Topic – Variable Contrast

My recent post Enlarger Beats My Scanner was about my newly acquired Durst M670 enlarger. As you get from that post I was pleasantly pleased with the paper copy of the Fujifilm GA645Zi 6×4.5 cm negative of the Piața Unirii.

As a newbie in this altogether there is much to learn and I decided to tackle the issue of enhancing contrast with the Kentmere Variable Contrast (VC) paper I had bought.

In the old days there were several different grades of a photo paper. A grade 0 (or 00 even) paper gave a visually very flat greyish result with low contrast, while a grade 5 paper gave a more dramatic contrasty effect. Today almost (all?) photo paper is of the VC type, where the actual contrast is controlled by the colour of the light used during copying. A yellow light corresponds to low contrast and a magenta one to high contrast.

It sounds too good to be true, and in some aspects perhaps it is. Anyway it turned out to be not as practical as, I at least, had hoped for. The colour head, colour mixing box, of the Durst has wheels for setting the amount of yellow and magenta  on a scale from 0–170 (it also does cyan but that is irrelevant for B/W-printing). The unit is unknown and some enlargers use a 0–130, 0–199 scale, or something else… I am also not certain that a “maximum” magenta of 170 is the same as a “maximum” one of 130 or 199 and so on. It could be, and it could not.

In very general terms the two colours can be mixed to achieve a certain grade of contrast of the paper at hand. If this mixing is done in, not only a preferred ratio, but also at some preferred total amount, the enlarger’s exposure time could be kept constant! That is very cool! Imagine first finding the amount of exposure needed, make a test exposure/development etc, and then set the amount of contrast and a second exposure at the same number of seconds and you’re done.

The Durst manual contains this table for constant exposure timesm607

Unfortunately reality rears its ugly head at this point as the values in each column depend not only on the enlarger’s mixer box but also on the photo paper used: What amount 0–170 should be used for yellow (Y) and magenta (M) for Kentmere Fineprint VC?

Using the data sheets for some photo papers I found on the web the following diagram could be sketched:

m607 (1).jpg

The diagram shows some collected data for a few VC papers  I found on the web. The x-scale indicates equivalent contrast grade 00 to 5. A general tendency of yellow (Y) at lower contrast and magenta (M) at higher contrast is obvious. But the devil is in the details, note how the red lines are all over the place… Quite disconcerting!

For some Ilford paper the amount of Yellow and Magenta can be read from above (Y ILFO, M ILFO). Yet another VC paper (Kodak Polycontrast) are sketched with wider lines (POLY). Kodak Glossy is in green (K GLOSS) and, finally, and unfortunately, Kentmere in red (K W).

It appears that Ilford’s paper behaves in a predicted way, as does the Kodak. However, even if the ratio of Y and M for a certain contrast is established, there remains to determine the total level of light for a constant time of exposure! We can imagine this as finding a path in a 3D-volume, where the x- and y-axes are Y and M and the z-axis is exposure time in seconds. And we want the path to be at a constant level too!

(Here I wish I could include a sketch of a 3D-landscape with valleys and mountains with a path of constant altitude in it. I tried. But I can’t. So I won’t!)

Perhaps it can be done, but singling out individual points in a 3D-volume is… well, impossible… or, at least, involves a huge amount work! And, even if I did go that way, any change of either paper or light source (which is much too light as it is) would require a complete re-run of the entire process. There has to be another  way!

Enter the saviour: Split grade processing. Given that yellow controls the highlights of the resulting picture and magenta the contrast. Why not combine the two, not at the same time, but after each other? This is exactly how split grade processing is done. The method entails two test strips of exposure: One for the highlights and one for the shadows. Then, in a final exposure, combine the two.

Given that Y and M are linearly independent (excuse the linear algebra vernacula for the moment) then make a test strip in the highlight using maximum yellow (170) and determine the number of seconds required for the highlights to just appear. This is the base exposure.

