View Full Version : Is this the best DIVA video quality?
03-05-2011, 07:13 PM
The sample videos I've seen on the Chasecam website are encoded in MPEG2, whereas the DIVA seems to use MPEG-4/AVC/Main@L3.1 as its raw SDHC .mp4 format. Can you view this and let me know if this looks correct? It seems very poor quality (pixellations, lots of digital artifacts):
This is encoded "sd for youtube" since I'd imagine that's the ideal encoding for youtube, although I don't see any better quality using MPEG4/SD,HighQuality,LargeFile (and other output video profiles). This is with VLC and K-Lite codecs loaded. When I look at GoPro SD recordings (the old GoPro cameras) they're also using MPEG-4/AVC/Main@L3.1 (although progressive not interlaced) but are much better quality, so it makes me think either I'm doing something wrong or there's something wrong with my equipment.
Note that before youtube encodes the file, the output file does look better locally, but not by much. That is, all of the artifacts remain, they're just not as egregious (the gauges and text are especially much clearer pre-youtube).
Thanks for any feedback.
03-05-2011, 08:18 PM
Can you view this and let me know if this looks correct? It seems very poor quality (pixellations, lots of digital artifacts):
That's out of a DIVA? To me, it has the same "PDR look" as the PDR100. Since the DIVA encodes with a different codec, I wonder if the underlying issue is the video camera or the video capture hardware in the recorder. It would be interesting to know if the same capture hardware (or at least the same chip manufacturer) is used in both products.
This is encoded "sd for youtube" since I'd imagine that's the ideal encoding for youtube, although I don't see any better quality using MPEG4/SD,HighQuality,LargeFile (and other output video profiles). This is with VLC and K-Lite codecs loaded.
Any time video is re-encoded, quality will degrade. Whether that is noticeable by someone depends on the person and how the re-encoding was performed.
All of the video codecs people are likely to encounter are of the "lossy" type - they throw away information that is considered to be less important to the picture. Once discarded, you can't get it back. An example of lossless compression is the ZIP format - what you get out is exactly what you put in. There are lossless video codecs, but you're very unlikely to encounter them outside of certain professional video settings.
Each time a video is re-encoded, more information is lost forever.
That's why any demonstration videos should include a sample of the raw output from the recording system. That's what is needed for a user to determine if the quality provided is satisfactory for their application. Users can always re-encode for smaller files, etc. to suit their needs.
Of course, if a user's requirements include video overlays, captions, etc. seeing the output of that step is also important as it is re-encoding and degrades the video quality to some extent.
This also means that any "do you see this problem" videos should be the output of the component thought to have the problem. Meaning, if you think the recorder is causing the issue, a video that has been through the Youtube mangle isn't the best way to demonstrate the issue - Youtube's concern is storing as much video as possible in the smallest space, not quality.
And now, a rant about codecs for playback. I'm not singling out you for this - I've read a couple "I installed codec X" posts here today, and it sort of touched a nerve.
You need to install a codec if your system doesn't have native support for the particular format used. For example, Windows XP doesn't provide a codec that understands DVDs.
You might also install a codec if you feel it provides a better visual experience than the default one that comes with your operating system / media player / etc. or if you need features only available in a "professional" version of a codec.
But installing a codec to "fix" a video that plays but has problems is not a true fix. In almost all cases, the problems are with the source encoding, not the playback decoding. Fix the source, not the player. Just because one codec can deal with an error introduced by the encoder doesn't mean that all the other codecs for that format are broken. There are certainly some cases where getting a usable playback picture is most important. But the general case of installing a codec just because someone says it is "better" is a waste of time and can cause other problems. For example, many "codec packs" available contain many codecs. I've seen some which contain 3 or more codecs for the same encoding method.
There certainly can be cases where a widely-distributed codec has a problem. But if you've got a relatively modern operating system and media player, ask yourself if it is possible that millions of Windows users are playing this type of video without any problems, and you're the first one that's found a problem - or if it is more likely that the source was encoded improperly.
Most of the hard work is done by the encoder - it usually doesn't have to work in real time and can take as long as it wants (subject to limits on the user's patience) to make the best video it can. The player just needs to be able to decode and play the video in real time.
Obviously, the "take as long as you want" doesn't apply to things like the ChaseCam recorders, video security systems, and the like - those DO need to operate in real time. But they are generally running on dedicated hardware with specialized chips to speed the encoding.
None of this is intended to single out ChaseCam products - all of the above applies equally to all video capture and encoding systems.
For those who are wondering, I have a background in professional video - I was one of the co-founders of BTG which was the technical development arm of Broadway Video. I've also consulted on hardware and software development of audio and video encoding systems.
03-07-2011, 08:56 AM
Hi Terry. Thanks for the info, I appreciate it. I don't have a place to store the original .mp4 online so I used youtube as one place everyone could view it. I'm still stuck as to whether the poor quality I'm seeing is an equipment issue or I'm doing something wrong on my end, since I'm a new Chasecam customer and don't have experience with their products. I'll contact Chasecam customer service directly and if their email allows it, email them a snippet of the original .mp4 to see what they think.
03-07-2011, 05:10 PM
It looks to me to be very possible that you have a camera lens problem. If you can send us the camera, we can test that.
It would be interesting to know if the same capture hardware (or at least the same chip manufacturer) is used in both products.
No, not at all. Totally new and completely different.
03-08-2011, 05:17 AM
Thanks Randy. I wanted to upgrade my camera to HD anyway, so I'll be sending it in. I'll drop you an email as a follow-up.
06-01-2011, 09:17 AM
Just to complete the thread: after sending my camera in under warranty it was returned with no change in quality, so it appears the above is the best video quality to expect from a DIVA SD.
vBulletin® v3.8.2, Copyright ©2000-2013, Jelsoft Enterprises Ltd.