Abstract background art

Welcome connections


We’ve had good non-linear editors for a while, but tape is still dominant in acquisition. Phil Rhodes takes an in-depth look at Panasonic’s P2 system, to see if solid state cards are the answer.

First published in Showreel magazine, Autumn 2005

Panasonic promotes P2 as the ‘move away from linear’. This is well said. To date we’ve had some very competent nonlinear editors, but far fewer nonlinear cameras. I first heard the term direct-to-edit with respect to Firestore’s hard-disk recording products, and it’s always been possible to record firewire to your laptop, but with P2, Panasonic intends to push file-based acquisition out of a niche.

The first camera to use the P2 system is the SPX800 – an on-the-shoulder standard definition ENG camera similar to the company’s much-lauded SDX900. This is soon to be followed by the HVX200 – a headliner-grabber featuring full Varicam-style HD recording and launching itself firmly into the dynamic new market carved out by the recent HDV camera releases from JVC and Sony. Indeed, the HVX200 should offer superior quality to HDV, as well as solid-state recording.

The camera

In use, the fact that Panasonic’s AJ-SPX800 camera isn’t recording to tape is quickly forgotten, clearly by design. It’s a nice camera. We expect decent images from the creators of the SDX900, and Panasonic claims that the camera head on the SPX800 is even better, by dint of higher-resolution DSP electronics – and in use the highlight handling is subjectively as smooth as I’ve seen. I’m impressed with some of the layout thinking here, and features such as a record button on the top handle and a mic level control where you can get at it are clearly the result of Panasonic employing an ex-DoP, retrained (almost unheard of in Japan) as a design engineer.

Two other things that immediately spring to the attention are more obviously the direct result of a solid-state recording system – the size and the weight, both of which are low for a camera of this class. Our Jimmy Jib operator particularly commented that even with stacked IDX batteries it’s a featherweight compared to tape-based DVCPRO50 or digital betacam.

The SPX800 has got to be the tool of choice for situations where you need 4:2:2 you can run around with, at least where you don’t need to carry dozens of hours of media. Panasonic even claims that they’ve offset some of the weight saving to beef up the chassis and make the layout of the electronics more rugged, potentially enhancing the hardiness of the whole device. These significant advances are realized by a single innovation – solid state – when usually the miniaturization of technology happens in smaller increments. The Japanese call this process of continual refinement kaizen, and generally, as applied to their technological output, it occurs in fairly small steps – but P2 is a bigger step than the usual.

The cards

So is there a price to pay for these advantages? It’s a compact, lightweight, power-frugal and relatively cheap camera which records images to the highest standards, but at the moment the downside is – as anyone who’s ever filled a 1Gb compactflash (CF) card with photos will know – storage capacity. A 4Gb card will hold only around 16mins of DVCPro 25 or 8mins of DVCPro50. There’s really not a lot of point in whining about this. I’m sure Panasonic is just as impatient to be able to sell bigger, faster and cheaper flash memory than we are to buy it.

That said, P2 storage does appear to be more expensive than its capacity or mode of operation would suggest – a 4Gb card is around €1200, whereas a hypothetical array of four 1Gb CF cards would set you back no more than €300 for similar technology. Panasonic deflects criticisms of the cost by contending that the cards are a really a more-or-less-permanent part of the camera, and I’m sure we’ll see package deals with camera plus storage. Added to this, €1200 is the list price – if you search around you should be able to get deals on that – and I believe there will be announcements at IBC that the price of P2 cards is going to drop significantly.

Treating camera and cards as a package is how I’d expect to operate, either as a freelancer (when you wouldn’t want to let the cards out of your sight), or as part of an ENG operation with laptop-based offload space constantly at hand.

It’s hard to get away from the price issue as it currently stands, though. We were supplied with three 4Gb cards, a rather startling €3600 worth at current prices, which at 25-megabit is around 40 minutes of footage. Even if you could get the cards for half this price, you’d still feel like you were paying ‘early adopter’ prices. This 40mins wasn’t enough for a day on the dramatic production we were on, let alone the permissive shooting mode of ENG. It will be interesting to see the price for the 8Gb cards that is announced at IBC.

