On Backup

I've written an old old note about the necessity of making backups of your photos before. And that hasn't changed. However, I think that there's a considerable amount of foul advice about backups spread on the internet today, especially in the wake of "cloud backup" (storing your files at an external hoster, uploading them over the internet).

Now, I've worked for more than ten years in IT, so I think I have a bit of insight of what's actually most likely to happen, and how a backup, and which backup, can save you from disaster. And what I'm about to write is just from my own perspective - that you're working with your photos on a single computer and are using a mainstream desktop operating system. I'm assuming that both the operating system and the data worthy of backing up are stored on one and the same computer, with a large enough internal hard disk.

I think that's actually a very desirable way of handling it, because, for the sake of convenience, if you can keep all the data together, from my point of view, you should. You might have an internal SSD as a fast boot device, and a large conventional hard disk for the data in your computer, or in other words, more than one single harddrive - that doesn't matter. I simply consider working with internal hard disks in a stand alone "client" type computer (desktop, workstation, laptop) preferable over using external hard drives for the permanent file storage. Drive letters and paths changing and whatnot are a pain in the brain. Internal hard disks keep things nice and easy.

Let's have a look at how disaster strikes first. In all these years that I've been working in IT, the by far most common scenarios where a backup was needed used to be:

1. "help, I've accidentally deleted a file!" (and it's on a server share and not in the Windows recycle bin; pre Windows Server 2003 era, ie. no server side shadow copies)
2. "help, my computer doesn't start anymore, it just makes a weird clicking sound all the time" (because the entire harddisk has failed)

#1 is where I do consider cloud-based backups useful. How useful? Depends. Operating systems have a recycle bin - if you notice your mistake soon enough, you can simply restore the file if you accidentally deleted it. And if you already emptied the recycle bin and are using Windows, you can probably go back and restore the file from a shadow copy (right click on the folder, "Restore previous versions").

But are cloud-based backups useful for #2? Not so. I have a simple question: how long does it take you to set up your computer, connect it to the internet, re-install all your applications, and then restore the data from a cloud-based backup? If you need to be back up and productive, the answer is: ages.

The most convenient way to keep the "downtime" short when disaster strikes in form of a total hard disk failure (reminder: it's the most common issue I saw on client computers) is a so-called "image" backup, and that's what I would like to recommend as a primary backup. A cloud-based backup for files is a convenience for the little bloopers like an accidentally deleted file, and I consider that a necessary, but albeit only secondary backup.

An "image" backup is a 1:1 copy of the computers hard disk(s) that is stored on another (most often external) hard disk. The software that creates these images either comes on a CD that can be booted, or it allows the creation of a so-called "rescue CD" that can be booted. Why? If your computers hard disk fails, there's no operating system to boot from of course. So if your hard disk fails you get a new hard disk, hook up the backup harddisk, boot the rescue CD, restore the image backup, and wait. For 400GB, the wait could be something like 5-6 hours. After that, you boot up your computer and are back on track. And that's the important part.

Do you think you can install your operating system, the drivers (for your screen calibrator), the software you need, etc. etc. etc. from scratch in 5-6 hours to then access your cloud-based backup? That's unlikely. And how long will it take to download all of your data from a cloud-based backup? Even if your internet connection is 50mbit or 100mbit downstream, downloading 400GB of data will take a veeeeery long time. It's just like that.

So, a local, image-based backup it is. For the sake of speed and convenience. So now about the how and when.

We're talking about photos, right? There is no way to get them back when the data is lost. Photos are not like your computers hardware that you buy again when it's toast. Photos are not programs that you can buy and/or download again. Photos are not digital downloads of music and videos that you can download again. Your photos are unique When you lose them and have to backup, they're gone. Forever!

With that in mind, it makes no sense at all to have one single backup of your computer's hard disk. Where do you store your backup hard disk? At home? Of course you do. What do you do if your home burns down, or burglarized, or flooded, or collapses because of an earthquake?

You need at least two sets of backups. Keep one at home for quick access. Store the other one in a drawer in your desk at work, or at the house of a trusted friend, that sort of thing. It's simple. It's convenient.

Cleaning Up

This post is not really photography related, but once more meanders off into the realm of social media and such. I find it quite remarkable how history is repeating at the moment with the discussion about Pinterest and how it is a site that "promotes" copyright infringement through their linking, thumbnails, copying of "pinned" content to their own servers, and last not least, their terms of use. Sounds familiar? Yes, we heard all that a little more than half a year ago when Google+ launched.

Regarding Pinterest itself... I got an invite and checked it out, but decided that it is not for me. It's another iteration of Facebook's "Like" button and/or Google's "+1" function, except that Pinterest allows you to put your "likes" into buckets, to which people can subscribe, interest-based. It's a very nice concept because it allows the followers of a person to decide what they actually want to see, instead of curating "circles" (Google+) or "groups" (Facebook).

But personally, I'm not really a big user of the (external) "+1" button or Facebook's "Like" button. I do not wish to contribute to even more profiling about me and my habits and preferences on the web (in fact, I'm using browser extensions like "Disconnect" and "Ghostery" in an attempt to block most of the tracking), and Pinterest is yet another service that at some point will have to monetize by selling the user data and profiles they accumulated, just like all the others.

It's also a question of time. I do spend a lot of time on social networks - too much time, actually. And honestly, a +1 or a Like or a Pin doesn't pay your bills or keeps your fridge filled, and I think the effect of social networks to generate revenue is grossly overestimated (read "The hollow emptiness in social media numbers"). It's just so nice to hang out there, right? Now, if Google would pay me 1 cent for every click on one of my photos, it would be quite interesting, but as it is right now, I try to cut down my activities there.

