M94 and galaxy cores. Balancing sharpness versus quality
Posted: Wed Apr 20, 2022 4:43 pm
Here - similar to the previous two posts - a ~ 3h OSC camera image of M94 from stacking 40s (F 4.0) exposures was sharpened by a process of a) making a 40 min total exposure Mono camera image of stacked selected 3s mono camera frames and b) extracting the luminance channel from the OSC image c) sharpening both the 3s image and 40s luminance by deconvolution d) combining these two into a single high dynamic range luminance image and finally e) transferring this HDR luminance back into the OSC image.
Equipment and Bortle 6 - 300mm F 4.0 VX12 and ZWO mono and OSC AS1294 cameras as decribed before and also detailed here (plus full picture) https://www.astrobin.com/in9vzz/C/
The overall impression that I get having now attempted to apply this process (or part of it) to 4 galaxy images, M94, M51, M106 and M63 is that for M94 and M51 it works pretty well -- but for M106 and M63 --- it didn't really add a lot.
Both the addition of 3s luminance and deconvolution can add sharpness to images and uncover otherwise invisible detail -- and used together even more so. BUT there are caveats with regard to wider applicability and image degradation.
Both tools are only really effective on bright areas with high SNR and where there is detail to uncover.
There is a danger of oversharpening - dotty appearance to the point of generating artifacts - and loss of depth of coloration relative to luminance. Also of just adding noise to lower SNR areas (the masking needs to stop the 3s luminance from encroaching on these areas)
Targets like M94 --(and probably some planetary nebulae) with compact bright centres can really benefit from this approach. M51 also benefited -- but in that case I overdid it and probably need to row back to regain some depth of coloration in exchange for some blurring.
Anyway here is the core of M94 --pre and post sharpening - a target the process did work well on. I like the way that it has uncovered some subtle details that were otherwise invisible via normal post processing.
Tim
Equipment and Bortle 6 - 300mm F 4.0 VX12 and ZWO mono and OSC AS1294 cameras as decribed before and also detailed here (plus full picture) https://www.astrobin.com/in9vzz/C/
The overall impression that I get having now attempted to apply this process (or part of it) to 4 galaxy images, M94, M51, M106 and M63 is that for M94 and M51 it works pretty well -- but for M106 and M63 --- it didn't really add a lot.
Both the addition of 3s luminance and deconvolution can add sharpness to images and uncover otherwise invisible detail -- and used together even more so. BUT there are caveats with regard to wider applicability and image degradation.
Both tools are only really effective on bright areas with high SNR and where there is detail to uncover.
There is a danger of oversharpening - dotty appearance to the point of generating artifacts - and loss of depth of coloration relative to luminance. Also of just adding noise to lower SNR areas (the masking needs to stop the 3s luminance from encroaching on these areas)
Targets like M94 --(and probably some planetary nebulae) with compact bright centres can really benefit from this approach. M51 also benefited -- but in that case I overdid it and probably need to row back to regain some depth of coloration in exchange for some blurring.
Anyway here is the core of M94 --pre and post sharpening - a target the process did work well on. I like the way that it has uncovered some subtle details that were otherwise invisible via normal post processing.
Tim