Digital Neg. Article in current View Camera Mag.

The current issue, July/Aug 2010, of View Camera Magazine has a article on creating digital negatives using QTR. It's written by Ron Reeder and includes some beautiful gum-over-pd prints by Ron.

The article is quite well written and explains the variables in QTR in a way that makes more sense, for me at least.

One question is raises for me is adjusting the Gamma variable. None of my QTR curves make use of the gamma variable and leave it set to gamma = 1

So is this new? Anyone else using the gamma for their digi negs?

I know that Ron is a member of this forum, so perhaps he will chime in as well, but here are my thoughts.

I'm not using a gamma function for my carbon negs, but based on my very early testing, I think I'll end up using a gamma curve as the base curve for my cyanotypes.

Keep in mind that the gamma function is just a mathematical tool for describing the curve shape. As you know, the closer you can get the original curve to the target (before you apply corrections from the step wedge for digi-negs in an .acv file, or before you apply linearization for prints), the more accurate those corrections will be, especially in the very light tones, where 0.01 logD error on your densitometer reading of 0.05 or 0.06 is a pretty big error.

QTR allows values for gamma between 0.1 and 10. Values <1 cause the ink curve to have a "hump", i.e. it rises quickly from 0%, and the seems to level off toward 100%. Values >1 cause the ink to rise very slowly from 0%, but increase more rapidly toward 100%. The closer the value of gamma is to the extremes, the more pronounced the effect will be.

For the technically minded, gamma is a power function, for which:

Output (ink) = Input (gray) ^ gamma

So for gamma=1: 25% gray = 25% ink, 50% = 50%, 75%=75%, etc.
For gamma=2: 25% gray = 6.25% ink (0.25 ^ 2), 50% gray=25% ink (0.5 ^ 2), 75% gray=56% ink (0.75 ^ 2), etc.

Looking at the graphs in QTR, though, it appears that Roy is modifying the function somehow, since gamma=2 for a curve does not give the expected ideal results, but the principle still holds.

Hope that all helps...

--Greg

gmikol said:

QTR allows values for gamma between 0.1 and 10. Values <1 cause the ink curve to have a "hump", i.e. it rises quickly from 0%, and the seems to level off toward 100%.

--Greg

Click to expand...

Thanks Greg,
This sounds to be consistent with what Ron says in the article.

"lowering this number has the effect of putting more light into the shadows of your print"

So that would mean that it is adding ink to the lighter ( 0 ) end of the scale, but is not proportional over the entire scale.

In the sample QTR curve in the article he uses a gamma=0.7

Interesting..... I'll have to think about how/if I should use this in my curves. I've started working on a silver gelatin curve (again) after abandoning the effort a few years ago.

R Shaffer said:

One question is raises for me is adjusting the Gamma variable. None of my QTR curves make use of the gamma variable and leave it set to gamma = 1

Click to expand...

Rob-- I also used to leave the gamma function set at 1, it's neutral setting. But, on Clay Harmon's advice I have begun lowering it to around 0.7, at least for platinum/palladium negatives.

In a two part gray system (only dark inks and light inks) changing the gamma value has the effect of shifting the crossover point either toward the dark tones of the negative (gamma values above 1) or toward the light tones of the negative (gamma values below 1). (The crossover point is the location in the ink curves where the light inks stop increasing and the dark inks begin contributing to negative density.) In writing a profile for palladium printing I often find that I need to get more ink into the light part of the negative to open up print shadows and make the final linearizing gray curve less agressive. I can accomplish this partially by increasing the light ink limits, but lowering gamma to 0.7 is also a big help.

Lowering gamma to 0.7 increases steepness of the light ink curves in the light end of the negative which has the beneficial effect of increasing contrast in print shadows. Conversely, it has a negative effect by flattening the ink curves in the dark end of the negative and thus decreasing highlight contrast in the print. But, fortunately, I can counteract the highlight flattening effect by setting black boost to about 50% higher than the other dark ink limits. This causes pK ink to increase sharply in the darkest negative tones and adds back the needed contrast. This, of course, gives the green negative guys fits but works for me. (and yes, I have switched from using mK to using pK ink but that is a whole 'nother story).

In sum, you can leave gamma set at 1, but lowering it makes for a milder gray curve for the final linearization.