So along with testing and abandoning Pinterest, I also gave up my account on 500px. The initial charm of the site wore off rather quickly, and looking at the "Popular" and "Upcoming" selections of photos doesn't really inspire me either. Compared to the old dinosaur Flickr, 500px's feature set is small, and while I wish them the best of luck with the "Market" that they're going to launch soon, I wonder how many photographers will upload their works full-size to the site (requirement: minimum 3600 pixels on the long side).

I see images being "stolen" (copied/downloaded) from 500px and other sites on Google+ (in the albums of hunters and collectors and self-proclaimed "curators" who in my opinion do little more than build a following by adorning themselves with borrowed feathers, drawing traffic and feedback away from the original artist) all the time, and I'll be damned if no one will find a way to rip the original size images off of 500px's servers.

I've been using 500px in the past as a simple portfolio site (with their "Awesome" option), but the choices of layout are limited. For just $20 more you can get your own domain(s) and a wonderful hosting package with Wordpress (or another content management system) - add a one-time investment into a nice photography theme, and you're all set and have more convenience than the limited choices a ready-made service like 500px provides.

Which is, needless to say, exactly what I've done for our photography business "Daylight Colors". Along with Pinterest and 500px went my own/old "portfolio" site www.alex-kunz.com, which was based on a mashup of Blogger, Picasa Web Albums, Dropbox and Fotomoto. While it was very convenient to upload a photo to my Picasa Web Albums and have it automatically appear on the portfolio site by setting a certain tag, being at the mercy of yet another service/provider (both Blogger AND Picasa Web Albums do not leave you in control of what's happening, feature and functionality-wise - they are free services, and you get what you pay for) was not too convincing in the long run. We're still in the process of building things, but everything will happen at Daylight Colors in the foreseeable future.

Tripods and center columns

A while ago I read a review Is this the world's best tripod?* (over at Luminous Landscapes), and especially the notion about the center column being a possible source for unsharp images got me curious.

Which is because we're using a carbon-fiber tripod from Feisol, the CT-3441S - with a center column (which itself has two sections). It is clearly the weakest part in our setup. I replaced the "kit" ball-head with an RRS BH-40 head and quick-release plate (and an L-plate for the camera as well - it's expensive stuff, but trust me, it's worth it; I knew the second I operated it for the first time), which turned it into a useful combo in my opinion - at least for travel and hiking, since it's very compact and lightweight.

The compact size of the tripod has one downside though, at least for me: in order to use it at or near eye-level, I have to at least extend the first section of the center column (Shuwen doesn't need to;-). I rarely use the second section, since using the first section introduces enough problems already.

Before I continue: I'm not putting a lot of weight on the thing - no camera/lens combination that we use weighs more than ~2kg/4.5lbs, the heaviest being the D700 (no battery grip) with the 24-70/2.8 - but we rarely use this (portrait) combination from the tripod since that is too static. During my hikes, I find the 24-70/2.8 is just too heavy, so it's the D700 with either the 16-35/4, the 24-120/4 and (far more seldom) the 70-300VR then - none of these combinations weighs more than ~1.5kg.

And it was with the 70-300VR where I noticed the apparent problem with the center column first: I could clearly see the image in the viewfinder wobble for a brief moment after an exposure from the tripod after the mirror flapped down again.

Now, there's a lot of advice "out there" in the internet on how to avoid the mirror flap affecting sharpness, and amongst them would be

1. using a remote release (cable or infrared) and/or the camera's self-timer (the latter being a crippled approach to getting one's hands off of the camera body when releasing the camera, it's better than nothing, but a remote is definitely better)
2. using the built-in "exposure delay" mode, where the shutter opens not immediately after the mirror flaps up, but delayed, and
3. using the manual mirror pre-release ("MUP" on the mode dial of Nikon cameras that offer this, more on that later)

But only the remote release and the manual mirror pre-release are actually helping in my case, because only that way I am able to delay the shutter long enough until all vibration from the mirror movement has settled: at least 3 seconds with both sections of the center column extended, and at least 1 second with just the first section out.

So how did I find these times? I mounted D700 with the longest lens (70-300VR) on the tripod, taped a laser pointer to the lens hood and pointed the camera to a white wall (zoomed in all the way to 300mm). In manual exposure mode, I set an exposure time of something like 5 seconds so that the mirror flapping up and back down would be easily distinguishable. I then used the self-timer to release the camera, and closely watched the tiny red spot on the wall as the mirror flapped up and down.

And yes, it moved. :P The D700's mirror flap is hefty: the little dot moved at least 2mm up and down, with the camera being at a distance of ~4m from the wall. And as I already mentioned: with both segments of the center column extended, it would not stop for 3 seconds! That of course means: the camera's built-in Exposure Delay mode doesn't help! Nikon only added the option to actually program this built-in function for a delay of up to 3 seconds in the most recently announced D800 - with all other cameras, it is a maximum of 1 second only.

So in theory, when I have to extend the first section of our tripod's center column, using a remote and manual mirror pre-release, I should be fine if I wait for 2-3 seconds after locking up the mirror before I open the shutter. But what if it's windy? Here's a 30 second exposure at 120mm with wind:

Ooops! Of course, that is a rather challenging situation, even for the sturdiest of tripods, but it nicely illustrates my point (Mike Spinak has a useful tip for increasing tripod stability [YouTube video] in such situations).

And the remaining problem is: the smaller cameras like the D3x00 series, D5x00 series, Dx0 series don't even have the - IMHO, trivial but absolutely essential - manual mirror pre-release. You're completely out of luck - the feature is only available from the D7000 on upwards...

---

*) needless to say, the RRS TVC-34L has become the object of my desire. You can add some baby steps to getting closer to it by clicking on the ads around here... ;)