Where is the error in this explanation of light collection/etendue?

Started 1 month ago | Discussions thread
Joofa Senior Member • Posts: 2,564
Re: Mosaics and SNR

Jack Hogan wrote:

Joofa wrote:

Jack Hogan wrote:

Joofa wrote:

tony field wrote:

Jack Hogan wrote:

tony field wrote:

Jack Hogan wrote:

tony field wrote:

Jack Hogan wrote:

tony field wrote: ... There are gains from the larger image. Certainly resolution will be one benefit.

Seems like a reasonable assumption for lenses of similar costs.

Possibly there will be a reduction in perceived noise.

Why would that be if the playing field were level, with the same number of captures, f-number and total integration time for both the mosaic and the stack? Ignoring non idealities in the lenses, the number of photons captured, hence SNR, should be about the same.

Apparent noise of the Mosaic image should be lower when image sizes are equalized

If you are referring to non-idealities like vignetting then I can see where you are coming from. If not, then SNR should be the same, because in such an ideal case the number of captured photons would be the same aotbe.

I think the actual signal-to-noise ratio Remains the Same with image size changes

Once the digital images captured, the noise components become part of the image. The interpretation of the image allows us to state that certain bits constitute noise and other bits are really the image component. Very often if we reduce reduce the size of the image, the size of the apparent noise components is reduced and we think there is lower noise.

When one resizes an image (say by binning four pixels into one) the signal to noise ratio improves and the image will look less noisy. So if you take two images of the same subject (FOV), one with 9 times the number of pixels than the other, when you downsize the larger one 9:1 to make both have the same number of pixels, SNR will go up.

My point is that in the case under discussion the 35mm/2 image has 9 times the number of pixels as the 12mm/2 image. But if we are keeping the playing field level the 12mm image will be a stack of 9 images taken in sequence with the same Exposure Time, so 9 pixels are also being binned into 1 and SNR will go up by the same amount.

Certainly your point is totally appropriate. Trying to work with the same number of protons will certainly level the playing field in both situations and should definitely result in "identical" images.

No, not really. It appears that Jack H. seems to be hooked on using binning in both cases. However, in the larger, mosaiced image, binning is not the only option for size reduction, and also not really the best. Several, linear and even non-linear methods of both size reduction and noise reduction are available to help achieve better SNR than binning to a lower resolution.

Both, SNR and Sharpness should increase, in general, while size reduction using oversampled images using a good technique that is different than binning.

Question: How many people use binning for size reduction in Photoshop? Does Photoshop even offer the option of binning?

Give it up Joofa, you are OT. Start a thread on downsizing algorithms and I'll chime in.

I don't see a reason for your bitterness except perhaps you are out of your arguments here.

My point does not rest on binning, it's just you trying to take the discussion OT. No hard feelings, just trying not to waste the last few posts here. Bye.

LOL. Here you go:

Jack Hogan wrote:

".... so 9 pixels are also being binned into 1 and SNR will go up by the same amount"

(From: https://www.dpreview.com/forums/post/61154730)

-- hide signature --
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow