if trad, or 1024x 576 if widescreen, and export. In your edit app, import your material with source interpreted as square pixels - the images will squish (a little or a lot, widescreen or not) into the standard 720x576 - but the video will display correctly in video output.
You also have to watch your colours. Flash will allow a full RGB output - you need to be make sure that your import settings recognise the material as RGB source, and alter the dynamic range to keep the video in the "video safe" range - something like 16-235 as compared to the full 0-255 of RGB.
One more thing to be aware of is the thickness and sharpness of lines and fonts; your video probably has to be backward compatible with the mass of CRT screens out there, so to avoid flicker in the final images you'll want to keep all your lines at least 2 pixels wide, and your fonts reasonably sized ...
Thanks very much for your time and reply, Mike.
I've tried different workarounds in the meantime. The animation part is very simple, so I made it in an alternative version by doing the frames as psd's in photoshop, and imported them into final cut pro. And there I made a sequence with these frames. When I export the sequence as a selfcontained movie with current everything (so to speak!) I great a great looking mov file, that I can even enlarge, and still the graphics look good. But when I encode to m2v, either by using compressor directly from the timeline, or sending the self contained mov file to compressor, the end result isn't that good.
Could it simply be this particurlar type of graphics that isn't "made for" m2v?
I made the files in photoshop from scratch in 768x576, then scaled them to 720x576, before importing to fcp. As the manual says.
I normally don't have problems with this.
It sounds like your problem is in the encoding, since the uncompressed mov looks great. No doubt you've tried multi-pass, variable bit rate, whatever compressor allows you to adjust ...
Compressor doesn't have the best reputation as an mpeg encoder, though my direct knowledge on the Mac is limited. Is it worth asking on the compression techniques forum here?
Top end software compression - CinemaCraft is highly rated, has excellent customer support, offers very good control and settings to optimize compression for different kinds of footage (including graphics), gives very good output and is speedy. There is a Mac version I think. http://www.cinemacraft.com/en/index.html
I understand BitVice gives good results but is slow. http://www.innobits.com/bitvice.html
Digigame's Megapeg has long produced decent results with OK control, and now has a Mac version http://www.digigami.com/megapeg/
Ffmpeg is reported to produce decent results. http://stephenjungels.com/jungels.net/articles/ffmpeg-howto.html
In the Windows world, Procoder does a decent job, with a lot of control over the details. http://www.grassvalley.com/products/procoder_3
Thank you very much. I'll look into those options. I've seen many different outcomes of compressor encodings... Still "forest footage" is the most difficult to get right in compressor...!
I also think that the many different TV set options today is making it more difficult. I know that my client is watching this on an HD TV set, on which their DVD player probably upscales, so that the graphics are being "stretched" -> and thereby doesn't look good to them. Same thing with software DVD players going full screen. Very few choose to see "actual size" DVD's on their computers, which makes the graphic situation even more delicate.
Again thank you for your time.