Now find the area of the picture with the darkest shadow and make another test strip in this area to find the amount of extra exposure required with maximum magenta (170) for the shadows to just disappear. Finish off with a total image using these two exposure times and Y and M settings, and you are done. To be more precise, the M-test strip should ideally be ‘initialized’ with a base exposure of Y, so that the final M-time corresponds to the exact contrast wanted. (Y and M are not perfectly independent, the M also exposes the highlights somewhat.)

The idea is to determine, in each negative, the exact amount of exposure needed for the highlights and the shadows respectively. It really is as easy as that. This assures that the combination of negative (and its exposure) and paper (and its range of contrast) are perfectly matched. Clever eh?

Ok, lets go into examples. The first one, the cows, are first printed with Y170 at the base exposure (turns out to be 8s at f/22)…

cow0

…and it is of very low contrast all over as expected.

The next has the same base exposure at Y170 but also M170 for 20s:

cow1 (1).jpg

Ouch, this looks completely off (it was even worse when straight out of the scanner. I had to do some curves in Aperture before publishing).  Ok, let it serve the point that it is too shadowy ok?

This (obviously) has too much contrast (murky shadows) because I did not foresee that the 20s exposure at M170 also had some general exposure effect (look at the clouds and see that they are indeed more contrasty in the last image than in the first). Just for fun I tried reducing the M170-exposure to 8s:

cow2.jpg

Which is better, but in my opinion should have been a wee bit contrastier still. Perhaps a 12s exposure of M170 would have been perfect? At this time I was bored of the critters and went on to another negative…

In this picture of the Golden Gate in ‘Frisco I set the highlights as per instructions above (Y170 etc) and used a base exposure as mentioned, before going for the denser stuff with M170 (6s Y170, 12s M170):

goldegate.jpg

This one I am completely satisfied with! Yes, proud even! In fact, I have never managed to control the contrast this much and this well in any picture before. Clearly split grade printing is the way to go!

Circled below are the two areas used to the test strips. The wash had the highlights and the concrete in the shadows were, well shadows. All other levels of grey fall nicely between these extremes.

gg (1)

I usually increase exposure on the test strips in f-stops (eg. 1, 2, 4, 8, 16s) but I suspect that given the base exposure with Y170, the M170-exposure can be made in a resolution of 2 second increments. Future will tell.

——

All images in this post were scanned with a Canon Canoscan 9000f. The cows were somewhat trimmed (!) with Aperture to resemble the prints in real life.

——

UPDATE. I also scanned the test strips used for the above prints for reference. First the cows’ highlights at Y170-yellow:

cowtest2

The notation may require some explanation? I use a sliding piece of paper as cover and in increments move it to the right. First the timer is set at 1s, make the ‘1’ exposure, then set timer to 2s, move the covering paper and make the ‘2’ exposure and so on. This way the last exposure (‘8’) has been lit for 8 seconds, the previous (‘4’) for 8+4=12 seconds etc. This addition is the lower figures. Here the maximum exposure was 8+4+2+1=15 seconds. The left-most field was a mistake, the series proper starts at the ‘1’. Every field has one f-stop difference from its neighbor. There were some subtle highlights at 8 seconds but with hindsight a 12s exposure would probably have been better…

cowtest1

As above but with magenta in the shadows and a constant 5 second increment instead of f-stops, with a base exposure of 15 seconds. I believe the base exposure was also with magenta here, it shouldn’t be!

For Golden Gate these were the test strips: First Y170 and exponential exposure where a 6 second exposure (4+2) was chosen as the base exposure:

ggtest2

A 4 second exposure did not have any effect, but a 6 second had. So I chose 6 seconds as base exposure.

And then the M170 exposure for the shadows where I settled for an additional 6 seconds with magenta, 4 seconds would also have been fine:

ggtest1

I specifically wanted a visible structure in the pillar’s concrete to appear.

 

 

 

 

Advertisements
Gallery | This entry was posted in All, Topics and tagged , , , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s