C-41 lab processing - is there much variation?

[Chan Tran]

I don't get it. You're just repeating the same thing. Matt, everything you just described is what I was referring to (twice): the mismatched gamma. You already called it "crossover" earlier. It is trivial to understand, it is covered in Kodak's PDFs although it is not called "crossover" there. I still do not see the difference between color shift and color crossover. Looks like it's the same phenomena, but some call it "shift" and others prefer "crossover".

When you plot the density of each color channel it's 3 lines. If the lines don't cross then you don't have crossover but if they shift from their expected position then you have a color shift.

 

[afriman]

Without knowing what it's called, let alone having any understanding of the theory, I definitely experienced crossover when I started doing colour developing and printing as a school kid using very primitive methods of temperature control. I spent massive amounts of time, paper and chemicals trying to get the colour balance right, without being able to figure out why, just when I got rid of a particular cast in the mid-tones and highlights, a totally different cast would raise its head in the shadows (or in a section of the image containing different colors)... It was most frustrating!

 

[MattKing]

I don't get it. You're just repeating the same thing. Matt, everything you just described is what I was referring to (twice): the mismatched gamma. You already called it "crossover" earlier. It is trivial to understand, it is covered in Kodak's PDFs although it is not called "crossover" there. I still do not see the difference between color shift and color crossover. Looks like it's the same phenomena, but some call it "shift" and others prefer "crossover".

A shift is when the relative amounts of colour change - e.g. you get more blue throughout - which is easily correctible through a change of filtration at the printing stage.
A crossover is when the the amounts of colour change at different rates with different levels of illumination - which is only correctible if you apply different amounts of filtration to different parts of the image - e.g. add magenta filtration to the highlights, and green filtration to the shadows.
Process errors can also change the shapes of the three curves - a simple change in gamma won't correct the distortion of that shape, just the average slope.
Adrian is correct though, if you are experienced in manipulating (digitally) the three individual colour curves, you can correct out much of the problem. But do you want to have to do that - a reasonably automated approach doesn't easily allow for that?

 

[runswithsizzers]

[...]
in short, an out of control process is a problem if you’re going to print optically. If you’re going to scan it, as long as you know what is actually happening, it’s not so much of a problem.

Good to know, thanks!

The OP has got some very good advice already.
Control strips are a must for any lab.
I would also think of how the minilab looks, is it neat and tidy.
If you can see processing machines, do they have stains on the side (bad sign) or are they clean looking (good sign). If house keeping is poor, you can bet the processing is too.
How is their attitude to film customers? Welcoming?
And most importantly, if anything goes wrong, blame the lab (that's a joke by the way)

Thanks for your input. I will ask about the process control strips. I am reassured that my local lab meets most of your other criteria, except for the "welcoming" one. I don't know that film customers are treated any worse than the others, but some of the sales staff lose interest as soon as they decide I am not going to buy something expensive. I think I somehow got on their list of "timewasters" except for one or two employees who are usually very nice.

One practice I do not appreciate - several times they have handed me out-of-date film, at full retail price. When I notice, they act surprised, and exchange it for fresh film - but the old film goes back on the same shelf as the fresh stuff, ready for the next customer. I don't mind shooting out of date film, but I feel like they should be more up-front about it and offer a slight discount.

I worked in a hospital lab for 20-some years, so I am all too familiar with "blame the lab" - it was a popular thing to do among doctors and nurses.

 

[Chan Tran]

Good to know, thanks!


Thanks for your input. I will ask about the process control strips. I am reassured that my local lab meets most of your other criteria, except for the "welcoming" one. I don't know that film customers are treated any worse than the others, but some of the sales staff lose interest as soon as they decide I am not going to buy something expensive. I think I somehow got on their list of "timewasters" except for one or two employees who are usually very nice.

One practice I do not appreciate - several times they have handed me out-of-date film, at full retail price. When I notice, they act surprised, and exchange it for fresh film - but the old film goes back on the same shelf as the fresh stuff, ready for the next customer. I don't mind shooting out of date film, but I feel like they should be more up-front about it and offer a slight discount.

I worked in a hospital lab for 20-some years, so I am all too familiar with "blame the lab" - it was a popular thing to do among doctors and nurses.

I used to run the control strip in the 1 hr lab every day. But back at that time I had 100 rolls a day to process. Today many labs I heard couldn't get the 10 rolls for the smallest Noritsu processor and running a test strip for a few rolls is expensive.

 

[Adrian Bacon]

Adrian is correct though, if you are experienced in manipulating (digitally) the three individual colour curves, you can correct out much of the problem. But do you want to have to do that - a reasonably automated approach doesn't easily allow for that?

Ideally, no, you don't. It should be as close to correct as you can reasonably get it.

 

[AgX]

One issue to my surprise not even mentioned yet is the washing.

Applying a washless process for C-41 seems now industry standard in Europe at industrial labs.
Such is reasoned as being a means of reducing water consumption and is regarded as a major point in the sustainability of such lab.

 

[foc]

One issue to my surprise not even mentioned yet is the washing.

Applying a washless process for C-41 seems now industry standard in Europe at industrial labs.
Such is reasoned as being a means of reducing water consumption and is regarded as a major point in the sustainability of such lab.

Any of the Fuji C41 minilab processors I have seen in the last 10 - 15 years had a counter flow wash water system. They were very economical in water consumption.

 

[foc]

Good to know, thanks!


Thanks for your input. I will ask about the process control strips. I am reassured that my local lab meets most of your other criteria, except for the "welcoming" one. I don't know that film customers are treated any worse than the others, but some of the sales staff lose interest as soon as they decide I am not going to buy something expensive. I think I somehow got on their list of "timewasters" except for one or two employees who are usually very nice.

One practice I do not appreciate - several times they have handed me out-of-date film, at full retail price. When I notice, they act surprised, and exchange it for fresh film - but the old film goes back on the same shelf as the fresh stuff, ready for the next customer. I don't mind shooting out of date film, but I feel like they should be more up-front about it and offer a slight discount.

I worked in a hospital lab for 20-some years, so I am all too familiar with "blame the lab" - it was a popular thing to do among doctors and nurses.

Their attitude towards out of date film sounds like penny pinching. Why aren't they selling enough to turn over stock. The "surprise" sounds lame but well practiced. Not a nice thing. If you are happy with your processed negs then continue with them.
Do you have any other choice of labs locally?.

 

[cramej]

I am always offered the option to pick up my negatives later the same day, so the machine seems to be in constant use, rather than running periodic batches. I don't really know what kind of volume they do. I live in a city of about 150,000 population. There is only this one camera store, but of course there are also Wal-Marts, Walgreens, ect. where people can drop off negatives for processing.

When you say "...the quality will be poor" - one of my questions is, What does poor quality look like?

Bad color, incomplete blixing, underdeveloped, etc.

Having a lab for a city that size is nice. The metro area around me is ~600,000 and there is only 1 'pro' lab that runs C41 once a week, B&W probably less and shut down their E6 line. The last time I had E6 done by them, it came out with a magenta cast. They started dying a slow death about a decade ago when their costs had everyone draining their bank accounts. The last thing I took them was a 2 rolls of B&W and got machine prints. That'll be $50, please.

I have had them process at least half-a-dozen rolls in the past year or two. Two of those were out-dated film from my freezer, and may have slighty underexposed - so they were not reliable benchmarks. With fresh film, the 4x6 lab prints look fine. But if I understand correctly, these old Fuji Frontiers are pretty good at making OK prints even from sub-optimal negatives - yes? no? All of the lab-scans from my local lab, and also from Dwayne’s Photo in Parson's Kansas looked bad with excess grain / noise in the shadows.

Which highlights my problem - if I can't trust the prints because they may be hiding small errors in processing - and I can't trust the scans because they introduce a whole 'nother can of worms - then how do I tell if the processing is reliably producing accurate color?

+1

If they use control strips, you should see very little variation with C-41. They should be able to readily answer that question, and tell you how frequently they run a control strip.

Dwayne's is iffy on the scans. I live less than 2 hours from them so I tried them a couple times. Was not thrilled. North Coast Photo has been exceptional since I started using them some years ago.

Control strips are good to ask about.

 

[Mr Bill]

One issue to my surprise not even mentioned yet is the washing.

Applying a washless process for C-41 seems now industry standard in Europe at industrial labs.
Such is reasoned as being a means of reducing water consumption and is regarded as a major point in the sustainability of such lab.

Hi, I actually did make a brief mention of washing, although more as an example of something that doesn't really show up on a control strip, yet CAN affect the permanence of the film. So this is one of those little things that relies, to some extent, on the attention of the machine operator.

FWIW, when the so-called washless systems were introduced, they were mainly a way to deal with the very stringent effluent regulations just coming out. The existing large photo labs of the day could generally afford the new technologies being developed to deal with such. But the new one-hour labs springing up everywhere didn't have that luxury. Thus, the washless systems - they essentially reduced the "wash" (final rinse, or whatever they called it) to a volume that could be hauled off-site for disposal at an "affordable" price. So there was no need for a connection to the sewer system with the appropriate permit, setting specific effluent limits, monitoring schedules, and the like.

Any of the Fuji C41 minilab processors I have seen in the last 10 - 15 years had a counter flow wash water system. They were very economical in water consumption.

Yep, even the earliest one-hour labs, such as Noritsu System II and System IIIs, used countercurrent flow systems. And of course, minimizing water use was a primary driver - the more tanks the more the flow rate can be reduced. And long before one-hour labs existed, high volume finishers all used multi-stage systems. Everything is a set of tradeoffs.

 

[Mr Bill]

hrmmm… even when it’s in process, they’re not parallel. Each color channel has its own gamma. In my experience, when people refer to “crossover” or the “curves got crossed” that just means the gamma of each color isn’t what it’s expected to be relative to what RA-4 would need to actually produce the colors it should.

Hi, I've mentioned this before, but the color channels, as "seen" by the RA4 paper, ARE apparently parallel. If you've ever seen my posts on overexposing of Portra film I give a first-hand account of studio testing where we could color balance optical RA4 prints (on professional papers), matching flesh highlights to within 1cc color, and finding that we could get nearly dead on color matches from roughly 1 stop underexposed to around 3 to 4 stops overexposed. When I say dead-on, I mean that professional color correctors, viewing prints in a proper color booth, essentially cannot tell them apart. I don't believe this would be possible without parallel response curves. The reason that the plotted curves, as in film data sheets, are NOT parallel is because they are done with a certain narrow densitometer response, called Status M. For reference, Giorgianni, formerly of Kodak, explains the same thing in his book. So the densitometer curves end up not being parallel, and probably scanner curves are even different from this (unless the scanner has a Status M response).

Ps, I'm doubtful that the wide exposure range/matching color can be achieved in a Jobo processor, largely because of the sparse developer volumes. My experience has been in machines where the tank volume is large enough that it would take a considerable amount of heavily overexposed film to put even a slight blip on a control chart.

 

[Tom Kershaw]

Ps, I'm doubtful that the wide exposure range/matching color can be achieved in a Jobo processor, largely because of the sparse developer volumes. My experience has been in machines where the tank volume is large enough that it would take a considerable amount of heavily overexposed film to put even a slight blip on a control chart.

As I've mentioned on this forum previously, my experience with C-41 processing using a Jobo has lead me to be generous with solution volumes, 3 or 4 films per litre of developer and no replenishment. I do replenish the bleach as this works fine, bleach is expensive, and I don't seem to have any bleach related issues. 25ml per film, alongside plenty of aeration.

 

[Alan Edward Klein]

Bad color, incomplete blixing, underdeveloped, etc.

Having a lab for a city that size is nice. The metro area around me is ~600,000 and there is only 1 'pro' lab that runs C41 once a week, B&W probably less and shut down their E6 line. The last time I had E6 done by them, it came out with a magenta cast. They started dying a slow death about a decade ago when their costs had everyone draining their bank accounts. The last thing I took them was a 2 rolls of B&W and got machine prints. That'll be $50, please.





Dwayne's is iffy on the scans. I live less than 2 hours from them so I tried them a couple times. Was not thrilled. North Coast Photo has been exceptional since I started using them some years ago.

Control strips are good to ask about.

have you used North Coast for E6 and and BW and what has been your experience.

 

[cramej]

have you used North Coast for E6 and and BW and what has been your experience.

Yes, all of the above. No problems whatsoever. Their Noritsu scans are very nice and I've used the files for some larger prints in the past. I just started uploading some things to Flickr and everything that I have there so far has been done by NCPS. I just sent them 7 more rolls of C41, B&W and E6 so I'm sure I'll be getting some more on there soon. John Cramer | Flickr

 

[AgX]

FWIW, when the so-called washless systems were introduced, they were mainly a way to deal with the very stringent effluent regulations just coming out. The existing large photo labs of the day could generally afford the new technologies being developed to deal with such. But the new one-hour labs springing up everywhere didn't have that luxury. Thus, the washless systems - they essentially reduced the "wash" (final rinse, or whatever they called it) to a volume that could be hauled off-site for disposal at an "affordable" price. So there was no need for a connection to the sewer system with the appropriate permit, setting specific effluent limits, monitoring schedules, and the like.

Yep, even the earliest one-hour labs, such as Noritsu System II and System IIIs, used countercurrent flow systems. And of course, minimizing water use was a primary driver - the more tanks the more the flow rate can be reduced. And long before one-hour labs existed, high volume finishers all used multi-stage systems. Everything is a set of tradeoffs.

Yes, but today, when it is not only about legal or technical issues, big players need to show their committment to sustainability. So a washless process is presented in this context.

 

[Mr Bill]

As I've mentioned on this forum previously, my experience with C-41 processing using a Jobo has lead me to be generous with solution volumes, 3 or 4 films per litre of developer and no replenishment.

Yep, I do recall your experiments, which I think were significant.

But on top of those, here's the issue with overexposed film. From memory, a film like Portra has a density change of somewhere around 0.20 per one f-stop exposure shift. So an exposure increase of, say 3 f-stops, should increase film density by about 0.60, which translates to about 4X the dye formation, and thus about 4X the "demand" on the developer. So if one tried to process such overexposed film, using their standard routine in a Jobo, I'd expect significant deficiencies in the processed film. Whereas in a larger processor using replenishment, the heavily overexposed film would be fine. So what I'm trying to say is, someone trying to test the overexposure capability of Portra film by using a Jobo processor would probably NOT get good results.

 

[Mr Bill]

Yes, but today, when it is not only about legal or technical issues, big players need to show their committment to sustainability. So a washless process is presented in this context.

I understand. But I see it differently. (As an aside, when washless systems were first introduced, it was an inside joke that the way "pollution" was being handled was to have the customer take it home with them.) On a serious note I would say that it's possible to use plain wash water and still achieve wash rates as low as a washless system. The way you would do this is via more countercurrent flow tanks at a very low wash rate. As an example try looking up the Konica CPK-18 (as I recall) process. I doubt you'll find any actual data on the process, other than some examples in various patents. But one of their major features was the use of many wash tanks and very low flow rates. Now, the way this is environmentally better is that this concentrates the problem chemicals, silver for example, in the first tank. So it is possible to bring your recovery efforts against a relatively small volume of wash, which recovers a proportionally greater proportion. At the same time you are not introducing an additional chemical, as in a washless system, and the customer is not taking home additional residual chemicals in their processed material.

When this research was being done, mostly in the 1980s, the industry was looking at what they called "thin tanks," which would improve the concentration effect.

FWIW we never used a washless system in our main labs, and consequently never did image stability tests on such materials. But inside information from some of the manufacturers indicated that the stability was fine. The outfit were i worked did, at one time, own a large chain of one-hour labs. Sometimes we would lend technical expertise, so I know that they did run washless systems in some number of stores. It was mostly in cases where virtually NO effluent materials were allowed. (The Southwest US has some places where rivers simply come to an end, so they tend to be very restrictive.)

FWIW my department initially handled all of the effluent issues. At some point the company split off a separate department, for environmental and "industrial hygiene" and "compliance" issues (taking some of my best people, boo hoo). And this new group would oversee the effluent permits being handled by the one-hour lab division, etc., so I have some familiarity with this sort of thing.

Regards,
Bill

 

[Alan Edward Klein]

Yes, all of the above. No problems whatsoever. Their Noritsu scans are very nice and I've used the files for some larger prints in the past. I just started uploading some things to Flickr and everything that I have there so far has been done by NCPS. I just sent them 7 more rolls of C41, B&W and E6 so I'm sure I'll be getting some more on there soon. John Cramer | Flickr

Thanks. I do my own scans with a V850. What scans sizes do you get with them?

Also, How has you experience been with them with getting the development right and no dust and stuff on the negatives?

 

[cramej]

Thanks. I do my own scans with a V850. What scans sizes do you get with them?

Also, How has you experience been with them with getting the development right and no dust and stuff on the negatives?

I get the 'Enhanced Scans' that, for 6x6, are 4800x4800 approximately. If I ever need anything bigger, I send it off to be drum scanned.

Development has always been spot on and the negs are clean. I might find some tiny specks here and there, but that's about it.

 

[Adrian Bacon]

Hi, I've mentioned this before, but the color channels, as "seen" by the RA4 paper, ARE apparently parallel. If you've ever seen my posts on overexposing of Portra film I give a first-hand account of studio testing where we could color balance optical RA4 prints (on professional papers), matching flesh highlights to within 1cc color, and finding that we could get nearly dead on color matches from roughly 1 stop underexposed to around 3 to 4 stops overexposed. When I say dead-on, I mean that professional color correctors, viewing prints in a proper color booth, essentially cannot tell them apart. I don't believe this would be possible without parallel response curves. The reason that the plotted curves, as in film data sheets, are NOT parallel is because they are done with a certain narrow densitometer response, called Status M. For reference, Giorgianni, formerly of Kodak, explains the same thing in his book. So the densitometer curves end up not being parallel, and probably scanner curves are even different from this (unless the scanner has a Status M response).

