Noise Comparison

Leading up to my decision to put up the $$$, and carry the extra weight and bulk, I spent a little time reading through many camera reviews. The feature that interested me most was noise and dynamic range. I think I found an interesting way of summarizing my findings. It's too much stuff to enter here, so I made a page

http://uwdigital.home.comcast.net/camera_noise/

for it. Please have a look. I welcome your comments.

Interesting comparison, Herb. Thanks for posting that.

As you know I also have the Oly5050 for underwater use and I don't like the look of the chart for that camera!!! But then I already knew that it produces a very noisy image compared to some, and I almost always shoot at ISO 64 because of that.

For exactly the same reasons as you mentioned (noise performance and dynamic range) I have also dipped my toe into DSLR waters for the first time - I just bought a Canon EOS300D, and currently am having fun with my new toy. I also like the idea of getting past F10 for some more depth of field sometimes.

I'm not planning to use the DSLR underwater for the time being. One issue is the size & weight (and cost!!) of housings and ports. Another is that underwater I really like the fact the I can hold the Oly away from my face, and I can still see what's happening around me while I'm setting up for a shot. Also, some angles are much easier to get using the LCD display rather than a viewfinder. I recognise that there are some significant advantages of the DSLRs but it seems to me that there are comparatively more disadvantages underwater than on land.

(OK, I think I drifted far enough off topic now!)

-David

Herb,

Thank you, this is fantastic information!

Some things I noticed:

The Fuji has higher noise in the darker swatches and the Canons have higher noise in the lighter swatches. Interesting!

From the Sigma camera on down, the highlights were completely clipped. Does this mean they have 2 to 3 stops less dynamic range?

Cheers

James

A number of questions. First, where did you get the test image? A reference to that would be helpful so that others could get the same one. Second, how about a description of what an ideal result would be? A reference scan on a high quality scanner could be useful, too. Third, why did you apply levels (you said you did)? I would like to see each sample at the best exposure possible but without levels. I would use TIFF if possible, too.

I'd like to contribute the 14n results (and fixed D100) to your data but really should have the precise test image.

From the Sigma camera on down, the highlights were completely clipped. Does this mean they have 2 to 3 stops less dynamic range?

I believe this means that these cameras were simply underexposed relative to the others.

Herb, if you're able to add more..........I'd be curious about the cp5000 also

A number of questions.  First, where did you get the test image?  A reference to that would be helpful so that others could get the same one.  Second,  how about a description of what an ideal result would be?  A reference scan on a high quality scanner could be useful, too.  Third, why did you apply levels (you said you did)?  I would like to see each sample at the best exposure possible but without levels.  I would use TIFF if possible, too.

 

I'd like to contribute the 14n results (and fixed D100) to your data but really should have the precise test image.

I tried to put all that in the write up but probably did not state it clearly. Let me try this way.

1) All the shots used in the comparison came from http://imaging-resource.com . It is one of the standard test shot they call the "dave_box" which they take for all cameras they review.

2) To reach the image I used for this comparison: find the review page for the camera in question, select the "sample image" page on the scroll menu, move down the page and click on the "Davebox Test target" link, and click the targe thumb nail.

3) They take a number of shot of the "dave_box" under varied conditions. I used the one featured on the "Davebox Test Target" page. The WB on these shots varies from camera to camera, but it's aways the lowest ISO, and high quality JPEG, and fast shutter. The discription of the shots can be found on the site.

4) I used the "levels" tool in PS for display only. No adjustiments. I could have used the histogram tool for this.

A reference would be useful, but I don't have the chart. The lighting of the chart is also a factor.

If I had the equiptment to do this, I'd shot it in raw and convert it to 16bit TIFF with all the settings at minimum.

That's what I thought you meant on the levels thing.

I have a couple of similar charts if I could only find them!

From the Sigma camera on down, the highlights were completely clipped.  Does this mean they have 2 to 3 stops less dynamic range?

I believe this means that these cameras were simply underexposed relative to the others.

They may be underexposed. All the shots were taken with the using the camera's auto exposure setting, and there are other things in the shot besides the crop that I showed, including a metalic surface.