Lab supplied ICC

<p>Some labs (WHCC, MeridianPro, etc.) provide ICC profiles for soft proofing only, and specify that their profiles NOT be embedded in the image. These labs request either AdobeRGB or sRGB color spaces for the images. Other labs, (i.e. Costco) provide ICC profiles and require that the image color space be converted to the supplied ICC profile. Can anyone explain why the difference? It would seem to me that one way is right, and one way is wrong. I don't understand why it's different based on the lab...<br>

Thanks all!</p>

 

<p>It's different based on the lab because not every lab runs images the same way.</p>

<p>Speaking generally, the first way is the "right" way, and the option you should take at any pro lab and nearly any other lab.</p>

<p>The method recommended by Dry Creek for Costco and Walmart stems from the way those labs run (or, more accurately today, historically <em>ran) </em>their printers. The Fuji equipment they typically used had some, er, <em>eccentricities</em> surrounding ICC profiles. Converting worked around them to yield the expected output.</p>

<p>Today, any knowledgable tech at a Costco lab with updated equipment will tell you just to soft proof and make sure to request no adjustments. </p>

<blockquote>

<p>Can anyone explain why the difference? It would seem to me that one way is right, and one way is wrong. I don't understand why it's different based on the lab...</p>

</blockquote>

<p>Sure, the labs that provide a profile for soft proofing but ask for sRGB or Adobe RGB are basically color management clueless and hope you don't know this. They want you to believe they are color managed (they clearly are not or they would provide you the profile to use to it's intended and fullest capabilities). They want you to believe the profile will be used for the final output which may or may not be the case. What rendering intent will they force on the conversion? That can play a huge role in both the soft proof and the output. Do they tell you which they will use or do you get to guess?</p>

<p>IF you print in-house or use a lab that has a clue about color management, you have an output profile. That profile is used for soft proofing based on the rendering intent you visually prefer. Then that profile is used to convert the data exactly in that fashion. This isn't happening with the labs that force you to use sRGB.<strong> And sRGB is about the worst possible working space to use if your intended output is to a print! </strong></p>

<p>These <em>'just send us sRGB</em>' labs want to funnel everything into sRGB to speed up their production, not necessarily to make the best print for you!<br>

You can see this for yourself if you have access to your own printer or a lab that lets you use the profiles as they were designed to be used, here's proof of concept:</p>

<p><em>The benefits of wide gamut working spaces on printed output:</em><br>

<em> </em><br>

<em>This three part, 32 minute video covers why a wide gamut RGB working space like ProPhoto RGB can produce superior quality output to print. </em><br>

<em> </em><br>

<em>Part 1 discusses how the supplied Gamut Test File was created and shows two prints output to an Epson 3880 using ProPhoto RGB and sRGB, how the deficiencies of sRGB gamut affects final output quality. Part 1 discusses what to look for on your own prints in terms of better color output. It also covers Photoshop’s Assign Profile command and how wide gamut spaces mishandled produce dull or over saturated colors due to user error. </em><br>

<em> </em><br>

<em>Part 2 goes into detail about how to print two versions of the properly converted Gamut Test File file in Photoshop using Photoshop’s Print command to correctly setup the test files for output. It covers the Convert to Profile command for preparing test files for output to a lab.</em><br>

<em> </em><br>

<em>Part 3 goes into color theory and illustrates why a wide gamut space produces not only move vibrant and saturated color but detail and color separation compared to a small gamut working space like sRGB. </em><br>

<em> </em><br>

<em>High Resolution Video: http://digitaldog.net/files/WideGamutPrintVideo.mov</em><br>

<em>Low Resolution (YouTube):

 

<blockquote>

<p>Speaking generally, the first way is the "right" way, and the option you should take at any pro lab and nearly any other lab.</p>

</blockquote>

Isn't the first way this: <em>Some labs (WHCC, MeridianPro, etc.) provide ICC profiles for soft proofing only, and specify that their profiles NOT be embedded in the image. These labs request either AdobeRGB or sRGB color spaces for the images.</em>

<em> </em>

<em> </em>

<strong>That's the wrong way. </strong>

 

 

The OP states the 2nd way is this: <em>Other labs, (i.e. Costco) provide ICC profiles and require that the image color space be converted to the supplied ICC profile.</em>

<em> </em>

<em> </em>

<strong>That's clearly the right way for the reasons I've outlined. </strong>

<p>Thanks Andrew. Can you recommend any labs that do it "the right way" as you've outlined (OP #2). I've always thought of WHCC as a higher end printing house - maybe I've been wrong. </p>

<blockquote>

<p> Can you recommend any labs that do it "the right way" as you've outlined</p>

</blockquote>

<p>Most Costco locations. Profiles are available through Dry Creek Photo.</p>

<p>I was thinking other than Costco - but maybe that's truly my best (and least expensive) option. <br>

I just got off the phone with WHCC and they use perceptual rendering for everything, which I guess is why it's not a big deal that the profiles are not embedded. They accept AdobeRGB, and will only convert to sRGB in the event that an image is sent without a profile. So I guess perhaps it's not that big a' deal that WHCC does not allow their ICC profiles to be embedded into the image.... </p>

<p>I know from my experience with a fairly new Fuji Frontier Drylab inkjet printer that the drivers that provide a GUI interface have been severely simplified with a Fischer Price push button design as shown to me by the manager of my local Walgreens.</p>

<p>I saw there were two buttons to push for output one was labeled "sRGB" space, the other "Printer" space. All I wanted was less over saturated and high contrast prints which occurs when the "Enhancement" button is selected.</p>

<p>I downloaded the ICC printer profile for that model of printer from a Fuji European website and converted to it and got noticeably reduced saturation/contrast over sRGB but not by much but I also had to have the tech turn "Enhancement" off which they sometimes forget to do.</p>

<p>Couldn't tell whether embedding the profile in the image made any difference which I tested for both tagged and untagged.</p>

<p>Carl, I'd take Jeff's suggestion and try Costco first, hard to beat the prices and depending on the location, you'll get a full color managed workflow. Check out any labs supported by DryCreek:<br />http://www.drycreekphoto.com</p>

<blockquote>

<p>I just got off the phone with WHCC and they use perceptual rendering for everything, which I guess is why it's not a big deal that the profiles are not embedded.</p>

</blockquote>

<p>Depending on who built the profile and the image itself, it can be a big deal. So what happens if you soft proof and prefer RelCol (which I find is the case more often)? <br>

The other question to ask WHCC is why they can't simply accept the data in an output color space and, is the profile you've downloaded the identical profile they use for conversions? </p>

<p>Thanks all... </p>

<p>Andrew,<br>

Good question... But Dry Creek and Costco seem to be out of the norm. I've checked as many "Pro" printing houses as I could find websites for this afternoon and nearly all of them say that all images must be embedded in a standard color space. As far as I can tell, if you want relative rendering, you're SOL, at least from WHCC. <br>

It would be nice to just embed the profile as many of the experts recommend. Does make me wonder why I'm having a hard time finding a non-Costco printer to do it this way.</p>

<blockquote>

<p>I've checked as many "Pro" printing houses as I could find websites for this afternoon and nearly all of them say that all images must be embedded in a standard color space.</p>

</blockquote>

<p>Under that circumstance, I'm glad you put quotes on "pro" <g>.</p>

<blockquote>

<p>As far as I can tell, if you want relative rendering, you're SOL, at least from WHCC.<br /></p>

</blockquote>

<p>And again, that's a far from ideal color management workflow! And again, we don't even know if the profile they stick up on their web page is even used to convert the data. If not, using that profile for a soft proof is just a lie and feel-good approach. <br>

<br>

I used to work with a lab that was fully color management savvy (Pictopia) but they went out of business a few years ago when the recession hit Calif hard. Or maybe they went out of business because they had a fully color managed approach. </p>

<p>I use mpix.com for prints. They have an ICC profile hidden somewhere on the website that I probably couldn't find again if I needed to, but I saved the profile into Lightroom and export everything that I send to them using that. Then when I order, I make sure to check the box that says "Do not color correct" and then double-secret affirm that I really, really don't want any correction. You actually pay less for their "proof prints" this way.<br>

The color has worked great for me this way so far.</p>

<p>This is what I could find on mpix's rather confusing web site:<br>

http://www.mpixpro.com/help/Help.aspx?id=21#anchor_168<br>

How to prepare your files<br />Please save your files in sRGB color space in 8-bit color, not 16-bit, to achieve the best print results. Also, please do NOT embed any profiles.</p>

<p>Both Bay Photo and MPix prefer sRGB and state as much, I believe, and yet these two companies provide a lot of print output for professionals: so what is this telling us?</p>

<blockquote>

<p>Both Bay Photo and MPix prefer sRGB and state as much, I believe, and yet these two companies provide a lot of print output for professionals: so what is this telling us?</p>

