Charles Simonson
Forum Replies Created
-
Basically, that is Compressor’s way of telling you that the bit rate you have set is too low for it to encode to and it is going to set its own bit rate that will keep the file size low and still provide good quality. At 2 hours and a target file size of 3GB, you must be setting your bit rate to about 3200kbps. Anything under 3500kbps (4500kbps really) chokes Compressor, and the results you are receiving seem logical (to Apple).
Here are a couple things that you could try. Download an app like DVD2OneX. I often use this app if my initial encodes and project size ends up being just a little too large to fit on a dvd-5 and I don’t want to go back and re-encode the assets. With DVD2OneX, just simply build your disc like normally, allowing the data overflow to occur, and then use the VIDEO_TS folder generated to fit your target size. This is a fast and simple solution. The second option, because you may find that the quality of MPEG-2 at 3200kbps to be too poor, is to re-encode your assets at 352×480. This may sound strange, but if your target playback is going to be on a television (SD), then you will very likely not notice much of a difference between 352 encoded video and 704 or 720 encoded video. Most of the resolution an encoder cares about is stored in the vertical lines, and thus horizontal resolution is very expendable when it comes to attaining low bit rates and small file sizes.
-
On BitVice Helper, I would suggest two things. First, convert the AIFF to WAV with QT Pro and see if that helps. If not, then it likely does have something to do the frame size. Could be BitVice Helper looks for a certain GOP structure, and if you’re using 720p60, then it may not support that. I don’t use BitVice Helper too much these days, as there is a newer program that generally handles more formats. This app is MPEG StreamClip. Download it and then try and save a program stream (Convert to MPEG). If the audio and video files have the same name, then MPEG StreamClip will usually recognize both and automatically combine the two when you open up the video stream in the app. If it doesn’t automatically do so, then you will need to just select the audio and video tracks at the same time and bring them in to MPEG StreamClip.
Most computers can easily handle up to 18Mbps these days for MPEG-2, but if 10Mbps looks good then I would stick to it. Per your problems with Flip4Mac, I find that pretty surprising. I have some issues with it, in that it is slow and the color conversion that it uses isn’t perfect (more a by-product of having to use QT though), but the overall quality of the encode is generally considered quite good. Apple’s H.264 and MPEG-2 encoders are quite good these days (although neither are the best), but I don’t think either is that much better or worse than the F4M encoder in terms of core quality at each’s best settings.
-
Best thing to do is to go to Apple’s QT Developer site or the third party developer’s site for external codecs. There are so many, that your best friend in this area is usually google.
-
Sounds like a color conversion issue. You are probably editing RGB images in AE, and when you go to compress, the color spaces for the codecs are generally YUV. And the conversion the apps and encoders are doing is not good it appears. On the mac, Compression Master is fantastic at this sort of stuff. On the PC, VirtualDub or AVISynth with the right filters is a good route to go as well.
-
I have almost no clue as to what you are really asking, but this is what I could gather and answer:
The QT MPEG-2 encoder is very good (the version included with Compressor 2), although it isn’t the best out there. But for most purposes, it is sufficient. Better encoders can be found with ProCoder and MainConcept IMO. Basically, the software MPEG-2 encoders are so robust and fast these days, especially for SD, that with the right tools, a hardware encoder is almost not needed for most situations.
If you have a 4:3 projector, and you have a 16:9 source, then the best option if you want to maximize space is to crop from the sides to a 4:3 image. Just make sure you know this in the editing stage though, as you may crop some material you don’t want to. Also, if all you are going to be using the encode for is the projector, and you know the resolution of the projector, perhaps 800×600, and you are working with HD, then instead of cropping and scaling to 720×480, crop and scale to 800×600.
-
Squeeze is optimized for multiple processors, but not all of the encoders it includes are. For instance, the encoders optimized for dual procs in Squeeze are the Sorenson Video 3 codec, their MPEG-2 encoder, their Real Video encoder, and the licensed Windows Media encoder. But the H.264, Spark codec, and On2VP6 encoders are not. On2 not being optimized is out of their hands, but the H.264 and Spark encoders should be. In any case, I wouldn’t use Squeeze for H.264 encoding, but the lack of an SMP aware FLV encoder is disappointing.
-
I would actually suggest you not do that and just export to QT without the conversion.
-
There’s Compression Master, MainConcept, ffmpegX, and Digigami.
-
BTW, I think their issues were less to do with poor color and more to do with poor overall quality in the final encode. If they were trying to deinterlace and scale with Compressor, that would likely cause more problems than color space conversions.
-
Because just exporting to QT beforehand (assuming they do an export to QuickTime and not an export using QuickTime conversion) is only going to net a movie that has the same color format as the source. DV is 4:1:1 so Compressor needs to convert that color space to 4:2:0 for MPEG-2 encoding. The same goes if the source is 5:1:1, 4:4:4, 3:1:1, or RGB. If you use Compression Master to convert from DV 4:1:1 to BlackMagic 8bit 4:2:2 and then to MPEG-2 4:2:0 (with either CM or Compressor), the resulting MPEG-2 will generally have better and more accurate color than just going directly from DV 4:1:1 to MPEG-2 4:2:0 with Compressor only. QT (the base core of Compressor) does not do a good job with color conversion and the image processing filters in Compressor are in general very weak.
-CS
“I hate your signature.”