To illustrate I attach three ink curve graphs. The top one is with gamma =1. The red arrow shows the crossover point where the dark inks begin to contribute.

The second graph is the same except gamma=0.7. The crossover point has moved to the left and more ink is being deposited in the light areas of the negative. Note also that on the right end the dark ink curves have flattened somewhat.

The bottom graph is with gamma=0.7 plus I have applied a gray curve to linearize the tones on the final print. To accomplish linearization gray curve has shoved the crossover point further to the left and has altered the slope of the dark ink curves on the right.

Cheers, Ron Reeder

Ron-san said:

To illustrate I attach three ink curve graphs. The top one is with gamma =1. The red arrow shows the crossover point where the dark inks begin to contribute.

The second graph is the same except gamma=0.7. The crossover point has moved to the left and more ink is being deposited in the light areas of the negative. Note also that on the right end the dark ink curves have flattened somewhat.

The bottom graph is with gamma=0.7 plus I have applied a gray curve to linearize the tones on the final print. To accomplish linearization gray curve has shoved the crossover point further to the left and has altered the slope of the dark ink curves on the right.

Cheers, Ron Reeder

Click to expand...

Thanks Ron,

I see I see......... the graphs are a big help. I recall early on in working with QTR trying to move that point by changing the ink limits.

I use a two part curve as well for all my negatives, so I will take a look into how this may be of use. My linearization curve is where I am opening up the shadows.

On a some of my kallitype images I get too big a jump between subtle shadow values. It looks linear in the test strip, but can show up in the print.

The thought of changing to pK from all my mK curves is just too much to contemplate.

Yeah, the whole gamma adjustment thing was one of those 'aha' moments one time when I was playing around with all the parameter using the Windows-based QTRgui to evaluate visually on the resulting ink distribution graph what changing each of the parameters would do. When I dropped the gamma from 1.0, it was immediately apparent that it drops more ink on the areas that will be shadows in the negative (i.e. the thinner areas) Which, in my experience you were going to have to do anyway with your gray curve. So now I routinely drop the gamma to 0.5 to 0.7 for most of the pigment ink printers and compensate by boosting the black for the highlights for one or two iterations to get as close to a decent looking step wedge as I can before I even start futzing around adding a gray curve.

Ron and I carried on an email correspondence about this and other stuff related to QTR and I learned a lot. He's definitely the man. Everyone should buy his new book. It is really good. I have a pdf <here> on my website describing some of this, but Ron's book is much more detailed and complete. I humbly bow down to his steely resolve in cranking out the book. It will clarify a lot the mystery surrounding QTR for the newbie.

We do disagree about one thing, though. I am a total believer that even us Mac users can benefit from using the Windows QTRgui to help build QTR profiles. With the VMware and Parallels virtual Windows environments, it is dirt-easy to use even if all the rest of your work is done on the Mac. I suspect Ron thinks I am a bit of a lightweight for using this as an aid, but hey, I love visuals.

If you don't have Parallels yet, the new Parallels 6 is amazing. It basically allows you to open Windows apps in a way that they look like native Mac apps on your desktop. Dragging data, files, whatever between them is totally transparent. I still like printing from the Mac since I can print from within Photoshop or Lightroom, though, instead of having to fire up a separate printing app like on Windows.

clay said:

We do disagree about one thing, though. I am a total believer that even us Mac users can benefit from using the Windows QTRgui to help build QTR profiles. With the VMware and Parallels virtual Windows environments, it is dirt-easy to use even if all the rest of your work is done on the Mac. I suspect Ron thinks I am a bit of a lightweight for using this as an aid, but hey, I love visuals.
.

Click to expand...

Clay-- This is a most minor area of "disagreement". If the Windows QTR gui helps folks use QTR, then hallelujah brother, more power to them. Having become so adjusted to looking at a textfile I am too lazy to figure out how to emulate a Windows environment.
Let's just keep optimizing this lovely tool. I, for one, cannot imagine how I ever made prints in the dark ages prior to digital negatives. And I am still firmly in the camp that thinks QTR is the ultimate app for making them.
Soldier on!

And thanks for the kind words about the book.

