I think they will get cheaper, but I can't see others starting to sell them - are they not proprietry to Panasonic? And as they do get cheaper/bigger, then the same economics will apply to CF/SD. The market will also be affected by the plans to phase out their architecture from laptops.
They may not have been designed for video in the first place, but it's wrong now to just label it a still camera format. The Grass Valley Infinity proves that, and the recent press release from them that it is being trialled by the BBC as a contender for their future production platform must mean we should take it very seriously. What I like conceptually about Infinity is that material may be recorded in parallel to CF for immediate use in an NLE and/or to Rev Pro as an archive or for a cameraman to give away to producer, reporter etc
Well, I'd love 4x ingest speeds and have never said otherwise. But that's a loaded question. If you asked instead "would you want 4x ingest speeds if it meant paying many times more for the media?" I suspect many people would forego the 4x. And as previously stated, most people are unable to download off P2 at much more than real time (for 100Mbs DVCProHD) anyway at the moment.

...And this confuses the @%&# out of me. What kind of slow-ass systems are people using that they can only ingets video at a paltry 12.5MB/sec or so?

The worst I've heard of is the Panasonic P2 store. According to Panasonic themselves ( https://eww.pavc.panasonic.co.jp/pro...ore/index.html ) a ".....P2 store can copy 4GB of data in only about four minutes...." or "a paltry 12.5MB/sec" as you put it. In practice I've heard real world tales of users being pushed to even get that. My own trials with a fairly powerful laptop and external hard drive managed better than that, but only about 20% or so, and I believe these figures are fairly typical. (Anybody got any other experiences?)

No doubt a far more sophisticated (and expensive) system could be engineered to do better, but all my comments have referred to an "average user", whatever that may be.

No, 20 MB/sec to be sure they have enough overhead to record DVCProHD reliably at full bandwidth, which would be 12.5 MB/sec (plus more for the audio?). Don't ask me why, but 20 MB/sec is the bandwidth a Panasonic engineer stated he needed at WEVA last August, and that's the bandwidth which high-end flash cards now offer.

What you're overlooking here is that P2 was developed before standard flash memory reached the current level of performance, so it made sense back then. Now that flash memory has improved there's no need for the P2 design to get the desired bandwidth, so it's become an anachronism. It's ironic that P2 is finally making headlines just as it's becoming effectively obsolete, and it's too bad Panasonic engineers didn't plan ahead for this. But as others have commented, if you can make P2 work for you today and turn a profit doing so, who cares what will happen in the future? Just don't buy any more P2 memory than you really need, because it's headed for museum status sooner or later.

not quite... using one cf card will always require paying for 20 mbps write speed, but with p2 you can use the slowest sd cards on the market, which will save some $$ over that single cf card.

so it's pretty funny to be calling the p2 design an "anachronism", when the potential is there to save you money over cf.

in addition, the last numbers that i saw credited sd cards with just over 40% of the production memory market share, so i guess that sd is bigger than cf? i don't know if that's reflected in current market pricing, and i'd sure feel better about all this if we knew of any 3rd party mfg's of p2 cards.

i'll be sure and put that tidbit right next to your prediction that we won't be seeing any h.264 cameras for at least the next 2-3 years.

An interesting point, and I confess it's one I wondered about. Just why should one fast card cost so much less than four slower ones of 1/4 the capacity packaged together? I can't give the answer, but will say that I'm less interested in "potential" than the actual state of the market. Whilst the price differential between P2 and SD/CF cards of the same capacity is so huge, "potential" seems irrelevant, and I'd tend to agree with Kevins use of the word anachronism.

One point that a colleague brought up that hasn't been covered here is the relative desirability of many small cheap cards versus one big expensive card - same total memory. For a small or one man operation, if the cost/GB is the same between the two options, there may be little in it, but for a large organisation the "many cheap" option potentially has big attractions.

Regarding transfer speeds, a little searching has thrown up this thread: http://www.dvinfo.net/conf/showthread.php?t=67007 , and posts 3 and 4 may be of interest in this context. In the former Barry Green states that: "If you plug straight into a USB 2 port you should get transfer speeds of around 1GB per minute." and in the latter David Tames gives some practical comparisons, with a P2 store here taking no less than 6min 24sec to transfer a full 4GB card!!! (A full 4GB card only appears to be about 3.51GB?)

