i1 profiling blues

Ive been involved in the study of color management since March of this year endeavoring to improve my image output through better equipment and understanding of operation. Ive visited several forums and read a lot of posts dealing with the subject. However, I cant seem to produce a profile that exceeds the canned profiles that Epson and Ilford offer with their products (with one exception).

I purchased i1 Photo in late March and begun printing targets and building profiles. The first profile I made for Ilford Classic Pearl indicated that it was bad. The Epson papers that I profiled didnt offer any better results than what was provided with the printer. I profiled Epson HWM, Premium Lustre, and Premium Glossy photo papers.

The i1 proved to be defective and I had to return it over the summer. The lamp was bad and needed to be recalibrated. I had high hopes of getting the unit back and thought that the rectified problem would help me to produce better profiles. Thats not exactly so, however. I did create one better profile for Ilford Classic Pearlby mistake. I followed all the prescribed instructions that appear in Bruce Frasers Real World Color Management, Andrew Rodneys Color Management for Photographers, Dry Creeks specifications, and lots of other advice from forum posts and reads. The mistake I made on the Pearl was in the Photoshop CS2 Print with Preview window. In the Options> Color Handling section of the dialog box, I selected Let Printer Determine Colors instead of No Color Management. That step is contray to all instructions I’ve read. That profile has proven to be the best that Ive created with the i1 to date.

Ive been holding off on purchasing some Ilford Classic Gloss paper because Im not sure I want to fail attempting to create a better profile for another media type. Ive pretty much determined that Im sticking with Classic Pearl for anything I want to look real good printed on my Epson 1280. I would like to have a gloss paper for output requiring fine detail and I have read that Ilford Classic gloss is a good choice.

I really am discouraged that Ive spent a lot of dough on equipment that doesnt seem to be helping me. Ive killed so many trees and used so much ink printing test targets that Ive become a target of environmentalists.

If anyone responds to this post, let me say this: I have tried the 3-4 second pause on the white at the ends of scanning each row of the test target. Ive used the 918 patch target exclusively until someone suggested using the 288 patch target for the heavyweight matte paper and fine art papers. When Ive utilized both of these strategies, I still havent produced a better profile than what came with the original installation disk with the Epson 1280.

The bottom line is that Ive very frustrated that I cant seem to make any gain by following the direction that my studies have pointed me in. The more I look for answers, the more overwhelmed I seem to become when results show no improvement. I guess the reason Im preparing this lengthy dissertation is because Im hoping some expert will spot something in my report that I might be doing wrong. I sure hope so. These days, its even difficult to get a response to some of my posts.

Maybe you could list out step by step the processes your are using for 1) printing the target, 2) printing with the profile you create. Double check all steps…are you ensuring that that you are printing the TC9.18 with all color management disabled (particularly at the driver, as this is the least obvious) and with the ideal paper type selected at the driver. When printing with your profile, are you sure your using the exact driver settings you used when printing the TC9.18? Sounds like you’ve done a lot of research with all the right sources, but its sure easy to miss something along the way.

Thanks for your response Michael. The following is the step-by-step process I use for printing the targets for four different papers that I’ve used.

In the print with preview window I have more options showing and the following settings when printing the GretagMcBeth 918 patch target:

Color Management
Print> Document (Profile: RGB Untagged)
Options> Color Handling> No Color Management

In the Epson 1280 Printer Driver I use the following settings when printing the 918 patch target:

Paper size: Letter
Advanced Tab
No Color Adjustments

For HWM paper I use:

Media Type> Heavyweight Matte > Print Quality > Photo (1440 dpi) > High speed off > all other options turned off

For Epson Premium Lustre I use:

Media Type> Premium Glossy Photo Paper > Print Quality > Best Photo (2880 dpi) > High speed off > all other options turned off

For Epson Premium Glossy Photo Paper I use:

Media Type> Premium Glossy Photo Paper > Print Quality > Best Photo (2880 dpi) > High speed off > all other options turned off

For Ilford Classic Pearl Photo Paper I use:

Media Type> Premium Glossy Film > Print Quality > Best Photo (2880 dpi) > High speed off > all other options turned off

For the Ilford Classic Pearl profile that has been the best performer and least changes between on screen display and actual print output while using soft proofing, I created that profile with a mistake in the Print with Preview window. This is the error I made:

Options> Color Handling> Let Printer Determine Colors

Instead of:

Options> Color Handling> No Color Management

Oddly enough, this profile is the best one I have to work with.

I"ve been over this so often, I see the procedure in my sleep. If I’ve made a mistake, I’ve dreamed about it first.

I don’t see any blatant flaws there…so how are you printing with your profile. Something like this …

Color Management
Print> Document (Profile: Your Embedded Profile)
Options> Color Handling> Your 1280 profile

…using the same driver settings in which you printed the target?

How’s your monitor…you’ve profiled it right?

Yes, I print by either selecting my output profile in the “Print with Preview” window in Photoshop, or lately been experimenting with Qimage and selecting the correct output profile there. My monitor is good, but due for calibration tonight. I used the i1 Display last time instead of the spectro and It seemed to be a slight improvement. At any rate, the display does a good job of matching the “Mistaken” Pearl profile. Can you begin to understand why I’m getting a little frustrated? lol

Another thing to watch out for, if you are trying to match your printer to your monitor, is that the monitor profile and printer profile both match the lighting you are looking at your print under. Profiling your printer to 5000K and viewing your print under Cool White Fluorescents, while your monitor’s white point is 9300K, is bad news.

