Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Post-processing & Printing | Join Upload & Sell

  

Archive 2012 · Managing near-duplicates (with Lr)
  
 
Alan321
Offline
• • • • •
Upload & Sell: Off
p.1 #1 · Managing near-duplicates (with Lr)


It's possible that I've asked this before but I'm getting forgetful lately. This is a long post but I think that within the next couple of paragraphs you'll know whether or not it is worth reading the rest.

I want to know how you handle bundles of nearly identical images in general, and more specifically in Lr. I realize that storage is cheap in terms of the hardware but it tends to gobble up time in terms of subsequently reviewing or managing the images.

Imagine you're at an old fashioned slide show and out comes 20 slides of basically the same thing - boring Every one of them might be great images in isolation but there's not enough variety to make them all interesting at the same time. So, what to do ?

An example of what I'm thinking about is this: A whale and her calf swim buy and you take a buffer full of shots from the cliff-top. All are good (if you're lucky and/or skillful). All are very slightly different from the previous one but the last one will be most different from the first one. Some are better than others in some way but worse in other ways, depending on exactly what the two whales were doing in each image. Twenty shots of a whale and calf is probably too many to keep - better to have shot a movie if the motion sequence was important. Two thousand shots taken over several hours is way too many but it was exciting at the time. Some shots are not particularly good but until you get back there to take more photos with more appropriate equipment they're the best you have.

In general the basic options are to keep them all (if they are individually any good) or to keep a representative selection and scrap the rest (in which case you hope you kept the best and that later on you'll still think they are the best). e.g. you might randomly keep every 3rd or 4th image in a long series on the assumption that any subject motions has caused at least some change throughout the series. I tend to go for the former but it's getting messy as my collection grows. A variation on the culling approach is to apply it to all photos rather than just the groups of near-duplicates. i.e. Only keep the very best. That is rather savage and you could discard many ok but imperfect images that become irreplaceable before you get a chance to shoot them again.

I can imagine that what you might do as an event or sports photographer would be quite different from what you might do as a nature photographer, largely because most most events are old-news and become stale relatively quickly unless and until something happens to make them significant again - or an old client wants another print.


So now to using Lightroom as the management tool:

One option is to keep all of the near-duplicate images - especially those that are individually up to standard - and then group them into stacks to minimize clutter. That works but if you apply ratings and labels to them do they all deserve to get the same rating because they are all much the same, or do you assign a higher rating to some and a lower rating to others just to separate them ? Do you use colour labels to identify some as surplus to requirements ? It seems to me that something needs to be done to prevent a simple search bring up heaps of much the same same thing, but I don't know what that something is. I'm inclined to think that ratings and labels should be applied on a per-image basis for consistency.

Another option might be to have a second catalog for holding the near-duplicates. Workable, I suppose, but not easy to manage.

A variation on that option is to have a folder for holding the near-duplicates but that folder will be subject to any searches and filters and those images will always be popping up when you don't really want them.

Another option might be to rate and label the best images in each stack and leave the rest unlabeled but that invites a lot of subsequent unwanted reviews whenever you do a search for unrated images to bring your library up to date.

There's a Survey feature in Lr but that is designed to help us decide which of a group of images is best and reject the rest. Not good if you don't want to actually reject the rest, and especially not good if "reject" is normally a precursor to deletion; These rejects might get confused with other rejects that really were meant for deletion.

Another option is to approach things differently - print your best images and don't keep reviewing the rest in Lr. This will not work for me because I do little printing and if I did more then I'd have trouble managing the prints.

I'm guessing that using collections might help (with enough time spent on building them) but collections are hard to manage when you have lots of them. The ability of Lr 4 to nest collections might make this far more useful than in previous versions of Lr. Ideally we would be able to keyword and rate the collections themselves independently of the images within the collections.


So what do you do ? And what would you want to do differently if you had the time ?

- Alan



Feb 02, 2012 at 09:00 AM
sbeme
Offline
• • • • • •
Upload & Sell: On
p.1 #2 · Managing near-duplicates (with Lr)


Personally, I keep it simple.
With a series or slight variations on an image, I initially use survey mode to cull out those with unfixable technical flaws and reject and delete. If there is a winner, I rate it high. With the others, I try to rate them, but I still am often left with a number of images I am not sure what to do with.
I leave them be.
Then I return, weeks, months later. And cull them again. Usually I am now less attached to the images and can reject and delete more and promote a small number.
Generally I move to saving less images over time. After all, if its merely a variation on other images and I still see nothing to make it stand out, why keep it? I do have a number of single images or very small sets I hold onto, sometimes reprocessing them years later with more than positive outcomes. But if its 20 shots of the same thing and none stand out, why keep them. And, like most of us, I suspect, we hate letting things go.
So, initial survey, rate/reject. Come back later, less attached, survey, rate, reject. Maybe reprocess a few. One more round, rate/reject. I might now have gone from 20 images to three or four. I am OK with that.

Scott



Feb 02, 2012 at 10:38 AM
Alan321
Offline
• • • • •
Upload & Sell: Off
p.1 #3 · Managing near-duplicates (with Lr)


Thanks Scott. That was a good point about leaving the images alone for long enough to reduce the attachment. Although the ability of photos to evoke an emotion is often a very good sign, here we are talking about the emotional attachment associated with the taking of the photo rather than what the photo evokes.

Perhaps I've been coming back to them too early and too often.

- Alan



Feb 02, 2012 at 01:51 PM
 

Search in Used Dept. 



mhayes5254
Offline
• • •
Upload & Sell: On
p.1 #4 · Managing near-duplicates (with Lr)


I do something like Scott describes but I try to be very aggressive with deletions on the first pass. I know that it usually makes little sense to keep slight variants but it is often hard to choose.

As for the ones that are kept, I like the option of keeping them stacked. I have done that but am not very consistent about it (or keywording ). That is much easier to manage than the other options you described.

Edited on Feb 07, 2012 at 11:59 AM · View previous versions



Feb 04, 2012 at 10:48 PM
sbay
Offline

Upload & Sell: Off
p.1 #5 · Managing near-duplicates (with Lr)


I delete anything that is technically not up to par and if I have two or more images that are almost exactly identical I keep the sharpest one and delete the other. However, this can still leave a lot of similars, in which case I pick out the best and mark them with stars / colors.


Feb 04, 2012 at 11:26 PM
Alan321
Offline
• • • • •
Upload & Sell: Off
p.1 #6 · Managing near-duplicates (with Lr)


Thanks Mike and Stephen. I missed your replies until now.

It looks like keeping just those that are technically ok is not enough to solve the problem and that more serious culling is probably required. I think I have to learn that what is left after culling is still far better than having nothing and that even though I can afford the storage space if I do not cull, the reduced clutter afforded by culling will make subsequent viewing and reviewing a much easier and nicer experience.

I do keep some of the duds too so that I can build a picture of what it is that produces the duds - equipment defects, equipment limitations, bad luck or user errors. I then try to learn from that. Many people won't have the time for this sort of thing and many either don't care or perhaps don't have too many duds, but I have a history of confirmed gear defects and increasingly of user error (mainly on days when my concentration wanes a bit or I've had too much or too little medication or caffeine).

- Alan



Feb 07, 2012 at 11:52 AM





FM Forums | Post-processing & Printing | Join Upload & Sell

    
 

You are not logged in. Login or Register

Username   Password    Retrive password