Gamut? 3800 EpsonStandard vs PhotoshopRGB

[jtk]

<p>PhotoshopRGB matches my monitor closely, looks great, but Epson Standard is lighter and very unsaturated... like 1/3 stop lighter, and like grade 1-2 paper in the old days vs grade 3. </p>

<p>It's good vs bad.</p>

<p>Is this the common experience? Is it just blind luck, having to do with my monitor? Does it have to do with raw/TIFF vs JPEG? For context, shooting raw I usually develop portraits lightly in LR2, switching to PS2 for retouching and printing. Printing a perfect file from LR seems identical to printing it from PS.</p>

<p>Similarly, 2.2 looks great and 1.8 is intolerable with skin tones (though can be OK with B&W). </p>

<p>Is there a way to default to 2.2 and PhotoshopRGB so I dion't have to confirm that with each print ?</p>

<p> </p>

[jtk]

<p>oops..I meant AdobeRGB...that's what works...not "Photoshop RGB"....</p>

[roger_smith4]

<p>Why aren't you using the actual profiles Epson includes with the printer and setting ICM to off? I find the driver controls more useful for tweaking B&W.</p>

[brad_smith8]

<p>Yes make sure you are using the proper printer/paper profile. There may be newer/better ones than the ones that come with the driver that you'd need to download (as with my 2880).</p>

<p>I develop into the ProPhoto RGB from LightRoom which is the closest match for my 2880. My monitor is only software calibrated, but the prints match and are gorgeous!</p>

<p>One note about ProPhoto RGB and Epson printers; the final Epson print preview does not know how to handle the wider gamut, so it looks way over-processed in the preview, but it prints out fine.</p>

[patricklavoie]

<p>Not sure what you trying to achieve? You want to print color to your epson 3800? well use the correct icc profile such as SPR3800_PLPP if you print on a Premium Luster photo Paper..</p>

<p>Want to print BW? dont use a profile period on any kind of inkjet paper..select NO COLOR MANAGEMENT in photoshop or Lr, then in the epson driver select ABW, go into the setting panel, put it at dark, and enter 3 and 3 in the little white box beside the color wheel.</p>

<p>Enjoy your amazingly good print.</p>

[jtk]

<p>Roger, thanks...I think you're recommending color controls "off" (box checked) using Epson profiles...is that right? </p>

<p>Testing that a moment ago, results are certainly saturated (probably way too saturated) but don't match (like 2 stops too dark) ... the method I've been using (Epson profiles in both cases) does match extremely closely...</p>

<p>I don't need to tweak color or anything else with the method (settings zero) using Epson profiles (Moab Colorado Satine Gloss and other Moab papers) I've described EXCEPT that I like to increase saturation (typically +10). </p>

<p>If I'm missing something, what is it?</p>

 

[patricklavoie]

<p>Then John, if you knwo what you need..do it : )</p>

<p>1_In Ps, use hue saturation and add 10% overall before printing.</p>

<p>2_In Lr, use the vibrance slider or the saturation slider to get what you need.</p>

<p>look too easy..i must miss something?</p>

[jtk]

<p>OK, good ideas, thanks...HOWEVER, forget LR for a moment ...think Photoshop...</p>

<p>I import raw to LR and work from there or, if I want to retouch I "Edit" in PS and print from there (or re-import to LR and print from there...same, same...</p>

<p>Please just think about PS because that's what I'm printing from at the moment.</p>

<p>I do get exquisitely good (yikes, it's true) color and saturation using the process I described...</p>

<p>(I always evaluate prints in all sorts of light and I do have a good color eye ...years of color wet darkroom and slide duplication).</p>

<p>AM I MAKING A MISTAKE, making things too hard somehow? : Since I'm getting GREAT prints, I'd like to use my current method as a DEFAULT ...does one of the suggestions above do that?</p>

<p>ABW or, sometimes, the "Black" setting (check it out) is wonderful.. One often overlooked but very important feature of QTR was "ink loading" : ABW gives even more control, allowing slightly better Dmax than defaults (if you don't go overboard).</p>

 

[jtk]

<p>Patrick, I don't "know what I need" (that'd be dumb) but I do know what's working...I'm hoping to find a way to create new defaults<br>

Its easy to get the color/density/saturation I want in LR or PS on monitor..reasonable photography gets me there in the first place :-) </p>

<p>The annoying part is 3800's defaults, which want to desaturate and print light...default is 1.8 gamma and Epson Standard, but if I use 2.2 gamma and AdobeRGB I'm very close to perfect...so I'd like to default to that.</p>

<p>I do add +5 or +10 saturation most of the time, but even with zero (0) this method is better than the defaults.</p>

