Storage for Photographers
paulstamatiou.com
paulstamatiou.com
- 40-60gigs shot in a weekend, need to offsite the RAWs as soon as possible. In six or seven years we've never needed offsite retrieval. Fine for it to be slow.
- Only need around 100-200g of raws locally for jobs in progress
- Once processed into final jpgs files are approx 8-10x smaller. Offsite these as well, and need access to last 1-2 yrs of jpgs randomly/sparsely w/in 1-2 days of a print order coming in
I definitely think some well thought out service that uses maybe a hybrid of glacier and S3 could be really awesome. Especially if the pain is taken out of getting the bolus 40-50g of photos "uploaded" in the first place.. We'd happily pay for a service that just sends us an external drive and a postage-paid box.. we drop it in the mail after a job and get another drive sent to us right away.
[EDIT: Occurs to me you might wonder what this looks like today. Today we get home, and immediately back up onto an 8-bay Drobo Pro (http://bit.ly/1diZUHD). It's a fancy proprietary RAID-like system. I think ours has 7 1TB Drives in it currently, and it's at about 70% capacity. Then we back up the same jobs to an external hard drive which gets driven offsite (used to be my office, but now that I work at home it's the in-laws place) and copied to a hard drive for permanent offsite storage. Once that happens, the cards from the camera can be cleared. After processing, the final JPGs get copied to the drobo and make it into the sneakernet offsite process as well.]
I finally hit a sweet spot of performance, archive accessibility, and offsite safety, by using internal SSD for latest import workspace, direct attached WD Velociraptor (Thunderbolt) for current projects (can reconnect between Mac Mini server and laptops as needed), and 2 direct attached LaCie 4big Quadras (daisy chained via FW800) for archived projects (up to 32 TB).
Interestingly, because these are all direct attached, BackBlaze will happily back them up offsite at ridiculously low cost if that's your thing. Initial backup of several terabytes will take several months, though.
For fast cheap offsite backups, partner with a colleague with similar storage needs who uses the same ISP via the same head end. Just backup to each other. You can enjoy near LAN latencies at maxed out link speeds. This trick will save you both time and headaches.
Finally consider a product like the CyberPower OR2220PFCRT2U (this one also blends nicely in an AV cabinet) to keep an Airport Extreme and Mac Mini along with all that storage running for about 80 minutes of power outage.
Really, I need something that gets a copy of the images away from my home within a day to have something I'd be comfortable with.
Really, I think it makes sense for someone to prepare for the future and assume this will be feasible for upload in the not-crazy future.. unless more hobbyists keep deciding we need exponentially larger RAW files.. It's a crazy race :|
These guys talk about doing that with a RAID 0 setup: http://www.youtube.com/watch?v=cKaXZVHcdGM
The Joplin, Missouri tornado, Moore, Oklahoma tornado, Hurricane Sandy and Katrina.
I don't think you should rely on someone who MERELY uses the same ISP. I think you need to be a thousand miles away from each other.
I also store full "originals" (JPEG only, not RAW, but at least its something) on an unlimited Pro Flickr account, again, until their policy changes.
Of course, the challenge is to make sure the peers are online when you need the files, and for this reason the network may need 5-10gb of storage for 1gb of data to duplicate the content across multiple users.
The difficulty is ensuring enough clients have a copy of the file that it won't disappear, especially with people connecting and disconnecting a lot.
Do you want a one-time fee, or some yearly fee?
If you could get me to _really_ trust your service, I'd consider up-selling permanent archives to our clients. Lots of our clients would pay 50-200 bucks for the re-assurance, I'm sure. I don't know that I'd want the ever-growing liability though (right now we contract for 1yr of backup/online-gallery, but keep everything permanently without committing to doing so).
One time or yearly fee doesn't make a ton of difference to me, but with other services we've definitely been tempted into fixed "Unlimited" pricing so that we don't have to think hard about math, or end up tempted into picking and choosing what we backup. Simpler is better.. Managing a store of assets where some are backed up and some aren't sounds like a really bad accident waiting to happen.
4 TB of usable space divided by 60 GB times $20 is $1,365. That's the kind of thing a consumer or prosumer could do on their own, but it would be pretty hard to make a real business out of it...
EXTRA CREDIT: Anyone who considers tackling this and finds themselves interested in the niche of wedding/event photographer should look at combining this service with shopping-cart/gallery solutions that currently exist. The marketplace for those is fucking awful. Feel free to email me, but googling will get you the big players. I just don't wanna bash them by name in a public forum. All well meaning folks I'm sure, but every one of them are terrible flash-based, no-mobile software. Do backup and gallery/print-ordering management correctly and we're a customer at four figures per year.
Simply, they are two radically different businesses with entirely different feature sets. You might very well need both.
They're not radically different—one just uses the other. You could easily do both and continue to expand the use of the backend to enable useful applications. It's ridiculous to shun this specific consumer model just because it happens to be only one way you might use the original idea.
I don't have near the volume of pictures that you have (around 1.5 TB) but here's my setup: - one disk attached to the Lightroom computer, with all images on it; it's local, so it's fast - one NAS (NAS1) with a 4-disk RAID that backs up this drive (3 TB capacity) - another NAS (NAS2) with just one 3 TB disk that backs up the other NAS
On NAS2 I rotate disks; disk A is in the safe when disk B is in the NAS.
I think my setup is quite similar to yours, except driving is replaced by "going to the safe in the basement".
If a disaster happen I should have at least the disk in the safe intact, which holds the data since the last rotation; and since backups from NAS1 to NAS2 are quite fast I try to have both disks pretty much in sync (after an important shoot I back up to disk A AND disk B).
Of course I won't know if this setup is good until my house does burn down (I hope I'll never know!) but it does give me peace of mind, which is at least something ;-)
We pulled our bootstraps hard and tried to get it going, but soon realized that there's a pretty big flaw with the idea: most photographers have no idea that they should backup their photos. They know hard drives crash, but they never expect it to happen to them.
This meant that we'd essentially have to become insurance salesmen, convincing people that bad things will happen to them.
Eventually, we decided to shut it down, which I also wrote about: http://tghw.com/blog/well-that-sucks-what-else-you-got
The old homepage is still up at http://www.snaposit.com/.
[Edit]
Also note that OP wasn't really just pitching "backup". He wants you to reduce the footprint his images take up locally. I think the messaging to casual and hobbyist users probably ought to play on this. Note Loom's taglines "more room to play". Backup is probably boring for people not under the pressure of liability like pros are.
My undergrad was in USC's entrepreneurship program. One of the points they always hyped on was whether or not you were "selling vitamins or vicodin", so to speak. As you learned through Snaposit, most people when making chit-chat will be supportive of your venture, but you won't know how they really feel until they vote with their wallets.
Most of the professors were older guys who didn't get online startup culture, but one of their favorite tricks was to put up a landing page detailing a concept they were interested in, with a buy button. They'd spend some AdWords to get people to the landing page and measure the conversion rate. If enough people tried to buy to justify the service, they'd go out and build it.
In fact, that's how the Four Hour Work Week or one of those books was written -- ads first, and when the ads converted, the book was then written.
It looks like your product targeted consumers, but professionals seem like they'd be a better market. And it should have been priced way, way higher; the more it cost, the more secure it looks. I'd have a hard time trusting my money-making business assets to a $9/month service.
Of course, I'm not a professional photographer nor do I have your experience in trying to build this business, so who knows what my opinion is worth. :-)
Consider the some examples:
* An older photographer who grew up in the days of film who has just transitioned over to digital.
* A homemaker who is trying to add some much needed income for her family, with a passion but no formal training.
* A well-off software developer who moonlights as a photographer for fun and a little extra money.
Of those, only the third has been exposed, formally, to correct data backup procedures. The others are likely to keep their photos in folders on the desktop of their laptop.
Furthermore, photography is not a very lucrative profession for most photographers. That limits the amount you can charge. Our original pricing was much higher, but we realized that would be problematic, considering the average photographer brings in around $30k per year.
"Professional photographers know they need to backup their photos."
A homemaker or techie moonlighting for extra income aren't really 'professional photographers'. They may take photographs that look professional, but it's not the same thing.
One might say that professional developers use version control, then have counter examples of "well, this web design student doesn't ever uer version control, but she's the only person working on the code, so it doesn't matter". This is not a professional developer, even if she is capable of producing results that look like they were done by a professional.
A professional, imo, is defined as much by their habits and practices as by the output.
A friend of mine had her wedding photographer lose 6 months of data in a hard drive crash. She should have known better, and may have, but she didn't do proper backups.
What else would you call it, if not professional?
That said, I have to harass her about backups.
Everyone I know backed up on-site, using manual methods as and when they remembered. I did actually set up a good automated backup solution for one of them, but there's definitely a limit to their interest. Remember they're making pretty much all their money off commissions, and are only keeping backups out of some sense of pride/responsibility.
Another blow is that, at least here in the UK, we're only starting to get the kind of upstream bandwidth necessary to make this level of data transfer feasible.
So, it's a brilliant idea and, as long as confidence in cloud storage grows, as time goes on the market will grow and grow. I'm just not sure it's quite there yet.
As a web developer who works on a lot of solo projects, this sentence hurts me.
> It did this by compressing the photos to full resolution, high quality JPEGs, which saved anywhere from 50%-90% of the size. The downside of this, of course, is that photographers who shoot in RAW would only have a JPEG copy backed up. We figured this was a reasonable compromise. In a catastrophic event, wouldn't a photographer rather have a JPEG copy than nothing at all? It turns out the answer was no. But we didn't ask that question, because we thought the answer was so obvious.
I am no photographer... but from what I can see in this thread, what people are looking for is a solution to store RAW image files. I am sure photographers can find already good solutions to store good-quality JPG copies of their files.
If we care about decades, storing the raw NEF or CR2 file is almost certainly the wrong approach. Those files could easily be as difficult to open in fifteen years as a ClarisWorks document is today (just to choose one example - proprietary formats for professional software from the late 90's are all pretty hard to open today). Also, important metadata is kept in the lrcat file. Unless we expect to always be able to open our copies of Lightroom, we need to migrate that data too.
We used to think of "archiving" as putting something on a shelf and forgetting about it, sometimes taking environmental factors like humidity into account. With our digital data, we need a plan for periodically checking it, both for fixity and to make sure at risk file formats are migrated to current best practices. This will require greater awareness for the complexities involved (and the need for open file formats) from everyone from creatives to storage and service providers.
By the way, every creative industry is struggling with this right now. Recording studios have session files created in ProTools 4. Designers have really old Quark files, even though they've been working in InDesign for the past five years. Composers have Finale files. A lot of these file formats depend on plugins to be able to display properly. I don't mean to sound discouraging, I think there's plenty of opportunities for viable businesses here. We just need to start using the right terms.
That said, I used to be a pretty prolific hobbyist musician and I haven't pulled this off myself. I have tons of old songs I can't open because they rely on a fiddly collection of old freeware plugins for freeware music apps from 11 years ago. I also have songs I can't open because they only play in a music tracker I wrote, which is open-source, but only runs on OpenBSD.
It is a tricky problem and it seems hard to tackle because the market for professional software inherently favors closed-source, which has no incentive to adopt open, non-siloed formats.
...do we know, when you store something in Glacier, does Amazon keep multiple copies, or just ONE on some unspecified media? Do they take a hash fingerprint, and compare it to the copy for bit rot corruption?
Even without getting to format obsolescence, the strategy of just keeping the bitstream reliably and verifiably retrievable for 'decades to come' is worth discussing.
Rather, they charge you on a monthly basis, according to your highest hourly retrieval rate from any single hour applied to the whole month (standardized at 720 hours). So if your peak hourly retrieval in a month was 10 GB/hr above your pro-rated free quota for that hour, you're charged for 7200 GB of retrieval overage, or $72, even if you retrieved nowhere near 7 TB of data. For companies with relatively constant retrieval this doesn't matter, but for small users, someone who has a large overage one day will be charged as if they were incurring the same overage continually all month.
To put it concretely, if you store 200GB of photos and then do a bulk retrieval, you aren't charged $1.90 ((200-10) x 0.01). Rather, the xfer is counted as 4 hours, so you have a peak transfer of 50 GB/hr. Your free daily retrieval is allocated to the hours in the day you retrieved data, so in this case 10/4 = 2.5 GB/hr free. So your overage is 47.5 GB/hr. Multiply by 720 and $0.01, and the fee for retrieving 200GB of photos is $342. Plus $22.88 in outgoing AWS bandwidth.
Or at least, I think that's what their pricing page says. Maybe someone else can better interpret this explanation: http://aws.amazon.com/glacier/faqs/#How_will_I_be_charged_wh...
Arq (a backup app I wrote) takes a transfer rate from you (guesses your max transfer rate as a default) and then requests enough to be downloaded in 4 hours at that rate. 4 hours later, it requests another 4 hours' worth while it downloads the first 4 hours' worth, and so on. This mitigates the peak-transfer fee.
"SmugVault is an added service to your SmugMug account that lets you store almost anything for next to nothing. Including files not normally supported by SmugMug."
http://help.smugmug.com/customer/portal/articles/84562-more-...
Right now, most of them use Drobos or similar devices. You'd be asking them to add steps to their workflow which is always tough to do.
Also, I think your estimate of 18 mins to upload 6+GB is an order of magnitude off. We have tools that allow people to upload JPEGs and most customers claim it takes longer than that just for 100s of MBs.
My (wild) guess is that Adobe is going to get in this space too. It only makes sense to have your Lightroom catalog in the cloud. Maybe they wont offer an affordable solution for terabytes of data, but I'm not sure photogs want their images in two different "clouds".
