direct to Disk - looks like a drop out ???

[Peter Dolman]

This is an interesting one, albeit very frustrating ...

I was adjusting back focus this morning .... captured direct to disk on my desk top ... to better see the results
The attached stills appear to have drop outs ... yet these were captured direct to disk ... does anyone know how this could happen ?

I have been struggling with what I thought were tape drop outs for some time now ... even had the heads professionally cleaned yesterday ... still had what I thought were drop outs ... apparently a different issue

has anyone else ever experienced this?

http://www.vidprostudios.com/images%20page.htm


cheers

[William Hess]

Sounds very much like a firewire issue. Are you using a Mac or PC? What type of firewire card are you using? If there are other firewire devices plugged into the same card, unplug them. Also try a different (shorter?) firewire cable. Did you attach the supplied ferrite core to the firewire cable as described on page 54 of the manual? (don't know if it helps, but it certainly couldn't hurt). Also, if you're running Windows XP I'd recommend installing SP2 if you haven't already.

[Peter Dolman]

William
thanks for your reply
Unfortunately;
I can reproduce the same issue on both my desktop and my laptop, using 3 different firewire cables on each ... both are running WinXP Sp2

additionally; these "drop outs" are the same shape, size etc ... as what I have been seeing when recording to tape

thanks for your possible solutions
cheers
Pete

[Keith Wakeham]

I've never had experience with this but I'm guessing you might need to get it professionally looked at and I'd guess that it has something to do with the HDV codec chip.

The HD100 has two outputs from each ccd with 6 analog frontends, for the artifact to continue across the entire chip means that the problem exsists after the ccd, the timing devices and the analog front end.

The next things in the chain is dsp then codec then output chips (firewire encoding and drive mechanism)

I would suggest setting it to dv and see if you get any of the same artifacts but i will suspect you won't. The reason I suspect it will be fine in DV mode is because it looks more like an issue with the codec because the artifacts are 16 pixels high. Not 100% sure, but close to it that the DV codec is 8x8 DCT blocks while the HDV is 16x16 dct blocks (plus other encoding). Since both images show 16 pixel heights its more likely that its a problem within the hdv codec and I would suspect its a problem that isn't in the I frames but more the B and P frames. If you could post a short 2 second mpeg clip I could narrow it down a little more but its not like I can tell you how to repair it. Sorry.

If you can't get it to reproduce the artifacts in DV mode that rules out the DSP and the output device (port or tape) and leaves only the codec chip which means you need servicing from the manufacturer.

[Peter Dolman]

Keith

great help !

when recording direct to disk in DV mode I received no drop outs/glitches

As you requested; I have loaded a 2 sec M2t clip - same page link as the images ... scroll down to M2t hyperlink
this should give you lots to work with
thanks again for your help
cheers

[Keith Wakeham]

Thanks for posting the clip, i've just tried to review it and log the artifacted frames and have found theirs much less predictability than I expected. It artifacts though I, P, and B frames so I would suspect that the ram that the hdv codec chips uses has gone bad or part of the ram system. So long story short call JVC and get it looked at, they will pretty much end up swaping out the compression board for a new one and ship it back to you. If you're still in your warrenty period it should be free.

Here is what I think is happening. When it calculates B frames from I and P frames it needs to compare the current frame with the I and the P. As near as I can figure this occurs after the a DCT transform. So when it gets saved its memory is going out of sync. How bad is hard to tell but its fairly consistent. Memory requires a latch, a signal that when it changes states it samples the data line, this is a very very small instance. I would guess that its getting out of sync by a single clock cycle at most and less in most cases. When it latches things haven't fully changed state.

This could be anything from a bad clock to the memory to a PLL thats gone bad but its around the memory or memory codec area so not user servicable and most of these chips are very thing quad packs or worse BGA so a local shop probably will end up ordering the board from jvc anyway.

This is my take on it, I could be way off. The techs at jvc know how this works better than me.

[Peter Dolman]

Keith
that makes some sense ... good explanation
it's interesting too that although that 2 second clip had 4 or 5 glitches ... the next 12 seconds following are perfectly clean ... and then 3 more glitches in about 1.5 seconds and then clean again for another 8 or so seconds

The good thing is; this is a loaner cam from my dealer. My camera went in for service for some other unrelated (but equally baffling) issues about 5 weeks ago ......... by way of the south pole apparently! ...kidding - my dealer here in Vancouver has been nothing but helpfull so far
Unfortunately, I have a live event scheduled this coming weekend and as it is, capturing with this camera will be impossible ... hopefully my camera will find it's way home in time.

Thanks for your ideas, very much appreciated
cheers

[Edwin Beeler]

Hi Peter
Was JVC repairing the drop outs? Was it a HDV-codec-error? Because today I was first time shooting with this camera, and back home viewing the result - it was horrible - almost every minute, sometimes every 5 seconds, sometimes every 5 minutes, those "drop outs" did happen, similar to yours. Or is it a thing for head cleaning? Tape brand issue? I'm using JVC PROHDV-tapes, not Sony tapes.
Any answer will be much appreciated.

[Peter Dolman]

Edwin
unfortunately I do not know the real issue
I now have a different camera - no issues

My previous camera definately had dropout issues both on tape and direct to disk. I found in the end that the direct to disk issues were firewire related (must attach the provided insulaters on both ends of the cable) yet the tape related dropouts continued.
Are your dropouts on tape? If so, and you have cleaned the heads and you still have dropouts, I suggest that you send the camera in for servicing right away ... that issue is not going to go away by it's self
Sorry that I can not be of more help
cheers
Pete

[Edwin Beeler]

Peter
I was working with Tape only. So maybe working with Focus/Firestore would help, but still I want to work with tape, of course. - I go to my dealer today. Thank you for your help!

[Daniel Patton]

Originally Posted by Edwin Beeler

Peter
I was working with Tape only. So maybe working with Focus/Firestore would help, but still I want to work with tape, of course. - I go to my dealer today. Thank you for your help!

Edwin, I was having the same experience when going to tape (far too often). Head cleaings did not fix this issue. So on this last shoot I took two of the new tapes after opening them and forwarded to the end / reversed back to the begining before recording on them... night and day difference! I'm surprised at how even my start and stops are clean compared to before. This is nothing new and has been talked about on this board for awhile now, it's just the first time I have tried it and I can say it honestly worked wonders for us. Have you tried doing this yet?

[Edwin Beeler]

Hi Daniel
No, I just used head cleaning tape (sony brand) a while ago, may not use it too often. So yesterday I brought the camera back to my dealer... Maybe JVC-Switzerland will make a firmware upgrade, as it's representative told me. But if the service will not help, I'll try your advice. Thank you very much.

[Edwin Beeler]

Originally Posted by Daniel Patton

Edwin, I was having the same experience when going to tape (far too often). Head cleaings did not fix this issue. So on this last shoot I took two of the new tapes after opening them and forwarded to the end / reversed back to the begining before recording on them... night and day difference! I'm surprised at how even my start and stops are clean compared to before. This is nothing new and has been talked about on this board for awhile now, it's just the first time I have tried it and I can say it honestly worked wonders for us. Have you tried doing this yet?

Hi Daniel - one more question - did you forward and reverse those two tapes only or do you have to go through this process with all the tapes? And by the way, what kind of tape brand/stock do you recommend? - thank you very much

[Carl Martin]

Originally Posted by Edwin Beeler

Hi Daniel - one more question - did you forward and reverse those two tapes only or do you have to go through this process with all the tapes? And by the way, what kind of tape brand/stock do you recommend? - thank you very much

It is aleays recommended to take up the slack in all tape formats before recording. Using your finger does the trick. Fast fw/rw is more ware on the camera.

[Daniel Patton]

Carl, although I agree it might be more wear on the camera (although not likely for the heads since they are not even engaged), but considering the end result when this is not done like dropouts, etc., who cares if I put some wear on the system if that end result is better? And although not everyone has a deck, we save such general wear by using ours for this task now as well as digitizing (another good reason for a deck then). I'm not so sure that just removing slack is the only issue here. I have recently been wondering if when they produce each tape off the spool if they are not doing so at times to poor standards (or simply poor consistency). This might also explain why some people get better results with less drop-outs even when using cheap tapes. Could they be wound better?


Edwin, I'm not sure I follow the question.
If you are wondering if repacking a previously used tape (one you have shot with) helps before digitizing then I would guess the answer is no.
I have only started repacking with the last 2-3 tapes used, but due to getting better results and less drop-outs this way, I will repack every new tape before using it with this camera from now on.