Backing Up

You can get LTO tape drives as low as $1600 at B&H. Depends on your needs. But you also need to get the software to manage those backups. Nothing is as dependable as LTO for archiving. Bru is a leading company for LTO.
www.tolisgroup.com/products/hardware/

Hey BenB,

Which LTO drives are LTFS capable? That's the main feature I would like.

BenB wrote: You can get LTO tape drives as low as $1600 at B&H. Depends on your needs. But you also need to get the software to manage those backups. Nothing is as dependable as LTO for archiving. Bru is a leading company for LTO.
www.tolisgroup.com/products/hardware/

Another shout out for Bru (because it's the restore that matters!). I've been using Bru Producer for years and it just works.

LTFS is great too, and on the face of it seems like it would be more flexible (which it is), but I keep on using Bru because of it's ultra reliability, whereas I know someone using LTFS that lost some files and they don't know why.

LTO is the only reliable long term solution. LTFS is not reliable enough as mentioned before. Hard drives are not safe either no matter what people say. They might be a kind of short term solution but if you want to be sure to open your media in 20 years only LTO can guarantee this.

LTO is expensive to buy but because it is the industry standard for large data backup the LTO never really expires, no matter what version number. I bought a LTO4 drive and BruPE 6 years ago and still have no problem getting tapes and updates for the software. So its a long term investment even for a one man band.

My back up system:
Cards get backup'd 2x on Raid 1 on location. Effectively giving me 4 times safety. the 2x Raid 1 drives never travel together. One hand luggage, one check in.
Back home the card data gets backup'd to LTO4 via BruPE onto LTO4. Than I import the clips via FCPx onto local Raid storage as unmanaged data for editing.

While editing I let FCPx make backups but also have 'Backups' running in the background because FCP doesn't let me make manual backups. If your last change is 1 min past the last 15 min interval you do not have the latest version. But when closing down FCP I can run a last manual backup via 'Backups' to be safe. Because I have the original media already on the LTO I can tell 'Backups' to only copy the library data without media. Destination is a hard drive that also gets backup'd via Time machine. So I have something like 2-3 data sets of my latest editing stage.

Once finished I backup the media from the local raids onto LTO again, effectively giving me 2x the media on LTO (original card data plus converted media via FCP) I also include the corresponding FCP library data into the same LTO archive. I leave the incremental FCP backups on the local drives. These get backup'd to LTO ever now and then.

LTO's are stored in a diff physical location to my edit suit in case the house burns down

my 2 cents.
Happy editing

Nice one Carsten. What LTO drive do you use?

Darren,
There are companies that will back up you drives onto LTO as a service. Send in drive, get back drive and LTO(s). Not as convenient as having a machine, plus you then need to send the LTO back to add to it or to recover a lost drive. Still saves the initial cost, but at a huge hassle.
Guess the real question is how long do you need the archive? I do the back up to a couple of drives, exercise them at least once a year and back up to optical disks. The opticals are not permanent either, so you need to make new copies about once every five years. Five to ten years are as long as I need to keep backups other than finished media.
Some companies are toting high capacity long term optical disks, but I have no experience with them.
Whoever comes up with super high capacity/long term(forever)/low cost/fast storage will be a billionaire.
I have some Betamax and DAT tapes that no longer have a device to play them on, but I bet the files are still there.
Greg

VidGreg, thanks for backing up my 'taking the drives for a walk' comment by calling it exercising them. That's even better.

Projects I made in the 90s transferred to betacam SP are still alive and kicking. Tapes are still the best for the long term.

The Academy of Arts & Science put out a pdf called "The Digital Dilemma". There is a part 2 now, it's a great read.

www.oscars.org/science-technology/counci...ects/digitaldilemma/

It's interesting because with all the films that will now only exist in digital form, as in no 35mm IPs struck, they are at a huge risk for being lost. The pdf goes into many scenarios of having to keep leapfrogging data from one standard to the next.

LTO tapes are the standard because there really isn't anything else like it. I still have exabyte tapes of projects I made ten years ago with absolutely no way to restore the data unless I hunt down someone with a deck. And even then, exabyte only worked reliably with the deck you used to make the tapes.

BDR discs are a safer bet than DVDRs as they don't use dyes to record the information, it's copper and silicone which aren't as susceptible to 'disc rot', so the theoretical lifespan is 100 years. It's at least ten years since the discs that I burned ten years ago (for $70 each!) still work.

That said, they are still breakable, could melt, be used by someone as a coaster, anything. No one is safe!

But I agree with VidGreg, there is a huge need for a rugged reliable backup system for professionals and consumers. With all the 40-something mothers and fathers of now taking digital photos of all the precious moments, the box of Grandma's photos will be pretty disappointing in about thirty years if something isn't done.

I've been leapfrogging technologies with digital audio and photos and videos for almost 2 decades now. With audio for much longer than that with changing and evolving DAWs.

Hi Sean,

I have a HP 1840 SAS running off an Atto HB380 in my early 2008.

Not sure what the current models are. I'm extremely happy with the 1840 and hope to keep it running once I move to a new Mac with thunderbolt via the ATTO ThunderLink SH 1068 Desklink Device.

Happy editing

BenB wrote: I've been leapfrogging technologies with digital audio and photos and videos for almost 2 decades now. With audio for much longer than that with changing and evolving DAWs.

And I bet it's a big ol pain in the butt!

My Toast Spanned discs scare the crap out of me. With those, luckily the spanning software is on each DVD, but some spanned discs require an OS that can run Rosetta. Right now, I can safely say I have that covered, but not much longer.

Actually no, because as I migrate from Mac to Mac, ext drive to newer ext drives, the stuff is still just data I drag along with me. Rare occasion I had to transcode/translate something. But no, not a pain, really. It's not like there was some huge, drastic change overnight. It was a slow, gradual evolution in technology. Floppies copied to CDs, which cooped to DVDs, which copied to drives, etc.

Well this thread and my sometimes lax taking my hdds (25+) out for exercise made me research low cost back-up solutions.
There is a relatively new technology called M-Disk by Millenniata from Utah that uses a hard inorganic recording layer onto 4.7GB DVD or 25GB BD(Blu-ray). Guaranteed for a 1,000 years. Does not use organic dyes. Can be written by a growing number of BD M-Disc enabled burners and read by any player. The M-Disc laser is hotter than a standard burner.
Link here… www.mdisc.com
LG became the first company to support M-Disc, but now lots of companies make suitable burners. Internal drives are normally SATA , external SATA, USB2/3, and some have HDMI interfaces. Costs between $100-$200. All the M-Disc capable drives also burn standard DVDS/BDs.
M-Disk media goes for around $3.00/DVD and $5.00/BD.
Write speed is fairly slow. Limited to 25GB/ disk. Torture tests look very impressive.
For a low cost solution, not bad for archive.
Anyone here using M-Disc?
Hope this helps, Greg

Thanks for the heads-up on mdisc. Looks interesting.

I did a quick price check and calculation comparing mdisc to @cofe's HP 1840 SAS. It looks like the break-even point is around 10TB of total lifetime storage. That is, if you store less than 10TB over the lifetime of your equipment then the mdisc approach costs less in equipment and materials. If you store more than 10TB, the LTO approach costs less -- and that doesn't include any accounting for your time, which would probably move the break-even to fewer TB of total lifetime storage.

Even if you think you will have over 10TB of total lifetime storage, I could see mdisc as an inexpensive way to get started for the first TB or two while saving for the longer-term equipment.

Nice addition to the tool kit.