FotoDave@aol.com
Sun, 23 May 1999 16:45:36 -0400 (EDT)
David,
You mentioned your paper and suggested that I read it. I have read it, not
recently but I have read it about a couple of years ago, I think. I think it
is an excellent paper and have recommended others to read it. My comment to
others is that the paper shows a clear understanding of the process of
calibration for the most part. I said for the most part because there is a
small part that I think is a little not accurate.
That part says something like make sure you argue, debate, fight, or do
anything to convince your service bureau so that the optical density is the
desired value, not just the % dot. To maintain the optical density is of
course needed, but it should have been there if the imagesetter is calibrated
consistently.
This issue is, of course, related to what you just said recently.
It has been a few years already, but I still remember when I read it the
first time, what came to my mind was the explanation must have been given by
some service bureau that does not calibrate the system regularly enough. That
a patch can have the same percent dot but with different optical density is
their business BS (sorry but I run out of vocab to describe it).
<< Next you measure the OPTICAL density of the 1% box and find that the
reading is say, 1.8. O.K., now you crank up the lasers and run another
neg. This time the optical density of the opaque square is 7.0, and the
OPTICAL density of the 1% box is 3.1 EVEN THOUGH THE % DOT STILL MEASURES 1%
>>
You see, optical density measures opacity, which is 1/transmittance.
Transmittance measure how much amount of light that passes through. In fact,
it measure how many percent of light the film passes through!
Let's take an example of 50% dot (I use 50% because it is half the amount of
light, or one stop, so many are familiar with it already). Because the film
passes 50% of light, then density MUST BE 0.30 (+ base + fog of course).
Repeat, 50% dot MUST have optical density of 0.30. If it is not 0.30, it is
NOT 50%.
In your example of cranking up the laser and you get, say, a density of 0.60,
it is not correct to say that the percent is still 50% but the optical
density is 0.60. It simply is not 50% anymore, and it is the job of the
service bureau to maintain the value. It is accurate, however, to say "what
is intended to be 50% (or what is exposed as 50%) now becomes 75%," or 25%,
depending on whether we are talking about positive or negative.
That is because simply by definition, an patch with optical density of 0.60
does NOT pass 50% light that shine through it. It passes less light, so it is
not, it cannot be called 50%.
This is also the same as copying a patch that has 50% tone. Suppose you
calibrate the whole thing so that after the copying, the copy also has 50%
tone, or optical density of 0.30. Now, you overexpose by 1 stop or 2, and the
dots grow, so the optical density is no longer 0.30. You do not say that it
is still 50% dot but with different optical density. The fact is, it is no
longer 50% tone (because the dots has grown due to overexposure).
There is a direct, one-to-one conversion from percent dot to optical density.
You can find a table from any graphic arts or printing book. There isn't such
thing as 50% dot with different optical density. The only exception is when
Dmax is less than 2.0. Then the opague area must also be included in the
calculation of optical density because in this case, both the clear area and
the opague "Dmax" area pass light through.
Thus one can control the process by either dot measurement or by density
measurement as Adam is trying to do. They are the same thing!
What the service bureau was saying is that he still expose the same thing
(expose 10% as 10%, 50% as 50% etc.), but the fact that the optical density
changes show that the calibration is not maintained.
Of course we are not living in a theoretical world, so process/normal
variation and margin of error must be allowed, and we cannot have
unreasonable expectation from the service bureau, but assuming that in offset
printing, the usual range is 10% to 90%, we should reasonably expect 5% and
95% to be reasonably accurate... probably 2% margin on the low end and maybe
5% on the high end, but this of course depends on the printing requirement.
Or it is exactly the same thing to require some repeatibility (with error
tolerance) with optical reading on the high and low density.
>> As it turned out, for me anyway, the OPTICAL density was important.
Of course, I have no doubt about it; but again, the optical density measures
the percent light that passes through.
>> If you like, you can read all about it in a paper I wrote which is posted
on the Bostick & Sullivan web page.
As mentioned before, I have and probably have downloaded it somewhere. Again,
I think it is an excellent excellent excellent paper.
Best Regards,
Dave S
This archive was generated by hypermail 2.0b3 on Thu Oct 28 1999 - 21:39:35