Color retouching

[benjamin_kim2]

<p>Well, I've doing photography works for 3 years in school but I felt that my color retouching skill is not good at all. It was obvious to other people but not me. I wish to get train for color retouching but I have no idea what to do. I know experiences will grant me more skills and I would like to hear some advices. I basically seek color retouching skill for portrait, still life, and landscape.</p>

[Jochen_S]

<p>What problem are you talking about? - Plastic surgery in Photoshop?<br>

Or plain and simple getting the impression of an otherwise sufficiently perfect picture right, as in: "I photographed a slice of bread and it just doesn't look yummy"?</p>

For the latter I suggest:

<ol>

<li>Medical color vision check</li>

<li>Calibrating your screen, workflow, light & everything and doublecheck the results against another set of devices. - Discuss them with advanced laymen / professionals too. - Do <strong>not</strong> trust any <em>single</em> device opinion or result.</li>

<li>Try and discuss! </li>

</ol>

<p>There are 3 nastiest colors: neutral gray, skin tones and food.<br>

A nasty exercise could be having one event covered by 3 different cameras and trying to profile them to a similar look. <br>

Or try tweaking photos taken in different light to match.<br>

That journey is long and hard. I guess it takes about 8 or 10 years in the color printing industry to get pretty good, <em>if</em> things run smooth. - I've worked there longer and bought into B&W... <br>

Only way I see: Make mistakes. - Get them explained. - Understand. - Fix them. Rinse, repeat & remain obsessed about the issue.<br>

Giant step on the side: Understand the miracle of "good enough".</p>

[Bill C]

[benjamin] " ...but I felt that my color retouching skill is not good at all. It was obvious to other people but not me."

 

Hi, this kinda' suggests to me that you MIGHT have some sort of color vision deficiency, so might be worth looking into. I wouldn't trust the

computer monitor-based tests, unless possibly you test side-by-side with someone with known-good color vision. The obvious test used to

be (still is?) the Ishihara color-blindness tests - a set of cards printed with multicolored dots; you try to see a vague number in the pattern,

but with certain types of color-blindness, you see a different number. If you're still in school, they might possibly have a set?

 

Otherwise, I wonder if your "judgment" is drifting off after a while. This is a well-known issue in pro labs; the way to deal with this is to have

well-corrected reference images around, and to periodically use these to "reset" your internal references. If you don't do this, it's real easy

for your color judgment to get corrupted - if you look at bad color long enough, it starts to look okay, or even good, until you look back at

your references.

 

Aside from this, much of what Jochen mentions. Can you elaborate a little bit more about the exact issues you have?

[david_henderson]

<p>I can see four potential issues, and you have to determine which is your problem</p>

<ul>

<li>You don't have mastery of the tools you can use to change colours in a picture. So you know exactly what you want to achieve but don't have the knowledge or skills in the use of the tools to achieve it.</li>

</ul>

 

<ul>

<li>You don't see what other people see. I'm sure that most of us see differently to a small degree; and many see differently with each eye. But I think there's a core of broad similarity that most people have that means that they can have a workable agreement on what's broadly right or acceptable. Many apparent differences turn out to be semantic- eg. I put a "grey" label on something that you'd call blue. But maybe your vision actually falls further away, at least on certain colours. </li>

</ul>

 

<ul>

<li>Your judgement of colour is inadequate/unreliable/variable. Maybe you spend time editing a picture, think you've got there but then when you go back to it later it looks bad. Or maybe you think you've got there and other people don't agree even though you're probably seeing similar things. I do have sympathy with the view that if you agonise over something for too long, it starts to look bad whether it is or it isn't. And the people that say that without getting feedback from others you can go on interminably doing things that look more right to you the more you do them, but don't fit with other people's perceptions of "right". </li>

</ul>

 

<ul>

<li>Maybe you're working in different lighting that makes what you've done seem different when you look at it again. I find that I can't carry out serious editing in artificial light. My screen is calibrated, my office is north -facing and so I get no direct sunlight. But the lighting in my office isn't right for photo-editing, and if I change colours at night, you can be sure I'll be changing them again in the morning. Or are you working on a laptop in a variety of environments? </li>

</ul>

<p>I'm assuming in all of this that you're working on a calibrated screen using colour-compliant software. <br>

Some of these things are hard to fix, but you won't get too far without diagnosing which problem you've got. </p>

 

[digitaldog]

<p>I suggest you ask over on these forums: http://www.retouchpro.com</p>

[Spearhead]

<p>Share some examples of yours, it may help to figure out what the problems are.</p>

[benjamin_kim2]

<p>Thank you for responding.</p>

<p>I have a calibrated BenQ SW2700PT monitor and I work in pure dark room in order to see colors accurately. I guess I have lack of knowledges and skills of using photoshop. But I can't view colors exactly on LR too. Perhaps Im using what Im not familiar with. Canon cameras seem to be very different with Nikon, Pentax, and Sony base on my experiences. It wasn't that hard to edit images with Pentax and LR mostly. </p>

<p>But still, my friends and other people think that I have lack of color visions since they know and get right colors with LR and Photoshop in simple way. So I tried scanning and retouching color films which is the most difficult task to do. Not able to get what The Find Lab did. </p>

<p>I assume that I have lack of software skill, lack of color vision, and lack of keen eyes. I usually check and see magazines, posters, and other works for developing skills. Well I dont know exactly why am I suffering lack of skills even I spent 3 years of using LR and Photoshop.</p>

[Tim_Lookingbill]

<p>Like Jeff said post some sample images that drew criticism on your color vision and editing skills. Results someone sees vs what others sees is quite subjective.</p>

[Bill C]

I would suggest, first of all, stop working in the dark - your eyes will adapt differently and you'll lose your references.

