Telescope Reviews: New Camera Time

In general, the larger the pixel, the better the per-pixel signal-to-noise ratio is. However, I pulled some data from a post by rnclark on birdphotographers.net, and came to the following comparison of the 30D v 40D v 50D:

A better comparison of these cameras is to look at how they compare when producing the same image; for example, a 1.5 MP (1500x1000 pixel) image. For each of these cameras, we need to bin camera pixels together to produce an image pixel.

If we multiply the full-well capacity of each pixel by the number of pixels and divide this product by 1.5 million, we get the full-well capacity of the binned pixels. Here is what the table in the post by rnclark says

This means that per equivalent pixel, the full-well capacity has gone up.

Noise is usually given as a standard deviation from the mean. The variance is the square of the standard deviation, and when adding random variables, we add their variances. Thus, to compare the read noise in producing a comparable picture, we compare the square of the read noise times the number of pixels, divide by 1.5 million, and then take the square root to put it back into the terms of noise per image pixel.

The per pixel signal-to-noise ratio (as computed from full-well capacity and read noise) for these cameras is fairly close, but adjusting these for producing a 1500x1000 pixel image puts the 40D a bit ahead of the 30D, but puts the 50D clearly ahead of both:

Since the image signal-to-noise ratio is inversely proportional to the square root of the number of image pixels, the relative sizes of the signal-to-noise ratios will stay in the same proportions if we change the size of the comparison image.

So from the information given in that post, it looks as if the 50D is a definite improvement over the previous two xxD releases for producing comparable pictures (but about equivalent for producing full resolution pictures). Of course, the Digic IV chip in the 50D adds some improvements over the Digic III in the 40D and the Digic II in the 30D.

I no longer have any doubts that the 50D is the better camera. For producing the same image, the 50D has 25.0% better signal-to-noise ratio than the 40D and 33.0% better signal-to-noise ratio than the 30D.

I used the modified camera with my William Optics ZenithStar 105 ED, which is a very fine scope, but it is not capable of handling the extended UV and IR of a modified camera. I got lots of red and blue halos around brighter stars, which came from the red and blue not being focused the same as the green. I upgraded to a Takahashi FSQ-106 ED and now I am able to get much nicer pictures with the camera. I entered an image of the Heart Nebula (IC 1805) in the October DSLR Challenge.

Quote:

The biggest issue that gives me pause about the 50D are the numerous Err99 issues being reported online - but it's hard to get a grasp on how widespread the issues actually are, and it seems that Canon is trying hard to get them resolved quickly.

David

I got one Err99 message early on, and had to turn the camera off and on a couple of times. After that, I have not had any problems.

I have been quite happy with my 50D and look forward to being able to do some more astrophotography with it.

--------------------
Rob Johnson

Celestron 14" SCT | 8" SCT | 102mm f/5 | Ultima 8x56
Takahashi FSQ-106EDX f/5
ZenithStar 105mm f/7 | 80mm f/6
Vixen ARK 16x80
CGE | CG-5
Canon 50DH
Starfish autoguider

Post Extras: