Sub denoiser?

[mschuster]

Sometimes I capture one sub to survey an area, or to double check target framing. The result is noisy. So I have been wondering if it would be possible to quickly denoise a single sub. Here is the result of one scheme. Input is one calibrated sub and values for sensor gain and read noise. Output is the denoised sub. No user input required (i.e. no masking or anything else needed) it runs automatically. What do you think? Would anyone use something like this? I am thinking this might be scriptable.

Thanks,
Mike

[Philip de Louraille]

I'd say this is useful.

[georg.viehoever]

Useful. But how is this different from other denoising algorithms? How does the result differ from first stacking images? My impression is that denoising subs is much harder than denoising stacked images because of the lower SNR.

[mschuster]

Georg, sorry for the confusion. I am thinking of the case where I have nothing to stack, only one subframe. And I wonder if I can get a clearer view of the target even so. The price is a loss of resolution, which of course stacking avoids. It seems to me that whenever I denoise I loose resolution. It's like a law of nature or something. On what's different from other techniques, to me it seems easy to use, no masking, no tweaking parameters and also well matched to detector characteristics.

Thanks,
Mike