Lawrence Bansbach
Forum Replies Created
-
[toke lahti] “That’s even older codec than dv. Why stick with legacy codecs? This is afterall about tapeless workflow at least before the master and archiving. I expect some new codec from this millenium…”
True, but it’s an extant, mature codec, and like DVCProHD it needn’t be tape based. Unless I’m mistaken, Panasonic developed it and, if it wishes, may extend it with 4:4:4 chroma sampling and variable frame rates. If Panasonic can come up with a better codec — one that it doesn’t have to license — by NAB 2006, great. It seems that Panasonic has, up to this point, thrown its weight behind DCT-based codecs. Conceivably it could develop some codec based on the MPEG-4 Studio Profile. But the computing horsepower (ie, encoding/decoding chips) necessary to do so probably wouldn’t be cheap enough to implement for the $15,000 to $30,000 range within which the new HD offerings seem to be falling.
-
[Steve Freebairn] “I think that a DVCPro200 codec will be released for the Varicam”
While that’s possible, why not just use HD D5? It can record true 1,920 x 1,080 at 8 bits (using 4:1 DCT intraframe compression) or 10 bits (using 5:1 compression) at 4:2:2 chroma sampling. It supports 1080i and legacy 1035i at both 60-Hz and 59.94-Hz field rates, 720p24/25/30/50/60, and 1080p24/25/30. I suppose that Panasonic could extend it to support 4:4:4 chroma sampling and variable frame rates. HD D5, however, has, I believe, a data rate of 250 Mbps, not 200 Mbps.
-
[toke lahti] “So, maybe there’s no pixel shifting at all?”
Except that Jan has repeatedly said that the HVX’s CCDs will use pixel shift. In one DVXuser post, she said, “All pixel shift does is cover the non-photosensitive areas, the registers, and give more resolution.” And in another, “Without pixel shift/spatial offset, you would have vertical stripes in the pictures, which once the light became low you would see.” From this I gather she means that pixels are shifted or staggered so that there would be no visible gaps between them. Beyond this, it hasn’t been revealed how much the pixels are shifted.
-
There is no 1080p format supported in DVCPro HD except 1080p24 and 1080p30, which are encoded as part of a 60i stream. With 1080p24, the 24 progressive frames are converted into 48 fields, 12 of which are duplicated to bring the number up to 60. This means that 6 frames per second are each synthesized from one field from each of two successive original frames. With 1080p30, the 30 progressive frames are converted into 60 fields.
-
[Barry Green] “P2 can record indefinitely as long as you keep hot-swapping cards. There is a field device (the ‘P2 Store’) which automatically copies the contents of a card onto its internal hard disk and erases the card, ready for hot-swapping back into the camera. If you had enough storage available, you could record for hours and hours uninterrupted by continually hot-swapping cards.”
Because the AJ-PSC060G “P2 Store” appears to take as long to offload a P2 card as it does to fill it, if you’re recording 1080i60, 1080p24, 1080p30, or 720p60, you’ll probably need at least three P2 cards to shoot without interruption for more than 16 minutes (assuming 8-GB cards). And even then the constant swapping could get nerve-racking. Even if the camera doesn’t get jostled during swapping, either the operator will have to swap the cards while shooting (which could be difficult during a handheld shot or a complex pan/tilt on a tripod) or an assistant will. Especially because the slots are right below the viewfinder, I would think that the assistant would get in the operator’s way. -
I believe that I saw a rumored price of under $1,000. And no, it won’t have a P2 form factor — but if it’s a quarter of the price of equivalent P2 capacity but with comparable ruggedness and reliability, it’s a fair trade. And I expect capacities to climb (and prices to drop) more quickly than P2’s because Samsung is not targeting this solution only to deep-pockets markets like the military and industrial sectors.
-
Panasonic isn’t Sony. Before the release of the FX1 and Z1, Sony said it was considering supporting 24p in those cameras. Because it’s likely that the CCD used in those cameras, which is purportedly incapable of progressive scanning, was designed before these HDV cameras were even demoed, it’s pretty clear that Sony never intended to support 24p. Panasonic, on the other hand, has been forthright about its camera’s specs. The only real exception is the imager, and its possible that several candidates, with different resolutions, have been tested. Panasonic knows that even implying the inclusion of a feature, let alone stating it officially, and then not delivering could be damaging to the success of the product in question. Sure, the specs can still change, but probably only for the better. If anything, Panasonic will overdeliver.
-
Woops. What I meant to say is “1 to 2 hours of continuous shooting at 1080p24.” I also forgot to add that Panasonic could have included hard-disk storage as an interim solution for DVCProHD shooting until P2 capacities and prices became more practical.
After all, a DV tape drive is included, even though you could record 80 minutes of DV on two 8-GB P2 cards.
-
[Jan Crittenden Livingston] “Perhaps I don’t understand how your vision would be advantageous over that which is already possible.”
Not to put too delicate a point on it, but one vision would be 1 to 2 hours of continuous shooting (without swapping cards) on a medium (memory, storage, or whatever one may wish to call it — I don’t want to start another semantic discussion) that didn’t cost $2,000, but preferably less than $100 and certainly less than $500. -
But how do you measure “0% defectivity”? If there are bad gates on a card, wouldn’t it fail pretty much immediately? If so, then all functioning cards are 100% reliable, at least until they fail. And, as far as I know, no P2 card has been around long enough to outlive its estimated MTBF, which has got to be several hundred thousand hours — or 15 or 20 years. The edge connectors will wear out sooner.