Zone System/Drum Scanners - B&W film calibration?

Friday, October 28, 2005

Greetings All,

I’d like to learn how others are dealing with the following, thus I’ll start with a given set of input factors:

Black and White films – landscape/abstracts/arch subjects.

Hassie H1 and Arca Swiss 4x5 formats

Howtek HiResolve 8000 – scans down to 3 micros with dmax of 3.9 per manf. specs. and DLP Pro software

Epson 4800 with Red River Ultra Pro paper as norm

X-Rite 810 densitometer available if needed

Photoshop CS2 on G5/2.7GHz with 8GB RAM

Challenge:
Now, if I refer to all the articles/books (Adams – The Negative, etc.) I’ve found so far, they are making an assumption that the negative is taken into the wet darkroom and that environment is what the film density needs to be calibrated to – diffusion enlarger, etc.

That may work in the wet darkroom, but the drum scanner should allow me to go further than that; i.e. what works in the darkroom is NOT going to be the OPTIMAL for the photographer equipped with high res drum scanner.

The 8000 line scanner is capable of extracting a tremendous amount of detail from a VERY dense negative – the film density Mr. Adams was looking for is not what I’m pondering the scanner is capable of pulling out – ie; their Zone 1 density may actually have detail in it on the drum scanner, thus making it a Zone II for me.

In talking with a Mentor of mine, he suggest just shoot a given/standard scene – take high, mid, low readings, average, shoot at different ASA’s, develop, scan. Simple, yet logical – that’s a nice attribute to this Mentor – keeps me from overcomplicating the solution!

However, here’s what I’m thinking (yes, I know I just departed from the KISS principal that my Mentor has championed!), and this is an attempt to compromise between the KISS principal and my normal mode of thinking:

How to Define the ASA of a given film when a high res scanner is the “enlarger”/capture device -
Take a scene with broad range of latitude -
Set spot meter to film manf. ASA rating
Select a portion of the scene you want as zone 2 – faint texture/detail.
Shoot it –
Note in your shot record the other Zones that the mid Zone objects are sitting at and a Zone 8 object.
Now, repeat this for the 3 ASA ratings below the manf. ASA rating.

Develop at manf. time recommendation in your favorite brew.

Now Scan all frames – examine each frame on calibrated monitor and select one that has presented the best Zone 2 exposure – make a test print to prove the retention of texture that you want.

Once the films Zone II ASA is defined (densest negative that the scanner can still extract texture from), then work on the Zone 8 details through the development time.

I’m expecting that if I then took the density values of these negative, the values would differ substantially from the “guidelines” I’m finding in the published material.

Seems logical to me – and that should be the first clue that the proposed procedure is flawed!
I’m expecting to learn a lot from your feedback and hopefully produce a far better path to mastering the Zone Systems use with a high res scanner.

Thanks,
Jack

I have a couple of questions for you.

If you get all that 3.9dmax up to white into your scan, how are you going to be able to see it on a monitor which doesn't have the contrast ratio available to view it?

Then, when you consider that a print only has a max of 5 stops (thats 1.5dmax to you) when measured with a spot meter, i.e. how we see it, how are you going to handle all the tonal compression you will need to do to get the 3.9dmax scan onto paper?

The simple version is that once you get all that scene information on your negative and into your computer you can manipulate it to best advantage for your output device and chosen media.

But I don't think you will find it that simple for every scene out there. As Rob just wrote above, you still have the limits of your print to deal with. It won't give anymore range than it has and the ability to seperate all those tones just may not be available. And its 8 bit output at best and then there is human vision to deal with.

For some scenes it might be that leaving out some numerical values of print tone gives the most effective print. This will vary with scenes, paper and inks. I don't think you will find a magic bullet here that will suddenly alter how you expose and develop your film. So in two sentences - get all the scene information you need or want on your negative in a way that your scanner can record it all. Then make the print that looks best.

Film has finite limits at its thresholds of exposure. Below a certain level nothing is recorded on the film, above and it blocks with little or no differentiation of tones. Obviously you will stay within those limits. The next limit you hit is the range of capture of your scanner. You'll stay within that as well. ; >)

Perhaps you are thinking this out too far. You have the possibly finest gear available - an amazing arsenal of tools. If you know how to use them you should be technically capable of making fine and wonderful prints. Stop thinking and go shoot something. Try out your ideas while doing so but remember to go expose film and make prints - that is by far the most important thing to do - your questions will be answered by doing.

It's really simpler than you are making it. Or at least it can be.

The way I look at the Zone System is that it uses film as an intermediary between the subject brightness range (SBR) and the chemical photopaper's density range. The SBR is typically much greater than the paper can handle, so you use the Zone System to compress the SBR on film into something that is a much closer match to the paper, and thus easier to print. And that is the goal of the Zone System - to make it easier to print.

