is everything going to be in the studio? If it's an on location shoot - you'll have to combine the footages, and hope that there's not THAT much of a difference. :)

However - I would guess that uncompressed HD footage would look better than the HDV footage off the tape - but i could be wrong.

Dave

I guess I just assumed that he was capturing live - if you're capturing from a recorded tape then Firewire is better than component. If you record component out from live shooting, then you'll see improvement (however minimal), if you record component from taped footage, then you'll be dropping quality.

Dave

The quality difference between capturing HDV tape and capturing uncompressed 4:2:2 via BMD will be quite substantial in most cases. Worth the substantial hassle and expense? Only you can decide that.

I shoot this way in a studio environment, and have a 100' SDI cable I can stick out through a window or run from a hidden mains powered system for outdoor shoots. This is not a particularly portable solution though.

I should have mentioned that my idea is to capture uncompressed from tapes made in the field, therefore it woudln't be practical to bypass the hdv compression going to tape, not to mention the $38,000 for a broadcast HD portable deck.

I just recall Spot mentioning that there was quite a quality bump by capturing uncompressed (from the HDV tape) and was wondering if it's really enough to merit the extra clams.

It's better, yes. Because you're avoiding part of the decoder. But 3K worth? I don't think so. We happen to have the RAID and card, that makes it worth it. I'd never build a system specifically based around cap'ing uncompressed.
In theory, Bob's right, you're introducing an A2D in there which should make things worse, but if you capture a rez chart, you can see a slightly finer detail. Again, this is because you're avoiding a step in the camera.
If I remember right, Dr. Dropout had played with this too, but I can't find his post on the subject. In any event, it's minor, if you're post-tape. However, doing it like Coursedesign is doing...it's significantly better, of course.

Forgive me, but if you capture analogue HD, just from the camcorder perspective - which step are you missing out from that of capturing HDV directly?

Surely once you've turned the pixel shifted 1440x1080i 4:2:0 that has been compressed into 25Mbps, you can't make silk out of such sow material just by passing it though a slik lomb?

I can appreciate the HDV and DV benefit from a recompression perspective by not constantly processing and exporting in a prosumer format. So taking one step into uncompressed and then tunning with that to the target/delivery format has some purpose - indeed, uncompressed is a fairly appropriate digital intermediate format for slick editing on a workstation class Windows machine with fast nearline storage. Just not notable for fluidity with 4:2:2 uncompressed HD with Vegas (on that specific format as of V6c and earlier). Though I doubt too many folks with Vegas and HDV cameras do too much dipping into different applications when Vegas does do so much good stuff for the videographer within the single app.

Where a camera does truly present clean unfettered barely processed component YUV or SDI outputs, the placement of HDV or DVCProHD can take a backstage move towards being a backup/last-resort format. However as pointed out, 100Mbps HDV family formats (e.g, Panasonic DVX) and truly uncompressed HDc ouputs are on cameras which start at (or will when released) the pricepoint of a Z1 with all the YPbPr->SDI adaptation. Not even considering the large scale storage requirements and realtime processing required to rely on an HD SDI capture of a piece of work. These are achievable but you don't expect to lash this sort of kit together, it is a tailored build of outfit - or a $50k deck.

Capturing analog HD from the camera bypasses the entire video (and audio) compression for tape, which certainly boosts quality, but uncompressed HD is very resource heavy, as noted.

An HDV camera tape is of course compressed already, although I think it should be said that the usual "sky is falling down" doomsday predictions, made by pundits who had never tried HDV, have been very thoroughly disproven.

The early indications of DVCProHD are not fantastic, and for affordable camcorders it may be a worse choice than HDV, time will tell.

Every format is a trade-off, and I think far too much energy is spent on trying to squeeze out the last bit of technical video quality, at the expense of energy spent on content quality.

Look at your format choices in real footage, pick one format, *learn it well*, and stick with it!

If you do an analog capture from HDV tape or DV tape, you can get slightly better video quality than via firewire. This has been reported repeatedly, for nearly as long as digital formats have existed.

Why?

Because you can get some built-in chroma smoothing this way if you're capturing to a higher format, say 4:2:2. Ditto for sharpening, this type of capture can also give you some built-in sharpening.

Wouldn't it be easier to just capture via firewire and do chroma smoothing and appropriate sharpening in post? You bet.

I don't think Vegas has enough granularity in its chroma blur settings to get the very best quality this way, but Combustion, AE and several plug-ins exist to do it optimally.

A complete uncompressed 4:2:2 HD editing system can be put together for less than $10K for sure, with just a little bit of smart shopping. Forget the ridiculously priced video decks, they are just low volume high profit products for the "TV" manufacturers.

Use SDLT600, etc. computer tape drives for backup instead: 1/25th of the price, 25 times longer head life, no 1500 hour expensive maintenance, no dropouts, vastly improved reliability, they take up less space, etc., etc.

Avid certainly make a big thing about this on their Mojo systems and we have a DSR 45 for those who want to go down this path.
I've seen a lot of literature about this and although on the face of it it looks like a good idea I have some reservations.
Firstly, I've not heard anything about what's happening with the audio, it too is going through a D->A->D conversion and you are at the mercy of the quality of the D->A converters in the deck and the quality of the A->D converters in the Mojo (or whatever) box.
But my bigger concern is generational loss, I've no argument with the idea of using some form of smoothing to improve the look of DV originated material in a 4:2:2 environment so long as it stays in 4:2:2 land and a digital one at that.
But going back to DV25 and then ingesting that again and again subjecting it another round of quantization errors etc and things should start to fall apart. And lets not forget the fate of the poor audio track.

Moving all of this over to HiDef raises some new issues, a system capable of ingesting 1080 4:2:2 material needs to have a large amount of fast storage. Certainly with todays hardware it can be done for less than a kings ransom but still an order of magnitude more expensive that a HDV only solution and if you're ingestion material that's already recorded and compressed as HDV there's at best a very, very minor possible improvement and one that can be done in software, thus avoiding the cost of disk arrays that can keep up with the data stream.

What we are looking at very seriously for field recording from HDV camera is the Grass Valley Infinity solution. This avoids the problem of needing fast disk arrays and matching mobo busses etc, a simple USB port is all that's needed (if you're patient), or for only a few hundred dollars you can put a 'VCR' onto a IDE port in your PC. That's a quantum leap for HD acquisition, we've gone from needing a HDCAM deck in the field (You're going to use THAT, WHERE?) costing >$100K, to a field recorder costing $10K; affordable, easy to buy media and a 'deck' for ingest that costs <$1K. Because this is a tapeless solution there's no imperative for your system to keep up either.
Bob.

More is the pity that more camera manufacturers are not willing to give up the tape mantra (money making machine) - not without it being an aftermarket supplement. So each camera that comes out with YPbPr outputs has to be independently checked to see if the analogue outputs are truly pre-compression with minimal frame latency, etc etc.

Sure the marketing department want to make their contributions, but it seems reasonable for a videographer to be able to store 12 or 14bit RGB CCD data in SD or HD out to hard disc - via a luggable PC, if he wants to. Does the equivalent cheapening happen with film cameras? Surely not to this extent. As mentioned earlier - the form of the work is more important than the technical specification. Yet with camcorders, the engineering team are made as silent as possible so that the broadcaster studio cameras and pro cameras and dedicated deck appliances don't stop selling. Yet really, would they stop selling? Do they really need to be tape oriented to be dedicated?

Thomson (and their sub-company GVG) pave a way just as Western Digital have with 10k rpm SATA drives. But unlike WD, they do have a market that could be eroded but they see the lower end as where it is at. Boeing and Airbus have had similar dilemmas with whether they should back the micro-model of air travel or the gigantic super-scale flight vehicle.

I think this thread has shown that people will find their own way. Whichever way works for you is what I'd comment back with.

