My rating habit has changed over the time I have had Aperture: I now only rate images one, two, or three stars. No fours or fives. Why is that? It's down to lack of time, lack of purpose, and simply that three stars are enough. Besides, who decided that five was the right number?

Star Rating Takes Time
Fewer stars means that I spend less time going through images changing ratings. Once I get to three, then I am done. Rating is easier because there are only three buckets the images can fall into, so I think much less when doing it and can do it in fewer passes. With only three stars I am so limited that I can't dream up any clever systems that I have to remember to follow.

Rate For A Reason
I don't have a specific need for five rating levels, and I am not sure that most other people do either. There is nothing I can point to that all four-star rated images achieve that three-star images do not. Why five levels? Who decided that five was the right number, and why? It is kind of traditional I think, the way that it has often been done. I see rating as a much more useful system for excluding images than for including them and adding more rating buckets does not help that.

Three Stars Is Enough
Here are the rating buckets that I now use:

Reject	Junk	Throw it in the trash
One star	Keep	Too good to trash, but not good enough to show anyone
Two stars	Show	Good enough to show people
Three stars	Brag	Good enough to say Look At That, I Did That

That's all there is to it. I can rate in two passes with this system, instead of four or more with five stars. The first pass handles all the rejects and applies one star to anything I would ever want to show anyone. It's a very simple and fast decision to make. Then I select everything that has not been rejected and add one star, so I am left with one-star images marked as Keep, and all the Show and Brag images with two stars. On the second pass (filtered down to the two star or better images, now about 20% to 25% of the original quantity) I add one star to the images that I think are good enough to brag about -- a few percent.

I caption all the Keep and above images (but very broadly so I can do 50 at a time), keyword the Show and above images, and adjust the Brags.

Why Only One Rating Per Image?
Image ratings have a serious problem that is made worse by greater granularity: the current star ratings are only one dimensional. They allow me to rate either one aspect of the image but not indicate what that aspect is, or rate all aspects, but not give them weights. When I look at an image to rate it I have to compromise between all the qualities and come up with a single rating that takes everything into consideration. More stars mean that that is harder to achieve (particularly consistently over the long term) and takes longer to do.

Let me illustrate. I have an image that has excellent composition and color, but is out of focus. As a large image it would get a three, but as a small image, a five. Or the opposite, a technically perfect image with a piece of the subject out of the frame. If I had a specific use for a part of the image it would get a five, but as it is, only a two. How do I rate these?

I would like to see Aperture implement a much more flexible rating system. Instead of a single five-star system I would like one that gives me as many three-star (or even two star!) ratings as I can think up. This would not only be easy to use and very flexible, but would let each person personalize their ratings so that the ratings worked for them, instead of them working for the ratings as is now the case. Besides, this would be a wonderful tool for vendor lock-in: how would you ever move your metadata to another system if that system could not support your ratings?

How would I use a system like that? I would rate by subjects such as Focus, Composition, Color, Uniqueness, Noise, Happiness, Movement, Absurdity, Isolation, Roundness, Humor, Geometry -- good grief! There are so many things that I could do with this.

And don't forget that Unrated is valid for all of the rating subjects: I don't have to rate by Color if I don't want to or need to. I can record no opinion if I either have not had one or did not want to make one. So this system actually does not multiply the work because not thinking is a perfectly useful thing to do. If the color does not grab me as I look at the image, I don't rate it. I don't have to make a judgement about whether the lack of color appeal justifies moving the rating down from a five to a four, as in the current system.