I think the key is to do customer development from full-time photographers who aren't hackers. I'd be happy to send out a survey to some clients if you want do more research.
and I still think it will be hard to convince people to "put these files on Adobe's cloud and these files on this other cloud using X software"
I think you're totally right about Adobe too. Those Smart Previews they added for LR5 - those are not just for space-starved SSD drives...
You can't really have the current catalog in Dropbox because then each action and edit become really slow.
But if you set up Lightroom to do a backup every time you quit the program, and that backup is in Dropbox, you get the best of both worlds (strong backups + snappy edits).
Lightroom backups are really "backups", as they are named by date+time, so in Dropbox I have a history of everything I ever did in Lightroom, not just the current (last) version.
It's also fast because I suspect Dropbox is able to do some rsync magic and only re-upload new packets; my Lightroom catalog is around 150 MB and the upload to Dropbox is almost instantaneous (on a pretty lousy upstream bandwidth).
Of course this doesn't address the problem of backing up RAWs, but as far as Lightroom is concerned I feel pretty safe.
Your LR catalog contains your photo edits, metadata and history... but is not your original files. Backing it up is very important but is only part of the story. Putting your Lightroom catalog backups on Dropbox is a great and should be done by everyone. (Again just not your active catalog.)
There really is no good solution to automated offsite backup when it takes several days of uploading to handle a single day's worth of data capture. To make any internet-backup service work, you'd first have to pitch the photog on switching from their residential broadband package to something more expensive -- if that's even available short of their leasing office space somewhere.
I guess you could try to build out a reverse-CDN sort of network, with local relay nodes scattered across the various ISP networks, that might achieve faster uploads from the user (and maybe not get counted against bandwidth caps) which then use a fat pipe to send that data to larger regional storage nodes.
Neither seems particularly plausible at scale.
We want to take backing up, sharing, and monetizing and make them very simple and enjoyable to do. We've started rolling out slowly already, our goal is to be open to the public within a month or so.
Currently, I have several multi-terabyte drives that I erratically back up to...I just got a 4TB drive that I'm going to try to backup everything I've ever shot in the last 3 years...
But I do want to move it to online storage...and I think this requires triage. For nearly every photoset, I've quickly gone through them with Lightroom and starred the ones that I kind of like, and then for maybe 1% of them, have taken the time to fully process, label, and upload them (in JPG form) to Flickr.
So for online storage, I think what I'll do is write a batch script that dumps all the unstarred RAW files as JPGs, because they chances that I'll ever need these photos in RAW is very slim, and then upload them. For photos that I've given at least one star and/or taken the time to properly edit and categorize, I'll send up the RAW file along with the processed JPG.
If I have about 3TB of RAW files...and maybe 10% of those are RAW images that weren't disposable...that's 300GB right there. And then the remaining RAW "keep em just in case" files would end up being compressed from 24MB to about 2-3 MB (let's say 15% as an average)...so 400GB.
700GB is still quite a big footprint. However, the biggest problem will be...let's say I give up local storage all together and rely on the cloud...what's the best way to browse any photoset at any arbitrary time? When it's all local, it's trivial to pop open Lightroom and do some quick browsing and filtering. For online storage, I'll have to put a ton of work into proper folder-naming, at the very least
If I could browse locally with smaller images (as you describe) like I currently do with Picasa (galleries, tags, etc) and then have the ability to download the original images when and if I need them, I'd be in heaven.
Of course, Apple's iCloud is very little like an actual usable cloud for most pro users.
1.) Personal: This doesn't mean I don't delete, but I generally tend to keep photos even though they might not be really good photos, especially if I only have a few of that day or person. I know I might enjoy looking at them decades later anyway, a blurry photo is better than nothing.
2.) Archival: like taking a photo of a flat before moving in. These aren't very many, so I tend to keep them no questions asked.
3.) Siblings of Keepers: You could call these "proof I really made the keeper photo" photos. Let's say I come across an animal I never photographed before, and like one of the photos enough to publish; then I'll still keep the others around. Or keeping a wider crop in case you change your mind later, etc.
4.) Keepers: here I try to be as harsh as I can. Photography is about painting with light for the pros, for me it's mostly about selection, that's the part of it I enjoy most. I walk around, and from the thousands of things I see each day, only a few make me take out the camera. Then I select keepers (partly before the images even leave the camera), and even years after I might revisit some keepers again, decide that my standards improved, and delete them.
I don't want to tell others how to pursue their hobby, but I think the web is full of galleries not even the owners look at, and that generally, a lot of people need to learn how to delete. The same goes for blogs and whatnot, the idea that everything has to be written on this indelible roll of toilet paper seems silly.
On one extreme, there is being OCD and deliberating endlessly what to keep and how to rank it, on the other there is just making more stuff to throw behind yourself on a pile you never investigate, because it's too awkward to wade around in. Productivity lies somewhere in the middle I think.