[patricklavoie]

<p>Ok, let start from the beginning, and concentrate ourself on Ps;</p>

<p>I assume you have a calibrated monitor with a device, so let start with, heres the correct method. The correct method for a complete color managed workflow.</p>

<p>1_you work on your image to the best of your knowledge, and you are ready to print, knowing that your calibrated monitor show you what you are suppose to get.</p>

<p>2_you then hit print, select Photoshop determine color, select the apropriate icc profile.</p>

<p>3_Now you are in the epson driver, you select the paper, the color ink, the quality of your print and you turn OFF color management in there; so no more density, gamma, saturation slider..they are gone.</p>

<p>4_using epson ink and paper you should then get a pretty good image. A image that should look pretty close to what you have on screen. A image that will look good anywhere because you use the proper way to create it and you use the correct icc profile.</p>

<p>If you dont like what you get from this *only one good way to print* that mean that you dont have a good color managed environment; bad monitor calibration or uncalibrated / not using the genuine epson paper and ink or custom icc profile for a different paper manufacturer / or you use the basic epson method where you cant use a icc profile.</p>

<p>If you like the way you print now using the basic epson method, be aware that your print migth only look good on your 3800 only, not mine or your friend one..only yours. Its call a close loop environment. The day that you need to print with a external lab, your back to square one, test, print, test, print etc...</p>

[rnt]

<p>When you turn the Epson Color Controls Off you need to make sure that you have the Photoshop Manages Color selected.<br>

Then you need to select the paper profile (Patrick is correct about the SPR3800_PLPP profile for Epson Luster).<br>

I usually select Rendering Intent Perceptual and uncheck Black Point Compensation.</p>

[KenPapai]

<p><em>"One note about ProPhoto RGB and Epson printers; the final Epson print preview does not know how to handle the wider gamut, so it looks way over-processed in the preview, but it prints out fine."</em></p>

<p>Exactly! My Epson print preview window colors SUCK. I pretty much ignore it -- my prints close enough match what I see on the screen -- Epson 2400, 3 years old now, and still as good as new. Does anyone see accurately rendered colors in Print Preview?</p>

<p>ALWAYS check ICM as OFF and for sure use the correct .ICC profiles for the paper.</p>

[jtk]

<p>Patrick..thanks...OK for the moment...one question: isn't Epson's canned profile for its paper the same as the "correct icc profile" for that paper?</p>

<p>Are you saying that what I'm doing is "basic Epson method?" I thought it was trial and error :-)</p>

<p>You're right about closed loop. I'm delivering good prints that way now but will eventually have to play by the rules. Reluctantly.</p>

<p>The problem is that my prints do match my monitor closely...upload raw to LR or PS, adjust/retouch to taste, print the way I described=handsome results with one or two 5X7 tests in most cases, more if I'm dealing with subtle skin tones.</p>

[steve_swinehart]

<p>John - there are a couple of other techniques you can use other than bumping saturation. For "flat color" photos - you know, like the desert in winter where it's all tan with just a few rust colored areas....</p>

<p>Duplicate the base layer. Change the mode to LAB. Go into curves and set it up so you have 16x16 box layout. Go to the "a" channel. Pull the top of the diagonal line in toward the center horizontally 1-2 boxes. Go to the bottom of the diagonal line and pull in the same amount as the top to balance the colors. Go to the "b" channel and do the same thing - you can control the saturation much better this way. You can tweak even further using the layer controls for transparency. When you're done - change back to RGB mode and continue editing.</p>

<p>Another method is to duplicate the base layer. Click the layer icon on the upper left side to open the layer menu. Select "soft light" - this will boost the color saturation way over the top. Use the layer controls (transparency, etc.) to tweak to satisfaction.</p>

<p>Combine both methods and you can get subtle control over the entire image. However, if you want to use the LAB method, do that as the very first step after opening the image. Otherwise, PS will want to flatten the layers if you have a lot of them. I think you'll find you can get better results by using / combining the techniques I've described rather than just boosting saturation.</p>

<p>If you're printing on the 3800 and have your monitor calibrated, you should be able to get a direct color match to the monitor by setting the Epson printer dialog box so that "color management" is turned "off" (let the application manage color), gamma is 2.2. You have to select the correct paper in the Epson dialog to match the paper ICC profile.</p>

<p>Once you do that, select the correct profile in the PS printer dialog box for the paper. Use "perceptual" color rendering with "black point compensation" turned "on." If you select either "Relative Colormetric" or "Colormetric" reproduction, there will be colors that get mapped to the same colorspace point. That will compact out-of-gamut colors and reduce the color reproduction range. By selecting "Perceptual" you maintain the color relationships as they get mapped from the image to the profile color space. Black point compensation turned "on" helps maintain the color saturation.</p>

[brad_smith8]

<blockquote>

<p>I usually select Rendering Intent Perceptual and uncheck Black Point Compensation.</p>

</blockquote>

<p>I agree about perceptual especially for wide gamut, but black point comp. comes into play when colorspaces are converted to others, and I believe the printer has its own 'physical colorspace' (based on ink combinations), and even if converting from a wide gamut colorspace, black point comp. comes into play.</p>

<p>I've always seen that it's recommended to turn it on.</p>

[teeuwen]

<p>> I've always seen that it's recommended to turn it on.<br>

From Adobe:<br />"Color conversion using Perceptual intent already maps source white to destination white and source black to destination black. Because this mapping preserves the relationships of the shades, it is unlikely that a whole shadow section will be mapped to the same black value. Therefore, BPC should not be necessary. BPC is available, however, for this rendering intent, to be used with malformed profiles. For a given picture, the user can decide whether using BPC improves the color conversion and can select it or deselect it accordingly.<br />BPC is available for color conversion using the Saturation intent. As with Perceptual intent, the user may or may not find that selecting BPC improves the conversion of a given image."<br /> <br />h**p://www.color.org/AdobeBPC.pdf<br /> <br />See also Eric Chan's (Adobe engineer) pages on the 3800 and especially his "Epson 3800 ICC Profiles for the Advanced B&W Photo Driver"<br /> <br />h**p://people.csail.mit.edu/ericchan/ <br /> </p>

[jtk]

<p>OK..I yeild...I'll get with the program.</p>

<p>I'll get the necessary gizmo.</p>

<p>For non-professional (low volume) use is there anything more cost-effective for monitor calibration than Eye-1 LT? </p>

[patricklavoie]

<p><em>isn't Epson's canned profile for its paper the same as the "correct icc profile" for that paper?</em><br>

<em><br /> </em><br>

Yes, if you use the one in Photoshop its the same, plus with this method you could install other icc profile from other manufacturer and acces them.</p>

<p><em>___</em></p>

<p>I always use relative coorimetric and black point ON personnaly.<br>

<em>___</em><br>

<em><br /> </em><br>

<em>is there anything more cost-effective for monitor calibration than Eye-1 LT</em></p>

<p>Spider3 pro<em><br /> </em><br>

<em><br /> </em></p>

[patricklavoie]

<p><em>The problem is that my prints do match my monitor closely...upload raw to LR or<br /> PS, adjust/retouch to taste, print the way I described=handsome results with<br /> one or two 5X7 tests in most cases, more if I'm dealing with subtle skin tones.</em></p>

<p>Oh, also, with the correct color managed method, you wont need test (at least i dont need test), as you know im dealing with skin tone everyday..and with a color calibrated monitor and workflow, what i see is pretty close to what i get : )</p>

[patricklavoie]

<p>Steve, just for your info;</p>

<p>1_theres is not one problem i cant fix directly in RGB, and without starting a long discusion on the subject, i never encounter a image that need to be fix in LAB. A lot of time when people talk about the L channel in LAB, they could do the same as doing the crrection using LUMINOSITY as the blending mode; same result, no information lost (Bruce Fraser in one of is book prove that going from RGB > LAB > RGB remove some information in the histogram)</p>

<p>2_There is no need to double the background and set it to softlight to acquire the effect you talk about, because doing so it will double the weight of the document..Simply create a empty curve and set it to softlight..same effect no add weight (in fact i use this method on almost all my shot, try all the blending mode to see the effect)</p>

<p> </p>

[steve_swinehart]

<blockquote>

<p>theres is not one problem i cant fix directly in RGB</p>

 

</blockquote>

<p>Well big guy....out here in the wild, wild west....we have lots of areas with little real color saturation. What you see (perceive) with your eyes is not what's recorded by the camera or camera and film (if you use film). Boosting saturation, or even picking specific colors in Hue and Saturation controls cannot get the same color change as using LAB and the process I've described. The difference being is what you're really doing is changing the color contrast and NOT the saturation while doing NOTHING to the luminance contrast.</p>

<p>You also might want to look at, "Photoshop LAB Color: The Canyon Conundrum and Other Adventures in the Most Powerful Colorspace," by Dan Margulis. That's where I picked up the LAB color adjustment method. </p>

<p>I found the book specifically because of the "Canyon Conundrum" which is a photo of a very flat colored canyon in the western US. He walks you through using RGB adjustments and then shows you the difference between the image in RGB and then when adjustments are made in LAB color. There is a marked difference and proof that, despite your preconceptions (and mine too prior to trying the method) - you really can't do everything in RGB.</p>