There is a reassuring roadmap all the way to 128Gb and beyond. As I suspected before I’d even seen the camera, this is a technology that will clearly come into its own as the storage situation improves. At the moment, a laptop with a large hard drive is a without-saying part of any serious P2 shooting kit. Your director might already be carrying such a beast, of course, but there are always those moments when you’re off on your own and know that you can’t realistically have another 184 minutes in the form of a tape in your back pocket.

Panasonic offers an alternative with the AJ-PCS060 P2 store drive, which costs €1800 and can hold 60Gb of data – the equivalent of 15 current P2 cards.

Solid-state acquisition

Moving away from tape acquisition is not a completely novel idea, and of course there’s direct competition in the form of Sony’s XDCAM range (about which more later). Soundies have been working this way for years – someone I regularly work with now records all of his audio onto CF cards, and the speed and convenience advantages are obvious. The question, though, is whether the technology that’s used to record a 1.7Mbit audio stream is mature enough to handle video – including HD video, now that we have the HVX-200.

Let’s be clear – the technology is fundamentally very similar to those CF-based audio recorders. A P2 card, the chunky credit card-sized storage device at the heart of the system, is really just a big, fast PCMCIA card, formed from an array of four or more SD format cards. PCMCIA is the kind of thing you used to plug into the side of your Amiga back in the old days, and which is still readable on more or less every laptop – assuming certain fundamentals.

A disappointment is that Panasonic has opted to develop a proprietary interface for P2, so it is not as ‘plug-and-play’ as CF cards for audio. I’m not convinced of this wisdom of taking this route, and we’ll be looking more in-depth at the whole issue of P2 implementation when we get chance to take a look at the HVX200.

Panasonic claims on page 4 of its P2 introductory leaflet that the cards are ‘practically indestructible’, and by page 9 they’re elevated to an unqualified ‘indestructible’. Indeed, it seems to be the party piece of the sales guys to beat them up to a quite alarming degree – to the extent that I was quite worried about using them on a production. But once we’d panel-beaten the dents out to the point where they fitted in the carriers, they worked without a problem.

Panasonic, as an LSI component manufacturer, is a huge player in flash memory, and so of course has a vested interest in solid-state storage making it big. However, regardless of this, solid-state storage has advantages over tape and optical disk systems that go beyond mere ruggedness and power consumption.

One particularly interesting characteristic is that the equipment to read it is very cheap – practically no mechanical components, and the rest is cheap-to-replicate solid-state electronics. Panasonic trumpets this only briefly in its literature, but I think it could become more significant in the future – it’s fairly easy to build a small high-definition TV camera, but much more expensive to produce the fine mechanism to record the resulting video onto tape.

P2 for ENG

The SPX800 is clearly aimed squarely at the ENG market. Given the sea change that’s required in both acquisition and post-production, it’s likely that the main early adopters of the format will be large-scale organizations doing ENG work rather than freelancers. There are two important reasons for this: first, the cost of the cards, and second, how you archive your rushes.

First, the card price. This will be less of an issue for ENG organizations than it may be for freelancers, for whom the concern is not just the cost of the cards (something that can be factored in when they make their purchasing decision), but the danger that the cards may be ‘borrowed’ by a client at the end of the day’s shoot, and then not returned. This frequently happens with cassettes, but is not so important because of the comparatively low cost of tape.

I can here myself plaintively saying, “you will get the card back to me then, will you? OK, I’ll trust you…”

But as I say, if you’re an upscale outfit that own both ends of the operation – a CNN or a Bloomberg – the advantages of what Panasonic calls ‘medialess’ acquisition will be very welcome when time is critical. On top of this, sideline features such as continuous pre-recording are extremely useful.

Storage

The second issue that makes solid state currently more suited to ENG organizations than freelancers is archiving. ENG is where the technology is liable to be stressed the least, since most people are still shooting 25-megabit standard def video in these applications. Crucially, an ENG application will often be supported by a server-based post-production workflow with voluminous long-term online storage, so the issue of storing rushes is less of a problem.

If you’re a freelancer without massive storage back-up, you have to solve this problem before the system will work for you. Despite the advantages solid state has over recording to tape, video cassettes have the plus point of being their own archive medium. When you’re shooting solid state, your material still has to be archived onto something, and €1000+ flash memory cards are only a short-term storage solution – in the long run something else has to be found. When Panasonic’s application of P2 to voracious storage-eaters like HD come to fruition, HD P2 cameras such as the HVX200 are likely to be used on ads, promos and features – an area of very expensive production where insurable backup is more critical than ever.