For example, you can't (and in the broad "you", aren't even qualified to) audit the provider for data-loss or SLA-impacting SPOFs. You just have to assume that they're there. You also take on non-technical failure modes: the provider can go out of business, get bought by an uninterested owner (think Delicious), change focus (Google Reader), or experience myriad other problems.
FWIW, the best success I've had with helping others switch to good backup policy happened once it became economically feasible to buy a new laptop and two bus-powered external drives as big as the laptop's internal storage. The externals become bootable backups, one of which can be preiodically rotated to an offsite location. This has prevented severe dataloss events for myself and others far more times than I care to count now.
(for local storage, once you go beyond individual USB drives, I'd go with a Synology 1813+ NAS ($999 chassis, 8-bay) or an Areca ARC-8050 thunderbolt raid ($1499, 8-bay).)
With that in mind we've started to look more on how photographers with larger photo volumes could benefit from our service. From the feedback from few photographers we've already included some features (RAW photos with smaller JPG thumbnails, workspaces and personal sharing for example) that will make the day to day workflow a bit easier. Currently our goal is to build something that would fit better for professional photographers (public for clients, ordering prints and possibly integrated payment flows, win/mac client + plugins for your favorite photo software).
Pricing wise we are currently not the cheapest but we have thoughts on how to make this better for high volume photographers. If you're one of them then please send me an email (martin at photry.com) and I would love to talk a bit further what we plan to do and if we would fit your storage needs.
For back-ups - just buy a huge disk or set of disks, and keep 'em off-site wherever you have a trustable location (friend, family, work - if allowed and bandwidth isn't an issue) and use CrashPlan's free software to sync a copy to the secondary location. It's free after the initial expense of the drives and gives you encrypted, but physical access to the content in case of an emergency. You could find a friend with the same problem, and just agree to dedicate a certain amount of disk-space to each-other for this purpose.
I'm the technical support for overall workflow, storage & backup engineer, etc, etc. I wear many hats.
Right now I've got a backup on import to a FreeNAS (BSD) filer running on a HP Micro Server. Lightroom backups go to the same filer and are synced as well. My local server rsyncs the delta every night to an identical machine at a friends' house.
This isn't perfect. The one place that isn't 100% backed up is current projects in post-production. That's an acceptable trade off for now.
The most important thing to note is that there are two kinds of photographers: Those who never lost data, and those who care about backups. As mentioned elsewhere in this thread, converting members of Group A into Group B is a tough sell. Video testimonials might be a good option.
A storage service for photographers MUST be invisible to their workflow. Make it as easy to use as an egg timer. At the MOST, an addition of a single screen with sensible defaults (just click next) on photo import.
Ignoring error messages that get in their way is to be expected. Care more about their backups than they do. Send me (tech support guy) an email to notify if there haven't been backups added in X days, along with weekly reports of data usage per day -- how many photos, how large, what root folders, usage trends, etc.
An ideal installation would be automatic detection of what program they're using, what places files are stored (hint: It's not just on a local drive). Detect new file locations and back those up too.
An inexpensive (<$25) option of "only backup current projects I'm working on and anything I shot in the last month", that actually works... I'd sign up in a heartbeat.
If your files are important to you - you need a back up. If not - just get rid of them in the name of simplicity :)
PS. Also, uploading many gigabytes over a typical ADSL connection is just painfully slow... even local NAS is too slow over WiFi for my taste, and if I have to have a wire - I can just as well connect an external drive.
But the idea of storing any files in the cloud is very distasteful to me ever since the NSA Snowden story broke. Even something as innocent as a photographer taking pictures of a skyline or some beautiful architecture could land him in hot water if the government is looking at metadata and figuring out that he took lots of photos of a government building or a bridge or a power plant and randomly throwing him on some watch lists or no-fly lists. People have been hassled by the feds for a lot less.
My preference would be for all of these cloud storage providers like Google, Facebook, Apple, Dropbox, Amazon, etc to unambiguously and fiercely challenge the government on this and somehow unambiguously prove to people that they're respecting peoples' privacy and not complying with this stuff. Their denials so far have been obvious non-denials that are very carefully worded around "no direct access" when that's enough wiggle room to do just about anything.
Using some kind of cloud photo storage is a bad idea IMO until this gets sorted out.
- Local daemon running that manages this process, on PC or Mac. It can't be done solely from the browser.
- The software manages the import from the SD/CF cards, you put them all in, it sorts everything out and wipes the SD cards on the way out.
- You then use Lightroom to do whatever you're going to do. The software would do well to have a Lightroom plugin that lets you select particular images to mark as "high priority" and "low priority."
- Now the daemon uploads all of your RAWs, Lightroom metadata, and corrected JPGs to cloud storage.
- Based on your settings, these RAWs get discarded after some set amount of time (maybe immediately, maybe never), and the highest-quality JPGs are also discarded after a longer amount of time for lower-quality JPGs just in case you need a emergency backup. After 6 months you lose the raws, 1 year the highest quality JPG, 2 years the medium quality JPG, etc. You are charged per gigabyte per month so you can decide how much you care about backup. Basically offload the cognitive effort of discarding old data that isn't worth its upkeep costs anymore.
- The files are available via an API that you can use to build your own gallery/shopping cart site, with the ability to call a certain quality, PROOF watermarking, etc. on-the-fly. But the service offers a default site right there for you to do the same, that's stylable and can be implemented on your own domain.
- You can get your entire catalog shipped to you on a hard drive at any time for a few hundred dollars.
You shouldn't be touching the directory structure. This tool should be either a plugin for Lightroom/Aperture/Photoshop or an entirely separate application.
I would lean towards separate application. It could be a background agent that detects when new files are added, would work for any known editor or user-defined directory.
In any case, building this sort of thing is a pain. The most obvious problem is the size of the images. Glacier is an interesting backup solution, and had just come out around the time I decided to stop working on my app. A less obvious problem is raw conversion. All the best converters are proprietary, meaning that I had to piggy back off the Mac OS converter engines. Again, I never got to the point where I felt the product was worth the trouble of buying racking up a Mac Mini to run a little market test. Maybe I'm wrong.
I had a (frankly somewhat scattered) draft of a post mortem written, and since it seems like there's some interest in this subject, I posted it: http://scraps.bigbento.com/2013/05/30/hiatus.html
I think it's possible to have a either S3 or GDrive based system that has a Angular.js (or some other JS MVC) that can operate without a backend for very simple browsing features and connecting to the S3 via apis.
I think it's doable ...
I also trust Amazon to stay in business but mistakes can always happen (not just hardware but ID theft, card expiry while you're on a 6 month hike, etc) and unless you're prepared to say "losing these is OK" using one service would be risky.
So we are an ecommerce company and we do our own photography for products -- and we do a lot of it. We average about 2-3 photoshoots per month, and each photoshoot can be upwards of 40GB before any processing. We shoot raw images and are starting to experiment more and more with video. After we do the post-processing work and start the video editing we can push 70GB a shoot easily.
Our current workflow is like this:
Shoot tethered to a laptop each day (typically each shoot lasts two days). At the end of the day, we upload the files up to our local NAS (Netgear ReadyNAS, just shy of 3TB) which is redundant.
At the end of the shoot, we load a copy of the entire shoot on the local computers that will be editing so that they don't have to work directly from the NAS. Over the next few weeks while they edit the files, they will push the new files up to the NAS into specific folders.
(This is where it gets hairy) We also have an offsite NAS that is located in the business owner's house (about 6TB, but this one is not redundant). We used to attempt to upload files each night from the local NAS to an offsite NAS via rsync, but the pipe was just way too small. It would take days to upload a photoshoot, and everyone in the office suffered from slow internet while that happened. Now what we do is bring the offsite NAS onsite every few weeks and do a manual sync! Locally the sync only takes a few hours and then that night it goes back home with the business owner.
Now since the local NAS is less than 3TB, we can't store all of our shoots on there forever. When we get close to full on the local NAS, we grab an external hard drive and archive old shoots (typically 2+ years old) to it, and then store that in our massive safe (on-site though). If we need access to just one or two files, we can retrieve them from the offsite NAS, but if we need access to quite a few of them, we can dig them out of the safe.
So that's our workflow. Probably not bullet proof, but it has been serving us well for the last few years. If someone came up with a better solution, we would be all over it!
The thing about Amazon Glacier... unless they are themselves making more than one copy, and taking hash fingerprints, and storing a couple copies of the hash fingerprints, and using it all to check for bit rot on a regular basis...
...and I don't think they're doing all (any?) of these things...
...then people are sadly going to find that their stuff _isn't_ neccesarily safe for decades to come. One copy on a single piece of media in an Amazon data center somewhere does not actually make for 'safe for decades to come'.
> Amazon Glacier is designed to provide average annual durability of 99.999999999% for an archive. The service redundantly stores data in multiple facilities and on multiple devices within each facility. To increase durability, Amazon Glacier synchronously stores your data across multiple facilities before returning SUCCESS on uploading archives. Unlike traditional systems which can require laborious data verification and manual repair, Glacier performs regular, systematic data integrity checks and is built to be automatically self-healing.
http://www.kickstarter.com/projects/cloud-guys/plug-the-brai...
