Interpreting C-41 control strip results

I finally acquired a color densitometer and scratching my head looking at my results. I'm using freshly mixed Flexicolor chemistry. The reference strip is above, my readings are below:


So it looks like my B values are pretty close, but R and G are too low. Fuji has a document called C-41 technical bulletin, their version of Kodak's Z-131. When I go through the troubleshooting/correcting actions table on page 28, I do not see a use case like mine.

Low/high temperature or low/high developer activity lead to the Blue value deviating first, which is the opposite of what I'm seeing.
Any ideas what's going on here?

In case this helps, I'm posting a photo of the control strips on a light table.
Fuji reference strip is on the right. The one I developed is on the left.

The orange mask on the reference strip visibly has more red and less green in it. When I white-balance on the transparent area of the reference strip in a RAW converter, it reads 150/150/150, my strip reads 145/161/150.

I was digging through process control documentation (Kodak and Fuji) all day, and there's nothing even close to my situation, where all of my variables are within spec, except the LD value of the green layer. I have reformatted the tables above to make it more obvious (included the delta vs reference in bold in the bottom table):



How it is possible to be a whopping -0.10 below target on the middle emulsion layer?! My blue is pretty close, and the red isn't bad, yet the green in the middle is fucked up.

FYI there was another recent instance of seemingly inexplicable C41 densities (although of a different nature and more problematic) with Flexicolor chemistry. I don't know what the subtle differences are between Flexicolor C41 and Fuji's color negative chemistry. It's conceivable they produce slightly different densities, especially since exact matching of C41 with other analog processes isn't required anymore in today's marketplace. A slightly different density and different gamma in a single channel can be corrected for in digital post, after all.

PS: when I was at Fuji, they emphasized process parameters and chemistry as the main challenges in getting decent consistency in color processes. I think it's pretty hard to achieve the kind of consistency of a high volume operation with stable processing volumes, replenishment regimes, chemistry monitoring, process control etc.

Are you running a replenished system or one shot?
How do maintain developer temperature?
Do you keep the box of test strips refrigerated?

Sorry for all the questions but sometimes small fluctuations in readings can send you off on a wild goose chase. Better to wait till you have a few readings and draw a graph (in coloured pens). It may make it easier to access.

A quick story. Back in the 1990s I used to send out c41 control strips to Fuji for reading (I had a minilab). One week the guy doing the reading would suggest 10% reduction in developer rep rate. The next week he would suggest an increase of 10% in the developer rep rate. This went on for a month. It was only when we drew a graph that he saw the silliness of his comments.

Steven, I assume you're using 'status M' filters on your densitometer ?

John S

Since everything seems to be on the lower side, I'd start with increasing the developer concentration a bit.

Otherwise, reading the Kodak Z-131, middle layer (G) going more astray than B and R in LD parameter doesn't seem like a thing that couldn't happen.


Sure, it would be ideal if your Dmin measured parameter would track well corresponding to the Chart 11 (Y and HD-LD seem to track well), but it may not be that you need to adjust only one of your developing parameters to get all the parameters within spec...

Please note that this is coming from a guy that never tested his C-41 processing with control strips, so...

But, this seem like fun. Anyone know if C-41 control strips are available in EU in home user friendly quantities?

Thank you everyone who responded!

John Salim said:

Steven, I assume you're using 'status M' filters on your densitometer ?

Click to expand...

John, yes. Although my model doesn't have a switch for this, X-Rite 880 is an automatic strip reader: I select CN-16 so it knows the location of patches on the strip, then the machine ingests the strip and gives me the reading. Mine was refurbished and calibrated last week by http://www.acurad.net/

foc said:

Are you running a replenished system or one shot?
How do maintain developer temperature?
Do you keep the box of test strips refrigerated?

Sorry for all the questions but sometimes small fluctuations in readings can send you off on a wild goose chase. Better to wait till you have a few readings and draw a graph (in coloured pens). It may make it easier to access.

Click to expand...

One shot chemistry. Temperature is maintained by a rotary processor with a water bath, I tested my process with water and it's within ±0.3F inside the tank. The strips just arrived in the mail from uniquephoto.com, the day before. They're in the freezer now.

While I was waiting for your responses, I have re-bleached and re-fixed the strip but that did nothing, just lowered all densities by additional -0.01 or so.

My current working theory is that maybe I made a small mistake measuring the developer+starter+water combination for the 500ml run. I am using the LORR developer replenisher which I mix all 5L at once. Then for 500ml working solution I use 382ml replenisher + 15ml starter + 103ml water and maybe was a bit off.

The 2nd thing I'm going to try is to raise the temperature by 1F, just to see what happens. I've spent hours doing test runs with water after getting the film processor to dial in the inputs, but will play with temperature again anyway.

And my final idea is to investigate how to run strips in a JOBO. In this case I had a 1540 tank loaded with two 120 rolls and the control strip, running with 500ml of chemicals. I wonder if running the strip alone would have been better, but then... what's the point?

TLDR: getting a densitometer + control strips is a nightmare for perfectionists. All my photography is now put on hold until I solve this.

Thanks for the detailed response.

Not to be negative (pun intended) but control stirps work best with a replenished system IMO.

The positive is that you appear to be doing everything by the book.

If you are mixing the developer replenisher to 5Lt at once, then why not mix it to a working solution (rep + starter+water)? Because of the higher volume, you may find that the starter measurements are easier.

Replenisher and working solution will have the same shelf life.

It is good to be a perfectionist but don't be obsessive and over think it.

foc said:

Replenisher and working solution will have the same shelf life.

Click to expand...

That's a great insight, thank you! Will definitely do this with my next batch. Adding starter to the current one will be tricky because I don't know how much replenisher I have in the wine bag.

I get your point regarding the replenishment. My primary motivation behind getting control strips was: confirming my temperature settings, detecting early signs of developer oxidation, but most importantly - experimentation. Once I start getting results within spec, I plan on running a few experiments comparing:

• Bleach+fixer chemistry vs blix-based C-41 kits
• Champion C-41 chemicals vs Flexicolor
• Fuji Hunt CN-16 chemicals vs Flexicolor
• Stated capacity of C-41 kits. Cinestill claims 16 rolls per 1L with time compensation, I want to see if that's possible

But first, I need to produce a strip within spec...

Quick update: I have played with temperature some more. By raising it by 1F I was able to land within spec, but the green and blue continue to maintain a solid density gap vs red:


And the orange mask in my strips is never as pink-ish as the control strip. The LD numbers describe the difference perfectly. At this point I am giving up, and I am questioning the validity of my approach of using Fuji control strips with Kodak chemicals.

Steven Lee said:

Quick update: I have played with temperature some more. By raising it by 1F I was able to land within spec, but the green and blue continue to maintain a solid density gap vs red:

View attachment 336109
And the orange mask in my strips is never as pink-ish as the control strip. The LD numbers describe the difference perfectly. At this point I am giving up, and I am questioning the validity of my approach of using Fuji control strips with Kodak chemicals.

Click to expand...

Pardon me, it's an honest question, are these densities supposed to be equal across all channels? Because every characteristic curve I have seen has this mismatch between channels. If these patches are exposed to white light, they seem reasonable.

Anon Ymous said:

Pardon me, it's an honest question, are these densities supposed to be equal across all channels? Because every characteristic curve I have seen has this mismatch between channels. If these patches are exposed to white light, they seem reasonable.

Click to expand...

Not equal, but match the reference strip. For example, you measure LD values for RGB using your control strips, and subtract the corresponding LD values from the reference strip. Ideally the R/G/B difference should be 0/0/0.

In my case, no matter the temperature or amount of developer starter, there's always the gap between green densities and blue densities. So the RGB deltas I get always look like 0/-3/+3. Red is always in the middle, the greens are suppressed and the blues are boosted.

Well... 8 test runs later, I am now convinced that nothing I can do to bring the green and blue LD and D-Min values closer together. I have tried varying temperature, amount of starter, wet warmup vs dry, rotary vs inversions. Red, green and blue densities always respond in unison to those changes and don't diverge relative to each other. My green is consistently suppressed vs blue.

Z-131 mentions diverging densities just once, suggesting it's a developer mixing error (too much or not enough part A vs B vs C). That is odd because I mixed the entire 5L kit all at once. Maybe Sino Promise did not dose the parts accurately enough during manufacturing?

Anyway, in the process of chasing this, I was able to answer several questions for myself:

1. Wet pre-soak (PE style, two 101F baths) produces exactly the same result as a dry method, if all other variables are controlled.
2. Similarly, keeping the developer at exactly 100F during the run produces the same results as starting at 101F and finishing at 99F. This means the water bath is unnecessary for typical indoor / ambient temperatures. In my case it was 70F.
3. I finally understand where Kodak's strict ±0.25F temperature accuracy guidance comes from: this delta causes density numbers to move by 0.01. Given that control limits for density are under ±0.05 for D-Min, this means that in practice the accuracy should be within ±1.25F
4. Agitation method plays no role. I compared both JOBO speeds as well as inversion (4 inversions every 15 seconds) and, when controlling for all other variables, got identical control strips.
5. When a control strip is outside control limits, indicating that the process is out of control, quoting Z-131: "customer film should not be processed and developer should be replaced" - you can't tell with a naked eye! Even holding two control strips next to each other, one within limits and the other's not - they look the same. This means that online comments like "I did XYZ and it worked fine" should be taken with a grain of salt.

Back to my problem: the solution seems to be to move away from Flexicolor. Diverging densities leave me no room for even the slightest temperature deviations. Even under the optimal conditions, my D-Min values are -0.03 for green and +0.03 for blue, right on both sides of the action limits.

Thanks for updating, Steven. Your conclusions are interesting and very valuable to have access to! They also make perfect sense to me. The one question still out there is if switching to different chemistry will indeed resolve this anomaly. I hope it will, and in any case I hope you'll keep reporting on your findings.

What do you have in mind in terms of chemistry to try next?

koraks said:

What do you have in mind in terms of chemistry to try next?

Click to expand...

Fuji Hunt 5L kit. This will be a while though, as I still have plenty of Flexicolor to finish. As much as the deviations bug me, with precise temperature control I am still within action limits.

If the Fuji kit is unavailable (which happens frequently), my plan B is Champion.

At the risk of stating the obvious, could the problems be with the recently acquired densitometer, and not the developer?
Is there anyone around who could use their densitometer and read the same control strips.

MattKing said:

At the risk of stating the obvious, could the problems be with the recently acquired densitometer, and not the developer?
Is there anyone around who could use their densitometer and read the same control strips.

Click to expand...

There are several thermometers involved here There's the $300 certificate + traceable master with 0.05F accuracy, the Paterson spirit thermometer, built-in sous vide thermometer for the tempering bath, and another traceable high-accuracy digital thermo-pen with 0.1F accuracy for extra convenience. As I said, my process is accurate within 0.1F inside the tank, I had a lot of spare time during COVID. Besides, I tried raising and lowering temperature on purpose, and densities go up/down together maintaining the same blue-dominates-green spread.

The densitometer was calibrated just a week ago, and has the current certificate. But it doesn't actually matter, because I am comparing against the reference strip (which I measure using the same densitometer), not against some reference numbers.

Besides, I can see the difference simply by looking at the orange mask, that's why I attached it.

Steven Lee said:

As much as the deviations bug me, with precise temperature control I am still within action limits.

Click to expand...

Referring back to an earlier comment, I imagine this is indeed pretty annoying, because the practical implications of this minor deviation are actually marginal at best - and yet, it's there, begging to be resolved!

Good timing on this thread. I'm about to jump back into this and see if I can tighten up my own process and get my control strips in order. My biggest issue (not only one) is getting my HD up. Below is my last control strip from last year. This was Fuji developer but kodak fixer and bleach. Control strip was Fuji.

I'm going to increase my chemical flow over the film to try and increase the HD. Thanks for sharing your results and efforts.





-Eric

Interesting... you're also seeing the pattern of the weaker green. TBH I am puzzled by this, but I am no chemist. How is it possible for the middle layer to be the most problematic?

Steven Lee said:

In case this helps, I'm posting a photo of the control strips on a light table.
Fuji reference strip is on the right. The one I developed is on the left.

The orange mask on the reference strip visibly has more red and less green in it. When I white-balance on the transparent area of the reference strip in a RAW converter, it reads 150/150/150, my strip reads 145/161/150.

Click to expand...

Steven, the two strips look pretty close to my eyes. Maybe they aren't as close as I think they are and the differences are clearer to others and are enough to give problems with the finished article?

I do for instance have much greater problems with spotting the likes of colour crossover

By way of an analogy, I, on several occasions went through the LICI 3000 colour analyser tests for getting to the correct grey reading but never quite managed the ideal target figures that LICI stated Despite that, my prints looked OK or so I thought and unless the viewers of the prints were just being polite so did they or so I thought

pentaxuser

@pentaxuser You can use color picker to sample the mask to see what I am talking about. In terms of practical implications of staying within control limits, I can only speak to the hybrid workflow. I do not wet print, but it should be obvious that correcting color problems stemming from development errors is easier digitally. However, even when working with CN scans, color balancing is just easier when film is developed perfectly. Tools like Negmaster give me a better starting point.

I also find the green/magenta shifts to be trickier to correct than cyan/red or yellow/blue for some reason.

pentaxuser said:

Despite that, my prints looked OK

Click to expand...

Getting OK results has always been easy with negative films, both color and B&W. But I always wonder (especially with the lucky shots) what is the difference between my OK result and the theoretically best possible outcome? Too frequently I find an old OK scan from years ago which I can significantly improve with my more modern equipment and experience. This is both fun and a reminder to question OK results.

I've gotten into control strips because it allows me to narrow the search for improvement. If I can't get good color out of a negative, I can eliminate development from the list of suspects and focus on my scanning, or just admit that the light was bad.

Thanks for the reply. What's colour picker?

pentaxuser

pentaxuser said:

Thanks for the reply. What's colour picker?

Click to expand...

A tool in Photoshop or similar image editors that allows you to examine RGB values of an image area.