Am I too stupid or is there a bug in the codec?
- Thread starter ThomasS
- Start date
HELLO, snicky. Sad for ThomasS but this is the first cam that shows these particular symptoms and may just be a 'lemon' as stated earlier. The real big issue is at this thread.
http://www.dvxuser.com/V6/showthread.php?t=172958
This is the number 1 issue. So take a couple of weeks, grab some popcorn and read. Informative AND entertaining.
Uwe Lansing
Veteran
Where one lemon is, there are more for sure. It´s pretty unlikely that pana has produced only one pal-camera with this flaw. I`d recommend to everyone in a pal-land to check it out immediately after purchase. Cheers.
Sorry guys, bad news.
I have bought two of these 8GB cards, only for using them with this camera, and the problem occured with files on both of them (I'm not that unexperienced with defective media to rely only on one card).
If there's not a general problem with these Sandisk cards, it'a bug in the codec, because DrBlaz (post #19)checked the MTS files and they are OK, which would not be possible, if the media is defective.
I have bought two of these 8GB cards, only for using them with this camera, and the problem occured with files on both of them (I'm not that unexperienced with defective media to rely only on one card).
If there's not a general problem with these Sandisk cards, it'a bug in the codec, because DrBlaz (post #19)checked the MTS files and they are OK, which would not be possible, if the media is defective.
I don't think your MTS file is ok. I looked at the original and saw the skipped frame issue in it. You should take it back and get it replaced. This happens to every camera when it's released. Some of them are lemons and that's just reality. Someone in the 5D forum got a camera that also skipped frames and he's getting it fixed/replaced.
I'd be careful about trying to make definitive conclusions without being certain about the problem when the easiest thing to do is swap out your camera for a functioning one.
Martti Ekstrand
Veteran
I beg to differ with your assessment that it's a bug in the codec. I'd venture that the read-out from the sensor is interrupted prior to the encoding for some reason. When viewing the file straight in VLC without conversion it doesn't actually skip frames but the stream is damaged yielding short strange interlaced sections. I doubt that this can be caused by the codec or a bad SD writer-reader / SD card for that matter. I've attached a screen grab from VLC as an example. When converting this in FCP the damaged sections are ignored causing the resulting file to have freeze-frames here and there.
So in all likelihood you got a defective unit. Unless of course your kid is possessed by a demon who causes a rift in the space-time continuum just to tease with you.
Attachments
That's VLC struggling to decode the video. It doesn't play MTS files properly. If you play it in Toast 10 it should play fine and you can see where the frames are skipped by scrubbing frame by frame.
Martti Ekstrand
Veteran
OK, but then shouldn't I see such artefacts in all files I play, not just in this one?
the h.264 encoder "chops" the image in macro-blocks and if these MB don't change simply mark them as "SKIP", in the MTS provided by ThomasS all MB in the repeated frames are marked correctly as SKIP.
What does it mean? at least the mem card is not corrupted! but it could be due to a lot of reasons, we should know more about GH1 hardware.
if the processing before the codec sends repeated frames then the output will be a perfect ACV file with repeated frames.
What does it mean? at least the mem card is not corrupted! but it could be due to a lot of reasons, we should know more about GH1 hardware.
if the processing before the codec sends repeated frames then the output will be a perfect ACV file with repeated frames.
Martti Ekstrand
Veteran
Make that all PAL files, I do see some odd stuff in NTSC files.
My assumption to take that as a bug in the codec comes from a different perspective.
In my day job, I'm working in the software industry. Seeing that panny managed to deliver this product excactly on time (beginning of june, as announced) can nowadays only mean, that they have put an enormous pressure on the developers to bring it on the market.
If your primary goal is to manage the given target date, you're willing to accept, that you deliver with a certain number of errors, or you simple shorten the test phase for e.g. unlikely test scenarios.
This has nothing to do with conspiracy theory, it's simply common business practice.
That can mean, that it (probably) can be fixed with a new firmware, but I cannot predict if or when, depending on the development schedule/ priority.
For my part, I'm not willing to give Panny a credit, put the camera on my desk and wait for an update that eventually fixes this problem. Or use it only in a way where I absolutely can control the environment where I'm shooting, restricting the freedom in the way I'm shooting, because I bought this cam exactly for these unpredictable run'n'gun or stealth shootings. Otherwise I would use my HVX, which in many situations unfortunately attracts too much attention.
In my day job, I'm working in the software industry. Seeing that panny managed to deliver this product excactly on time (beginning of june, as announced) can nowadays only mean, that they have put an enormous pressure on the developers to bring it on the market.
If your primary goal is to manage the given target date, you're willing to accept, that you deliver with a certain number of errors, or you simple shorten the test phase for e.g. unlikely test scenarios.
This has nothing to do with conspiracy theory, it's simply common business practice.
That can mean, that it (probably) can be fixed with a new firmware, but I cannot predict if or when, depending on the development schedule/ priority.
For my part, I'm not willing to give Panny a credit, put the camera on my desk and wait for an update that eventually fixes this problem. Or use it only in a way where I absolutely can control the environment where I'm shooting, restricting the freedom in the way I'm shooting, because I bought this cam exactly for these unpredictable run'n'gun or stealth shootings. Otherwise I would use my HVX, which in many situations unfortunately attracts too much attention.
In my experience there are always lemons. It sounds like your camera isn't functioning properly. I've looked at other PAL footage that looks perfectly fine.
Isaac, I don't doubt that most of the footage is fine, also 95 of 100 clips i shot were fine. It's the 5% corrupted ones I'm worried about. Actually, I' d prefer for my own peace of mind that I only got a bad copy. But considering this problem, the panning mud issue and the color rendition in critcal situations, I for my part find, that the cam is - not yet - a finished product.
I hear where you're coming from and the AVCHD issues are why I haven't placed an order yet. It's the reason I'm pushing hard for an mjpeg firmware update that bypasses the AVCHD codec altogether.
John Caballero
Veteran
Tell that to whoever shot these two.
http://www.youtube.com/watch?v=tHgAu...eature=channel
http://www.youtube.com/watch?v=Lvxah...eature=channel
John,
this footage surely looks good. Although, if you observe the shadow areas of the trees on both sides of the river you'll see, that it "muds" the details (not shure if it's youtube or the original footage).
I also did several terrific shots with this cam, but that's not the point. From a camera with these specs and price point - compared to other consumer AVCHD cams, I'd simply expect it to work not only in 95% of the clips I record.
this footage surely looks good. Although, if you observe the shadow areas of the trees on both sides of the river you'll see, that it "muds" the details (not shure if it's youtube or the original footage).
I also did several terrific shots with this cam, but that's not the point. From a camera with these specs and price point - compared to other consumer AVCHD cams, I'd simply expect it to work not only in 95% of the clips I record.