Musings on: deleting images too soon
I deleted the image below – several times.
My usual practice, after copying the captured images from a memory card to my computer, is to flick through the files and delete those which don’t work, or which have technical flaws which I’m unwilling to accept. This one just looked bland, as did the five other versions I’d taken in quick succession as breaks in the cloud allowed sunshine to sweep across this valley. They all experienced the delete key and both, duplicated cards were formatted when I returned the SD card to the camera.
A few weeks later, thinking over my Scotland trip, I recalled spending a couple of hours standing by the side of the road near the Rannoch Moor end of Loch Etive and imagining a gently-lit composition which highlighted the multiple triangles I could make out in this basin beneath Ben Starav and Glas Bheinn Mhor. I remembered visualising the image above – the raw material for which which I’d repeatedly deleted the day after capturing it. Considering it after so much time, I found it hard to believe that there really wasn’t something worthwhile in one of the captures.
Fortunately, several aspects of my file protection set-up cater for ‘deliberate, over zealous deletion’, rather than mechanical failure, accident, or software issues. In this particular case, every ingested RAW file is copied to two internal drives on my laptop, one of which I work on – deleting ruthlessly – and one of which I never touch, but which is itself copied to several other places on my network.
I’m glad I do this!
This may not be an especially spectacular image, but, having experimented with various DxO processing options for it and finally produced something quite similar to my visualisation, I do now like it, and I’m pleased to have stress tested the ‘idiot operator’ provisions in my backup strategy too.
Either don’t delete anything, or make sure you put an ‘I changed my mind’ solution in place
Let’s leave aside whether you like the image in question; that’s not the point. I like it, and you may change your mind on some of yours too. I’m sure we all capture the occasional image which, at first glance, is inadequate in some way, but which proves worth working on sometime later. I urge you to think of a mechanism to make sure that you can!
Of course, one reliable solution is to simply not delete anything, but I find it useful in my work-flow to reduce the RAW captures to a manageable few in the folders I’m working on; so, for me, deleting is good. That said, it’s fortunate that I foresaw the flaw in this approach some while ago and put in place a mechanism to avoid the obvious problem with this method of working.
I’m not saying that any of this is terribly clever – I’m merely suggesting that if you haven’t allowed for ‘over-zealous deletion’ in some manner, by making copies of everything very early on in your work-flow, do consider it. You never know when you might want to revisit an image-capturing session and make really sure that there was nothing in it worth working on.
An alternative to multiple, ideally automated, backups is not to review images too soon after capture. I know many people like to leave their files alone for a few weeks and then view them with more objective eyes. In this instance, the RAW files fell way short of what I’d envisioned and I more or less deleted them in a fit of pique; perhaps, had I left them a few weeks, I would have been more generous?
Whichever you do, make some provision to enable yourself to rectify the sort of initial mistake I made!
For anyone interested, I’ve written a page describing my overall work-flow and file protection set-up (also linked from the menu bar at the top). If you’d like to comment on this little story, and perhaps argue that I was foolish and ‘got lucky’, please feel free, though I may at least attempt to refute the suggestion with the view that working out the sort of foolishness I might be guilty of in future, and guarding against it through automation, ameliorates the fault to a large degree…
8 Responses to “Musings on: deleting images too soon”
I read your work flow description a while ago and was impressed (and frankly a little scared :) ) by its thoroughness – it’s more rigorous than most of the companies I work for. It pays to be prepared though. My own backups are a slightly more slapdash – I often don’t run a backup until after I’ve looked at the last lot of images which will invariably involve some deletions, but then I don’t bother formatting a memory card until I’ve moved on to a fresh one so I often have a couple of months worth of images on there. I do occasionally find things on the card when reviewing them and wonder why I deleted them. So I have a “best” drive with everything that doesn’t offend me, a backup drive which has some content on that has since slipped off the main drive and then another external which stays off site as frequently as possible.
Well I think three drives is pretty good going compared to the norm, particularly a [mostly] off-site one! I know what you mean; my backup solution is a wee bit OTT in some respects ;-) The key thing to me is that it’s completely automated though – it removes the need for me to think about having to do something.
I’m basically lazy so it’s that characteristic which primarily drives setting up lots of layers of backups ;-) I also like tidying things (sporadically!), so always formatting memory cards and removing electronic clutter from things I’m actually using is rather satisfying too.
Thanks for taking the time to comment :-)
A great article I agree with you. I too have a backup system designed to prevent this I find that processing shots a while after taking allows the emotion of the shoot to die down so deletions etc are more objective.
I also use lightroom so any deletions are just “marking for deletion” and I only purge a folder 1year after the shoot.
Thanks very much Chris. The key to any solution is that it exists! Usually, I’m too impatient to wait for a while before processing, hence my solution fits me. My point in this article was just to say “have one of some kind that suits you, whatever it may look like”, and I confess to being close to pleased that I had to actually use mine on this occasion ;-)
I tend to disagree on this one. My workflow is aimed at reducing the amount of images I store to only the ones that I consider worth holding onto (OK, that doesn’t always work quite as well as I might hope – we all hang on to stuff for sentimental reasons even if they aren’t that good). Even being fairly ruthless, I already store more images than I reasonably know what to do with and quite a lot of my stuff just gets an airing on Flickr and is rarely seen again.
I honestly can’t be bothered trying to knock an ‘also-ran’ into shape because I know from experience that the results rarely justify the time spent. Adages about sow’s ears are appropriate here, I think (not that this applies to your image, Mike!) That said, I do make several passes through my images weeding out duds and keeping borderline work at every stage – even stuff I was initially impressed with but, as my tastes change and my skills improve, now feel are inferior in some way. With each pass, the quality bar gets raised a little higher and fewer images make the cut.
I also try not to get too attached to my shots (unavoidable sometimes, I know) and I don’t worry that I might have deleted an image that might be a masterpiece with a bit of work. If nothing grabs me about the image first time round then it’s unlikely the image will grab anyone else’s attention – even with a bit of Photoshoppery. Of course, as your shot proves, there are exceptions to any rule… ;)
I’m not entirely sure that you are disagreeing very much here, Jools…
My workflow is also aimed at reducing what I store to, pretty much, only those which have made a whole series of ‘cuts’. In practice, I’m very ruthless: that means that my main file system on my laptop has two images from that day’s shooting (two RAW files, a TIFF of each, a PSD of each, and a JPG of each); just the two I like. The rest are gone from my primary storage.
The only place the ‘everything that was on the card’ set exists is on my secondary, internal laptop disk and its automated backups. My policy on that disk and its backups is that when it fills up (500GB) I’ll delete the lot, after which its backups will then fade away over a few months too and stop taking up space. The only persistent storage is for the 2 files and the various versions of them. In other words, I have different sorts of automated ‘protection’ for different eventualities, and this particular one is only really going to apply a few weeks or months after the capture (as, indeed, it did).
And I totally agree on not wasting time on ‘also rans’! My issue is that I like (at the moment) to post-process very soon after capture, so I’m insufficiently objective sometimes, as I was with this one. And I went back to it since I’d spent a lot of time and was convinced that I should have caught something worthwhile – I really just wanted to have another look at the files and was pleasantly surprised.
I go through the same purging routine as you do. I need to do this because I can’t work with a cluttered Lightroom library regardless of how clever you can be with collections and filters. Before deleting an image from Lightroom (and the hard drive), I export a copy to a folder called “deleted images” which I only purge on a rolling basis every now and then. I think Julian’s comments are quite valid, but there have been times when I’ve applied the “it’s not a good enough image” rule, only to find myself wishing I’d not deleted the image sometime in the future, even if only for reasons of keeping it for personal and emotional reasons.
Your approach sounds very similar to mine, Colin. Same motivation too: I really don’t like clutter ;-) I’m perfectly capable of being clever with filters and things, but I like to know that things I’m not going to work on are physically gone!
You’re also quite right about the personal and sentimental reasons aspect. Come to think of it, my motivation for having another look at the image featured here was more to do with wanting to keep it as a snapshot / memento than to make it into a ‘proper’ image. That’s a very valid point indeed I think!