How do I avoid video artifacts exporting for Final Cut Pro?

From my experience, Mp4 isn’t the greatest codec. The H.264 certainly helps though. How much does the file size increase if you crank the bit rate higher?

As it sounds like you know, each video is different so a static setting won’t look perfect for everything. It sounds like raising the bit rate even higher may be your only option. I’d certainly help if could tell us more about the video itself.

I’ve had similar problems with FCP encoding to mp4 but I disagree with Spargett – mp4 is great format when the encoding is done right. Therein lies the problem: I don’t think FCP does a very good job of mp4 encoding and it certainly doesn’t offer a ton of options.

You could export to DV or some other high quality format (if you’ve got the drive space) then encode to mp4 using something like ffmpeg where you’ve got a lot more options. A post on the discussion I linked to recommends using VisualHub though I haven’t tried it personally.

Hey, thank you. Good suggestions. yeah, I was surprised because the MPEG-2 came out so nicely. I don’t even remember how I did that compression though- it was awhile ago- I was compressing to put onto a DVD in DVD Studio Pro.

Well, here are some details if that will help. The project was shot on 16mm, went through telecine to mini DV tape, edited and outputted on DV (i.e. I never went back to film). This may be a dumb question, but could it have to do with the frame rate since I shot on film. I’m guessing No, since I assume the conversion to DV takes care of that.

@paulC I’ve exported it to a full quality Quicktime file through the export to Final Cut Movie, keep self contained option and have an uncompressed file about 2Gb. It’s on my external and I’m cool with it there, but I’d like to have a compressed copy to put on my laptop and itunes (and maybe a lower quality one to stream, later). So I’m assuming this will be the best quality video to compress with another program, right? I’ll look into some other programs.

The file size I’m getting for about 15 minutes of video is around 150MB and I’m okay with that. I’d even be okay with a 200 MB file I think. My worry is that I have the settings wrong ( the frame rate thing) or that I could be oversampling, if that possible, Spargett. I did find a varible bit rate option and am trying that now.

Could it be that I need to de-interlace the video? I’m wondering why the Mpeg-2 turned out fine. As a rough guideline, I tired to follow the compression specs for Hotel Chevelier, the free Wes Anderson film from Itunes, which looks great in Itunes, is around 15minutes, and 150 MB. They probably had some crazy awesome program compressing it though, not to mention much better sources.

thanks

I think the median you are aiming for, deinterlacing could be a really effective way of cutting file size without sacrificing (relative) quality.

Might I recomend using Vimeo to host your video? They’re light years ahead of YouTube, etc.

I deinterlaced and did Variable Bit rate and it looked better (artifacts gone) but I got a 600 MB file size. I’m going to try deinterlacing but limiting the bit rate to 1400 or so.

There are also a couple of jerky spots. I’m wondering if this is errors in the compression, or something to do with my slow processor (I’m running a 867MHZ G4 and dealing with the long time- not exactly the perfect setup for this kind of thing). Or I wonder could the jerky movement have to do with Keyframes- I did 1 per 24 frames.

More and more I’m thinking a third party will have a simpler solution.

As far as hosting, I’ll look into it- thanks for the suggestion but this is more something I’d want to stream from my own site.

No problem. You can simply embed the video uploaded to Vimeo on your site and spare yourself tons of bandwidth.

Try exporting using Compressor instead of FCP. It tends to do a much better job.

One problem you might be having is that there are no compression markers in your footage. If you were to right-click on the edited sequence in FCP and send it to Compressor, each edit point is automatically treated as a compression marker. Since you brought in your edited piece from a DV master tape, FCP doesn’t have any edit points in the file (assuming I understand your post correctly). You could add the compression markers manually at each edit point, which is kinda’ a pain, but it does improve things.

Compression markers tell the software: “This is the beginning of a new shot, so don’t compress it with information from the previous shot.”

@rory

thanks good thought… I didn’t know about that…. well more fun with compression to day… I have to go into my edit on another project late tonight and output newly synced clips to new QT files. I think I’m not going actually compress them though- keep them in the same codec. I really need to learn how to work compressor and all this stuff…