Greetings--
I'm using avidemux to edit some videos taken with a digital camera (Sony DSC-T30). The camera gives me MPG files. After I edit the file in avidemux, I change "copy" to a particular video codec and create a new file. The problem is that the new file is always of a poorer quality than the original. I have experimented with different codecs and settings, trying to get the highest quality possible. In fact, I don't really want to compress the video, but it seems that I have to do so in order to save my changes. Even the lossless codecs (FFV1 and Huffyuv, I think?) produce videos of a lesser quality than the original.
Perhaps there is a simple solution or maybe I'm doing something wrong (I'm new to video-editing, but that's probably obvious), but I think my question is "How can I save my changes without decreasing the quality?"
Thanks!
+ Reply to Thread
Results 1 to 8 of 8
-
-
What kind of editing are you doing? You can use copy mode if you are just cutting,joining but I guess you are using some filters.
-
I found a sample MPG video from that camera. It's 640x480, 30 fps, MPEG 1, all I frame, ~9600 kbps, constant bitrate, progressive.
http://www.dcresource.com/reviews/sony/dsc_t30-review/MOV00014.MPG
I don't know FFV1, but HuffYUV should not be of lesser quality -- but it will be much larger than your source MPG file. I recompressed the above sample with HuffYUV and the output was visibly identical to the source.
If you recompress with MPEG 2 at the same bitrate and similar settings as the original the quality should only degrade a little bit.
How are you comparing the the before and after videos? You can't use two media players playing the two videos side by side. The one you start first will be using your graphics card's video hardware, the other will be using the Desktop. The two usually display video differently.Last edited by jagabo; 6th Nov 2010 at 05:49.
-
Thanks for your help--
Baldrick: Yeah, I'm doing some simple stuff with the fade and subtitle filters.
jagabo: Good to know that about comparing videos. I had been using mplayer and switching between workspaces. Nevertheless, even without simultaneous comparison, the difference is still apparent. I took screenshots of each video (non-simultaneous playback) at the same frame. The HuffYUV image (hhhhhhhh.avi) is definitely better than the MPEG 2 (mpeg2_like_camera.mpg), but it is still not as clear as the original (silent_scenery.mpg). Of course, the original, as you'll see, is not very clear to begin with (it was taken through a dirty window on a train); I certainly don't want it to get any worse.
One note about the screenshots: they were all taken at frame 47, but for some reason the HuffYUV one is slightly advanced. Not sure why. -
The HuffYUV image is more blurry but it isn't exactly the same frame as the original image. Assuming it is representative of the overall video, then the blur is happening before HuffYUV, or possibly during playback. Are you applying any other filters (say, noise reduction) in AviDemux? Did you change the frame size?
When you provide sample images you should provide them at the frame size of the video, not scaled to full screen or scaled to correct for the aspect ratio. Otherwise you can't tell what artifacts are in the original frame and what artifacts are created by the scaling. Also, with screen caps from a media player you don't know what processing the media player has done. The player may be treating different sources differently. VirtualDub is good for examining video files (it displays them onscreen pixel-for-pixel by default) and making snapshots.
I just tried using AviDemux to convert the sample file I linked to earlier. The program appears to be performing some deblocking or noise reduction of MPG sources. I think that is what's causing your problem, not the HuffYUV compression.
Yes, here it is: In AviDemux select Edit -> Preferences, then go to the Video tab. Disable Horizontal and Vertical Deblocking, and Deringing. You may have to exit and restart the program for the change to take effect. Assuming that clears up your videos, go back and enable those options but change the strenth. Some deblocking is usually helpful -- and your source is pretty blocky.Last edited by jagabo; 7th Nov 2010 at 07:11.
-
I'm not sure why the HuffYUV shot is not in the same place--one can see in the bottom-right corner of each image that they were all taken at frame 47.
For these videos, I did not apply any filters. Between opening the original and saving the new files, I only set the A and B markers.
Thanks for the tips. I will check out VirtualDub. If a quality difference is still apparent, I'll try to get some shots from that.
[edit] I see that VirtualDub is a Windows program. Don't have Windows, so I'll be looking for a similar program for Linux... -
I added some text to my earlier post. I think it explains what is going wrong with your conversions.
Similar Threads
-
Best format to use to maintain video quality?
By Blackhawk1 in forum Camcorders (DV/HDV/AVCHD/HD)Replies: 10Last Post: 2nd Apr 2013, 01:17 -
Can't compress and maintain quality. Why not?
By justaguy23 in forum Video ConversionReplies: 7Last Post: 10th Jun 2010, 01:55 -
HFS10 - AVCHD how to maintain quality?
By Machoman in forum Camcorders (DV/HDV/AVCHD/HD)Replies: 30Last Post: 18th Jun 2009, 11:57 -
How do you lower bitrate but maintain quality?
By zanos in forum Video ConversionReplies: 7Last Post: 11th Dec 2008, 00:43 -
maintain quality output file after cutting video
By karebo in forum EditingReplies: 14Last Post: 24th Feb 2008, 05:47