You’re going to be shooting a lot of data – my colleague’s FR-2 compactflash audio recorder produces, in a day, a collection of files that’ll fill one or two DVDs. Ten minutes after the wrap you’ve burned them off onto two separate sets, handed them over to production and left. Not so with video – there’s a very good reason that video acquisition is the last real bastion of magnetic tape, and that’s sheer volume of data. A 184-minute DVCAM tape contains something over 45Gb of data, and until blue-ray disc writers become affordably available, freelancers are going to have to find other ways of storing their archived footage. Panasonic is right about the cheapness of the equipment to access P2, but this is a seesaw balance. Nobody likes the fact that a DSR-45 VTR costs over €5000, but 45Gb of storage then costs the price of a cassette.

Perhaps the greatest expression of the ‘expensive deck, cheap media’ paradigm is HDCAM-SR, a tape format which at 880Mbps is capable of outpacing the solid-state electronics of P2, but which of course costs a fortune to access. So where does Panasonic’s baby come on the seesaw? Well, clearly P2 storage is expensive – but this isn’t balanced by having a cheap deck – the P2 VTR (AJ-SPD850 ) is hardly a snip at €15,000. And it only holds five cards, so even this is not a long-term storage solution. There is the option to attach a DVD burner, but again you’re going to need a hell a lot of DVDs to archive all your footage.

So what’re you supposed to do with your rushes? Well, if you’re CNN, you have a giant broadcast server that you’ll probably store them on and you’ll be very happy with the system. You also own the cards and the system that’s going to edit the footage, so you don’t have to worry about people borrowing the cards and inadvertently not returning them.

And of course, these issues are only of concern for freelancers at present. When storage becomes cheaper and the cost of the cards falls they will become less of a barrier.

The competition

But hold on, wasn’t there a mention back there of blu-ray disks? That brings us on to the competition. Isn’t there a conspicuous camera format in direct competition with P2 that’ll give you something like that straight out of the camera?

OK, I’m sure XDCAM won’t offload data quite as fast as P2, and I’m sure the decks aren’t so power-frugal, low-cost and shockproof, but you do get the other advantages in terms metadata, file-based, direct-to-edit technology, plus a format that self-archives on a medium you can hand over to production. OK, I’m applying freelancer thinking to a format that’s clearly intended for end-to-end ENG organizations, but the point remains: if you don’t have a giant broadcast server, how do you archive P2 now?

Panasonic mentions blu-ray and DLT. Its sales literature states that DVCPRO-HD tape at $120/hour is more expensive than hard disk storage, based on a 15-hour (900Gb) production. Given that $1800 won’t yet buy you 1Tb of protected storage, such as Medea’s VideoRAID RTR series, this claim may be slightly premature.

Independent productions, wooed by the cost-savings offered by little HD cameras that record a format everyone can read, will find a way, even if that way is a DVD burner and shares in TDK, but it’s no substitute for a huge server farm.

This is no criticism of the potential of the format. But it strikes me that, whereas the system may be suitable for ENG applications already, it will take time before it is suited to all applications. I say this because Moore’s Law seems to fully apply to solid-state storage capacity, and the only fundamental difference between capturing video to flash memory and capturing audio to flash memory – which works very well – is file size.

Of course, before solid-state acquisition is suitable for all, someone has to pioneer the technology, someone has to be an early adopter, and there’s no doubt in my mind that some day we’ll all be shooting raw data onto solid-state storage. This might in itself stay my hands from an XDCAM purchase. At the moment, though, potential users have to weigh the undoubted advantages of file-based nonlinear acquisition against the cost of the storage. But as I say, this is largely an issue for freelancers, not for large ENG organizations.

The SPX800 is a very good camera and the system certainly demonstrates the potential of solid-state recording. What has to be remembered is that this is a pretty new system for video storage – and we’re still at the relatively expensive early adopter stage. P2 is already suitable for ENG organizations, but the cost of the media and the issues relating to archiving are likely to cause freelancers to consider long and hard before committing their own money to the format – at least until Moore comes to the rescue.

Go back button Back Go home Home Contact Go home