Re: How to translate log density readings to percent?

From: Loris Medici ^lt;>
Date: 09/07/04-02:52:54 PM Z
Message-id: <000e01c4951c$a644dbd0$bd02500a@Loris>

Thanks for the second answer.

Actually my primary intention is not to measure negative densities, but print densities (the 1.33 I was mentioning was the print's dmax, not negative's - for instance I remember S. Wang saying he gets around log 1.25 dmax with his double coated classic cyanotype prints, emulsion: 2A+1B, probably even gold toned vandykes or platinotypes don't go beyond 1.5 - 1.6 because of matte paper). I will use imagesetter negatives... I was planning to print a non adjusted 101 step tablet using a standard exposure time (which is the minimum exposure time to achive dmax of the emulsion/paper combination - determined by braketing exposure under clear transparency material - masking with fully exposed lith film; with a dmax figure not less than log 4 - probably around 5) and then reading densities of every patch so that I can "map screen/file densities to print densities". I thought this will give me the necessary values to design a "rough" curve (or "transfer function" if you like) and then I can repeat the cycle few more times (I guess this is the iteration you are mentioning) - from now on, applying the curve - for fine tuning (especially highlights). I think a densitometer will make all this struggle much more easier and precise (probably it will also serve to establish an exposure index/development time per process for large format in camera pinhole negatives - which is another story).

I know people are calibrating their inkjet printing systems (such as Piezography) using this method (or a very close one). Don't know, maybe I'm establishing a strategy which is completely wrong (because I misunderstood conversations between inkjet printer users @ DigitalB&WThePrint Yahoo group!?) or unappropiate for alt. process printing?


  ----- Original Message -----
  From: Dave Soemarko
  Sent: Tuesday, September 07, 2004 10:44 PM
  Subject: Re: How to translate log density readings to percent?


  We are getting into some details. For using densitometry with digital negative, the basic assumption is that the transmission density of Dmax is at least 2.0. If the Dmax is less than that, it is hard to relate density to % dots because some light is transmitted through the clear area but some light are also transmitted through the "opague" area because the Dmax is not that high. It is calculatable, but it is not very practical because when you expose the negative, the same transmission of both areas will occur, so the percent dots alone doesn't give you all the needed information.

  If your Dmax is 1.33 only, you want to use a short-scale process, or mix your chemical in such a way so that it is contrasty. For curve adjustment, you can use a different approach, print a negative with 5%, 10 - 90%, 95%, for example, and adjust your process so that when you look at your print, at 5% you see a very slight but visible difference from Dmax and a 95% you see a very slight but visible difference from paper white. Then you can adjust 50%. Say if at 50% the print is too light. How much do you want it to go darker? You can look at the patches and say that the right value is somewhere between 20% to 30%. You adjust the curve to achieve that (you can try 25%). In most digital printing, one would correct for 25%, 50%, 75% tone instead of just 50%. With a few iteration, you can get the curve you want.

  Dave S
Received on Tue Sep 7 14:53:43 2004

This archive was generated by hypermail 2.1.8 : 10/01/04-09:17:55 AM Z CST