Muddy Greens

kb3lms · Jun 23, 2014

After a hiatus, I've been trying to do some color printing but, have been having some problems getting the filter pack right and could use some help. My subject is a fake Macbeth chart that I inkjet printed and photographed. So, while it is not a real color chart, it is a constant and my intention is to settle on a good filter pack to reproduce the colors as best possible and then make measurements with a color analyzer so I should hopefully be able to find the proper filter pack to reproduce the proper color balance.

So far I have gotten to a decent appearing filter pack but the greens are very muddy and the "grass green" patch is very dark. All the green is more olive colored rather than a real green.

The image I am working with was shot on Ektar 100 (35mm) which I processed. The print is on Ektacolor edge and scanned on an Epson 3490 using VueScan. The current filter pack is C0/M40/Y70. Here it is:

Here is a scan of the original negative:

And for completeness, here is a scan of an inkjet print of the scanned negative. Scanned on the same 3490 with the same VueScan settings.

The scan of the print on Edge paper is pretty true to how the print appears. It seems I am "off in the weeds" with the filter pack (currently C0/M40/Y70) because I cannot make any substantial improvement. Any change seems to leave me with a cyan, magenta or yellow cast. This is the best I seem to be able to get the greens without spoiling another color.

I've got a lot that I would like to print if only I could get this pack right! Am I being too fussy?

Photo Engineer · Jun 23, 2014

Have you noticed the "bump" in magenta in the first print and the more linear gray scale in the other two? Something is going on here.

PE

Tom1956 · Jun 23, 2014

I think 1 and 3 are too magenta-y. And seems like a hair more exposure is in order.

kb3lms · Jun 24, 2014

Now that you guys say it, yes, I do see the magenta. Maybe that has something to do with walking away from it for a while.

PE, when you say a "bump in magenta" do you mean something other than tweaking the magenta a bit?

IDK why the third one would be magenta, but like you say, Tom, it is.

Other than tweaking the filter pack, what things might cause a color cast? The C-41 developer (Kodak) was basically fresh. It may have had a few rolls through it but no more than 2 or 3. The RA-4 (the Kodak RA/RT variety) was mixed a couple months ago last time I was trying to print. Wasn't used much and stored in glass. Processing is at room temperature, about 68-72 degrees F, for 2 minutes in the dev (as noted elsewhere in the forums) brief water rinse to reduce crossover and 2 minutes in the blix. At this point the liter of RA-4 developer has had about 25 4x6 size prints through it.

Sirius Glass · Jun 24, 2014

Consider magenta to be negative green.

Long time members will remember that the two letter troll insisted that magenta was not a color. How is that working for him now? Eh, PE?

Rudeofus · Jun 24, 2014

kb3lms said:
Now that you guys say it, yes, I do see the magenta. Maybe that has something to do with walking away from it for a while.

PE, when you say a "bump in magenta" do you mean something other than tweaking the magenta a bit?

IDK why the third one would be magenta, but like you say, Tom, it is.

I can't speak for PE, but looking at that bottom row, I see clean whites, clean blacks and magenta mid tones. I don't think you can fix this with filter packs, somehow your curves are bent. Since your negative scan looks fine I would suggest that something is less than perfect in your RA4 setup. Either that paper, or your chemistry, or your processing is off, or a combination of the three.

The biggest question to me is whether these small errors in color balance will be obvious in real images ...

Tom1956 · Jun 24, 2014

Rudeofus said:
I can't speak for PE, but looking at that bottom row, I see clean whites, clean blacks and magenta mid tones. I don't think you can fix this with filter packs, somehow your curves are bent. Since your negative scan looks fine I would suggest that something is less than perfect in your RA4 setup. Either that paper, or your chemistry, or your processing is off, or a combination of the three.

The biggest question to me is whether these small errors in color balance will be obvious in real images ...

I guess we have different monitors. I think the middle one is best, but needed maybe 1/2 stop more exposure at the camera, and be more accurate with chemistry temperatures.

pentaxuser · Jun 24, 2014

Rudeofus said:
The biggest question to me is whether these small errors in color balance will be obvious in real images ...

I too think that is key. I recall I was given a Fuji neg which for some reason was called "Lucky" for calibration purposes by the man from whom I bought my Jobo processor. I was also given the Fuji print which was the "match" which I had to duplicate on my print.The neg and print was a Japanese girl in a garden holding a colour chart

I could get close but never an exact duplicate of the Fuji print. Every change I made resulted in some things being closer but others further apart. I gave up in the end and while it was frustrating not to achieve a replica it really made no difference to my prints.

[Moderator's note: Reply to off-topic post deleted.]

pentaxuser

Photo Engineer · Jun 24, 2014

The simple answer to the problem is that there seems to be some crossover in the prints. When a neutral goes from magenta to neutral as you move up the scale, there is magenta crossover. (or minus green)!!!

PE

Tom1956 · Jun 24, 2014

I've never really understood "crossover", but I'm following this to get a handle on it. My simplistic cure to the OP's problem is a bit more exposure and more accurate chemistry temperature, and the filter pack will find itself. But this "crossover" bit is something that still eludes me.

kb3lms · Jun 24, 2014

Recall that these are all three from the same negative! The original scan of the RA-4 print looks less magenta on my monitor at home. It's still there but less so than it appeared at work. My monitor here at home has been calibrated.

The actual print does not show a magenta cast at all. Lighting for the shot was full sunlight from above. The shadow under the clip at the top goes a little to the bluish/purple. On the negative scan it is gray.