<blockquote>

<p> (Bruce Fraser in one of is book prove that going from RGB > LAB > RGB remove some information in the histogram)</p>

 

</blockquote>

<p>Nope...not true. With all due respect to Bruce Fraser, if you work in 16 bits there is no information lost. I've taken images from RGB to LAB and back 12 or more times just to test this exact claim... if I can't see it in my images on my computer it doesn't count. What Fraser tested was on older versions of PS. The latest versions don't do anything when converting from RGB to LAB that I can see in the final image when printed --- THAT'S what counts for me.</p>

 

<blockquote>

<p >There is no need to double the background and set it to softlight to acquire the effect you talk about, because doing so it will double the weight of the document</p>

<p > </p>

</blockquote>

<p>Who cares about the "weight" of the document in processing? I certainly don't as I'm going to flatten and convert to a TIFF, and save the TIFF before printing anyway - but, thanks for the tip - that's the nifty thing about PS there's all kinds of different ways to do things. </p>

 

[patricklavoie]

<p><em>Who cares about the "weight" of the document in processing?</em></p>

<p>Me. Because working om a P45 image 16bit is 250meg to start with, so by doubling the background just to pop the contrast will bring it to 500meg, then you need to double the background as a safty net when working in fashion so you can show the before and after to the photographer..now your image is 750meg...when it could only be 500meg is you didtn double the first one : ) The good thing is i work with Ps for such a long time that i have learn that even if there is multiple way of doing things, there is only 1 good way to do it fast, to the point, and with the same result as any other way that take more time or add weight. Im glad that now you know this one, it will make your file smaller ; )</p>

<p>As for LAB, again, even if you live on the moon i can prove you that i can get anything i want from your wild west shot with a simple 2-3 curve and blenfding mode ans mask most of the time; just a matter of knowing what does what. With all the respect i have for D.Margulis, i have the chance to learn from him and Michael Keiran back in the days, in Toronto, heavy multiple weekend formation where whe learn how to color correct a image on a bw monitor..by numbers, on a CMYK file! And i still stand to my point about i dont see up until today what can be done in LAB versus RGB<em>.</em> By using the hue saturation, color, selective color, curve and using the rigth blending mode you can achieve so much that you dont need to go into LAB..when you know how.</p>

<p>I will be more than happy to prove it to you if you care of sending me a raw that you already have work in LAB so i can have a reference of what you want, and give you the same result in no time just with RGB. For the record, i totally understand your point and fully understand LAB even if i dont see the point of using it.</p>

<p>I dotn want this thread to turn about rgb vs lab, so whe will keep it at that if you want, or create another thread about it?</p>

<p>Waiting for your wild wild west shot amigo : )</p>

<p>Big Mean Guy..grrrrrrrrr (is 6 feet - 290 pound qualify me as a big guy?LOL)</p><div></div>

[steve_swinehart]

<blockquote>

<p>By using the hue saturation, color, selective color, curve and using the rigth blending mode you can achieve so much that you dont need to go into LAB..when you know how.</p>

</blockquote>

<p>Whatever works for you...I also know all about hue / saturation, selective color, curves, masks, the "right" blending mode, etc. I prefer to use the LAB adjustments on some images as it is faster and gets better results. It's just another tool in the tool box and not needed or applicable for all images. As always ... everyone is free to use whatever works best for them. For me (and NOT for you) - I find the LAB adjustment method the best for some of the images I make....as the curves, hue / saturation, etc. do NOT get the same effect and are much slower.</p>

 

<blockquote>

<p>Because working om a P45 image 16bit is 250meg to start with, so by doubling the background just to pop the contrast will bring it to 500meg</p>

 

</blockquote>

<p>I understand the problem...really. I work on scanned 4x5 film...using a mega-slow, 8-year old computer...I just put up with the extra time to redraw the image.</p>

[jtk]

<p>Patrick, Thanks. I yeild. I'll get calibrated. </p>

<p>...and I'll buy Spyder Pro 3 because I've never lost by taking your advice. It does piss me off though...yet another learning curve!</p>

<p> Speaking of pissing...pardon me for interrupting the Lab/RGB exchange.</p>

 

[patricklavoie]

<p>With that last answer Steve i totally agree with you; to each is own way if it make your workflow faster for sure. It just when people say that this is better than this..without fully knwoing the whole portrait (not pointing you here).</p>

<p>I still want you to send me a wild west images, it could be fun for both of us to see the result? Take it as a image that you will have retouched by a pro, and i will take it as a image i retouched from someone in the US client!</p>