Keep or Delete?

2015.05.18

To keep or delete? is one of the two eternal questions in digital photographer. The other, unsurprisingly is RAW vs JPEG? which is wormhole of its own.

Of course, this is a personal question and it varies tremendously. Some situations, like fast-moving actions almost guarantee a higher hit-to-miss ratio, so there is no exact numeric answer to what percent of images to keep vs delete.

Finding Keepers

My motto for this is Delete is my friend :) First, I delete anything that is not technically perfect, making extremely few exceptions (less than 0.01%) for moments which give emotional attachment. Say, the first step of a baby, even blurry, is probably a keeper! Then, it is time to and then delete anything that has no point of interest or is too similar to another shot.

A great exercise is to give yourself a the challenge to simply not shoot the bad ones. It has been working for me, diminishing my deletion ratio is while augmenting the quality of my shots. Now, I am below 80% deletion. From what's left, only 10% get shown, either off or on-line, and about 4% get sold as prints or licensed to publications.

The most common reaction is that storage is cheap and I agree, only the cost of managing storage is not.

Workflow

Here is a workflow that works to efficiently reduce the number to store:

• Delete immediately in-camera missed shots. Things like people entering the shot at the wrong moment, forgot the camera was in MF, etc.
• Delete anything that is not technically perfect: sharp, focused, well exposed, well framed, correct WB, level, etc as a first pass on the computer. PMVIew Pro on Windows and Geeqie on Linux are most efficient for this.
• Delete everything that is too similar, keeping the best of course. This is the second pass on computer, also with the same fast viewers as above.
• Import into Lightroom, apply keywords and rank as the third pass. This is when documentary shots that lack interest get ranked low. Lightroom has a 0 to 5 start ranking system which is reasonable granularity. Most DAM Software offer something similar. My personal ranking is:
  • A) Zero stars for things that are not pictures such as panorama pieces or brackets for Exposure Fusion or HDR.
  • B) Things that are technically reasonable but lack interest get 1 star. Try to crop them and see if they get more interesting, in that case they can be upgraded to 2 stars. If they show a technical error, also rank then with 2 stars.
  • C) A technically perfect and interesting photographs gets 3 stars unless:
  • D) It is also evocative and would make an appealing print in a visible location. Album prints do not count, those are more for souvenir-type images. In this, rank with 4 stars unless:
  • E) There is NO way the picture could have been improved by a change of position, framing or camera settings. In that case, it is ranked a full 5 stars.

Improving

The goal, while repeating this process after every shoot, is to stop shooting the 1-3 stars images. This is what will make the ratio of keepers constantly improving. Always ask yourself why something came out poorly and what were you thinking at the time. Always tag anything that got cropped (or worse ;) ) to known each time when you failed to do things properly.

Closing words from Jay Maisel:

"If you are not your own harshest critic, you are your own worst enemy."

• Digital Photography