The magenta cast may be a red herring. Maybe edge can't reproduce that green very well?

Will try a real picture tonight from this same roll if I have time and see how it looks.

DREW WILEY · Jun 24, 2014

No, you're not being too fussy. In fact, you're so far off in the weeds that I can't tell what's going on. Improper exposure, dicey scanning, development issues???? Maybe all the above.

Photo Engineer · Jun 24, 2014

Tom, in color films, all three curves must be identical in slope and evenly spaced in speed. If they are not, they cross over each other creating odd color casts.

PE

MattKing · Jun 24, 2014

What PE said.

Think of a film that was exposed with even, consistent light.

If there is a cross-over issue, the shadows may have one colour cast, while the highlights have another.

If you correct for purple shadows, the highlights go too green.

If you correct for green highlights, the shadows go too purple.

I once worked doing for a lab doing proofs and machine enlargements for wedding and portrait and commercial photographers. One of our customers tried to save money by doing his own colour negative processing, when he shouldn't have.

Every negative had cross-over issues.

There were a lot of weirdly coloured, unhappy brides.

Tom1956 · Jun 24, 2014

MattKing said:
What PE said.

Think of a film that was exposed with even, consistent light.

If there is a cross-over issue, the shadows may have one colour cast, while the highlights have another.

If you correct for purple shadows, the highlights go too green.

If you correct for green highlights, the shadows go too purple.

I once worked doing for a lab doing proofs and machine enlargements for wedding and portrait and commercial photographers. One of our customers tried to save money by doing his own colour negative processing, when he shouldn't have.

Every negative had cross-over issues.

There were a lot of weirdly coloured, unhappy brides.

Thanks fellows. Got what you said. Just need to look in these archives to get a better grasp. Now that I got this F2S I'm happy with, I'm liable to get into color more than I have in the past. I'm seeing some 100 ft roll prices of color neg film that stack up with the monochrome pretty favorably.

kb3lms · Jun 24, 2014

I'm seeing some 100 ft roll prices of color neg film that stack up with the monochrome pretty favorably.

Umm, Ok, I'll bite. Like what and where from?

Off to tweak with the crossover.

Tom1956 · Jun 24, 2014

I don't know where I saw them. I was jumping around places from idas I was getting from the thread about Tri-X reducing prices a couple days ago and I saw some. Now I need to start all over.

kb3lms · Jun 24, 2014

I decided to see what I had to do to the filter pack to get green.

A pack of C15/M85/Y90 gets greens in the printed color chart that are right on. Now as I understand it, we are not supposed to use all 3 filters at one time. Doesn't this add 15 units of neutral density? Unfortunately reducing the magenta factor (adding magenta back in) reduces the green, as it should.

The red patch, though, now tends to orange. Pinks, magenta, lavender are very de-saturated.

Is there a method to convert the pack to remove the neutral density?

pentaxuser · Jun 25, 2014

If the original but "wrong" pack was C0/M40/Y70 and the new one(better) is C15/M85/Y90 then simply remove the 15C and the same amounts of M and Y so you get C0/M70/Y75.

I have no way of knowing this but it seems to me that your new pack with no C of M70 and Y75 might be about right or has overdone the reduction of magenta. The jump from M40 to M70 I'd have thought is quite a reduction on the magenta balance and based on red going towards orange may still be too much.

On my Durst dichroic filters an increase of 30M makes a very large difference to the magenta balance. I can detect a difference with as little as a 5M change

pentaxuser

Photo Engineer · Jun 25, 2014

I would have to agree that the filter pack is wrong, but with crossover, nothing can just quite cure the problem in the negative unless you dodge with color filters.

And don't forget that if you remove 15C the exposure will be 1/2 stop less to get the same density.

Small color shifts are often present when you use or remove cyan filtration. This is due to the fact that the cyan is the least "pure" of all of the filters. In fact, all of the filters have some impurities that overlap so that M corrects a tiny bit of Y and Y corrects a tiny bit of M and so on.

PE

kb3lms · Jun 25, 2014

Then it looks like the problem was with the original negative, yes? Can crossover be thought of as uneven development through the different layers of the color film?

What sorts of things might produce this crossover in C-41 hand processing? I can think of two off the top of my head,

1) Temperature
2) Agitation

Green sensitivity in the film is in which layers? I'm trying to reason out why this might be so I understand it, and maybe someone else will learn something, too.

Photo Engineer · Jun 25, 2014

Your suggestions are correct. It is a layer wise effect caused by some uneven condition during development. (Bleaching and Fixing can cause a vague similar effect due to retained silver). Green sensitivity is split between 2 layers and depending on film, they may not be adjacent. They form magenta dye (ooops, magenta is not a color! ).

Anyhow, there is a fast and a slow magenta layer. Some films even have a fast, medium and slow.

Every layer must have the same curve shape or you get crossover.

PE

Tom1956 · Jun 25, 2014

I wonder it there are any filter fading studies out there.

Tom1956 · Jun 25, 2014

I guess I shouldn't mention that I use a 40 year old ink book every day. Well... not every day any more, since I already know most of the formulas. But, you know.

Photo Engineer · Jun 25, 2014

Yes, there are fade tests, and yes the correction filters fade in the dark.

The thing is, they generally do not introduce crossover. That is a process fault.

PE

Muddy Greens

Moderator

Privacy & Transparency

Privacy & Transparency