Having a low-comp or no-comp HD editing solution is one thing. Having the means to acquire in that/those formats is remarkable. However running out HDV into an uncompressed format for the benefits of the chroma signal ringing into a a line signal that fills what would be gaps if digitally acquired as 4:2:2 seems to me to be approaching the realms of awkward justification. Running into uncompressed because the workflow is fragmented over different apps, systems or simply you don't use too many straight cut edits seems more plausible to justify a spend over. Whenever we edit HDV with much more than a straight cut, we are introducing recompression with anything but uncompressed. So pretty well all edits that are anything other than the equavalent linear-edit/straight-cut ie scaling, cropping, fading, PiP, colour correction, virtual-sets etc are all bashing the source data's natural compression algorithm through recomp. Many folks said how DV was not worth lowering oneself to, and were proven wrong. HDV is a slightly different animal as lenses, CCD/CMOS, prisms, tape-dropout and publishing formats are all very much new territory at these price points. We have to live with them and use our artistic and perceptions and broadcast grade monitors (or 23" WS HD LCD) to assess how well we've managed.

2006 will surely bring more changes and some great opportunities for selling HD oriented equipment to folks who previously paid premiums for SD equipment. Video publishers will possibly turn less frequently to using film (16mm/35mm) for the acquisition format. For a new set of reasons including price but not discluding quality and look.

I'm generally more of the same opinion or taking that same tack as farss (Bob). I'd recommend venturing into > ~150MByte/sec formats when the camera is itself able to present and you can acquire in that format. Otherwise consider such a spend to be in preparation for when you can do more than 25Mbps long-GoP HDV. Going back to what I said about using full rate HD as the digital-intermediate format, as there are benefits to be had, even if HD WMV9 or more interesting still - DVD-Video is your final target. What you archive to is then also more flexible. Whether you ever acquire from an analogue hose because you have to, to get lower compression HD out without undoing the camcorder case and hacking it, or because you like the contribution this makes to the aesthetics is secondary I think.

As for audio, well, audio should have a backup that (incidentally) doesn't apply compression at all that is made at source. Even the more lowly videographers should have that on their radar as things to achieve. So tape might continue, but ought not to be with us in the long term for much more than a failsafe backup (or to stop the camcorder going into power-save mode!).

Interesting thread this. I hope I've asked the right questions to understand the earlier contributors rationale sufficiently?

Please, not another conspiracy theory.
Cameras will be made that record to tape for a long time to come, not because of the value of tapes sales, that's a nuts theory. Canon just released a HDV camera that records to tape, yet to the best of my knowledge they've never made an inch of tape. For years we sold Panasonic tape for use in our Sony cameras and I'm certain there's plenty of Panasonic cameras out there that have only ever had Sony tapes go over their heads.
The transition away from tape will be a long drawn out process, it's so ingrained in the broadcast world that I think it'll outlive my children before the last piece of tape rolls over a head, there's still viable businesses retieving material off 2" quad tape and there's still stations using 1" R2R video tape formats. This isn't some conspiracy. Even if the magic bullet that'll make tape obsolete appeared tomorrow none of this would change much, it's just that things take time, a lot of time. Just to go through the millions of hours of tape and move all that content to the new medium will take armies of technicians decades. Only today I'm copy audio from 1/4" tape, and there's still millions of reel of the stuff to be converted to digital. The budget to convert one local networks analogue tape library to digital is $30M.
I've no doubt tape is on it's last legs but there's many challenges ahead, tape is still the cheapest and most robust solution there is. Trying to sell a consummer or prosummer camera without a tape transport is no simple task, the manufacturers be they Sony, Canon or Panasonic are heading in the right direction, all offer HD prosummer cameras that give you the choice of using an external non tape based recording solution that's of higher quality than the internal tape drive.
There are cameras that are tapeless, the Grass Valley InfinitiCam and Sony XDCAM being two examples that come to mind however record time on both formats is around 20 minutes, that's fine for ENG or large budget drama. The industry is used to that, from the days of film until now. It isn't an issue for sports or live event coverage for them either, they've got OB vans with racks of recorders that take 90 minute tapes. But for us lesser mortals it's a deal breaker, even the 60 minute limit of DV is limiting for many at this section of the maket, that's why tape survives.

And unremarkably enough, yes the same compromises are forced on those who shoot film. 35mm print stock is of significantly lower quality than camera stock, we could have much better images in our cinemas if we didn't mind paying $100 per ticket to cover the cost of prints. We could even have movies with mind blowing temporal resolution by running film at 60fps but what would that make movies cost. And low budget productions are shot on S16, parts of some movies are even shot on Pro8 because they're the only cameras that'll fit in the space available.
Bob.