When you switch to printing digitally, you loose the need for the full Zone System methodology because you loose the need to use the film as an intermediary between the SBR and the paper. The scan becomes the intermediary, and it makes an *exact* match to the paper. IOW, a value of 0 is ink black, and a value of 255 is paper white. Every time.

How do you optimize a negative for the hybrid workflow then? Very simply: expose for the shadows and let the highlights fall where they may. You no longer need N- and N+ development times. Why? Because the scanner takes whatever density range you give it, and neatly fits it into the range (8 bit is 0-255, 12 bit is 0-4095, 16 bit is 0-65535).

This doesn't mean you can toss accuracy and precision out the window. Far from it. The scanning workflow will take great advantage of you getting your exposure exactly right, and of your developing the film repeatably.

And you can take the film a little closer to the edge. What I do is develop for what used to be about N-1.5 in my wet darkroom days. This is now my N development time, and it's the only development time I use, regardless of the SBR for an individual exposure. This decreases graininess and increases sharpness a bit without sacrificing tonality (at least for 5x4 Tri-X in XTOL 1:3). These flat and thin negatives drum scan beautifully, but I wouldn't want to take them to the wet darkroom.

What I'd warn anyone away from is making the negatives even more dense. The idea is interesting because you'd think that the scanner can read thru just about any density you can throw at it. It can. But developing a B&W negative to give you a Dmax of 3.4 does bad things to the negative. Bad things. A fully cooked negative has an ugly and large grain structure. You'll also find some interesting artifacts that I can only describe as "density echos" in areas of high micro-contrast. It's not a pleasant sight.

If you are thinking of using the same negative for wet darkroom and scanning, then optimize for the wet darkroom. This will scan just fine. If you are only going to scan, then optimize for scanning as described above. Works beautifully.

Rob - to address your questions;

My mission is to extract as much data from a negative, even if it is beyond the capabilities of what can be utilized at THIS particular moment in time by devices downstream from the scanner.

If I have the data today, I do not have to rescan it later to bring it up to the "new" capabilities of printers that are going to be coming down the pipe in the near future.

I've spent many years in software development, primarily on the Silicon Graphics line of computers, and learned that the technology moves very fast.

Thus, my logic for purchasing an 8,000 line drum scanner - I consider it as data generator - everything down stream of the "capture" device is dependent upon the quality of the capture, even if it can't match the device in bit, dmax, etc. at this moment in time, the data is there when the technology catchs up.

Thus, my negative must be optimized for the scanners abilities.
Jack

at 8000dpi 16bit colour channels for 4x5 neg = 7.3GB files. I hope you've got lots of disk space available. You better double up on your RAM as well.

You should also review your Shannon on noise versus information - at 8000 dpi on LF you are way past picking up information and are way into grain noise. The drum scanner is good for the Dmax, and I think there is some interesting work to be done on the use of high dmax negatives for scanning - as the alternative process guys remind us, their is useful info in them. I will be interested to see what you get. My first test, however, would be set of scans from 2000 DPI to 8000 DPI with a real world scene to hone in on the best DPI that picks up info without excessive grain noise.

IMO, GIGO- garbage in=garbage out. First you have to maximize the info in the negative, and this includes sharpness, halation, edge effects, and grain. The least important thing is probably tonal distribution, as you can remap it any way you want it. The highest quality negative will be exposed so there is information as far down in the shadows as you deem necessary, and will have a limited density range. The last thing you want is densities up over 2.5+, even if the scanner can deal with them just fine. That's not where film performs at its best. It's true that optimum for the scanner will be different than wet process, but you're going the wrong way. Having a scanner with infinite range and resolution doesn't change the game as much as you think, because it wasn't the weakest link to begin with.

"Then, when you consider that a print only has a max of 5 stops . . . "

With traditional silver paper you mean. One of the reasons I believe I'm able to make better prints digitally than I could in the darkroom is the tonal range available (which I believe is what Bruce is saying). I can easily print a 21 step wedge on my Epson 2200 printer and see 21 distinct steps (i.e. 10.5 stops). In a darkroom with an enlarger on silver paper the most I could ever get was about 12 steps. I could choose where I was willing to accept the loss of detail - whites or blacks - or I could get nice midtone separation and have no decent blacks or detailed whites at all - but somewhere I was going to lose potential detail depending on the contrast at which I printed because of the limitations inherent in the materials.

Brian,
Please do a little test for me. Take your printed 21 step wedge and place it on a normally lit wall in your home. Then take a spot meter and meter the darkest step. Then meter the lightest part(paper white). I guarantee you that you will not get more than 5 stops of range. If you were to shine two bright spot lights at 45deg and measure you could possibly get 6 stops from a very glossy print but for the majority you will only get 4 to 5 stops. Under a densitometer you may 7 stops or at a push 8 stops but unfortunately you cannot see like densitometer.

Being seduced by digital and sensitometric numbers doesn't alter the fundamental properties of paper in normal viewing lighting.