Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Photolab Silver Halide Color Management  (Read 3281 times)
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1311


WWW
« on: January 08, 2010, 01:18:08 PM »
ReplyReply

This thread concerns the color management of silver halide machines found in photolabs and is a response to questions raised in a previous thread: http://luminous-landscape.com/forum/index....showtopic=40660

I've been managing color on Lightjet, Durst, ZBE Chromira, Agfa, Kodak and Noritsu machines since the mid-nineties. As a color management consultant with a background in photo lab work, this is close to my heart and an area of specialty for me.  I've traveled across several continents consulting with some of the most demanding photo labs in the industry. I've also enjoyed a close relationship with Noritsu for over 10 years and have lectured, consulted and trained with them internally. I'm privy to a number of color handling issues that their techs aren't familiar with. I've worked with the guys in Koki (Japan) and Buena Park (USA) and know know exactly how they make their profiles and understand the challenges with communicating across cultural boundaries. So perhaps I'm in a decent position to say a few things here.

Quote
If you want a fully color managed workflow I don't think any silver halide process will provide that.
Safe to say I've implemented fully color managed workflows with every silver halide machine on the market. It's totally doable and has been for many years, but unfortunately some exceptional know-how can be required. Some lab owners don't have the patience to get there and settle with "good enough".

Quote
having tried soft proofing with our Lightjet & Noritsu printers, it just doesn't reflect what the finished print looks like.
Getting an excellent screen to print match is harder than just buying a calibration package and hitting the calibrate button. Understanding which calibration package to get, the intricacies of the process and grasping the role of high quality lighting all go into this. High quality lighting is the most commonly overlooked aspect, IMO. Some brands of profiles really do allow for better print to screen matching than others. One of my long-time clients (15 years), digitalprolab.com is bold enough to have a print to screen match guarantee and supports it with on-location lighting and display calibration training for their customers.

As for profiling silver halide machines there are several unique challenges that aren't found in other printing processes.  Achieving optimal linearity prior to profiling, black point handling (which varies per profile creation software), front end integration (with products with Labtricity and DP2), machine color handling variations (different model Noritsu's manage color differently for example), and process control are all issues to understand.

In addition, some silver halide machines implement a Grey Component Removal (GCR)-like technology in their RGB profiles, believe it or not!! To elaborate a little; if you expose your red laser at it's max calibrated level you get at great red without light bleeding. If you expose maximum RGB light to achieve max black there is so much light that it bleeds which results in color fringing on prints (especially with less expensive thinner papers). For this reason some manufacturers use a RGB version of GCR technology to cut back on light exposure in the shadows, but not for solid colors. This one issue has been my biggest challenge for many years now as there aren't any profiling packages on the market that will implement this (currently). I use a technique on Noritsus where I insert my own profile and let Noritsu apply the RGB CGR on top of it.

Let's look at achieving optimal linearity prior to profiling. One would think that you could just perform the printer's on board calibration (linearization) process and then print your profiling target. Some machines actually have several ways of performing a calibration and some are better than others. On some machines if you perform this process several times in a row you'll get a better calibration than it you had run it only once. These systems create curves to neutralize the grayscale and while some processes simply tweak the previously made curves, others start from scratch to make a fresh set of curves. Understanding these differences can make the difference between being able to make gorgeous, perfectly natural looking B&W prints and making disappointing ones. It's essential that one get an excellent calibration prior to the printing of profiling targets. If an iterative profiling process is implemented it's important to work quickly before the machine starts to drift.

Some silver halide systems calibrate for maximum black density sacrificing black neutrality, while others prioritize black neutrality over density. DIfferent profiling packages handle what I call black point handling, differently so one's choice of profiling software can produce surprisingly different results in the shadows and blacks. ProfileMakerPro and EyeOneMatch use a relative black approach which doesn't attempt to neutralize the blacks at all, which can be a problem for silver halide processes that have a colorful DMax. MonacoProfiler (MP) will attempt to neutralize the blacks, sometimes sacrificing density. MonacoProfiler can only go so far with this black correction so sometimes profile editing (as much as I hate it) can be in order. There is new profiling tech that isn't to market yet that will handle this better and help avoid editing (thank goodness). For now, Monaco Profiler's RGB profiling process the real champion for silver halide profiling. Their perceptual rending is particularly notable as it renders better skin tones than PMP/EOM's pink skin tones, and the gray balance is a tad warmer which I found to be universally preferable. Monaco Profiler lets you choose between a 1 step or iterative 2 step profiling process and it's important to know when which process should be used.

As for front end integration, this is where the color management handling comes in. Most of these systems rely on front end software like DP2, Labtricity, Darkroom Pro, or their own custom software to perform color management. While some newer Noritsu systems will do this internally, older Noritsu's won't. Putting a front end between your customers files and your silver halide machines has, for a long time, been the answer to implementing color management in a photolab. Having your customer convert their images to print space profiles in Photoshop before sending files to the lab is a lousy workaround that's loaded with problems! So, it's important to be familiar with these front end solutions and know when to recommend one over another.

Understanding machine color handling variations is another issue to look at. Noritsu's have historically have two modes: QSS mode and Net Order mode. The QSS mode allows operators to stick a memory card or CD directly into the machine and make prints. While this process wasn't color managed with ICC profiles it did implement some pretty impressive color corrections to the image with served the same purpose and worked surprisingly well. Unfortunately, checkboxes like "apply digital camera correction" often confused operators and the final results weren't always optimal. NetOrder mode is the mode one puts the machine in when a front end package is sending files to it. NetOrder mode doesn't alter the color at all as long as the channel corrections are zeroed out. NetOrder mode turns off color management and color correction so that the front end package (DP2, Labtricity, etc) can handle the color space transformations. Todays Noritsu's have a new front end called EZ Controller that has it's own ways of handling color management that I won't go into here. Newer Noritsus can also make their own ICC profiles onsite by printing out color patches and reading them. They are pretty complex and extremely smooth profiles that include this RGB CGR tech. Although the tech usually does this, anyone can do this. They also have a second set of files that are called "profiles" that shouldn't be confused with ICC profiles. Their real ICC color profiles don't use the ICC or ICM extension but my NDA prevents me from saying more about them or where they are located. Noritsu understandably doesn't want their customers hacking around and calling them for help when they mess up their system. If you haven't worked with a newer Noritsu, you'd be impressed with how much work they've done recently. They have two way communication with 3rd party front end software that tells it what paper is currently loaded in the printer so that the software can select the appropriate profile. Noritsu has some "big deal" stuff in development right now that's going to make lab-wide color management a lot easier for a lot of labs soon. And FWIW, Fuji minilabs are now just repackaged Noritsu machines.

Process control is the last aspect to address. Silver halide machines are subject to drift due to temperature, paper and chemical variation. All of these systems have a built in calibration process that corrects for this drift but not everyone is good about knowing how often it needs to be performed. One simple process control method is to print an evaluation print immediately after profiling. I encourage labs to store this "reference print" for future comparison. If a new operator prints this same evaluation image at the interval of his or her choice, s/he can can compare it to the reference print and get a feel for that device's drift characteristics and timeline. With this knowledge s/he can determine how often they need to calibrate the machine. With process control one can profile once, and calibrate often to maintain consistency. For those that don't like the thought of process control, calibrating silver halide devices every 8 hours is a decent rule of thumb.

Anyway, those are some of the issues I've learned to address when color managing silver halide machines. As you can tell the "color management landscape" is complex here and it's understandable why a number of labs don't have the patience or expertise to tackle these challenges. It can be done though, and the results and print to screen matching can be fantastic. Labs can have dozens of different silver halide machines printing to different papers and color management can be implemented to manage it all automatically. Inkjet's can also be made to match the silver halide gamuts when needed.

Knowledgeable color managed labs can accept files in any color space and will handle the print space conversation at the lab. There is a growing number of labs that are at this level and the knowledge is there to help the rest, as long as their customers request/demand it. Thanks for reading - cheers.
Logged

Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2808



WWW
« Reply #1 on: January 08, 2010, 01:41:20 PM »
ReplyReply

Quote from: Onsight
I've been managing color on Lightjet, Durst, ZBE Chromira, Agfa, Kodak and Noritsu machines since the mid-nineties. As a color management consultant with a background in photo lab work, this is close to my heart and an area of specialty for me.  I've traveled across several continents consulting with some of the most demanding photo labs in the industry. I've also enjoyed a close relationship with Noritsu for over 10 years and have lectured, consulted and trained with them internally. I'm privy to a number of color handling issues that their techs aren't familiar with. I've worked with the guys in Koki (Japan) and Buena Park (USA) and know know exactly how they make their profiles and understand the challenges with communicating across cultural boundaries. So perhaps I'm in a decent position to say a few things here.
Interesting.  We have specifically asked for help in this area to Noritsu on numerous occasions, including while meeting with engineers both in Koki Japan as well as several times when they sent teams to our locations, and never have they mentioned any solutions or offered any indication as to a resource such as yourself.