 

[benjamin] "... my friends and other people think that I have lack of color visions since they know and get right colors with

LR and Photoshop in simple way."

 

If they're doing this on the same monitor this you are using, it really seems like it may be a color-vision deficiency, aka

color blindness. It's probably worth checking it out. This doesn't necessarily mean that you can't do this sort of work, but

you would probably want to learn to rely more on "the numbers," rather than your visual judgment, for "correct" color. And

perhaps have someone else double check your color edits.

 

I don't know where you find someone in the real world to check your color vision, but the Ishihara tests seem to be selling

on Amazon for about $70 or $80; this may be cheaper than a specialist, I don't know. In the very large lab where I spent

some years, my department screened all of the candidates for color correction jobs - probably near a hundred or so

people during my time as the manager. We considered the Ishihara tests too crude for our purposes, using instead the

Farnsworth-Munsell 100-hue test for "fineness" of color discrimination. But I doubt you'll find it in general use - it's more of

a specialist tool for certain industries. Plus it's fairly expensive - 5 or 10 times more than the Ishihara set. Anyway, the

Ishihara tests will find any significant color blindness. Best of luck.

[digitaldog]

<blockquote>

<p>I have a calibrated BenQ SW2700PT monitor and I work in pure dark room in order to see colors accurately.<br>

I would suggest, first of all, stop working in the dark - your eyes will adapt differently and you'll lose your references.</p>

</blockquote>

<p>Actually the OP is spot on, one should <em>attempt</em> to work in as dark a condition when editing images on-screen as possible. You can't go to low (you can go too high). Any ambient light that strikes the display affects our perception of black and calibrating black on a display (when possible) is super important! <br>

http://digitaldog.net/files/BlackisBack.pdf<br>

You can have your color vision check using a similar process found on this X-rite web site which would be a more accurate indicator but the site is kind of fun and not entirely without some small merit:<br>

<br />http://www.xrite.com/en/hue-test<br /><br>

Going to a vision specialist who has these actual tiles and having the test done is the way to really evaluate your color vision. </p>

[benjamin_kim2]

<p>I checked both and I got 2 for color vision test and no issue with a monitor.</p>

[Landrum Kelly]

<p>I don't mean to sound flippant, but perhaps you could still become a very good black and white photographer.</p>

<p>--Lannie</p>

[Jochen_S]

<p>I did the test too on a laptop and scored 2 by getting something in the blue wrong (its a tiles sorting test).<br>

I assume if your color vision was severely impaired you'd have scored worse. - My doctor did excessive Ishihara testing on me before I signed up for my apprenticeship. - Testing was health insurance covered written result would have been charged, but who needs such?<br>

Back to "calibrated(?)" screen: using one i-one on it regularly might be enough but I'd try to check it against another. - They seem good but not necessarily flawless.<br>

Anyhow your problem is as it sounds:</p>

<ul>

<li>Lack of experience and</li>

<li>Unfamiliar tools. +</li>

<li>Attempting the hardest job of them all. </li>

</ul>

<p>Do yourself a favor and try to understand color tweaking by reverse engineering / analyzing automated miracles at hand.<br>

I was pretty blown away by watching how much closer X-Rite's passport colorchecker & Lightroom plugin got the results from a mix of cameras (Leica Pentax & Samsung) - 80 Euro/$ well spent!<br>

I'm not using LR seriously and would recommend trying to analyze & <strong>understand</strong>, what a Colorchecker generated profile alters compared to the default setting. <br>

Sometimes its simple things ruining an image that are beyond our basic thinking. - In offset printing we learn pretty early that the C/M balance is important, learning to spot the impact of Y is a different issue and seems to happen later. <br>

I'm just tooting Bill's horn to "go by the numbers" first, until you are somewhat savvy to understand what a subtle change can do. <br>

I would not dare working in a dark room. I'd get myself some half decent D50 room illumination as a reference and look on a good image or subject to compare my screen to it. - It is hard to tweak one's results towards a proof but it is really way(!) harder to work without one and generate perfection. <br>

Upon your scans and different cameras: I assume results are kind of rigged. - Software devs usually have great color tweakers who try to develop camera lens combo specific profiles for serious software (Capture One, DxO etc.)<br>

You start to stumble if you for example import Samsung RAWs into the same simple software as Pentax although the camera is technically the same and just rebadged (K20D / GX20) results look pretty different since the software uses a far off generic or universal profile for Samsung and might have something customized for Pentax.<br>

I am trying to stress: 2 shots with kind of identical hardware should <em>hold</em> the same result and its just up to us to figure out how to get there by adjusting curves and presets. Besides that the two of us don't know how to do it yet, it should be possible. </p>

I tried scanning film once. - It was pretty horrible. - Scanners are old "professional" technology. - Maybe a tad worse than *ist D RAW files. - They could be tamed. Its just: SOOC JPGs from that old camera were quite a nightmare and far from usable. And with film of course all individual flaws of the emulsion neglected filtering for white balance processing mistakes etc. start playing a role and you have to compensate those by hand. I'd really waste a frame on my colorchecker and if its too late to do so, I'd look up what "I feel lucky" in Picasa altered in my file. - Sometimes it works but unfortunately not always. Anyhow the more you analyze the more you'll know.

Since you have critical friends: Ask them to<em> show</em> you on your machine how they 'd get stuff right or closer to right.

There is a time when we need guidance while not yet knowing what we are doing.

 

Good luck!