I’m totally baffled by why ‘Let Printer Determine Colors’ would give you a better profile. Something weird is going on.


My monitor is a Sony Trinitron with the white point set to 6500K. I use the i1 Spectro or Display for calibration and I seem to get a pretty good match. I haven’t generated enough revenue to step up to proper viewing lighting yet, but I have read about some economical ways to achieve good viewing. I’m certainly aware of what you’re saying, but haven’t been able to take those steps yet.

I haven’t had the courage to intentionally repeat that mistake in the “Print with Preview” window to see if the results will be repeated. Yes, weird seems to follow me in my pursuit to printing better images. I wish it wouldn’t and everything would go as described in the books and articles I’ve read. However, the greater the challenge, the more I seem to learn.

So, your lighting is not 5000K? In that case, it is no surprise that your monitor does not match your print.

Here’s an experiment. If you have the option, read your testchart in as spectral data and change the viewing light source to match your lighting when you build your profile (In ProfileMaker it’s on the lower-righthand side of the window). With the i1 you should even be able to take a spectral reading of your exact light source (you’ll need Eye-One Share), otherwise simply choose a light source that approximates yours. You may even want to lower the white point on your monitor.

Granted, this profile will only be useful for acheiving good monitor to print matches under your specific lighting setup. But, it will allow you to see if you are doing everything theoretically right. My guess is you probably are and your lighting is causing the problem.

Let me know if you followed everything I said.

Two questions / comments:

  1. Matching your viewing white to your display white is an essential part of screen-to-print matching. A good test is to open a blank white document in Photoshop (enlarged to occupy most of your screen) and then take a blank piece of paper from your printer andhold it up beside the screen. If the white of the paper does not match the white of the screen in BOTH color and brightness then you will not be able to obtain a good screen to print match. Control your lighting by using Solux bulbs or a light booth.

  2. I may have missed it but what, exactly, are you not happy about concerning your prints. A clear description of the problem may lead us to a solution…



Steve, your test with the blank white screen is pretty close with my meager viewing arrangements. I simply don’t have the revenue to step up to a light booth or special lighting at this point.

What is daunting is “what else” that always seems to be added to the list of essentials to get great results. It was a big step for me to move from sharing a Monaco colorimeter with my brother to calibrate my monitor once a month to buying my own i1 Photo. In reading these forum posts, one can profile toilet paper. Well, maybe that’s true, but after reading how to do it over and over and following all of the steps impeccably, I find that there is just one more thing I need to do to acquire perfection. Now it’s buy some special viewing lighting.

I understand optimizing all elements, but just can’t meet those objectives at this point of my color development.

Let me explain my disappointment in profiling my own media. I’m printing with an EPSON 1280. I used to print mostly on Enhanced Matte and Premium Lustre papers. I would choose “Let Printer Determine Colors” in the Print with Preview window and ICM for color handling in the driver. The results were a close representation of what I was seeing on the screen and this was using the old monitor calibration before i1. Since I have “stepped up”, I create profiles according to all the formulas and results show posterization in some browns, greens, and the reddish orange of my golden retriever. When compared to the generic 1280 print driver, the profiled media that I’ve tried to improve using Epson and Ilford papers isn’t quite as good as the old approach.

I’ve been in contact with lots of folks that have offered to help and I’ve made small advances based on their guidance. However, if I want to represent color as seen on my screen, I know I can always rely on the old EPSON Stylus 1280 profile and the ICM color handling approach in the driver. The other profiles do work on some images, but the generic approach works on ALL images.

Its also confusing to me to have so many different answers to choose from. For example, Ill read by one color expert that the display should be set to 6500K and then the same expert recommends in a forum post to set display to native white point (which is closer to 9300K in my understanding). There are inconsistencies abound and I dont have the knowledge to sort it all out. I used my displays native white point until reading two books on color management and reset it to 6500K. Im not sure now that I did the right thing. I can tell you though that your paper test didnt show any obvious discontinuities.

So Im limited right now to viewing my prints in real daylight, but I can tell if a print is good or bad if I view it in the bathroom under incandescent lighting. I dont need a light booth to see posterization (that only has showed up since Ive begun to make my own profiles).

Things did seem to change for the worse though when I went to the 6500K settings when calibrating my monitor. I followed Frasers advice on adjusting the guns, but I have six settings; RGB gain and RGB bias. Its my understanding to set the red guns to max and adjust the other two to get the sliders as close to center as possible knowing that its probably never going to happen. Ive done this on both bias and gain controls. Should I do something different?

Heres my monitor info:

Dell Color Graphic Display
Sony Trinitron
UltraScan 1600 HS Series
Model D1626HT
Serial #2714149

Well this is a long response, but I appreciate your answering my post. I have read your responses to other posts and its obvious that you spend some time trying to help folks out. I dont know if any of this response on my part helps to identify any problems I might be having with my profiling blues, but I thank you for jumping in, nevertheless.

By the way, the people that work for you are top shelf.

Best regards

Steve asked

You don’t seem to have answered that yet.

There are two separate questions to answer

(1) The characteristics of the actual printed output that you are not happy with.

(2) Any aspects of the soft proof you are not happy with.

If you could answer these questions it might help to identify the problem.

Profiling Epson inkjet printers can be difficult, BTW. Don’t lose heart yet!