Specifically, the $129 level, where you get two plugs, which they claim, "to replicate your data in 2 different places." Talk two of your friends in remote parts of the country into letting you Plug in to their home networks, and this might be viable.
As a background, we now have thousands of users and customers. We also finished raising a series A in February. (Not that raising money proves anything) http://www.boston.com/business/technology/innoeco/2013/02/mo...
We are now growing at about 5%-7% a week. We recently peaked in the top 100 grossing US iPad Photo/Video Apps in iOS - although most of our revenue comes outside of iOS. (As they take a 30% cut.)
We target serious photographers (pro's and prosumers) who use Adobe Lightroom. For pure backup customers our sweet spot are those who have outgrown BackBlaze and Crashplan. (Both great services but geared at consumers.)
On the backup side, we help photographers manage their backups by using Adobe Lightroom metadata. Customers can choose to backup everything in Lightroom automatically. However, many photographers use metadata like stars or flags to designate their best photos. We allow users to automatically backup these "best" photos. This provides the automation of the best backup tools with more precision for larger customers.
We use Amazon Glacier behind the scenes to store the original photos. (Yes, WMF the restoration costs even out for us over all of our users.) (Sidebar - we actually ran our data center for months before switching to Glacier...very glad to have made the switch.)
However... we are not an online backup company. Online backup is a feature is our service... but not the entirety of it.
Once we have have photographers photos, we also want to help photographers better use, manage, share, and enjoy those photos. At one point I would have said we are a backup company, now I call us a workflow company.
We also offer a web and iOS App to privately view and share your Lightroom photos. (We take JPG snapshots of the RAW files so they load quickly and have the Lightroom edits. This part of the service runs in Amazon S3 not Glacier) This gives you access to your photos in the same way you are used to experiencing them in Lightroom. (Lightroom users try this for free - https://itunes.apple.com/us/app/mosaic-archive/id627973694?l...)
The next step which we are working on now is adding two-way sync where you can manage your photos in the App and it populates back to your desktop files and Lightroom. We are also working on more collaborative features so photographers can share this culling and editing process with their clients or family members.
The bigger problem we are solving is that photographers have loads of large files... backing them up is painful yes... but managing them all is more painful. Sifting through 000s of images to find your best stuff frankly stinks. This problem is only getting bigger as we shoot more and more.
For those interested in this idea, please consider joining us. We are hiring a couple more software developers this year. http://www.mosaicarchive.com/about-mosaic-photo-storage/care...
Thanks for the discussion and happy to answer all questions.
Best, Gerard www.mosaicarchive.com
Especially given "However... we are not an online backup company." I'd feel tremendously safer and more trusting of your service. I haven't used Mosaic so maybe you already do this, I'll take a closer look this week.
Also (in a nice way), your site could benefit from a redesign. Unfortunately I think quite a few consumers judge the trustworthiness and value of a brand by their site/landing design.
We have certainly thought about giving direct access to customers Glacier accounts and behind the scenes each customer has their own glacier "vault" for potentially this reason. But for many customers they would prefer a set-it and forget it type of service. We didn't want a barrier of first create an glacier account then call us...
We also then didn't know what to do if a customer then changes their glacier files... if we are syncing with Lightroom and this gets out of whack... then it defeats the purpose of the automated sync with your local files. If it just "read only" access, we are already providing this with our App.
I think the difference in what we are providing in online backup is that we are not a pure cloud where we encourage you to delete your local file - We actually want to sync with your local files. Our philosophy on the workflow side is that changes that you make to your files on the App, should be reflected locally in your Lightroom catalog file. This way you always own your own data.
I agree with you on the site redesign... have had our head down on this. Good news is that the design is approved and being implemented as we speak (although won't be out for a couple weeks.)
Thanks again and happy to answer any other questions. Best, Gerard
While the post mentions avoiding externals I think a handy feature would be to push to both an external drive, glacier, and then remove the RAWs locally. That way you have two copies, so the data is actually backed up. Trusting all of your photography to amazon not losing data doesn't sound ideal.
With high network speeds, I can quickly re sync directories, or just power on one of my old laptops.
In fact, "a plugin for Aperture/Lightroom that backs up to Amazon Glacier" is on my idea list. Great minds think alike, fools seldom differ.
One thing I'd like to point out is that photographers are unlikely to be as cost conscious as other consumers. We put a lot of effort into taking these damn photos, and it'd be nice if I could be confident that they'd outlive me. I'm willing to pay a premium for that.
Downloads add significant cost if you want to retrieve more than 0.17% of your data in a day.
Downloading 5% of your data increases your monthly rate to $0.097 per GB. 10% increases to $0.187 per GB 50% increases to $0.907 per GB 100% increases to $1.807 per GB
Downloads are also delayed by 4 hours after you make a request.