Ron Reeder

clay said:

He's definitely the man. Everyone should buy his new book. It is really good. I have a pdf <here> on my website describing some of this, but Ron's book is much more detailed and complete. I humbly bow down to his steely resolve in cranking out the book. It will clarify a lot the mystery surrounding QTR for the newbie.

Click to expand...

Ron's new book is certainly on my reading list. You've got some really good info on your site Clay. I was looking through your QTR instructions and my problem is apparent

I have not run my B.ase U.nderlying L.evel of L.ight S.ensitivity with a H.alftone I.nterval T.iming

How can I possibly hope to attain the pinnacle of my craft without?

A note: For those who don't want to *pay* for VMWare or Parallels, Sun's *free* VirtualBox software also does the job, and pretty neatly. I'm using VirtualBox myself - no issues, works like a charm!

Regards,
Loris.

P.S. Oops! Forgot that it's not Sun anymore, became Oracle: http://www.oracle.com/us/technologies/virtualization/oraclevm/061976.html

clay said:

...
We do disagree about one thing, though. I am a total believer that even us Mac users can benefit from using the Windows QTRgui to help build QTR profiles. With the VMware and Parallels virtual Windows environments, it is dirt-easy to use even if all the rest of your work is done on the Mac. I suspect Ron thinks I am a bit of a lightweight for using this as an aid, but hey, I love visuals.

If you don't have Parallels yet, the new Parallels 6 is amazing. It basically allows you to open Windows apps in a way that they look like native Mac apps on your desktop. Dragging data, files, whatever between them is totally transparent. I still like printing from the Mac since I can print from within Photoshop or Lightroom, though, instead of having to fire up a separate printing app like on Windows.

Click to expand...

I was just having a little fun overstating this so-called 'disagreement' . I know we both think that as long as you are using an Epson printer, QTR is by far the best way to go. I think your book is a must-read for anybody that is just starting to investigate this route to digital negatives. I was impressed with the improvement in the printing quality of LuLu book's images too. A few years ago, it was essentially a text-only option. But your new book's quality is vastly better than what you could get back then.

Ron-san said:

Clay-- This is a most minor area of "disagreement". If the Windows QTR gui helps folks use QTR, then hallelujah brother, more power to them. Having become so adjusted to looking at a textfile I am too lazy to figure out how to emulate a Windows environment.
Let's just keep optimizing this lovely tool. I, for one, cannot imagine how I ever made prints in the dark ages prior to digital negatives. And I am still firmly in the camp that thinks QTR is the ultimate app for making them.
Soldier on!

And thanks for the kind words about the book.

Ron Reeder

Click to expand...

Oh with all this new digi-neg info I just had to give it a try. So I remade my kallitype curve with gamma=.5 and photo black in lieu of my usual mK. And here is the result......

Just curious where you all are making this gamma adjustment? Is this in the advanced settings part of the QTR program window that says "Gamma (midtones)" or in the grey curve section of the Curve Creator? My guess is the Curve Creator...

PVia said:

Just curious where you all are making this gamma adjustment? Is this in the advanced settings part of the QTR program window that says "Gamma (midtones)" or in the grey curve section of the Curve Creator? My guess is the Curve Creator...

Click to expand...

Yes, curve creator, at least on a PC. The gamma adjustment in the 'Curve Creator' under the second tab called 'Grey Curve' is a box labeled 'Gamma' just above where I plug in my photoshop curve. Normally it is set to 0.

Thanks, but actually I think its normal default is 1.

I have had a new Epson 3880 for about a month now. I bought it mostly to do digital negatives. Now the learning curve. Does anyone have a link to a good tutorial on using QTR? I don't understand the information here, out of contax as it is. Even better, does anyone have a configuration file that can be used for cyanotypes that I could start with? I think I would like to start with the text based version first. I'm an old unix/linux hand, and don't have a problem using text configuration files.

Thanks, Jon

Get a copy of Ron's latest book. It clearly describes two methods of creating digital negatives -- Epson ABW and QTR.

A New Book on QTR and Digital Negatives

jon.oman said:

I have had a new Epson 3880 for about a month now. I bought it mostly to do digital negatives. Now the learning curve. Does anyone have a link to a good tutorial on using QTR? I don't understand the information here, out of contax as it is. Even better, does anyone have a configuration file that can be used for cyanotypes that I could start with? I think I would like to start with the text based version first. I'm an old unix/linux hand, and don't have a problem using text configuration files.

