Lab supplied ICC

[Robin Smith]

<p>I agree with you Andrew: re suboptimal, but as Carl points out re the comparison, it is often only in comparison that we truly notice the difference. For most of my shots I really don't require the wider gamut. I am happy. We don't really disagree, except, perhaps, as to how important it is for most photographers.</p>

[carl_s]

<p>Andrew,<br>

Original image was a 14 bit raw (D750). I'll try my experiment again and print your tif from Costco in different variations. I had watched your video previously (Thanks for producing these by the way) and just watched it again for an obviously needed refresher regarding OOG clipping.<br>

The differences seen in the dots portion of your tif image comparison were very similar to the levels of blue saturation I mentioned in the blue shirt of my original experiment. The print that had been embedded with the printer's profile showed more variation in saturation levels than just "blue" as was seen in the sRGB version. <br>

At the end of the day, I personally enjoy knowing that I'm getting the very best out of my own capabilities, hence my quest to learn the "right" way. That said, the sRGB version still looks pretty darned good. I think herein lies the true beauty of Raw photography and a LR workflow. If I don't like what I printed last month, year, etc. I can just go and reprint using different settings! No loss of original data! <br>

Thanks again for everyone's comments on this thread. It's been highly informative. I'll post back any interesting results after printing your tif from Costco (and probably WHCC as well). </p>

 

[digitaldog]

<blockquote>

<p>Original image was a 14 bit raw (D750).</p>

</blockquote>

<p>But converted after rendering into ProPhoto RGB? That first conversion is key to what you then have to send to an output device. </p>

[carl_s]

<p>No... I might be missing something then. I imported the NEF directly into my LR catalog, made my adjustments, and then used either the Print module or Export module to select the output profile and rendering intent. Then sent those files off to print. <br>

My understanding is that everything remains in LR's working color space until I select something different at export/output. ProPhoto never even came into play - I thought it was just bypassed. </p>

[digitaldog]

<blockquote>

<p> I imported the NEF directly into my LR catalog, made my adjustments, and then used either the Print module or Export module to select the output profile and rendering intent. Then sent those files off to print.</p>

</blockquote>

<p>OK, that's clear, you then used a ProPhoto RGB gamut (LR's underlying color space is based on that working space). That is if you selected the print output profile in Export. That would go:<br /><em>ProPhoto RGB</em> gamut color space (this color space has no name) > <em>Costco Output</em> color space. </p>

[carl_s]

<p>Thanks. That's exactly what I'm doing. On a sidenote, I sure wish LR's export function would all me to choose a rendering intent. It is a pain to use the print module for each image when it seems I could far more easily do a batch export. I wonder if LR Export picks one option be default? -- since I'm not presented with a choice.</p>

[digitaldog]

<blockquote>

<p>On a sidenote, I sure wish LR's export function would all me to choose a rendering intent.</p>

</blockquote>

<p>Initially one couldn’t export to an actual print, output color space, just working spaces and the rendering intent is always RelCol. A profile has a default RI when built, it's usually Perceptual. If you're on a Mac, you can see what is used or set with the ColorSync Utility. If the user has not control over RI, that's what is used. So if you soft proof and prefer RelCol, using Export isn't an ideal method to convert to the working space. Photoshop would be the place to do so. </p>

[carl_s]

<p>So to clarify that - if converting to a color space, LR Export always converts via RelCol. Photoshop gives an option regardless of conversion to color space or output profile. <br>

A LR Export conversion to output profile, however, is typically Perceptual, unless specified differently using LR's print module OR Photoshop's Convert to Profile?<br>

I'm assuming that LR's Print module and Photoshop's Convert to Profile module are doing the same conversions, provided that the same output profile and RI are selected. </p>

<p> </p>

[digitaldog]

<blockquote>

<p>So to clarify that - if converting to a color space, LR Export always converts via RelCol.</p>