</blockquote>

<p>I'd suggest they have more than a preference, they are pretty much <strong>demanding</strong> sRGB or perhaps a bit better, Adobe RGB (1998). What they are telling us in one sentence is they are a so called "<em>pro lab</em>" and in another, they don't understand color management. <br>

Again, about the least optimal working space for images intended for output to print is sRGB. I've provided a means to test this yourself, with your own equipment or equipment of any of the above labs, assuming they would even accept ProPhoto RGB. <br>

I've <strong>never</strong> seen the gamut plot of any output device, true RGB contone or otherwise that wasn't larger than sRGB, often greatly so. </p>

<p>I am suggesting, I suppose, Andrew, is that you are certainly correct but that, for the professionals who use these "pro" lab facilities, they presumably consider it relatively unimportant, as the result they get satisfies themselves and their clients. I have to say that I too have used them extensively and have never been unhappy, despite the limited gamut. Many things are better in comparison, but without a comparison one is none the wiser. Of course one can retort that all these people are deluded and should only be satisfied with the best available, but this is perhaps a rarified approach that is not very relevant for many, especially given the unreality of many images (posterized/highly saturated/HDR/black and white etc. etc.).</p>

<blockquote>

<p>I am suggesting, I suppose, Andrew, is that you are certainly correct but that, for the professionals who use these "pro" lab facilities, they presumably consider it relatively unimportant, as the result they get satisfies themselves and their clients.</p>

</blockquote>

<p>I'd agree if and only if the client tried using a proper color management path and compares the output to the lab not using it. I've provided a file and proof of concept anyone who wishes can use, URL's above. Have you viewed the video?<br>

What satisfies a lab and what satisfies a client are often two completely different things. <br>

Using an sRGB, non color management path isn't going to necessarily produce poor results. It's going to produce sub optimal results. </p>

 

<blockquote>

<p>The color has worked great for me this way so far.</p>

</blockquote>

<p>My experience too from Bay. I have a calibrated monitor and their prints essentially match what I see on screen, I too request no color "correction", so I have no complaints to date, although I usually print out at home.</p>

<blockquote>

<p>out of business because they had a fully color managed approach.</p>

</blockquote>

<p>No, but it does rather speak to my point: for their users their system works.</p>

<p>Robin, I'm not suggesting an sRGB workflow will produce poor results. I'm stating it's <strong>suboptimal</strong>. You're not getting the best possible output by neutering the data to sRGB. But unless you're willing to test this yourself, if you're happy with what you have, be happy that you're happy with what you have!</p>

<p>Some “real world” sRGB at the photo lab…<br /><br />The overwhelming problem is the lower gamut of the photo paper process. “sRGB only” is misleading. The lab output profile is necessary in order to provide a clue, via soft proof, of what particular bright colors will not reproduce as desired. Fortunately, many labs provide the profile. <br /><br />Imagine the high school senior who arrives at the studio wearing a day-glow colored sweater despite being advised to wear muted, non-distracting colors or patterns for her photo session. Her sweater presents an out-of-gamut scenario. Nowadays, the photographer can show the client a preview of the surprising color results. <br /><br />Other challenging scenarios…<br /><br />1) Commercial photographer shooting certain fluorescent colored bath towels for Big Box Bath & Yonder. <br /><br />2) Landscape photographer shooting certain (bright) flowers at sunset (more bright colors.)<br /><br />3) Fine-Art re-photography attempting certain colors of paint, inks, media, etc., natural or synthetic. <br /><br />4) Forensic photography can be challenging. <br /><br />You can probably imagine many more scenarios of your own. <br /><br />Unfortunately, these conditions are still out of gamut and the best one can do is know why it happens and how to best control it using the technology available. <br /><br />Perceptual intent is the most useful for general photography because it helps retain detail. <br /><br />Some color photography experts have suggested that sRGB contains most of the colors found in the natural world. That may be not far from true, however, the photo paper process is VERY different than sRGB. Therefore, the photo labs have not done anyone any favors by suggesting their process is sRGB.<br>

I believe the photo process is VERY good for what it is. It works great 90% + of the time. There are just many differences between the two models. </p>

<p> </p>

<blockquote>

<p>The overwhelming problem is the lower gamut of the photo paper process. “sRGB only” is misleading. The lab output profile is necessary in order to provide a clue, via soft proof, of what particular bright colors will not reproduce as desired. Fortunately, many labs provide the profile.</p>

</blockquote>

<p>Lower gamut than what, sRGB? Nope, I've never seen a print output color space, plotted 3D that wasn't larger in many areas of color space <strong>than</strong> sRGB. You're forced to provide sRGB. If you're concerned about gamut clipping, deal with conversion TO sRGB and move on. It's pointless to soft proof and presumably take some editing action on anything but what you the image editor can convert <strong>to</strong>, in this case sRGB. Again, about the worst color space one could convert TO for output to a printing device. Converting from sRGB back to the larger gamut output color space does not expand the gamut of that image! </p>

<blockquote>

<p>Perceptual intent is the most useful for general photography because it helps retain detail. <br /></p>

</blockquote>

<p>Unless you've gone out of your want to find and download a V4 sRGB profile (that's NOT being installed or used by Photoshop), there is no perceptual table in sRGB, the color space you are forced to convert the source, hopefully wide gamut data, into. You always get RelCol. </p>

<blockquote>

<p>Some color photography experts have suggested that sRGB contains most of the colors found in the natural world.<br /></p>

</blockquote>

<p>That's nonsense and easily disproven! What color photography experts made such a claim? <br>

The sRGB color space was built based on a CRT circa 1993. Just view it within the CIE chromaticity diagram which does define the gamut of human vision and you can see how silly this concept is. Colors are not 'found' in the natural world. Color, is a perceptual property. So if you can't see it it's not a color. Color is not a particular wavelength of light. It is a cognitive perception, the excitation of photoreceptors followed by retinal processing and ending in the our visual cortex, within our brains. As such, colors are defined based on perceptual experiments. Those experiments provided the CIE chromaticity diagram of which sRGB represents small subset. </p>

<p>To test the differences between sending Costco an sRGB file vs sending them one with the printer profile embedded, I sent them three copies of a sample image (Granted, my test is limited by the colors in the specific image I chose). The three versions I sent were: 1) sRGB, 2) Profiled using Perceptual, and 3) Profiled using Relative. <br>

The results are quite interesting. All three images look great. It would not be surprising to me if 99+% of users would be more than happy with what they get out of the sRGB image. Both images that had been converted to the printer's profile look nearly identical. The biggest difference between the profiled images and the sRGB image is that the profiled images appear to reproduce more shades of certain colors (ie A Blue shirt that contains multiple shades of blue prints the shading more accurately with the profiled image whereas the sRGB version just looks "Blue). <br>

The final test came when I gave all three images to my wife (who doesn't know or care about color management). She humored me, and after looking at all three images she picked the profiled ones because she felt the coloring more closely matched the shirt that the subject was wearing. She said "All three are so close in color that if I didn't know exactly what this shirt looked like, I wouldn't have been able to pick one over the other."<br>

The argument to not use sRGB seems quite clear. Costco won't accept AdobeRGB, and "most" of the other online labs won't accept a profile embedded image. Either way, as long as I stick with either AdobeRGB or embed the Costco profile, I think my images will look great. <br>

One question though... When soft proofing with the printer's profile, I'll get print gamut warnings by clicking the triangle on the right side of LR's histogram (To be technically correct, I think it's called something else when soft proofing, but you get what I'm talking about...). If I deselect the gamut warning, is the proof showing me how the printer will reproduce that out of gamut color? In other words, if I'm happy with the color shown in the warning area, I need look no further, right? This would be as opposed to using the HSL panel to try and desaturate some of those out of gamut colors. </p>

<blockquote>

<p>To test the differences between sending Costco an sRGB file vs sending them one with the printer profile embedded, I sent them three copies of a sample image (Granted, my test is limited by the colors in the specific image I chose).</p>

</blockquote>

<p>What was the original working space from what (raw, camera JPEG)? Note, here's the file to use for testing next time:<br>

http://www.digitaldog.net/files/Gamut_Test_File_Flat.tif<br>

Check it out, even by converting and viewing a soft proof, I think you'll see why it was built for such testing. </p>

<blockquote>

<p>When soft proofing with the printer's profile, I'll get print gamut warnings by clicking the triangle on the right side of LR's histogram<br /></p>

</blockquote>

<p>Ignore it. The Out of Gamut (OOG) overlay is both buggy and inaccurate, it treats OOG that's a tiny bit and hugely out of gamut the same, the output profile will handle this for you. Better to soft proof using the actual profile for conversion, in LR make a proof copy and apply output specific edits <strong>if</strong> necessary based on the soft proof, not an ugly OOG overlay that blocks your image. <br>

Illustration of this seen in this video:<br>

http://digitaldog.net/files/LR4_softproof2.mov</p>