OK, so not quite up to the numerical standards of tonight's Autumnwatch, but whenever I see a fair few Starlings I try to get a count, given that I know indigenous numbers have declined dramatically.
On November 16th, at dusk, I was particularly pleased to see a flock wheeling several times over Thames Road Wetland (of which I am Site Manager), and I was able to get a 'side on' mobile phone camera picture - albeit a rather distant one.
The speed of the birds and size of the flock meant that I'd been unable to get a visual count estimate of the numbers of individuals at the time, so I cropped the photo down to the margins of the flock and printed it out. Whilst the image of each bird was not particularly sharp, I was able to number all the birds and, with reference back to the somewhat sharper on-screen image, resolve how many birds were present in the surprisingly few image 'overlaps'. I came up with a total of ................
231 birds (or possibly 230)
Interestingly, this is considerably more than my 'on site' guesstimate, which I rejected at the time, and also for an estimate based on application to the photographic image of the sort of rapid, rather 'jumpy' and 'finger-pointing' count I've used in the field before, which came up with about 80.
So if you want to get a more accurate count for a flock of birds that you can get within one frame - and still blow up to a decent size and reasonable resolution - you don't need a fancy camera to do it.
Here, as a matter of interest, are the original photo, and the cropped version that was printed out to do the numbered count :-
Flock of Starlings over Thames Road Wetland at dusk, 16th November
Starling flock cropped and enlarged out of the preceding photograph, printed out, then counted by numbering all the individuals in the resulting image. It still doesn't look like 231 to me ....... but cameras don't lie!