Joakim, For understanding: - ECIrgb has some advantages in highlights and midtones (barely visible for normal printing processes) - ECIrgb has some trouble with shaddows and darker colors (3/4) which are visible, but often superseeded by wrong separation settings on CMYK printers. They will be visible when using direct RGB printing processes like those being used for diapositives. - ECIrgb, like other RGBs with gamma, have trouble to reproduce neutral grey, as it is allways interpolated LStar-RGB has the same colorspace but tries to correct the problems stated above - neutral grey is defined in the profile. There is only little interpolation involved - It has more details in dark tones and shaddows with a barely visible expense in the highlights and midtones (RGB 127:127:127 is 50% in LStar-RGB but 60% in ECIrgb) - It's internal structures are closer to the way how color management software does it's magical math. This way there is less guessing and more prediction. For practical work: - Both LStar-RGB and ECIrgb deliver better performance than AdobeRGB and sRGB when printing in Europe. - You will notice differences in output only when the whole color workflow is using professional color management at a high quality level. The differences will be superceeded by other printing issues when color management and printing is not handled professionally The issues you experienced before might be due to a wrong setup of your printer (like sending LStar-RGB, but using ECIrgb in the printer...) Hope this helps Karsten P.S.: Markus, thank you for your help to solve this Am 31.05.2005 um 11:54 schrieb Joakim Eskildsen:

On May 31, 2005, at 7:09 AM, Karsten Krüger wrote: I can't say I agree with this statement. Matrix profiles always produce neutral gray (a*=0, b*=0) when R=G=B. It's a mathematical requirement of these profiles. If anything, they are more reliably gray balanced than a table based profile can be because you can't put an entry for every single point along the gray axis in even a 33x33x33 profile. Chris Murphy Color Remedies (TM) www.colorremedies.com/realworldcolor --------------------------------------------------------- Co-author "Real World Color Management, 2nd Ed" Published by PeachPit Press (ISBN 0-321-26722-2)

Am 01.06.2005 um 00:27 schrieb Chris Murphy: Chris, I agree with you that a large colorspace has benefits. But you have to be very conscious on what you are doing. Using ProPhoto RGB leaves you with just one choice of color rendering: colorimetric. If you are using photographic rendering your resulting image will be very small compared to any other RGB colorspace. But colorimetric rendering is often not sufficient, because there are likely out of gamut colors in digital photography (i.e. in blue tones). For my experience the most predictable and accurate reproduction of an image will happen when you are as close as possible to the final output colorspace, which for ISOcoated / european CMYK is LStar-RGB / ECIrgb and for internet related work sRGB. An other drawback of ProPhoto RGB is that you will hardly find a monitor that is able to reproduce the whole colorspace. So you end up editing data you are not able to see on your monitor and not able to print out later... Thanks for your comments anyway, Karsten

Am 01.06.2005 um 09:59 schrieb Chris Murphy: When using perceptual rendering, the whole source gamut has to fit into the destination colorspace. Rendering does not look for how much of the source gamut your image is using. If it would do so you would receive different results when moving from image to image. So when your source gamut is huge, it gets compressed a lot when rendering to a small CMYK output gamut. This does not happen with colorimetric rendering (either absolute or relative). In this case everything outside the smaller output colorspace is just clipped to the edge of it. If your image is small enough, you don't see any problem. If it is bigger, you end up loosing details. I put something online a couple of months ago. It is written in German, but I believe you will still get the idea by looking at the pictures: - all pictures are at the same scale, looking from top onto the colorspace - the first section shows the relation of the gamuts to each other (Wide Gamut RGB is close to ProPhoto RGB) - the second section shows how RGB workspace gamuts may fit into ISOcoated (with colorimetric rendering). Dimmed colors are unprintable. - the third sections shows what happens to your data moving from scanner gamut to color workspace gamut. All but Wide Gamut are rendered with photographic intend. Since Wide Gamut RGB is big enough it is better to use colorimetric rendering - the forth sections shows what happens to your data moving from color workspace to ISOcoated using photographic rendering. For reference there is also a direct rendering from the scanner gamut to ISOcoated using a photographic rendering intend. Take a picture of neon advertisement at night or flowers with vivid colors in bright sunlight with an EOS 20D. Render them to ISOcoated. Using photographic rendering intend you get decent results. Using colorimetric rendering intend you end up loosing a lot of details and the emotions from your pictures are gone. I like to use an other setup. Why should I keep and manage data I am not able to reproduce. I start at the output process available for a given job. Do I output to a film printer - LStar-RGB. Do I output to a CMYK press - LStar-RGB or ECIrgb (depending on customers requirements) plus Photoshop softproof set to ISOcoated or whatever the printing service needs. Do I output for internet or PowerPoint presentations - sRGB. My archive: camera raw, settings, delivered data to customer. This covers 98% of my needs.

To all The discussion are very interisting and I am learning a lot. Thank you for taking time and answering Chris! Karsten Krüger I could not find this online web place you talk about in you las mail. I would be intrested in seing it! I have read Bruce Fraser´s articles on the advargse of editing on large spaces and its very interisting. I like to ask you if there are any down sides of all thise colores we cannot see when converting to the small CMYK color space or the ink-jet color space, and if you make hevy editing on your 16 bit file will you sooner get into banding in the big ProPhoto RGB than in LstarRGB? Karsten Krügers writes about the color rendering problem: colorimetric. It is confusing. Does it mean thet if you use perceptual rendering or relative colorimetricit the picture will be compress so the colores change as there are so much out of gamut and therefore the look and realitionship in the image will not corispond to the screen anymore? Will this convertion be more price and eacyer using the smaller colorspaces like LstarRGB? I thought that LstarRGB was a relativly large colorspace slighly bigger than Adobe 1998 and bigger than any printed colorspace. Are ther any chance that one will develop Ink-jet printeres colorspace so that it will have a bigger colorspace than LstarRGB? About my work flow: I start out from a raw 3f scann and start the editing from that in photoshop so the editing is large. Practely all most all my pictures has too much detail, most often in the shadow when saving the from the scanner (3f files) and I there fore whonder if the extra detail is importent wich one would get fro ProPhoto RGB. I have the raw files without an colorspace arcived so in the future I can save a new file with the possible improved colorprofiles and printing comes. My screen is calibreted with BasiCColor in with lab mode on with eye- one. I am fully awere thet there are no such thing as "best" colorspace for everybady but I am trying to find the one most good for my workfload. I do also understand that the diffences maby are relative small but Is a good start when printing to know that the bace is good! (My screen is calibreted with BasiCColor with lab mode on with eye-one.) P.s Here are some of my works: www.30bytaik.com (look under "artist" and "Joakim Eskildsen") P.p.s Here is an artickel on the ProPhoto rgb: http://www.luminous-landscape.com/tutorials/prophoto-rgb.shtml On Jun 1, 2005, at 1:27 am, Chris Murphy wrote: (I did safe a 3f file from the scanner with no profile and sepreatly assign the profile to 2 no profiled images). I have read Bruce Fracer´s artickels on the advargse of editing on large spaces and its very interisting. I like to ask you if there are any down sides of all thise colores we cannot see when converting to the small CMYK color space or the ink-jet color space, and if you make hevy editing on your 16 bit file will you sooner get into banding in the big ProPhoto RGB? Practely all most all my pictures has too much detail, most often in the shadow when saving the from the scanner (3f files) and I there fore whonder if the extra detail is importent. Anyway I shall defentely give it a try! I am fully awere thet there are no such thing as "best" but it helps to get some comments! Thank you very much for your thoughts and insight! I do consider my selv as a serius picture printer, I have being doing b/w and color prints in traditionel wet darkroom for years and have since 1998 only been doing it digitaly and are in the last few years abel to make much better prints digitaly comparet to the older cemical darkroom prints! But it is including much teckinaly considerition like this color space issue!

Very good job you did with this article, congratulation. Actually perhaps this may be a bit off topic, but still as well related: We are always talking about coulours, but what if one uses gray pictures? More me this "dot gain" is more or less an effect, which basically has something to do with print, an effect based on blurring or of the mapping of rectangular pixels to round ones. Some how I never found a connection to gamma or gammut. So helplessly I end up with using L-Star or some other colour space, just because of the lack of knowledge for gray. Olaf ----- Original Message ----- From: "Karsten Krüger" <kk(a)kkrueger.de> To: <eci-en(a)lists.transmedia.de> Sent: Wednesday, June 01, 2005 11:22 AM Subject: Re: [ECI-EN] ECIRGB versus LstarRGB 3

Am 01.06.2005 um 11:20 schrieb Joakim Eskildsen: Rendering intends: - photographic (same like Perceptional, Foto). Shrinking source gamut until it fits into output gamut. All colors change, but details are kept. Grass is grass, wall structure is wall structure, but may be the grass is greener, the wall whiter etc. - relative colorimetric: Mapping source and output gamut 1 to 1. Everything inside the output gamut is 100% identical, everything outside is clipped to the edge. Grass and wall appear to be flat areas. - absolute colorimetric: Mapping source and output gamut 1 to 1 like relative colorimetric. Paper has a color. Most papers are not "100% white", some are yellowish (cream color), some are blueish (bright white). Printing 10% magenta on the first tends to look a little bit orange, printing 10% yellow on the later one a little bit greenish (cold yellow) because the color of the paper adds to the printed color. Relative colorimetric ignores this, absolute colorimetric tries to simulate this behavior. For samples: http://homepage.mac.com/karstenkrueger/rendering.htm

Am 31.05.2005 um 11:54 schrieb Joakim Eskildsen: Dear All, I have made a somehow similar experience and after some testing came to the conclusion that you would want to make sure your monitor is profiled (and the software is capable of doing so) towards a linear L* in order to compare the prints created from Lstar-RGB on your monitor accurately. The monitor profile would make the colors look "more colorful" on the monitor simply because of the nonlinear luminance aka. gamma applied. When looking at the LAB "horseshoe" a higher L* is capable of producing more saturation rather than a lower "darker" L*. As described before in this thread by Chris the effects were subtle but noticeable. @All Sorry for this OT question in advance but due to the coherence; does anybody know the english translation for the german term "visuelle Gleichabständigkeit" ?? Thx, Jan-Willem