C41: Home processing and scanning vs lab processing and printing

This question is mostly about how labs deal with color correction and white balance when making prints.

 

So through user error, I over exposed some Fujicolor 800 by 2 to 3 stops. I processed the film at home and scanned with a Coolscan 8000. The negatives were definitely on the dark side but the scanned images were not awful, at least the outdoor ones. The flash photos were hideous though, but adjusting the white balance made them OK.

 

This is from a cheap point and shoot with built in flash. I've never used the camera before and I typically shoot film in natural light. Why I chose to use this camera is a long story covered in a different thread. Basically I'm looking for a cheap but decent P&S camera to give to a novice. They will be bringing the film to a lab, hence my question.

 

So while I can get a passible photo by making some adjustments in post, I'm curious as to what might happen if a lab had done the work instead. Do they color correct each frame before printing or is it something done for the whole roll?

 

I could see the person who ultimately gets the camera using the flash fairly often.

Color balance with electronic flash should not really be a problem since both the flash and most film are daylight balanced. 2 decades ago I sent all my color print film to local film labs and it is my understanding that each frame is analysed for exposure and color by the automated equipment. That said, I don't remember flash images to have any real problem (but for deer in the headlights look or overexposure of close-up subjects or underexposure of dark backgrounds). All my bounce flash images looked real nice though indoor available light shot under incandescent lights still showed up too as too warm for my taste, that today just moving the color temp control to the left (in Lightroom) would solve.

It would depend on the lab--some are great and others not so much. I once had a Vivitar 285 on full power manual for one shot and forgot to switch it back to auto exposure until I realized that the recycle time was way too long for the distance that I was shooting at. The resulting couple of machine prints looked fine for both exposure and color balance despite really dense negatives.

Ken and AJG: Thanks, that's helpful.

Here in Vancouver I simply tell them how many stops over (or under) I exposed at and they typically did a good job of getting

what they could from the negatives . (ABC Photo). Peter

Tradition is that color negative films do fine with 2 or 3 stops over exposure.

It does make them harder to print or scan.

 

It is usual to balance so that the scene averages to white.

Often enough this works, even when much of the scene is one

color, such as grass.

I'm curious as to what might happen if a lab had done the work instead. Do they color correct each frame before printing or is it something done for the whole roll?

 

As some have suggested this is generally done with an automated system nowadays, at least for the most economical photofinishing. Fwiw the printing is almost invariably from a scanned image so just about anything can be done in the correction process. If you were to go to more expensive finishing, the sort that used to be called a "pro lab," you could have skilled humans hand-adjusting each image; this would typically give a significantly better overall quality, but at much higher cost.

 

It is usual to balance so that the scene averages to white.

Often enough this works, even when much of the scene is one

color, such as grass.

 

Glen, that used to be the standard method, known as "integrate to gray," until roughly around 1990ish. It worked somewhat decently, at least for amateur work, but DID have big problems when the scene was primarily one color. A worst sort of situation would be a person photographed on a red background. The system sees a lot of "red" and tries to cancel it out which makes other things, such as the skin tones, go cyan-blue (this is the worst thing you can do with skin tones). In the original one-hour lab world this took a skilled operator observing each negative at the printer gate, then punching in color overrides on a keyboard. This would greatly improve the results, but still not adequate for professional work.

 

Agfa made a breakthrough with their MSC family of mini-labs which did a low resolution scan of the roll, and for the first time such a system could make "intelligent" adjustments to the color. Roughly around the same time the APS photo system (yeah, I know it's redundant) came on the scene, introducing low-resolution "index prints" to the amateur processing world. Once this ability to make "intelligent" judgment of a scene became viable, the older "integrate to gray" method pretty much fell out of favor.

Somewhat related, here is a comparison I'd been thinking of posting, but as it was more of a rant than anything else, I hadn't done so.

 

Here are two crops, the film is Ilford XP2, shot sometime last year and processed by a local minilab, on, I think, an Agfa MSC machine.

 

I asked them to just process and scan, naively assuming that the scans I would receive were those straight from the minilab, as it happens, they charged me €6 for the scans as 'they had to do a lot of work'.

 

 

I really can't begin to understand how they did such a bad job, the images they gave me are ~2MP at best, heavily compressed and just plain bad. Did they use the low res 'Index scan' that Bill talks about above?

 

For comparison, here is a scan from a 3x5 wet print, on 20 year old paper, scanned at 600dpi with my cheap flatbed. It's not a good print, I was struggling to get any sort of contrast at all from the paper. Scan was on auto, with no corrections and looks considerably better than the print.

 

 

Please ignore the scratches and marks, it was my first time in the darkroom in over 15 years and I was really struggling with the tiny paper size, not to mention the tiny size of my darkroom/bathroom. A 'serious' attempt at printing will occur soon with some fresh film and paper..

 

To my eyes, the skin tones & texture are clearly better in my print, it's far more obvious on the face, but I don't have a model release, so you'll just have to imagine, the child fills about 1/4-1/3 of the frame, these crops are both 1:1 from the respective images.

 

Needless to say, I won't be going back there, though I would have been curious to see the original scans before they ran them through Photoshop, firstly I've found a much better 'Professional' lab that is actually cheaper (also where I buy my chemistry) and secondly I'm now setup to process my own film again.

 

So, in short, minilab quality depends greatly on the operators and it's somewhat unlikely that the current operators are those who were originally trained on the machine - I think the girl who served me was younger than the minilab...

I really can't begin to understand how they did such a bad job, the images they gave me are ~2MP at best, heavily compressed and just plain bad. Did they use the low res 'Index scan' that Bill talks about above?

 

Hi, no, the sort of scans I'm talking about would be more on the order of 2 MP for the entire combined roll. I really don't remember the details but it was probably something on the order of 200 x 300 pixels, maybe as high as NTSC television of the era. Those machines were still making optical prints. High-quality digital exposures on an affordable mini-lab of that era would have been a pipe dream.

Even the same lab will return variable results IME. Even if that lab calls itself 'professional' and charges accordingly.

 

Most minilabs are very variable. Pro labs occasionally variable.

 

I believe the latest minilab printers give the operator an electronically generated preview on a monitor. That's if the lab has invested in such machines, and if the operator can be bothered to look. It's all down to process control and human skill/integrity.

Edited September 25, 2020 by rodeo_joe|1

Professional processing labs will specifically require input from the photographer as to the metrics of the finished result — this has always been the case, especially when negative film ruled the streets for weddings, commercial, portrait, advertising and sport. For them, it is not a case of just dip the film in and process to standard parameters — there will be specifics to follow, for under or over-exposure, push or pull, developer type, temps, proofing etc... But for High Street consumer labs, it is pretty much true you get what you pay for: a fast, automated result with default corrections and a default colour space from the machine, unless you've said 'No Corrections'. Much can be touched up at home by a knowledgeable user in Fauxtoshoppe or Lightroom, but with negative film, a working knowledge of profiling and colour spaces for that type of film is pretty much essential, moreso if you are prepping for print rather than just web viewing.

 

Scanning of negative film frequently causes flare-ups on fora with for and opposing views (I ignore such things). Pointedly, there is no one-size-fits-all approach to scanning one or any work — it must be tailored, and negative film especially requires specialist care and in many instances, combometric profiling (two profiles combined e.g.from the ICC) to get the best results. I would not expect this professional level of processing and scanning service to be cheap.

 

Off the top of my head, here in Australia there are no professionals actively using C41/neg stock or scanning it for production, but there is a substantial legion still using E6 over multiple formats with lab processing, chiefly MF and LF — very little in 35mm, and just about all of that E6 processing and scanning is print-destined (giclée or hybridised combometric RA-4). As standard practice, exposure is finished in-camera — there is no correction at processing, and profiles/colour spaces and metrics are applied per spec at the lab — quick and cheerful!

 

 

[uSER=2403817]@rodeo_joe|1[/uSER]

Most minilabs are very variable. Pro labs occasionally variable.

 

True, mini labs are not consistent, I can say that much. Pro labs...the onus is on you to explain how things will be (as discussed above).

 

I believe the latest minilab printers give the operator an electronically generated preview on a monitor.

 

Yes, usually a very small, uncalibrated, unprofiled monitor which does not differentiate palettes or tones. Their business is about speedy throughput and money in the tank. They're dealing with mums and dads and students just looking for a record, as we all did eons ago.

Pro labs...the onus is on you to explain how things will be (as discussed above).

Apart from push/pull, I have never had a pro lab request information on how I like my film baked, or my prints grilled. And occasionally, even the bog-standard processing that I expected has been screwed up. Like un-requested over-developed C-41 that I once received - something that I would never specify.

I would have the lab processes the film in C41 process and then scan or make my own RA-4 prints myself. I can't beat either the price or consistency of the lab processing C41 at home.

Apart from push/pull, I have never had a pro lab request information on how I like my film baked, or my prints grilled. And occasionally, even the bog-standard processing that I expected has been screwed up. Like un-requested over-developed C-41 that I once received - something that I would never specify.

 

Things must be different then in the USofA! We've been filling out job sheets at labs since at least 1994.

Really though, pro labs here in AU aren't occupied with negative processing now, but lots of E6, scanning and digital-to-print production. Lots of Chemists accept negative film from mums and dads for processing, and that's as much as they need or want. I never bothered with negative film — coming, as I did decades ago as a pictorial editor, negs were far too time wasting to examine on the light table compared to slides.

Things must be different then in the USofA!

Except I'm not in the USofA!

 

All I can add is that my experience with all grades and prices of lab has been disappointing. And unlike BeBu, I took to processing my own C-41 and getting far more consistent results using a simple Jobo rotary processor than I did from most 'labs' (a misnomer, since calling them that implies some degree of scientific precision). It was also slightly cheaper in chemicals than paying 'professional' processing prices, although not in time.

Edited October 15, 2020 by rodeo_joe|1

It was also slightly cheaper in chemicals than paying 'professional' processing prices, although not in time.

 

Professional facilities will give quantity discounts for processing, scanning, proofing, printing: the more you do, the cheaper it gets. Yes, I understand this won't sound good to everybody, but it is an economy well worth supporting, and for me, ,it is one that frees me up to increase my productivity. We don't all have time or desire for home processing (I'm actually painting my house and sleeping in a tent in the backyard!) — time that can definitely be put to good use in travelling about in the endless pursuit of the production of images. Maybe I will retire in 2021, which appears to be just around the corner...

(snip)

 

Glen, that used to be the standard method, known as "integrate to gray," until roughly around 1990ish. It worked somewhat decently, at least for amateur work, but DID have big problems when the scene was primarily one color. A worst sort of situation would be a person photographed on a red background. The system sees a lot of "red" and tries to cancel it out which makes other things, such as the skin tones, go cyan-blue (this is the worst thing you can do with skin tones). In the original one-hour lab world this took a skilled operator observing each negative at the printer gate, then punching in color overrides on a keyboard. This would greatly improve the results, but still not adequate for professional work.

 

(snip)

 

Yes before about 1990 might be when I last thought about it.

 

Starting in 1991, I did a lot of family pictures on VPS, and mostly sent to Qualex labs.

I don't remember any with color balance not close enough for usual family shots on 4x6 prints.

 

I only remember once wondering about it, on a scene that was mostly grass (green),

and that it didn't try to fix that. But as you note, that doesn't make green skin.