Ps, I'm doubtful that the wide exposure range/matching color can be achieved in a Jobo processor, largely because of the sparse developer volumes. My experience has been in machines where the tank volume is large enough that it would take a considerable amount of heavily overexposed film to put even a slight blip on a control chart.

I have no way of knowing what RA-4 paper sees. I only know what my scanning system sees, and there, even kodak’s reference control strip presents as non parallel. Each color channel has its own gamma.

all that being said, it doesn’t really matter what the gamma of each channel is if you’re going to digitize it. All you need to know is what the gamma of each channel is going to be. If you can do that and still match the reference so that RA-4 will still print, then you have no problem.

with respects to jobo processing, yes, you need to be careful about how many rolls you process per tank, but these days that’s a bit less of a problem as most people actually under exposed their film, not over expose their film.

 

[Mr Bill]

I have no way of knowing what IRA-4 paper sees. I only know what my scanning system sees, and there, even kodak’s reference control strip presents as non parallel. Each color channel has its own gamma.

I can, more or less, understand where you're coming from. My issue is that I periodically see you saying that the color curves are not parallel, whereas this is an really an artifact of your system. In the place where these systems were designed, but specifically for Portra films/professional papers, the "color curves ARE parallel. (To be more correct i should only say the NEUTRAL color response curves.) The curves become non-parallel when the film is viewed by something other than the appropriate paper, or when processing is "out of spec," which will almost certainly happen with sparse developer usage.

If you don't want to hear me repeating this issue, all you have to do to is put a small disclaimer, "in my system," when you say the curves are NOT parallel.

FWIW I can't say anything about the "parallel or not" situation for films other than the pro portrait films of Kodak (and a couple of other makers), as I don't have (much) first-hand knowledge.

 

[Adrian Bacon]

I can, more or less, understand where you're coming from. My issue is that I periodically see you saying that the color curves are not parallel, whereas this is an really an artifact of your system. In the place where these systems were designed, but specifically for Portra films/professional papers, the "color curves ARE parallel. (To be more correct i should only say the NEUTRAL color response curves.) The curves become non-parallel when the film is viewed by something other than the appropriate paper, or when processing is "out of spec," which will almost certainly happen with sparse developer usage.

If you don't want to hear me repeating this issue, all you have to do to is put a small disclaimer, "in my system," when you say the curves are NOT parallel.

FWIW I can't say anything about the "parallel or not" situation for films other than the pro portrait films of Kodak (and a couple of other makers), as I don't have (much) first-hand knowledge.

That's totally understandable, and I have no problem with that, but to be clear, I don't generally say "with my system" simply because my system is not unique. Every system that is not RA-4 (including commonly used digital minilabs) are going to see basically what my system sees, which is that each color channel has a unique gamma. Heck, that wouldn't even matter for RA-4 if each of it's color channels matched that in opposite directions, which would result a print that visually looked correct.

With respects to other films, they all basically behave the same way with respects to the gamma of each color channel. I've characterized pretty much every color negative film that I can get my hands on and they all fall within a fairly narrow set of gammas for each color channel, which makes sense, because they have to if you only want to have one printing standard. You can't deviate much from that without totally hosing up being able to make an RA-4 print that is acceptable. That's just a fact of life with color negative film.

 

[btaylor]

Super interesting exchange, thank you @Mr Bill and @Adrian Bacon[/USE]

Obviously this luxury did not exist for transparencies, and that probably explains why they're more expensive and inferior in quality to print film.

Uh- that is a very broad statement. I would argue that the Cibachrome prints I made from Kodachrome slides were certainly NOT inferior to how prints from color negatives onto RA4 materials looked at the time. KII, super sharp, grainless- Ciba: color that makes you say “wow”

 

[btaylor]

@btaylor I won't get into comparing prints. All forms of reflected light medium are uninteresting to me. Scanned Ektar on a high quality display is what defines "wow". I hated prints even back in the 90s. Went from projectors to monitors, never bothered with dead trees.

Interesting perspective. Prints are what are valuable and meaningful to me. If you are in the digital realm then negative or slide really makes no difference because the digitization process and manipulations through software are entirely flexible. Why do you bother with film?