How I scan
- Thread starter zk-diq
- Start date
zk-diq
Senior Member
Control unit in the conventional scanner has a few part,
A: colour computer: Selective colour correction for various of printing ink/display unit requirement. Tone/luminosity compress/control etc
B: Scale computer: setting of final scan size
C: USM Computer: a separate computer to generate USM signal
among all, Colour computer is the most complicated unit
A: colour computer: Selective colour correction for various of printing ink/display unit requirement. Tone/luminosity compress/control etc
B: Scale computer: setting of final scan size
C: USM Computer: a separate computer to generate USM signal
among all, Colour computer is the most complicated unit
Cheers!
zk-diq
Senior Member
I have been ask a few time why do we need such complex colour computer unit? Because we need to compress and correct our tone/colour to fit into the publishing/printing limited colour / gradation world, even up to today. In fine art high quality reproduction work, Print density is approx from D1.8 - D2.4. Where colour trans usually is around D3.2 - D3.9
You lost D1.2 - D1.8 dynamic range.
And you only retain not 50% of the information, but not more then 20% of the original info.
Just like enlarging your negative, in B/W we have different grade of paper to compensate contrast/dynamic range, for colour film, we only have one grade of paper, negative colour film has a built in orange mask for basic colour/contrast correction. A correct exposure of neg film has a dynamic range of 1.1 to 1.4, and it fits into normal photographic paper print range. You may easily make a good print without dodging or burning.
Unlike slide/chrome reversal film, a correct exposure slide film hold a dynamic range of 3.2 - 3.6 where kodachrome goes even to 4.2. To make a good print from slide is nightmare to me. R14 process from kodak or cibachrome, I can only have good reproduction from highlight to mid tone with poor result from mid tone to shadow, or the other way round. Eventually I tried Kodak Dye Transfer Process. It take a week to make a print and cost a bomb. With all the highlight mask/contrast reduction mask at least I make a print that can visual from highlight to shadow.
You lost D1.2 - D1.8 dynamic range.
And you only retain not 50% of the information, but not more then 20% of the original info.
Just like enlarging your negative, in B/W we have different grade of paper to compensate contrast/dynamic range, for colour film, we only have one grade of paper, negative colour film has a built in orange mask for basic colour/contrast correction. A correct exposure of neg film has a dynamic range of 1.1 to 1.4, and it fits into normal photographic paper print range. You may easily make a good print without dodging or burning.
Unlike slide/chrome reversal film, a correct exposure slide film hold a dynamic range of 3.2 - 3.6 where kodachrome goes even to 4.2. To make a good print from slide is nightmare to me. R14 process from kodak or cibachrome, I can only have good reproduction from highlight to mid tone with poor result from mid tone to shadow, or the other way round. Eventually I tried Kodak Dye Transfer Process. It take a week to make a print and cost a bomb. With all the highlight mask/contrast reduction mask at least I make a print that can visual from highlight to shadow.
Last edited:
zk-diq
Senior Member
Measure Negative density at 0.2 shadow and 1.7 highlight, scan according to film density range
scan with lower density setting in shadow to open shadow area, increase highlight setting to capture more highlight tone.
increase more mid tone setting, some dodging was carry out with photoshop over shadow area
some detail for visual. some fine detail loss due to jpg.
zk-diq
Senior Member
Resolution is inherently lower and deteriorates due to compression of density range in offset/litho/dye transfer printing. A certain degree of unsharpness in the mask help reduce register problem and increase sharpness of detail in reproduction. USM mask is just optical illusions. When higher the contrast in a light and dark border, the sharper it appear. Technical term is Mach Band Effect
zk-diq
Senior Member
Between Photomultiplier and CCD sensor, I prefer PMT. CCD has improve drastically since the very first day, but still has limitation.
1) Density range, PMT can see more then 4.0 where most CCD can't. Consumer product usually can only read till D2.4. (commercial grade go up to D3.6 to D4.0 as claim by manufacturer)
2) Controlling highlight and shadow point, PMT like rubber band, you can stretched 2 point within very narrow and wide density. CCD work like rubber ruler, it read your original density and compress according to an algorithm or lookup table.
3) Excellent detail, PMT read by point to point and line to line depend on aperture, CCD read in area and flare is common if you exam it carefully.
If you have Minolta mirror/ Nikon annual / hasselblad forum / Linhof technic / Pentax family or any high quality publication, you see the difference compare to what we have today.
Basically my experience is I have less tone separation detail in shadow area and highlight point. (those days, I did not try on today new product)
1) Density range, PMT can see more then 4.0 where most CCD can't. Consumer product usually can only read till D2.4. (commercial grade go up to D3.6 to D4.0 as claim by manufacturer)
2) Controlling highlight and shadow point, PMT like rubber band, you can stretched 2 point within very narrow and wide density. CCD work like rubber ruler, it read your original density and compress according to an algorithm or lookup table.
3) Excellent detail, PMT read by point to point and line to line depend on aperture, CCD read in area and flare is common if you exam it carefully.
If you have Minolta mirror/ Nikon annual / hasselblad forum / Linhof technic / Pentax family or any high quality publication, you see the difference compare to what we have today.
Basically my experience is I have less tone separation detail in shadow area and highlight point. (those days, I did not try on today new product)
Last edited:
Similar threads