Thanks, Jon

Click to expand...

pschwart said:

Get a copy of Ron's latest book. It clearly describes two methods of creating digital negatives -- Epson ABW and QTR.

A New Book on QTR and Digital Negatives

Click to expand...

+1 Ron's book is a great start and a bargain.

Also, Clay Harmon has a nice QTR digi-neg tutorial here
http://www.clayharmon.com/techne/?page_id=2

And if you shoot me a PM I can send you a starting point for cyanotype, a QTR .qdif file and .acv curve.

Thanks! Just what I need. I'll be ordering it this week.

Jon

R Shaffer said:

+1 Ron's book is a great start and a bargain.

Also, Clay Harmon has a nice QTR digi-neg tutorial here
http://www.clayharmon.com/techne/?page_id=2

And if you shoot me a PM I can send you a starting point for cyanotype, a QTR .qdif file and .acv curve.

Click to expand...

Thanks Rob, I sent you a PM. I'll also take a look at the link.

Jon

R Shaffer said:

+1 Ron's book is a great start and a bargain.

Also, Clay Harmon has a nice QTR digi-neg tutorial here
http://www.clayharmon.com/techne/?page_id=2

And if you shoot me a PM I can send you a starting point for cyanotype, a QTR .qdif file and .acv curve.

Click to expand...

I must say i found Clay's QTR tutorial really helpful for windows users and he has also been able to help me personally which i am very grateful for. Am nearly there with it, needs a few more days calibration, however initial results are better than I expected, this seems to be one of the best digital negative workflows I have tried.

I have one question relating to Chart Throb. When people are scanning in their print to be analysed are they scanning as an RGB file? On the Chart Throb help page it clearly shows it being scanning in as RGB, however it then mentions it should be ‘scanned with the full grayscale range’ Maybe just my interpretation is wrong??

I get a curve with the RGB scan but Chart Throb will not to accept greyscale images and flags up an error.

Davec said:

I must say i found Clay's QTR tutorial really helpful for windows users and he has also been able to help me personally which i am very grateful for. Am nearly there with it, needs a few more days calibration, however initial results are better than I expected, this seems to be one of the best digital negative workflows I have tried.

I have one question relating to Chart Throb. When people are scanning in their print to be analysed are they scanning as an RGB file? On the Chart Throb help page it clearly shows it being scanning in as RGB, however it then mentions it should be scanned with the full grayscale range Maybe just my interpretation is wrong??

I get a curve with the RGB scan but Chart Throb will not to accept greyscale images and flags up an error.

Click to expand...

Been years since I did Chart Throb. But I went back to some old scans and they are all RGB and Chart Throb read them and made curves that looked right. If I converted the scan to greyscale, then it would error out as you say.

Here is a good site for digi-negs using Chart Throb & HSL or RNP Arrays.
Dead Link Removed

R Shaffer said:

Been years since I did Chart Throb. But I went back to some old scans and they are all RGB and Chart Throb read them and made curves that looked right. If I converted the scan to greyscale, then it would error out as you say.

Here is a good site for digi-negs using Chart Throb & HSL or RNP Arrays.
Dead Link Removed

Click to expand...

Hi Rob

Can you put up one of your palladium/na2 scans with the curve as it would be nice to compare. Have attached what i am getting at the moment.

Davec said:

Hi Rob

Can you put up one of your palladium/na2 scans with the curve as it would be nice to compare. Have attached what i am getting at the moment.

Click to expand...

I was not doing pd printing back in the Chart Throb days. These are kallitypes from 2008. One is the original contact print, then the curve and then the contact print with curve applied. These are from an HP B9180 printer I was using at the time.

R Shaffer said:

I was not doing pd printing back in the Chart Throb days. These are kallitypes from 2008. One is the original contact print, then the curve and then the contact print with curve applied. These are from an HP B9180 printer I was using at the time.

Click to expand...

In looking the curved contact print, it looks a little underexposed. I poked around and could not find any of my old chart throb test prints. But I ran an analysis on the curved print and it came up pretty darn linear, just the blacks don't seem quite up to snuff.

jon.oman said:

Thanks! Just what I need. I'll be ordering it this week.

Jon

Click to expand...

I just ordered Ron's book. I'm looking forward to getting it!