</blockquote>

<p>For Adobe installed, version 2 ICC matrix <strong>working space profiles</strong> (sRGB, Adobe RGB (1998), ProPhoto RGB) yes. There's no perceptual or saturation table in these profiles. <br>

For output color spaces, there are three tables but you have no control over the RI so the one hardwired into the profile (usually Perceptual) is automatically applied. NOTE: there are no rules in how a perceptual rendering intent is created in a profile! It's based on the software used to create this rendering intent. Each product will do this differently, some better than others. So you do need to view the differences with a soft proof and pick the one you prefer. Comments like: <em>Perceptual intent is the most useful for general photography because it helps retain detail,</em> isn't a hard and fast rule by a long shot. It depends largely on the profile maker's software, color engineering etc. </p>

<blockquote>

<p>I'm assuming that LR's Print module and Photoshop's Convert to Profile module are doing the same conversions, provided that the same output profile and RI are selected.<br /></p>

</blockquote>

<p>Yes, at least the last time I checked this. They use the same color engine. Note that LR <strong>always</strong> uses Black Point Compensation and Dither. Those options are optional in Photoshop although it's best to use them for all conversions. </p>

[carl_s]

<p>"For output color spaces, there are three tables but you have no control over the RI so the one hardwired into the profile (usually Perceptual) is automatically applied."...."It's based on the software used to create this rendering intent. Each product will do this differently, some better than others. So you do need to view the differences with a soft proof and pick the one you prefer."<br>

I realize this is a highly technical discussion, but the above statement seems contradictory. In the first sentence, you say that I have no control over the rendering intent. In the second sentence, you recommend picking the rendering intent that I prefer. </p>

<p>I understand that I'm stuck with the hardwired method for LR export, but if using LR Print or Photoshop, I have the option to override the hardwired perceptual RI?</p>

[digitaldog]

<blockquote>

<p>I understand that I'm stuck with the hardwired method for LR export, but if using LR Print or Photoshop, I have the option to override the hardwired perceptual RI?</p>

</blockquote>

<p>Yes for LUT based, <strong>output</strong> profiles that have the tables. You just select the RI you wish, you'll see the preview update based on this selection. Not the case with working space profiles, you can select Perceptual or RelCol, you'll get RelCol, you see no update on-screen for that reason. </p>

[carl_s]

<p>Very good - thanks. I think I'm flat out of questions. For now at least....</p>

[skip_kempe]

<p>OK, now I have a question. If labs are requesting that I not embed my images with the ICC profiles (and I assume that by this they mean, for example, don't export from LR using the profile), I'm supposed to simply use the ICC profile for "soft proofing" which I didn't know about in LR before. According to this workflow, I'm supposed to edit my actual image specifically for the printer profile of a particular printer (say, Costco or North Coast Photo or Mpix or wherever). I had been operating under the assumption that I should use my calibrated monitor to get the photo to look exactly the way I wanted it to look, then output using the profile for the specific printer I was using.</p>

<p>The problem with the soft proofing seems to be that I now have to specifically adjust my photo for each particular printer I might use, instead of doing it once on a calibrated monitor, getting it to look how I want, and then having the export/publish engine in LR apply whatever specific adjustments are necessary to keep the final print in gamut with whatever printer I'm aiming for at that time. Am I understanding this correctly? If so, this seems like a real problem, as it will require me to go back and readjust images anytime I want to print them based on new or updated profiles for whatever printers or output media I might use in the future.</p>

[digitaldog]

<p>The idea is edit the master image, in a wide gamut, high bit condition to look as possible in that condition. <em>Then</em> invoke a soft proof which simulates the output and produce at this point,<strong> if necessary</strong>, output specific edits based on the soft proof. In LR this is done on a <em>Proof Copy</em> so the original master isn't affected. </p>

[stephen_ray]