I would be curious to see some type of actual description on how you do this, and wonder why Noritsu doesn't offer some documentation prepared by you .. it would be most helpful.  As I mentioned I gave up after banging my head against the wall for years.
Logged

Scott Martin
Sr. Member
****
Offline Offline

Posts: 1311


WWW
« Reply #2 on: January 08, 2010, 02:00:39 PM »
ReplyReply

Quote from: Wayne Fox
Interesting.  We have specifically asked for help in this area to Noritsu on numerous occasions, including while meeting with engineers both in Koki Japan as well as several times when they sent teams to our locations, and never have they mentioned any solutions or offered any indication as to a resource such as yourself. I would be curious to see some type of actual description on how you do this, and wonder why Noritsu doesn't offer some documentation prepared by you .. it would be most helpful.  As I mentioned I gave up after banging my head against the wall for years.
While the Enterprise level sales and tech guys recommend me to their customers quite often, some of the regional guys aren't as familiar with me. Mike Allen, Steve Cushing, Dave Edwards, Jon Layden all know me well - if any of them are your contacts feel free to bring it up. It is a big company and sometimes the right connections don't get made. You may have seen me speak on color management for them at their private mini trade shows. I did write a color management white paper on their EZ Lab product a while back...

Color Management is a challenging topic for a company like them. While they want to offer consulting for their customers it's hard to build the expertise, and if they cross the line and talk about color management on a lab-wide basis (which is what's really important) then they step outside of the company, tread on other's products and bring liability issues.
Logged

Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1142



WWW
« Reply #3 on: January 09, 2010, 12:42:23 PM »
ReplyReply

Scott's inside info certainly confirms why Wayne was having so much trouble. And pretty much explains my experience I've been having with Noritsu's in central Texas using DryCreekPhoto profiles from their database. The GCR and saturated primary mismatch behavior explains a lot why Soft Proofing wasn't working for me.

Just wonder how far Scott's NDA draws the line on how much he can divulge. Why now?

I mean Scott's been contributing on the web on the subject of color management for years over at Adobe forums and here for about as long as Wayne as been banging his head on the issue of silver halide. How and why he's now privy to all this information and giving it away freely here when he's made it known to one of my clients I sent to him that he commands a pretty penny for such information.

This is why I don't trust anyone in this biz anymore and why I got out just as Wayne except I had to rely on far less resources.

Why don't you start a digital imaging school, Scott, so I can get financial aid?
Logged
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1311


WWW
« Reply #4 on: January 09, 2010, 03:23:33 PM »
ReplyReply

Quote from: tlooknbill
Scott's inside info certainly confirms why Wayne was having so much trouble. And pretty much explains my experience I've been having with Noritsu's in central Texas using DryCreekPhoto profiles from their database.
Try sending working space images to a fully color managed lab like digitalprolab.com. No need to mess around with converting your images prior to sending. Have a chat with the owner Phil is you like - he's brilliant. Again, IMO, converting images to print profiles prior to delivery to the lab is full of problems.

Quote from: tlooknbill
Just wonder how far Scott's NDA draws the line on how much he can divulge. Why now? I mean Scott's been contributing on the web on the subject of color management for years over at Adobe forums and here for about as long as Wayne as been banging his head on the issue of silver halide.
I'm usually traveling like crazy and rarely have as much time to even look at forums like I've had this past week. A planned break in my work schedule is what's allowed me to make this post this week. I'm not trying to hold back info or anything.

Quote from: tlooknbill
How and why he's now privy to all this information and giving it away freely here when he's made it known to one of my clients I sent to him...
The information I've given here doesn't violate my NDA. I haven't violated my NDA with any of my clients either. I have permission to do certain things (like replacing Noritsu's profiles with my custom ones) on clients machines but I can't tell them how they can do it. I'm not making or breaking the rules here.

As for decimating information, one of my goals is to find the time to put more information on my website and newsletter to get that information out there and hopefully cut down on the heavy amount of correspondence I deal with today. If 100 people each want to spend a few hours talking me about these things I have to charge for my time. If I spend one or two hours putting it online, then everyone can read at their leisure. Please don't make me out to be the bad guy here - I'm just trying to spread the word. I'm sorry if I didn't make this post years ago.

Quote from: tlooknbill
Why don't you start a digital imaging school, Scott, so I can get financial aid?
I been teaching workshops for over 15 years, some of which (like those through TPS) are extremely affordable. You're always welcome.
Logged

Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1142



WWW
« Reply #5 on: January 10, 2010, 10:34:36 AM »
ReplyReply

I'm not making you out to be the bad guy.

Just trying to distinguish between those who truly want to help folks in this industry from the social marketers mining for those frustrated enough with this technology to buy their way out of the confusion like my friend Mike Wilson over at "River's Edge Gallery".

Never saw so many with their hand out (excluding you of course) wanting to solve a problem that shouldn't have existed after signing the contracts with the "business machine" vendors.
« Last Edit: January 10, 2010, 10:46:09 AM by tlooknbill » Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad