kevin_standlee (![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
kevin_standlee) wrote2015-07-10 01:20 pm
kevin_standlee) wrote2015-07-10 01:20 pm
Entry tags:
Westercon Business Meeting Videos
The Westercon Business Meeting was just over twelve minutes long this year, including site selection results and the initial adoption of a motion to lower the meeting's quorum from 15 to 10 members. This made the video a good test-bed for the different file types and resolution rates in MovieMaker. Accordingly, here are three different versions of the video, in increasing order of quality.
This is the "quick-and-dirty" MP4 that the proxy card generates directly in the camera. The video is relatively small, so if you full-screen it, you'll see the degradation in the video quality.
This is almost certainly the quality/size you can expect to see of the WSFS Business Meeting videos as we upload them as fast as we can get them off of the camera at Spokane.
Converting the native MXF files to medium-resolution video and combining the video and audio results in this file. It takes a double conversion, because MovieMaker doesn't like the converted MXF files very much. You have to first import the converted files into MovieMaker, use that to make a WMF file, then start a new MovieMaker project with the WMF file as the input; you can then add the titles and credits. This also lets you bring in the second audio channel, which is the camera's onboard microphone. That lets you get a little bit more of the audience sound, instead of being completely dependent upon what came through the head-table microphones. But it's slow. Call it more than an hour of futzing around to do a 12-minute video, not including the upload time. The file itself is about 28 MB.
This is the higher-definition (2.1 mbps) version of the file, which runs the file size up to 186 MB. The video quality is much better, but it roughly triples the amount of time it takes to make the video in the first place. Whether we are ever able to edit together a full-length, higher-quality video from this material is doubtful, as it would probably take days just to generate the files.
There is at least one higher-quality setting in MovieMaker, but that predicts a 2.5 GB file for a mere twelve minutes of video, which seems a little extreme to me.
Lisa is still unhappy with the results, because she's not been able to get the camera's back-focus just right, and because this is a pro-grade camera, it tends to be far less forgiving of tiny errors.
Note that the sound is excellent on all three versions and is not significantly affected by the video resolution. That's crucial. We want people to be able to hear what's going on.
ETA: If you feel ambitious about doing the conversion/editing yourself, I've put a link to the raw files in a comment to this post. Be sure to read the warning about the file extensions.
This is the "quick-and-dirty" MP4 that the proxy card generates directly in the camera. The video is relatively small, so if you full-screen it, you'll see the degradation in the video quality.
This is almost certainly the quality/size you can expect to see of the WSFS Business Meeting videos as we upload them as fast as we can get them off of the camera at Spokane.
Converting the native MXF files to medium-resolution video and combining the video and audio results in this file. It takes a double conversion, because MovieMaker doesn't like the converted MXF files very much. You have to first import the converted files into MovieMaker, use that to make a WMF file, then start a new MovieMaker project with the WMF file as the input; you can then add the titles and credits. This also lets you bring in the second audio channel, which is the camera's onboard microphone. That lets you get a little bit more of the audience sound, instead of being completely dependent upon what came through the head-table microphones. But it's slow. Call it more than an hour of futzing around to do a 12-minute video, not including the upload time. The file itself is about 28 MB.
This is the higher-definition (2.1 mbps) version of the file, which runs the file size up to 186 MB. The video quality is much better, but it roughly triples the amount of time it takes to make the video in the first place. Whether we are ever able to edit together a full-length, higher-quality video from this material is doubtful, as it would probably take days just to generate the files.
There is at least one higher-quality setting in MovieMaker, but that predicts a 2.5 GB file for a mere twelve minutes of video, which seems a little extreme to me.
Lisa is still unhappy with the results, because she's not been able to get the camera's back-focus just right, and because this is a pro-grade camera, it tends to be far less forgiving of tiny errors.
Note that the sound is excellent on all three versions and is not significantly affected by the video resolution. That's crucial. We want people to be able to hear what's going on.
ETA: If you feel ambitious about doing the conversion/editing yourself, I've put a link to the raw files in a comment to this post. Be sure to read the warning about the file extensions.
no subject
The other thing I notice from your post (in the light of day) is that it looks like you're doing a format conversion from 4x3 to 16x9, based on the resolution that you quote. Anything that involves a format conversion almost always takes longer and eats more CPU power than something that runs effectively pixel for pixel.
This sample seems to be at lower resolution than the original test clip that Lisa shot. If I recall correctly, she's shooting at lower res now to try to accommodate the proxy card in the camera. Is that the case?
Good luck!
no subject
Lisa is indeed shooting at a lower resolution because it seems to make the MP4 that feeds into the SD card (via the proxy card) come out better. We're going to buy a faster SD card today if we can find one and see if this makes a difference. We have a theory that the speed of the SD card has been one of the limits on how good the MP4 comes out; that is, if the card can't accept the data fast enough, it gets a worse file. The camera also stores a copy of what should be the same MP4 on the P2 card itself, so we should be able to do some side by side comparisons.
no subject
The extra cores appear to be a good thing.