I'm interested what kind of system Jeff Kilgroe is using if he is consistently achieving far faster times than this, especially if he is achieving 4x?

for one thing, david tames is on a mac, and the problems with usb and the mac o.s. have been discussed before:

"THE RESULTS:
1 straight 10 min clip shot at 720P 24pn recorded to 4 gig P2 card
Card copied to P2 store (approx. 4 min.)
P2 store connected to G5 USB 2.0.
P2 "no name" Imported straight into FCP. TIME 5 Min. 39 Sec.
THIS IS GREAT AND ACCEPTABLE
However, same P2 "no name" copied to a new folder on G5 24 Min.
SAME USB Cable, SAME EVERYTHING!"
http://www.dvinfo.net/conf/showthread.php?t=67592

what we need to see is the transfer speed of the p2 port on the panasonic toughbook... got any numbers on that?

so we agree on some of the potential advantages of the p2 design over the single cf, but it's panasonic's pricing that's causing all the grief... i don't know that anachronism is the proper useage of the word when it comes to prices, but i'm with you on the cost of the p2 cards!

if i had a p2 camera i'd be using it with the firestore, i wouldn't buy any extra p2 cards at all.

Dan, the figure I quoted has nothing to do with use of a Mac. "..... a P2 store here taking no less than 6min 24sec to transfer a full 4GB card....", according to post 4 of the thread I quoted refers to the time taken for a P2 store to download a P2 card to within itself - no other equipment, no Mac, involved. The download to Mac times he refers to are FASTER than that.
Not for a Panasonic toughbook, but for a high specced laptop PC a little faster than that, but nothing to write home about. By and large, real time downloads for 100Mbs seem par for the course, maybe a bit faster, maybe a bit slower.
I've a lot of experience with a Firestore with a 2/3" SD tape camera, and will be using it tomorrow and all week. Used like that it's fantastic, but I would always run it in parallel with the tape, when it is my primary means of import to the NLE and the tape is the backup and archive. It normally works well, but isn't 100% reliable - with the tape backup that doesn't normally matter, but I wouldn't use it as the only recording media.

It's obviously only useful at all for material I edit myself. When rushes have to be given away, tape takes a lot of beating.

Thats what so appeals to me about Infinity. A cheap archive medium and a more transient solid state "NLE medium" running in parallel - just like the Firestore and tape - but no extra boxes and cables. Fantastic.

In theory perhaps, but we're not seeing that happen. P2 is a complex, proprietary product with limited volumes selling at high prices, while CF has already undercut it by a factor of 2 for an identical level of performance. That can't do anything but get worse as more and more CF manufacturers reach this level of performance at volume prices, while P2 continues to be a specialized low-volume product. Unlike our debate about H.264 cameras this matter is already settled, and it would make no sense for the use of P2 to spread given the current pricing situation. You don't have to look any further than the new Grass Valley HD cameras to see how this is going to play out in the marketplace.

I'm ingesting my cards through the following system:

A MonarchComputer Hornet Pro SFF
AMD X2 4800+ Dual Core CPU
2GB RAM
nVidia 7800GTX 512MB
2 * WD Raptor 10Krpm HDDs in a RAID-0 stripe
Operating with WinXP Pro SP2 and all current updates
Spec-Comm PCI to PCMCIA adapter
Connected to our LAN via its primary gigabit ethernet port.

Speed is most definitely limited by the speed of the P2 card itself and by the gigabit ethernet port (which is theoretically 10X faster than DVCPROHD anyway). Our primary storage is a fiber backbone SAN, but that's a non-issue at ths point as I can ingest at 4x or so on this system to its local drives

In re-reading this discussion it strikes me that the extra potential transfer speed of P2 can be relevant at the editing end even if it's not necessary at the recording end, so maybe P2 isn't rendered obsolete by advances in standard flash memory after all. I still think more people would be interested in an IT workflow using inexpensive media than P2, and it will be interesting to see what Panasonic (and others) do in the future using AVCHD discs and so on. These are interesting times in terms of equipment options, and hopefully we'll see even more new options in the future.

That's true enough, but I've very carefully used the words "average user" throughout this thread, and I don't see P2 ever really offering enough advantages to such to make it viable to them if such as a CF solution was available. What will be available obviously depends upon how manufacturers respond, whether any one sees a business opportunity and a chance to undercut the opposition.

For the reasons given above, I feel P2 is more likely to carry on in expensive pro cameras than in the prosumer market such as for the HVX.

Theoretically, even higher end users could get speed benefits at download (if their systems are capable) by techniques such as downloading 4 CF/SD cards in parallel - so 4 8GB CF cards could be downloaded in the same time as 1 32GB P2 card. say.

Practically, I suspect the biggest threat to P2 comes from the architecture upon which it is centred currently being phased out. Increasingly, laptop computers may have inbuilt CF slots, but need an external reader for a P2 card.

Ah, good point - will be interesting to see if such a solution develops. As far as 'average users' are concerned, it's a foregone conclusion that they'll take an affordable mainstream solution over a more expensive and obscure one - hence my earlier comment that P2 is becoming an anachronism. But P2 obviously works for some people for now, and until those people feel they have a better alternative it doesn't matter what happens in the future. If I was buying stocks I'd pick a company which makes CF over one which makes P2; if you're buying a camera pick the one which does what you need at a price you can afford. (HDV works for me, but I keep an eye out for something I'll like better.)