Kodak Characteristic Curve - Log H Ref equals middle gray?

[markbarendt]

I don't disagree with the thought that the "Ref" falls in the mid-tones.

I found this, this morning. http://wiki.magwerks.com/wiki/images/b/bb/NIST_SP250-37.pdf

It does appear to be a measure of exposure time, in relation to a standard illuminant, and in the Kodak cases the scene is a step wedge.

What I can't find is a way to convert it to an "EV" or lux value.

 

[ic-racer]

What I can't find is a way to convert it to an "EV" or lux value.

The curves in the quoted Kodak PDF files look to have the X axis as the log of lux-seconds, so without knowing the exposure time of the sensitometer you can't get the lux value.

 

[Bill Burk]

I can break down my system and give you a practical example of how much lux there could be.

For example maybe you are trying to adjust your enlarger to a certain brightness to make a test setup...

My sensitometer is basically an electronic flash aimed up at a piece of glass with a Stouffer scale taped to the top. There's an ND filter in between, this helps make the graphs fit on paper for the films I use most ( 400 to 32 speed film).

The sensitometer has 3 choices of exposure times: 10 to the minus 2, 10 to the minus 3 and 10 to the minus 4...

So that's 1/100th second, 1/1,000th second and 1/10,000th second.

At 1/100th second (the "shutter" speed I use to test film) my unit delivers 100 lux to the film under the step that lines up with the "0" on my graph x-axis.

I have three x-axis markings on my graphs. There is the lux-seconds log marking at top (mcs = lux-seconds), the corresponding film speed just under it, and at the very bottom is the density of the step wedge.

http://www.beefalobill.com/images/sensitometry-CP1.pdf

An important thing about my top scale: I calibrated it by a simple practical method that is so obvious it's almost stupid. One of my graphs of TMY-2 happened to hit the ASA target. So I lined up the 400 where that curve crossed 0.10 above B+F.

---
Back to the "Ref" - it seems geared towards color negative material. I think it is gray card 18%, because so much literature suggests shooting a gray card and measuring its density.

There isn't much emphasis on shooting gray cards in black and white.

I'll guess it's because color film is pretty much developed for a "standard" time for the best color balance. While black and white is developed for various times to increase/decrease contrast.

So the place you check for proper exposure for black and white is down in the shadows, while the point that you can easily check for proper exposure of color negatives can be in the middle gray.

ps I work for Kodak but the opinions and positions I take are not necessarily those of EKC. In other words the stuff I do at work has nothing to do with film, when it comes to film I am a hobbyist...

 

[markbarendt]

I'm still not buying the idea that H (exposure) = the grey card reading. The grey card, even with all it's accolades doesn't read "exposure" directly.

My impression is that log H is an expression of the "real" exposure not an arbitrary reference point.

The biggest reason is that any reference point in the scene can be used to find the camera setting, with equal quality; it is simply a matter of knowing what offset to apply.

A grey card for example, given it's known status in relation to exposure settings; is truly the equivalent of an incident meter reading when done properly. The only difference in practice is the need to do a bit of Maths, applying the proper offset.

 

[Photo Engineer]

You are making too much of a deal over this guys!!!!!

Put it in a camera and expose it!

PE

 

[Bill Burk]

I'm still not buying the idea that H (exposure) = the grey card reading. The grey card, even with all it's accolades doesn't read "exposure" directly.

I think it's the gray card landing!

According to the instructions, you don't read the gray card - you use an incident meter.

Then you set your densitometer onto the negative image of the gray card picture you took... after the film is developed.

If it's too dense or thin according to the guidelines you adjust your EI accordingly - and that puts all the mystery of the 18% to rest.

 

[markbarendt]

You are making too much of a deal over this guys!!!!!

Put it in a camera and expose it!

PE

I think you are right.

 

[ic-racer]

An important thing about my top scale: I calibrated it by a simple practical method that is so obvious it's almost stupid. One of my graphs of TMY-2 happened to hit the ASA target. So I lined up the 400 where that curve crossed 0.10 above B+F.

This is the portion of the thread where it veers off topic...

I spent a while looking for a thread I thought I had started on using film to calibrate a sensitometer. Both APUG and over at Large Format, but I could not find it. Anyway, I agree it is a good method especially when PE tells us that Kodak film is trimmed to an exact ISO number.

Also a few papers on using film to calibrate sensitometers:

Kodak patent to cailbrate one sensitometer to another using film: http://www.freepatentsonline.com/5543883.pdf

US Air Force Sensitometer Calibration with "Sensitometer Calibration Package consisting of two rolls of film..." Dead Link Removed

 

[Photo Engineer]

This is a tale from about 50 years ago.

When I was at Cape Canaveral, (CCMTA), a group was established that was charged with testing all incoming films to assure that they met the specs for grain, sharpness and speed. There were, at that time, published specs by Kodak and GAF our primary suppliers.

Well, I bucked heads with many department heads over this and I lost and a lot of taxpayer dollars were spend on testing, but in the end, no one found a discrepancy! Kodak and GAF lived up to their specifications that were published.

A hind!

Kodak still is as is Fuji and Ilford. So, what is the fuss? Why worry about these plots? This is 50 years of experience telling you that those curves are correct and a small math convresion will reduce it to your local tests. However, if you just set your meter at the correct value, you will get a good picture.

Bests wishes.

PE

 

[Bill Burk]

I have to admit defeat on Log H Ref: - several data sheets of films have different values for this even though the speed is the same. So it can't be tied directly to the speed point - otherwise two films of the same speed would have the same Ref.

Log H itself, and 0 being 1 lux-second. Those things are OK.

But what the ref is reference to, I don't know.

 

[markbarendt]

This is a tale from about 50 years ago.

When I was at Cape Canaveral, (CCMTA), a group was established that was charged with testing all incoming films to assure that they met the specs for grain, sharpness and speed. There were, at that time, published specs by Kodak and GAF our primary suppliers.

Well, I bucked heads with many department heads over this and I lost and a lot of taxpayer dollars were spend on testing, but in the end, no one found a discrepancy! Kodak and GAF lived up to their specifications that were published.

A hind!

Kodak still is as is Fuji and Ilford. So, what is the fuss? Why worry about these plots? This is 50 years of experience telling you that those curves are correct and a small math convresion will reduce it to your local tests. However, if you just set your meter at the correct value, you will get a good picture.

Bests wishes.

PE

That's good advice PE.

When I came back to film four short years ago my eyes were full of stars and wonder.

I wanted the "best".

Adams' was on a pedestal, tweaking was too; I was spot metering and pushing and pulling and semi-standing and...

I learned a lot.

The preponderance (IMO) of talk here at APUG and at the local photo club and at workshops and at other web sites is about tweaking to get the best. To a great extent this talk is (again IMO) mostly about finding magic bullets. How many "what's the best camera, film, developer, lens, meter, method, time, aperture, or format" threads have we seen?

We are in many ways like kids at the drag strip talking about tweaking our cars but forgetting that we don't drive very much on the drag strip. We go home and "make the mods" to our everyday cars and end up putting up with the side effects for 99.9% of the time just to shave a 1/4 second off a 1/4 mile run once every two or three months.

It took me some time to get over that starry eyed phase for photography.

I can honestly say that the biggest improvements to my craftsmanship/repeatability/quality in photography have come when I moved away from specialized tweaks and toward the norms in the instruction sheets; when I moved away from the specialized, finicky, subjective spot metering and EIs of Adams, to the dispassionate incident meter at box speed.

I learned to trust Sekonik, Kodak, Fuji, Ilford, Mamiya, Nikon, and Schnieder.

I learned not to listen to landscapers when I am trying to shoot a portraits.

Still it is also fun to know or figure out the technical bits.

 

[Stephen Benskin]

This is just a guess, but the log-H reference could have to do with the equivalent neutral printing density. SPSE Handbook defines this as "a form of analytical density that applies to color films intended to be printed rather than viewed. The system is analogous to equivalent neutral density except that it refers to a printing neutral rather than a visual neutral. A printing neutral is one that has identical red, green, and blue printing densities; that is, it prints as if it were a nonselective gray absorber."

 

[markbarendt]

Good morning Stephen, I was wondering if you were going to chime in.

 

[Photo Engineer]

I have to admit defeat on Log H Ref: - several data sheets of films have different values for this even though the speed is the same. So it can't be tied directly to the speed point - otherwise two films of the same speed would have the same Ref.

Log H itself, and 0 being 1 lux-second. Those things are OK.

But what the ref is reference to, I don't know.

Bill;

This post makes my point for me. You are an expert and from EK and still have problems.

Part of what you describe may be due to changes in illuminant for different films with the same speed and some may be due to old fashioned typos as mentioned earlier. IDK. In KRL we used a fixed exposure (Daylight or Tungsten) and ND filters to change the amount of light. And this the exposure for a film might be 1/100th of a second, but the intensity might have centered the step wedge for an ISO 25 film. For an ISO 50 film, we added a 0.3 (1 stop)ND filter and for an ISO 100 film we added a 0.6ND filter. Thus, it the speeds were correct, there was no reciprocity, and the curves should be superimposable. Calculating the light from the data was straightforward. The method used in the Kodak on-line data is kind of like removing an appendix via the mouth! Possible to do but not desired for best results. That is my opinion.

BTW: Give me a call someday and we can talk.

PE

 

[Bill Burk]

Thanks Stephen,

That seems likely, at least searching for "equivalent neutral printing density" brings up interesting reading. Unlike searching "href" which brings up every web page on earth.

I no longer believe there were typos. There is enough variation in the references to suggest that they refer to something unlinked to film speed point. Like maybe the amount of exposure that will give you a patch that prints 1.0 "neutral" - again I don't know what the ref is, but at least I trust the numbers (whatever they are) are correct as documented.

Because I read through every reference in my library on this, I learned a couple things by this exercise. You can calibrate your camera and light meter by shooting a gray card (and then taking densitometer readings off the negs). And (completely off-topic) I learned about the parallax focus method using a clear spot on a ground glass (I had seen the pictures before but didn't "get" why all three pictures had an "x" on them).

PE, I'll take you up on that, think it would be great to chat.

 

[Photo Engineer]

Send me a PM or e-mail Bill.

PE