<p>More “real world” sRGB at the photo lab…<br /><br />Many know that sRGB is a so-called standard profile for computers monitors. The monitors are emissive devices. The Type C photo paper is reflective. Not really a fair comparison. <br /><br />A more fair comparison could be a Type C translucent clear (DuraClear) or milky-back (DuraTrans) print film. At least the film in its lightbox would be somewhat emissive but even then, the photo dyes are just very different than phosphors, LCDs, LEDs, etc. <br /><br />The Type C paper print fails especially in the bright colors of sRGB as seen on-screen by photographers and graphic designers. Again, very misleading for a lab to say “sRGB is the goods.” Their output profile is “the goods” and unfortunately, that’s as good as it gets. The photo paper simply will not reproduce some of those bright sRGB colors. The photo emulsion dyes have a lower gamut in those areas. <br /><br />Regardless if you begin your file editing process with a wide gamut such as ProPhoto and soft proofing with the best Type C output profile, many bright colors could be reduced to a disappointing lower gamut. However, again, Type C from sRGB, Adobe98, or ProPhoto works great 90% + of the time, and again, by using Perceptual intent to preserve details. Experienced photographers and graphic professionals know their tools and the limitations of their resources. The photo lab profile for soft proofing is the tool to show the limitation of the resource. <br /><br />1) In the 1990’s the “common” interface between photo printer engineers, customers, sales, etc., was the PC monitor. <br /><br />2) In the 1990’s the photo printers were delivered to market just as sRGB was believed to be the “standard.”<br /><br />3) In the 1990’s photo labs were told by the photo printer manufactures to use sRGB. <br /><br />4) In the 1990’s some Photoshop users needed an expensive plugin from Kodak to soft proof. (PS 6 made things better.) When did PS 6 arrive? <br /><br />5) In the 1990’s LaserDiscs had a better picture than the suboptimal VHS, just not as practical as it turns out. <br /><br />I hope this helps. <br /><br />Question(s) for Andrew: <br />With your color tools, can you count how many colors are in the two given spaces of sRGB and any Costco Type C profile?<br />From the difference, can you imaging which group(s) of those colors are the most problematic in reproduction? </p>

[digitaldog]

<blockquote>

<p>Question(s) for Andrew: <br />With your color tools, can you count how many colors are in the two given spaces of sRGB and any Costco Type C profile?</p>

</blockquote>

<p>I can produce what is called a<em> gamut volume</em>. The so called number of colors (Device Values) has nothing to do with color gamut. http://digitaldog.net/files/ColorNumbersColorGamut.pdf</p>

<blockquote>

<p>Regardless if you begin your file editing process with a wide gamut such as ProPhoto and soft proofing with the best Type C output profile, many bright colors could be reduced to a disappointing lower gamut.</p>

</blockquote>

<p>One reason we need big RGB working spaces is that they are based on theoretical emissive devices (ProPhoto being very theoretical when you look at what falls outside human vision). Necessary because of their simple and predicable shapes. So while there are many more colors that can be defined in something like ProPhoto RGB than you could possibly print, we have to deal with a significant disconnect between these simple shapes of RGB working space and the vastly more complex shapes of an output color space. Simple RGB working space matrix profiles when plotted 3 dimensionally illustrate that they reach their maximum Chroma at high luminance levels which makes sense since they are based on increased Chroma by the addition of more light. The opposite is seen with print (output) color spaces where this is accomplished by adding ink: a subtractive color model. One reason we need such big RGB working space like ProPhoto RGB is due to its simple size and to counter the disconnect between mapping to the output space without potentially clipping colors. There can be issues where very dark colors of intense Chroma (which do occur in nature and we can capture with many devices) don’t map properly with a smaller working space. Many of these darker colors fall outside Adobe RGB (1998). When you encode using a smaller color space, you clip the colors to the degree that smooth gradations become solid blobs in print, again due to the dissimilar shapes and differences in how the two spaces relate to luminance. I suspect this is why Adobe picked ProPhoto RGB primaries for the processing color space in their raw converters.<br /></p>

[Robin Smith]

<p>Stephen,</p>

<p>A very good point about C-type papers.</p>