HVX-200 Archiving

[Jeff Kilgroe]

Originally Posted by Kevin Shaw

Not bad, but note that it's pretty easy to find bare hard drives at prices of ~30-35 cents/GB or $300-350/TB, so the break even point is closer to 6 TB. If you shoot a lot of footage that could still work out in favor of tape storage, but hard drives aren't a bad deal these days.

I suppose... Some of the lower capacity drives have come down in price a lot -- I was looking at pricing for some of the top tier 500GB drives there. 250GB 7200rpm units and even a few larger drives can be had in the $0.35/GB and under range if you do some digging.

But even at the prices I was figuring, hard drives aren't a bad deal at all and they have a lot of things going for them that tape doesn't. Tape has its advantages too... Mostly I was trying to point out that there are some pretty affordable tape options since most of the tape naysayers complain about how expensive it is compared to hard drive.

Realistically, no media out there is perfect and anything can fail. The best is to have a good backup strategy that produces redundant copies of across 2 or 3 volumes in case a tape or hard drive fails. It's the old addage about not putting all your eggs in one basket. If you're smart about how you archive your data and you can make it work with hard drive or tape or even something like DVD-R or BluRay/HDDVD, then that's great. In the end, what works for one won't work for another

Tape is a tough nut to swallow if you don't shoot tons of video and don't want all the extra cost up front -- that's where most of the cost of tape archiving is - in the drives. Media does cost, but tape media is probably still the cheapest media out there in terms of $/GB. You also have to figure that most all the moving parts in a tape backup workflow are in the tape drive itself. Now how many hours on that tape head do you think you'll get before you have to buy a new drive? 500 hours of video and 200 tapes down the road, you may be all happy about having a $0.18/GB archival system when suddenly you have to spend another $2K to replace your drive. Now that tape system is just as expensive (or more) than HDDs would have been.

[Dan Euritt]

hard drive prices continue to drop like a rock, but dlt pricing has always been absurdly overpriced.

having spent 10 years as a pc network admin, i can tell you that tape drive interfaces and the operating systems that the tape drive software requires will always become obsolete... how well i remember the panic phone call from my old employer, as they tried to recover a dos-based tape backup that used a parallel port :-/ how many of you still have a parallel port on your computer, or a scsi interface, for that matter?

buying into a home-built external raid with removable drive trays, and multiple interfaces like usb, firewire, sata, etc., will work in your favor as time goes on... you don't need software to control it, hdd pricing will continue to drop, and if you keep the drives in sets as they fill up, so you should be able to recover the data if one drive actually dies, which is doubtful.

as pc operating systems go, the upcoming microsoft vista is a major upgrade, whereas winxp was actually considered to be a minor upgrade over win2k... and win2k was a major upgrade over win98.

so if dlt is what you must have, make sure that the control software that it uses is future-proof... don't limit yourself to a winxp-only dlt solution.

[Steve Mullen]

Originally Posted by Matthew Groff

Well said, Steve. It's this very issue that is the sole determination whether or not this camera is usable in the field. If you're on a distant location and shoot for ten hours (which is the equivalent of what? 2.5 cineporters?) and then, after a long day have to sit up and write the data to tape at best at 2X realtime... it just doesn't add up.

Tape/XDCAM media is here to stay until Flash/P2 is throwaway media under $50/hr.

mg

Frankly, to compete with DV tape and XDCAM HD it really needs be under about $25/hour.

It would be helpful if someone created an Excel spreadsheet that allowed one to enter the type (24p, 1080p, etc.) of video shot, number of minutes continuous shooting, number of minutes per day, number of days shooting in a row, etc. Then the sheet would determine the number and capacity of P2 cards needed. Plus the options for achieving and their cost in terms of money and time.

Once a basic sheet was created, it could be modified into TYPES of shooting: ENG, DOC, etc.

There's two ways such models could be used: one, for those who have already decided to buy an HVX200 -- what should they buy "in addition."

For those who are evaluating tapeless vs tape shooting -- the sheet could be expanded to include XDCAM HD and HDV. There are only a few inherent reasons to exclude any of these alternatives (Adam's tests showed they all produce great HD video) so a cost analysis should, IMHO, be part of everyones buying equation. While "image" may be subjective, total cost of production is not.

For geeks -- the sheets could be enhanced to consider falling P2 and optical/HD prices and increased optical/HD capacity and optical disc write speed. And, that would tell one WHEN P2 will be cost effective vs other options.

This should be done by someone who favors P2 so there can be no claims of bias.

[Barry Green]

But that kind of thinking just doesn't apply in the real world of shooting.

Right now I've only got two 4GB cards. It's all I need, and it works fine. I paid off my complete HVX package on the first job I landed with it. Free and clear. Anything from here on out is just gravy, and my situation is far from unique.

It doesn't matter whether some other system offers cheaper cost per gigabyte, because I got this specific tool to do a job, it does it, and it pays rental fees to me every time I use it. I deliver the footage on hard disk to the client, and they love it. Owning the HVX has proven to be a positive-cash-flow generator, regardless of whether HDV tape is $4 each or blu-ray disks are $30 or whatever. The client pays me more than it cost me, so how can it possibly get more cost-efficient than that? It's positive cash flow.

Everyone I know of who's using the HVX professionally will tell the same story. It's not a case of "how many tapes do I have to save on not buying before P2 costs break even"... it's a case of "oh, you can shoot high-def? So I don't have to rent a $1200/day camera? Shoot this project for me, I'll pay $600/day for the camera package, plus your normal rate."

That's how it's working in the field. And yes, the clients do ask to see footage before voting for the HVX over the VariCam and CineAlta. And yes, every one of them so far has voted for the $600/day HVX vs. the $1200/day VariCam and $1400/day CineAlta. Not that there isn't a difference in the footage, of course there is -- but on a cost/benefit ratio, they've chosen the HVX every time. Peter Kagan'll tell you about a national commercial for Subway, starring Heisman trophy winner Reggie Bush, where he offered to shoot it on his own Aaton 35mm camera, and instead they opted for the lower cost of the HVX, and they were thrilled with what they got. He told us at the HD Bootcamp that he's now ready to sell his Aaton.

I mean, I guess the cost-per-gig-storage thing is relevant to some people who are closed shops, one-man-band type of things, weddings and whatnot, but -- if you're working with clients and agencies, doing advertising, infomercials, or other types of paying work, it's just a total non-issue. It isn't even a factor. You add a $150 hard disk into the budget, rather than a box of HD tapes and $150/hr deck rental time, and you go from there.

[Steve Mullen]

Originally Posted by Barry Green

The client pays me more than it cost me, so how can it possibly get more cost-efficient than that? It's positive cash flow.

Can you image CNN buying 200 HD camcorders because some shooters like the picture? (Especially, since contrary to your assumption that everone loves the HVX -- I would say in an organization there would be many who would not like the HVX200. Moreover, once the tech evaluators look at SD CCDs scaled-up to 1080i -- who knows if they'll vote for the HVX. It's not the slam-dunk you say it is.) In fact, image is the likely the least important issue. It's ROI that they are looking at. And, various measures of profitability. The bean counters will do exactly what I suggest.

Bottom-line, positive cash flow is not how a business evaluates itself. If it were, all companies that had a cash flow that was positive would have the same stock price. They don't, because that is not a useful measure of how succesful a business is. It's like saying you have "passing grades." That won't get you a great job.

[Jeff Kilgroe]

Originally Posted by Dan Euritt

hard drive prices continue to drop like a rock, but dlt pricing has always been absurdly overpriced.

This is exactly the naysaying I'm talking about... DLT prices drop too and currently, DLT media has the cheapest per GB cost of any media available. Tape pretty much always has been the cheapest, most accessible media type. Seriously, 400GB SDLT3 tapes can be had for < $95 where can you buy 400GB of HDD space for that? Yes, it's true that you'll have to spend $2K for a drive to use one of those tapes, but that's beside the point. Do the math and you'll see that at some price point such an option as this makes sense. Tape is not the best option for everyone, neither is HDD. If you don't shoot a lot of video and won't be archiving multiple TBs on a monthly (or even weekly) basis, then a large tape archival system may not be for you. OTOH, if you are archiving TBs of data every month, maybe HDD seems like a silly proposition.

having spent 10 years as a pc network admin, i can tell you that tape drive interfaces and the operating systems that the tape drive software requires will always become obsolete... how well i remember the panic phone call from my old employer, as they tried to recover a dos-based tape backup that used a parallel port :-/ how many of you still have a parallel port on your computer, or a scsi interface, for that matter?

There are so many little flags standing out in the above paragraph and most I wouldn't touch on with a 10' pole. But I will say that everything in the IT industry evolves and is bound for obsolescence. When comparing hard drives to tape, how is attempting to recover archives from a 17 year old Travan I tape system with a parallel or SCSI interface any worse than recovering archives from a 17 year old stack of ESDI or RLL interface hard drives?

And parallel ports and SCSI ports are still widely used and are industry standard. Not to mention that all current SCSI systems are backward compatible to the old 25pin SCSI-I, provided the proper connectors/adapters are used and the devices are properly placed on the SCSI chain. Next to the latest fiber channel implementations, UWD SCSI is still the highest performance storage interface out there. Parallel ports are still standard fare on just about every PC system out there as so many parallel devices are still in use. USB to Parallel or Serial adapters are pretty commonplace too and cost about $35.

There's a good bet that (S)ATA hard drive interfaces will be non-standard, obsolete crap within 10 years. Is this a big deal? No, because there will still be plenty of adapters and interface cards out there to help us out.

buying into a home-built external raid with removable drive trays, and multiple interfaces like usb, firewire, sata, etc., will work in your favor as time goes on... you don't need software to control it, hdd pricing will continue to drop, and if you keep the drives in sets as they fill up, so you should be able to recover the data if one drive actually dies, which is doubtful.

You don't need special software to control a properly implemented tape system either. Both tape and HDD pricing will continue to drop.

as pc operating systems go, the upcoming microsoft vista is a major upgrade, whereas winxp was actually considered to be a minor upgrade over win2k... and win2k was a major upgrade over win98.

And this has what to do with backup solutions?

so if dlt is what you must have, make sure that the control software that it uses is future-proof... don't limit yourself to a winxp-only dlt solution.

As I said, a properly implemented tape system doesn't need special software. There are already ANSI and IEEE defined standards for tape and archival storage systems. This is how the backup utility included with windows or the multituded of such generic tools within unix environments all know how to talk to your tape drives. If you're using a proprietary backup software that encodes the data or places it inside some sort of file container, then you're using a short-term backup solution instead of a true archival system. Most backup software out there and what typically ships with tape drives is just that backup software. Proper archives that are intended to last years should be made only of the data itself, exactly as it exists on the source storage volumes and placed on a media that has an industry standard I/O spec. If compression is used, it should once again be an industry standard that won't be a future issue, use formats like ZIP, LZH, etc... Most backup softwares do use common standard algorithms, but they often foolishly still place the data into proprietary file types to force you to continue purchasing and using their software. Stupid. What happens 10 years from now when you need something off an old tape volume and it uses Company X's proprietary backup format and Company X is no longer in business? The only copy of their software you have is backed up on one of their proprietary volumes, oooops. You're pretty much toast unless you can find someone out there that has the tools to help you out.

Same applies to HDD. Keep the data natively raw. If you must compress, use a standard method.

[Barry Green]

Whatever, Steve.

NASA's buying the HVX, and I dare say they stringently evaluated its performance. The BBC is buying the HVX, and there's nobody who evaluates a camera more thoroughly. Independent producers are doing great work with it and making money with it. Lots of it.

We'll continue to do so, whether you think a blu-ray disk is cheaper per gigabyte or not.

For those of us who are actually in the business, it's kind of hard to beat having gear that pays you to use it, vs. using gear that you have to pay to use.

[Steven Thomas]

After having compared the HVX to other cameras in its price range, I'd say the HVX feature list is what's selling the camera.

[Gary L Childress]

As a long time editor, I have used DLT frequently for backing up media from completed projects. The DLT's I have used were very slow, perhaps newer models are faster. However, I never had a DLT fail. The actual process didn't impact operator hours that badly because the accepted workflow was to set it up at the end of the day and let it backup overnight. I restored many projects that needed updates years after the original edit. For large projects we restored overnight. Saving hundreds of gigs of media via DVD or even Blue Ray is going to require alot more operator time than the various data tape systems. Down the road, new archival technologies like holographic storage will become available. In the meantime data tape formats are useful and cost effective. Hard drives are fine as well but if I was backing up media that there was no original tape source for such as P2, then I would double backup to two hard drives for safety. I suggest that those who put all their eggs into one hard drive have probably never faced the prospect of a hard drive that has died with valuable media that can't be replaced at any cost. These cost-per-gig comparisons do not factor in the value of the media itself if you had to reshoot media that was somehow lost.

[Dan Euritt]

Originally Posted by Jeff Kilgroe

This is exactly the naysaying I'm talking about... DLT prices drop too and currently, DLT media has the cheapest per GB cost of any media available.

that is not correct, but the fact that you are even preaching an ancient dead format like dlt, instead of lto, pretty much sez it all.

make no mistake about it, what we are talking about here is utilizing the best i.t. practices we can afford for backing up our data, and in the real world, some people feel that tape is dead:

"But when it comes to backup and recovery of essential data, tape is fading from the lineup of usual technology suspects. It's glitch-prone, for one thing. A Yankee Group survey of IT executives last year found that 42 percent of respondents were unable to recover data from tape in the previous year as a result of tape unreliability."
http://www.businessinnovation.cmp.co...eature_4.jhtml

Originally Posted by Jeff Kilgroe

Next to the latest fiber channel implementations, UWD SCSI is still the highest performance storage interface out there.

but why buy ancient technology like scsi, when you can get fiber channel? the best bang for the buck is sata, by far, and it's currently headed for 600MB/sec and beyond.

