Monday, November 1, 2010

Extreme Color Management with the X-Rite i1XTreme and ProfileMaker 5 – Part 2 of 2

This article is a continuation of Part 1 found here.

i1XTreme Continued…

This section resumes my prior discussion of the i1XTreme Color Management Solution….


Match 3 was a snap to use with only one scan error

When working in Eye-One Match 3, I was able to easily scan the patches and only hit one error when scanning  row 5 shown above, but a simple re-scan solved the problem. In no time flat I had the measurement data I needed and a new profile.


Match 3 Measurement Success

Trouble on the way to Paradise

When scanning Bill Atkinson’s targets RGB 800 WIR Eye-One 2 Page targets shown below (see previous article for more info) I had a much more difficult time.


Bill Atkinson’s RGB 800 WIR Eye-One 2 page target

I’ve tried to establish credibility on this blog by calling things like I see them rather than just doing gloried press release reviews like some other reviewers often do. In this case I have to fully disclose that I hit a speed bump during my review that made me want to toss the i1XTreme system out the window and never do another profile again. Let me explain why I hit these problems, assure you that there are solutions to the problems I hit and why I’m glad that I didn’t give up.

When scanning the color patches you simply hold the button down at the beginning of the row while you are on a white spot, wait for a beep, and then glide the i1 across the page until it stops and release. The software will tell you if you failed (which requires you to rescan the row) or if you can continue (by notifying you to scan the next row).  The system seems simple enough and with the limited test patches (which happed to be much larger too) in the Match 3 product I had no problems. In fact, I thought it was rather fun to use!

With my first success behind me, I decided I’d take on Bill’s big chart because 800 should be a heck of a lot better than 288, right? I could just see this amazing print rolling off my Epson with this new and improved profile that I was about to create, so I scanned with much excitement.

Well, I was excited until I got my tenth “Too Many Errors” alert (see below). I became annoyed at the fact that the user-interface never bothered to tell me what I did wrong or how I could do better next time, so I would scan again and get another failure. I’d scan slower, and I’d scan faster, yet both resulted in the same error. I’d adjust the paper and I’d take my time scanning. All of these efforts seemed to work on at least one occasion, but none of them worked reliably. However, when I got the nod from the MeasurementTool to advance to the next line, I was thrilled to move on – and move on I did.


Too many errors. Please read strip again.
You will hate this error message and
wish evil things up those who created it!

In a little over three hours I scanned the 32 rows of these charts (approximately 400 row scans), so to say I had a few errors is an understatement. I was tired and frustrated with a blister on my right index finger – but I was done. I saw the completion page below and thought I had some great data for ProfileMaker to do its magic and create a killer profile for me.


After 3 Hours of scanning, I thought I was done.
I thought wrong!

After I generated my profile, I quickly examined it in a cool tool called ColorThink Pro (review coming in November 2010 which also shows how to generate and use the charts). Upon inspection (see below), I immediately noticed it was @%#ked up!!!!


3D Graph of my first ProfileMaker ICC Profile
Never try to print using a profile that looks this nasty!!!!

I didn’t need to print to know that I would get horrible results, so I scratched my head and went back to the drawing board.

In doing my analysis, I noticed that my scan chart didn’t look as clean as Bill’s charts, but I thought ProfileMaker would use that data and magically make a great profile. However, what I really discovered is that the tool let me advance to the next slide when it got a row of data – not necessarily a correct row of data (see below) – WTF????


Scan success, but ignore the directions
to move on to the next row!

I felt like an idiot, so I did some more research and proceeded to scan again. This time knowing that a row not only had to be successfully scanned, it also had to be the correct pattern, I found myself investing 5 hours more of my life in getting 32 accurate row scans. FIVE HOURS!!!!

Not the birth of my children, nor my first magazine cover, nor anything I could imagine in the future (i.e., winning the lottery) had brought me more joy than to see the dialog below and know that wouldn’t be scanning any more of those @#king patches again that day!!!!!!


All rows scanned successfully.
One of the greatest accomplishments of my life!

Now I understood why the scan table had an advertisement that read “Need Automation? Add an Eye-One iO!” Pardon my language, but no shit!!!! Holy cow, I looked at that ad in disgust as I scanned for hours – once taking a 12 hour break out of frustration with the device.

With my frustration behind me I was ready to get my reward – a great paper profile, so that meant a visit to ProfileMaker 5.0. While there, I chose Bill Atkinson’s source chart (a text file) as the reference data to compare to my new data (also a text file) as shown here:


Generating the profile in ProfileMaker 5.0

The whole Gamut Mapping stuff confused me, but I found this i1 Gamut Mapping information to be very helpful. In the end, I did all three choices and graphed and printed using all three and got what appears to my eyes to be identical results. I’m sure there are differences and maybe eye isn’t trained enough, but the graphs in ColorThink Pro indicate they are very similar if not identical.

The Reward of Perseverance


An excellent profile generated from my 800 patch scan

To demonstrate the value of profiling I’ve printed the same source photo multiple times using various paper profiles – most of which I’ve generated myself. This test will show how one device can differ from another, as well as the importance of the software in this process.

NOTE: The images shown in this section are photos of prints taken using identical camera settings in identical conditions, but photo reproduction this way is quite tricky so they are subject to human error. The photos of the prints do not truly express the quality of the prints for a variety of reasons, so no conclusions about the paper or print should be made by using these photographs. Instead, they should just be used as a tool to demonstrate significant differences in the background gradient and tones on the tires in one image versus another. Sharpness or alignment may vary slightly due to my sample photos error, not the actual prints.

LexJet Sunset Photo Metallic Profile Test

For this example I am LexJet’s Sunset Photo Metallic paper where they do not provide a paper profile for my Epson Stylus® Pro 3880 printer. I started by trying using the Epson 3800 paper profile (which I know doesn’t work, but I did for illustrative purposes) and got horrible results (mouse over below or click here). To improve upon those results, I created my own profile using the ColorMunki and you’ll notice a huge improvement! Score one for the ColorMunki!!!!

Mouse over to see before, mouse out to see after
Mouse over (or click here) for LexJet 3800 Profile vs.
Mouse out (or click here) for
ColorMunki 3880 Custom Profile

Mouse over to see before, mouse out to see after
Mouse over (or click here) for the Eye-One Match 3 Profile vs.
Mouse out (or click here) for the MeasureTool/ProfileMaker Profile

Now I probably don’t have to tell you but you’ll notice that the image above (mouse out or click here) was the image with the most tonal range and is the print that most closely represents the stellar results I was getting on my Canon iPF6300 printer using LexJet’s phenomenal profile. If you mouse over above (or click here) you’ll the Eye-One Match 3 profile which is a clear improvement over the ColorMunki profile, but still falls short of the MeasureTool/ProfileMaker profile created using Bill Atkinson’s 800 RGP 2P test patches. Now to be fair for the Eye-One Match 3 I only used the i1 RGB Target 1.5 test patches which only had 288 patches – but you have to go to MeasureTool if you want large or custom patch sets.

What’s the net result here? Naturally, the more patches you test – on your specific model printer – the better your prints will be. You’ll extract more tonal contrast out of the image and more accurately represent the colors you are seeing on your calibrated display. In fact, if you create your profile properly the print you create should blow your display image away.

Conclusion

The scanner and display calibration with this product are both simple yet offer the flexibility for advanced calibration as well. This was great because this was more advanced than what is possible with my beloved ColorMunki. However, what really makes this product worth the money is its ability to generate a exceptional printer profiles.

Fine art printing can be expensive and it is the final representation of your work, so if it doesn’t look amazing then you are selling yourself short. For those who care about the details of their final result, an excellent printer profile is a must! Sadly the profiles generated by others (including paper and printer makers) aren’t always the appropriate choice for the best results possible from your specific printer.

For those who wish to control the process of creating the best possible profile, I’d recommend you consider using an X-Rite i1XTreme UV Cut Color Calibration Solution (which includes Eye-One Match 3) or ProfileMaker (until i1Profiler becomes available). I must also HIGHLY RECOMMEND the i1iO Automated Scan Table. Alternatively, you may also consider the i1iSis as an even faster solution for reading your test charts, but either way automation is critical in my opinion.

For Photography, I do not recommend purchasing the non-UV cut version. I questioned a panel of 5 printing experts featured in this series and they all insisted that the UV cut is required for photographers due to the growing use of papers with Optical Brightening Agents (OBA).

Ordering Information

The X-Rite i1XTreme UV Cut Color Calibration Solution is available from Amazon, Adorama, B&H and other resellers.

ProfileMaker (which includes MeasureTool) is currently only available i1iSis bundles which I could only find on Adorama at this time.


Adorama and B&H both carry the i1iSis


i1iO Automated Scan Table

Save yourself from frustration of manual scanning by purchasing an i1iO Automated Scan Table to complement your i1XTreme system from Amazon, Adorama, or B&H. NOTE: You only need an i1iO or an i1iSis for automation, not both. The i1iSis is the fastest solution from X-Rite.

Special Offer

X-Rite does not offer any special offers on its products, but I’m happy to announce that LexJet (maker of the Sunset Photo Metallic paper featured in this article) does! Click here to learn more.

Disclaimer

X-Rite provided a loaner i1XTreme color management solution for me to use so I could write this article, as well a dongle for use with ProfileMaker. If it were not for X-Rite’s generous support, I could not have brought this article to you. I receive no compensation from X-Rite, but if you purchase products using links in this blog to external sites I might get a commission. Thank you for supporting this blog by returning here when you make your purchases.

If you enjoyed this article, please support future articles like this by making a donation or saving money by using my discount coupon codes. Either way, your support is greatly appreciated!

This blog is intended for freelance writing and sharing of opinions and is not a representative of any of the companies whose links are provided on this site.

The opinions provided are of Ron Martinsen alone and do not reflect the view of any other entity

No comments: