[straton]

While I appreciate the finesse Pentax Digital Camera Utility (DCU) offers for tricky images, as a very new digital user, I am wondering if there is any advantage to using it instead of Photoshop levels for expanding the range of tones in a file neither of whose highlights nor shadows are blocked nor is otherwise problematic.

Suppose, for instance, I have a file whose tones run from 15 to 142 out of the possible 0-255. If I set the level correction in DCU (I'm using version 5) to 15 for the shadows and 245 for the highlights, I suspect I am essentially doing three things: (however DCU actually does it, I can reproduce the result for this particular choice as) (1) doubling the tone number of each intensity point on the histogram (2) subtracting 30 tones, and (3) making the next higher tone number on the histogram equal to the one just calculated. For instance, the current shadow point in tone 15 becomes 2*15-30 = (tone) 0 and we also set tone 1 to that same shadow intensity (call it intensity 1). Likewise, the current highlight point in tone 142 becomes 2*142-30 = (tone) 254 and we also set tone 255 to that same highlight intensity (call it intensity 1). If the top of the histogram curve is in tone 62 it becomes 2*62-30 = (tone) 94 and we also set tone 95 to that same peak intensity (call it intensity 7).
Done graphically and roughly, a current histogram of shape
_____________7
____________6_6
___________5___5
__________4_____4
_________3_______3
________2_________2
_______1___________1
0000000_____________000000000000000000000

becomes
____________________77
_________________66___66
______________55_________55
___________44_______________44
________33_____________________33
_____22___________________________22
__11_________________________________11
0______________________________________0
.

But one could just as well output the original narrow-histogram-image as a tiff file unchanged in DCU and do this process of expanding the range of tones in Photoshop levels by setting the shadow slider to 15 and the highlight slider to 245 and ignoring the grey slider. Is there anything about DCU that produces a better result than Photoshop in such a nonproblematic case, such as interpolating the intensity of each "next higher" bin number step size, in the above histogram as 1, 1.5, 2, 2.5 … instead of 1, 1, 2, 2, . . .? (On my drum scanner, by contrast, setting the highlight and shadow points before scanning the film does cause the analog photomultiplier tube (it is not a CCD) to read intermediate light values that the analog to digital converter faithfully reproduces.)

Since Photoshop is so much less clunky than DCU, is there any reason I should not save myself some aggravation and do so for the easy images?

Thanks, Jack

P.S. DCU5 has a Highlight Correction slider in addition to the DR Expansion slider. Can anyone explain how to use these in tandem (or should I just use one of these)? By trial and error I have found that a DRE of 1.0 shifts the highlights leftward slightly and a HC value of 9 often but not always reduces the peak height of that spike a bit. I would love to know how these really couple.

[sbh]

Not sure wheter I understood the middle part, but I get your point. Imho its less about the app rather than the source data raw vs tif. Doing the adjustments from a raw file gives the app much more data to re-calculate the values than an already „developed“ file.

[stevebrot]

Each tool works differently in regards to mapping sensor values to pixels (RAW processing) or managing existing pixels (TIFF or JPEG image editing). Creation of tonal ranges from expansion of discrete data requires a bit of dark art and the degree of success and utility varies from tool to tool. I tend to rank PDCU fairly low in regards to utility and usability, though it carries the advantage of being able to reproduce the full range of in-camera custom image settings should that be the user's desire. I have also found that its processing is often difficult to reproduce with other tools.


Steve

[BigMackCam]

Each tool works differently in regards to mapping sensor values to pixels (RAW processing) or managing existing pixels (TIFF or JPEG image editing). Creation of tonal ranges from expansion of discrete data requires a bit of dark art and the degree of success and utility varies from tool to tool. I tend to rank PDCU fairly low in regards to utility and usability, though it carries the advantage of being able to reproduce the full range of in-camera custom image settings should that be the user's desire. I have also found that its processing is often difficult to reproduce with other tools.

I agree completely with this appraisal.

If you want post-processing to closely match the camera's own raw-to-JPEG processing, Pentax's DCU is the way to do it. But otherwise, flexibility and range of processing are limited compared to most established third-party offerings. This is especially so in terms of noise reduction for higher ISO images, but also things like shadow and highlight recovery where tools such as Lightroom and Darktable (and possibly others) do considerably better...

[dms]

When I was learning photoshop (photoshop5) I was told by the instructor (International Center of Photography in NYC) that curves should be used, as levels will degrade the image more. Also if you look you will find curves is said to be far more versatile/capable; however, as I don't use levels, I cannot say more about it.