Home of Professional Cinematography since 1996
class="style5" HDCAM Tape Dropouts And Aging
>Published : 29th September 2005
>Has anyone noticed errors or dropouts on older HDCAM tapes?
>I recently have seen some problems with some 24P material played back on a JH-3.
>On pause mode the image on screen had a blockiness in a very small area.
>I am concerned that older archived tape will not be stable in the long run.
>Maybe converting it to optical (Disc or Film) or hard drive storage may be a better way to keep the program around in the long run.
Mark Forman Productions
class="Paragraph">>Has anyone noticed errors or dropouts on older HDCAM tapes?
>You may want to try it on a high end Sony studio deck. The JH-3 is a cheap utility playback deck and may only have minimal error correction.
>I agree with Bob... I've never used the JH-3 but we use J3 decks as feeders, and they're not designed for "stunt" modes - freeze frames, off-speeds, etc.
>They're basically intended for screening, dubbing, and feeding into non-linear systems, and don't have the more sophisticated electronics required to give you a perfect image off-speed.
>This used to be a lot more obvious on analog decks - the less expensive VTRâ€™s would give you a nice tracking bar across the image that left no doubt in your mind that the deck wasn't designed for stunt modes.
>Mark Forman writes:
class="Paragraph">>I am concerned that older archived tape will not be stable in the long >run. Maybe converting it to optical (Disc or Film) or hard drive storage >may be a better way to keep the program around in the long run.
>Probably the most archival way to store digital data would be to encode it as a 2-D matrix (like those 2-dimensional "checkerboard" barcodes or digital optical soundtracks) and record it out to the full width of 35mm B&W film. It would use a lot of stock, but would have the archival permanence of silver-emulsion film and could be easily duplicated. With color emulsions you could probably multiplex your data and achieve a much higher data density, though you'd lose a bit of robustness. I think this approach would be ideal for long-term storage of expensive DIâ€™s.
>No new hardware would be needed -- just the software to translate the data stream into the 2-D matrix, which you'd then feed to any laser (or other) filmout recorder. Or maybe a custom recorder that could record between the perfs and almost out to the edge. (Mr. Kreines, are you listening?)
>I've been suggesting this approach for about two years, and so far nobody has so much as *commented* on it -- even to shoot it down. So never mind.
>Anyway, I think the next most robust archival approach is to archive to at least two different types of media, on the theory that the one that starts deteriorating first will sound the alarm for you to immediately start cloning (reduplicating) the other one.
>Two different brands of hard drive might actually fill the bill.
>It's too bad magneto-optical discs are so slow -- As far as I know they're about the most archivally permanent medium around. They're immune to magnetic fields, and, if I'm not mistaken, even their reflective coatings aren't necessarily 100% critical to retrieval of the stored data, which exists as tiny bubbles INSIDE the medium, not on a coating.
>Dan "archival is good" Drasin
Marin County, CA
class="Paragraph">> (Mr. Kreines, are you listening?)
>We've actually been working on something like this, very intermittently, for several years.
>However, the promises of holographic storage would immediately kill off this on a cost per gig basis.
>I have some of the first HDCAM tapes introduced by Sony. No dropouts, no tears. Perfect condition. All tapes are run through the HDW 500 series decks. Can't speak to the J-3 decks.
>We have 3/4" (1980) tapes that still look great.
>Jeff Kreines writes: (Re: 2-D digital data storage on B&W film)
class="Paragraph">>We've actually been working on something like this, very intermittently, >for several years...however, the promises of holographic storage would >immediately kill off this on a cost per gig basis.
>So how close to production is holographic storage? With 4K cameras already rolling off the assembly line we're gonna need SOME kind of archival medium yesterday!
Marin County, CA
class="Paragraph">> We have 3/4" (1980) tapes that still look great.
>---It is entirely possible to practice good archival procedures and expect this kind of longevity.
>Hats off to you.
Freelance Shooter and Editor
> We have 3/4" (1980) tapes that still look great.
I second the above. The original writer who wrote in about tape "Dropout" was most likely viewing some other artifact. As a practical matter, dropouts are a thing of the past. I routinely witness:
>- Ancient tapes like noted above that play back without dropouts, with no special storage or preservation technique.
>- Tapes that are used for months or years for dubbing and replay which exhibit no dropouts, even though they have hundreds of passes on them, and you'd think the oxide is falling off!
>- The advent of metal tape made dropouts essentially a thing of the past. We rarely see it anymore.
>It would seem therefore most unlikely that modern HD tapes with a few passes on them would exhibit any dropouts.
>Lew Comenetz - Video Engineer
class="Paragraph">>We have 3/4" (1980) tapes that still look great.
>It took me a couple of weeks last year to dub all my 3/4" tapes from the early 1980's to DV (DigiBeta seemed like expensive overkill), mainly because it took a couple of run-throughs on each tape before it would play without clogging the heads. My "archival" procedures had been to put them in a cabinet in the garage for half that time, so mouse droppings were an issue.
>I also had some EIAJ 1/2" tapes from the Porta-Pak era which basically melted. More to the point, a fair number of my data tapes from the early 1990's (mostly DDS-1) have failed. Data tapes (TAR) have a tendency not to play through dropouts, which is a virtue of the video recording model. Error concealment is better than non-recoverability, for pictures anyway.
SFD Vfx & creative post
Santa Monica, CA
>It is easy to have a tape drive read through an error when using tar ( using "e" switch) but an issue will be the file format - if you stored the material as a sequence of images (image.00000.cin, .tiff, etc.) then you'll only have one or a number of individual frames corrupted. If you stored the material as a single movie file (QuickTime, avi, mpeg, WM9, etc.) then it's more difficult to recover.
>Some applications will try to play it and bomb out when they get to the corruption, some won't even try. We've run into that with some material on "vintage" Exabyte tapes. Usually, we are able to recover the material (approx 10 or more yrs old) but when we do the artists realize that it doesn't look anywhere near as good as they remembered it and they're better off redoing it using 2004/5 technology then trying to reuse 10 year old computer imagery.
50 East 42 Street
New York, NY 10017