Originally Posted by Jeff Kilgroe

There's a good bet that (S)ATA hard drive interfaces will be non-standard, obsolete crap within 10 years.

wrong again, jeff:
"According to consulting firm TheInfoPro, 90 percent of companies plan to move from tape to Serial Advanced Technology Attachment (SATA) disk drives as part of their backup/recovery and data lifecycle management plans by 2006."
http://www.businessinnovation.cmp.co...eature_4.jhtml

Originally Posted by Jeff Kilgroe

You don't need special software to control a properly implemented tape system either.

so a tape drive just plugs into the computer to be accessed like a hard drive? and you can do incremental backups to it without a software program? that's certainly news to me :-)

[Dan Euritt]

Originally Posted by Gary L Childress

However, I never had a DLT fail.

i had a dlt tape backup fail to restore, and even the data recovery service could not bring it back.

that was about 10 years ago, which gives you an idea of how old the dlt format is... we lost mission-critical data off of a file server because of it.

i'd like to believe that modern dlt tape is more reliable.

[Steve Mullen]

Originally Posted by Dan Euritt

t"According to consulting firm TheInfoPro, 90 percent of companies plan to move from tape to Serial Advanced Technology Attachment (SATA) disk drives as part of their backup/recovery and data lifecycle management plans by 2006."

Something being missed in this thread is that when we erase P2 cards having copied them to X -- we are not doing a classic BACKUP of data that we restore in whole or in part by filename.

We really want the P2 contents to go to another random access media that we can easily view the contents to find something we can only see or hear. Something we have a memory of.

That means either optical or hard disk that can be mounted like a P2 card. Once you decide this:

1) You can debate optical vs HD cost-per-bit

2) You can debate optical vs HD reliability

3) You can debate optical vs HD storage life

But, there is one thing beyond debate -- SPEED. Only a HD can move data at the rates possible with P2. And, while some who shoot only a couple of 4GB cards a day may not care, to replace tape or challange XDCAM HD, FASTER THAN REALTIME transfer is a must.

IMHO this means a HD in a nice package. Which means a visit to the Grass Valley booth to see how they use the REV HD.

Lastly, if you really, really want to insure data safety, HD can be run as a RAID. That's a unique option.

[David Heath]

Originally Posted by Matthew Groff

Tape/XDCAM media is here to stay until Flash/P2 is throwaway media under $50/hr.

There is a third option, and that is the Grass Valley Infinity. Cheaper than P2 flash memory for immediate ingest into the edit system, with the possibility of making the backup as you go along on Rev Pro. OK, it's early days yet, the camera has yet to be proven and let's at least wait until NAB, but if it lives up to hopes then it could give P2 a very hard time.

It also remains to be seen if Grass Valley will produce a 1/3" variant in the future. Such a beast would have the potential to give the HVX200 a real hiding, certainly from the archive/storage viewpoint.

And yes, any discussion of archive costs must take into account time as well as media costs.

[Barry Green]

Originally Posted by Steve Mullen

Only a HD can move data at the rates possible with P2.

LTO3 data tape moves data at 80MBps,the same speed as a P2 card and six to 15 times realtime.

[Steve Mullen]

Originally Posted by Barry Green

LTO3 data tape moves data at 80MBps,the same speed as a P2 card and six to 15 times realtime.

If you read my post, you'll see I had already rejected ANY data tape based system and was only comparing optical vs HD.

If folks want super fast random access -- the major virtue of P2 -- because they don't want to seach though video tapes, then they certainly don't want -- the moment they erase a P2 card -- to place it's contents on a many, many times bigger data tape that can't be visually searched.

And, it's crazy to copy once to a HD and then later copy again to data tape. Time is money.

It seems very clear to me that only HDs offer super fast random access (as does P2), super fast copy to, plus current huge capacity, and rapidly decreasing cost-per-bit. Thus, I reject both data tape and optical disk.

Which, of course, is why GV has gone with HD right from the get go. And, why so many folks want to attach a HD to their camcorders. This indicates to me that if Sony and Panasonic weren't so interested in selling propritary media -- they would have simply gone with HD.