Hi! The reason it works that way is that there is a big difference in the amount of red/blue captured by the sensor under different lighting conditions.
For example, when you capture images in a low-kelvin environment (like a lightbulb), there is plenty of "red", which your sensor picks up. On the other hand, there is very little blue. When you do the whitebalance adjustment, you amplify the "weak" channel - here the blue. So any inprecisions are amplified much more. Glad to help! Regards, Klaus Post http://www.klauspost.com On Wed, Apr 13, 2011 at 23:27, David Vincent-Jones <[email protected]>wrote: > Very interesting Klaus; > > Did some tests on contrasty high altitude mountain snow scenes and by > setting the red channel to zero the 'grain' all but disappeared, further > I was able to use greater sharpening and the denoise function worked > exceptionally well. Very big difference! > > On the same imagery the blue at zero was not so effective but I will now > make some tests with other imagery and try to standardize some settings. > It would be interesting if, with Saturation set to zero, the Colour > Temperature somehow directly related to thy Channel Mixer settings > (probably just too complicated). > > More work .. I probably now need to rework a whole bunch of data ... but > seriously thanks again. > > David > > On Wed, 2011-04-13 at 22:14 +0200, Klaus Post wrote: > > Hi David! > > > > On Wed, Apr 13, 2011 at 17:47, David Vincent-Jones > > <[email protected]> wrote: > > I am finding that sharpening of my monochrome images appears > > to create > > fairly heavy noise, even when done in moderation. I have tried > > the > > 'denoise' feature and find it makes little or no change. > > > > The "denoise" also functions as a noise floor for the sharpening > > algorithm, so in general you should try to > > > > Secondly, when doing BW, you can sometimes reduce noise by eliminating > > a channel that is "stretched" due to white balance, so if you have low > > WB temperature, eliminate the blue and if you have a medium to high, > > eliminate the red. > > > > I just did a few tests on high contrast, high ISO BW images, and I can > > see your issue, since color noise is converted to luma noise. I will > > do a few tests, but it seems like it could be a solution to heighten > > the maximum value for denoise, to 150 for instance. I don't want to > > set it too high, since we begin to get artifacts in some cases. > > > > In part this may occur since I tend to manipulate the tone > > curve fairly > > extensively with monochrome. > > > > As I understand it sharpening needs to be applied subsequent > > to all > > other functions ... does the program know that? > > > > Yes - sharpen/denoise is the last function to be applied in the chain. > > > > > > > > Is anyone out there also finding this same situation and is > > there a > > solution? > > > > David > > > > Best regards, Klaus > > _______________________________________________ > > Rawstudio-users mailing list > > [email protected] > > http://rawstudio.org/cgi-bin/mailman/listinfo/rawstudio-users > > > > _______________________________________________ > Rawstudio-users mailing list > [email protected] > http://rawstudio.org/cgi-bin/mailman/listinfo/rawstudio-users >
_______________________________________________ Rawstudio-users mailing list [email protected] http://rawstudio.org/cgi-bin/mailman/listinfo/rawstudio-users