Photographers with the problem described in the article are within the target audience, even if we didn't implement some of the key features for them yet (e.g. storage of images in RAW format).
We'd love having some early feedback from you guys! (Didn't officially launch yet.)
Unfortunately it is quite a niche and most photographers are gadget hounds and don't mind having 2-3 drobos (including an off-site one).
I'd love to hear more about this service from some users. I am bumping up against Dropbox limits with my photos as well.
searching is also a breeze there.
not sure how you'd go about downloading them all though...
Sorry but what? I actually had to read that a couple of times.
Kind of hard to take anything seriously after reading that. He buys a seriously expensive camera and shoots it in point and shoot size. That's just insane.
What's so hard with a simple home NAS. I replace the drive in my laptop every year for whatever is then biggest (sweet spot) and ue a NAS for backups. RAID5
I want everything local.
Regardless, that seems a nice setup. I'm not opposed to the cloud (as long as I treat it as a dumb storage system), but it's certainly more expensive than buying your own disks.
Personally, I prefer to use both local and cloud storage (encrypted, bien sûr).
Drives/tape/flash/dvd is pretty damn cheap and easy. These S3 type solutions would have to be much cheaper to be interesting.
Besides, how do you verify that those backups can be restored when you need them, if they're offline in some remote location? That seems rather untrustworthy to me. I run checks against parity files regularly to verify their integrity.
And can you download 47GB faster than your mom can mail you a DVD?
And can you download 47GB faster than your mom can mail you a DVD?
Probably, though I'm not quite sure about the speed of S3. I use a cheap VPS instead (much more expensive than Glacier, but also much, much more useful), from which I can easily get 3MB/s, so that would definitively be yes.
"Which is Better: a 10MP camera dialed down to 3MP, or a 3MP camera?
Easy: when set to the same resolution, a higher-resolution camera set to a lower resolution always looks better than a camera with that native resolution.
This is because digital cameras cheat and use Bayer Interpolation. They don't really have that many color pixels on their CCDs. CCDs are actually black-and-white and use colored stripes to see color. Bayer Interpolation spreads the data around to fill in each pixel as if it could could see color. Because of this the images of all digital cameras at their native resolutions are more blurred than they could be.
If you halve the size of the image, the pixel-to-pixel sharpness will increase because there is now enough data for each and every pixel. See my Bayer Interpolation page for examples.
Therefore it's always better to use a higher resolution camera set down to a lower resolution than a camera with only that much resolution."
You might want to read this, it will help you understand how sensors work.
http://www.cambridgeincolour.com/tutorials/camera-sensors.ht...
I recently compared a Nikon D3200, D5200, D7100 and a D700, all using the same (or close to same) lower megapixels count of around 10. The higher end cameras made a huge difference in quality. The lower end also could not support higher end flashes.
I own a proofing company with thousands of users, and ran a print lab for years. You don't need more than 200PPI for larger prints and thus can save a lot.
In fact, considering most people order prints 11x14 and smaller, having the larger megapixels actually hurt quality because the printers (Fuji, Durst, etc.) downsample the files and you lose sharpness if you send them the large file (which most photographers do).
For my wedding shooting, I never need a larger file than 12 megapixels as that's large enough for a 30x40 at 300 dpi. I prefer to save the hard drive space.
My point was that shooting that wonderful camera in 5mp mode is stupid.
"The higher end cameras made a huge difference in quality" - even when you down-sampled them to 5mp? And I think you're being ridiculous when you say "huge". There will be more intra-shot difference in quality than between any two modern DSLRs. The light will make far more difference to whatever it is you think is quality.
For weddings of course you don't need much. That was my point really.
You do realize right that you can put those same better lenses on the entry level canon right?
You, and everyone else, it seems, misunderstood. My point was that down-sampling to 5mp a canon 5dMkIII's images is a waste of that camera, and any entry level DSLR will produce indistinguishable images at that down-sampled level.
Also the standard for lab work is 300PPI, not 200, so I am not sure if I lend very much credibility to your comment.
Which is Better: a 10MP camera dialed down to 3MP, or a 3MP camera?
Depends on the lens. 10MP vs 3MP is one thing, 10MP vs 1000MP another... at some point you're just blowing up a fixed amount of signal, and with most lenses that point comes much earlier than one would think.
But even then, doing things like straightening or color adjustments before scaling it down looks better than doing it afterwards. That's what the camera does when you shoot in JPEG, but once that's done, it's done, no going back.
Would you buy CD, transcode it to MP3, and then throw the CD away? Maybe not all photos are that important, but sometimes you only find out where a particular one falls after some time passed.
It's all about the importance you place on the data, and your level of technical skill.