Pages: « 1 ... 3 4 [5]   Bottom of Page
Author Topic: PK Sharpener Question  (Read 13875 times)
Jack Hogan
Sr. Member
Offline Offline

Posts: 300

« Reply #80 on: June 28, 2013, 08:24:45 AM »

When you re-read the great article by Charles Cramer

A good article indeed.  It also shows how far ACR/LR's terminology has drifted from both a technical and practical frame of reference (e.g. 'Exposure' appears to control (mainly) midtones, non-linearly).

« Last Edit: June 28, 2013, 10:43:10 AM by Jack Hogan » Logged
Full Member
Offline Offline

Posts: 247

« Reply #81 on: June 30, 2013, 11:21:12 PM »

You can construct a 32 bit per channel step wedge in PS, over an arbitrary range of exposures, save it as a floating point TIFF and open it directly in ACR to see what kind of range-adaptive and automagic adjustments are going on in PV2012 compared to PV2010.  Because your step wedge will be rendered in the ACR histogram as very thin spikes, add a little bit (Amount 5, Size 10) of film grain to soften the spikes into a more Gaussian-looking profile at each step.  I also make a 50% gray (i.e., brightness = 50%, 0.00 exposure in the 32bit color picker) bar across the entire image - this will give the 0 exposure a larger "hump" in the ACR histogram, so you have an idea of where your middle value is.  Try making a -4 to +4 step wedge in 0.5 exposure increments for starters.  Then open the resulting TIFF in ACR and try making adjustments to the tone sliders in PV2010 and PV2012 with the clipping indicators on to see how the histogram changes.  You can also see the way Recovery / Highlight, as well as Clarity, in PV2010 and 2012 changes across the histogram.  You can also change the output color space to see how that affects the histogram.

While PV2012 may give more pleasing results, the controls are not intuitive in terms of how tonal range is manipulated with each slider adjustment - for example, how deeply one slider's effect reaches into the adjacent tonal range values (it appears that the effects of each slider are transitioned into adjacent tonal ranges to a degree which will prevent artifact, but how far into the range is hard to appreciate).  Instead of adding a "raw" histogram to ACR (we can always see our raw data in Raw Digger) why not add a graphical representation of the tone(mapping) curve that is being applied in the auto-adaptive PV2012?  If there are auto-shoulder or auto-roll-off operations that are being applied to the raw data, then why not display that curve behind (superimposed on?) the histogram?  At least that kind of indicatrix, sort of like a film curve, will give the user more feedback on which area of the full tonal range the particular wang bar is operating and how that slider is mapping that tonal range.

The step wedge obviously has very well-defined, discontinuous areas of tone, so the effects of the sliders in each PV may not necessarily be the best representation of the perceived effect in a photograph - however, it is easier (for me) to get a sense of how tones are being [re]mapped with a contrived image like the step wedge.  I imagine that if one varies the range and extent of the various step wedges, one may also get a sense of the range-adaptive effects inherent in PV2012.  I have not bothered to go this far at this point.

« Last Edit: June 30, 2013, 11:42:16 PM by kirkt » Logged
Pages: « 1 ... 3 4 [5]   Top of Page
Jump to: