Luminous Landscape Forum
Equipment & Techniques => Digital Cameras & Shooting Techniques => Topic started by: on June 29, 2005, 11:03:21 am
-
There is a better way Jack, but more on that in a few weeks.
Michael
-
If you're using Adobe Camera RAW, you can optimize the calibration settings for your camera by following the procedure outlined here (http://www.visual-vacations.com/ColorManagement/cm_101/04camera.htm). While not the same as a true custom profile, it will make a very significant improvement in the accuracy of the colors you get from ACR.
-
A Color Checker doesn't cost all that much, and I think you'll be surprised at the difference it makes.
It's something I would definately like to own, maybe I'm being a cheapskate or maybe I'm looking in the wrong places but £50/$100 seems a lot for a piece of card. Does anyone know of any cheaper alternatives or should I just suck-it-up?
-
You folks have a valid point. Here's my game plan. Next week I have vacation. I am going to read Fraser ("Camera Raw with PSCS" or "Color Management"? I have both.) and absorb this technique. I'll order the color checker and when it arrives have a calibration event. I am quite open to improvement. Everything I learn here has helped.
I figure I'll have this all accomplished in a few weeks. I'll ask questions if I need to, OK?
I'm off to the Alps in a couple of days (happy happy!) and won't be back for a couple of weeks, but I'm happy to answer anything I can when I get back if you get started before then. Using the instructions and script from Jonathan's web site was relatively simple, though, so I don't imagine you'll have any problems there (at least I didn't) - if you have any questions beyond Jonathan's instructions, he's much more likely to be able to answer them than me (but I can try if he's not around).
Lisa
:cool:
-
there is something i dont get. if we profile our cameras, then what's the use of the WhiBal? i mean, after all, isn't that WhiBal that gives correct color? what's the point of using a WhiBal when you'll leave the White Balance as As Shot anyways?
White Balance has only one (in camera) or two (in ACR) parameters you can adjust. ACR's camera calibration has a bunch more (I don't have it in front of me right now, but it's something more like eight). One or two helps get closer to "right" (for some definition of right), but it isn't enough to cover all bases, color-wise. Ideally, you should use calibration to get a reasonable baseline image where the colors don't look "wrong" in some ill-defined way, then you can use white balance to adjust them to your taste.
Lisa
-
Where in the world will I find a neutral location to take the test shots? Don't you need a predictable light source?
All I did was to pick a bright sunny day when the sun was high, lay a white sheet out on my deck (to have neutral-colored surroundings), put the color checker in the middle of the sheet, and stand over it with the camera pointing down at it (without getting any shadows on it). It's true that not all images are taken under sunny conditions, but that's the most common for me, so that's what I went with for my calibration. Maybe not rocket science, but it worked OK for me. :cool:
Lisa
-
For a comprehensive explanation, check out this link: http://www.photoshopforphotographers.com/p...ad/movie-10.pdf (http://www.photoshopforphotographers.com/pscs2/download/movie-10.pdf)
Here is the setup I used. Grey seamless background, (2) 400WS Studio Stobes with softboxes at 45 degrees to the color checker plaque. The exposure that worked was f/11, stobes at 1/4 power.
(http://boku.smugmug.com/photos/27626334-M.jpg)
The RAW file, before the profile creation, rendered as follows:
(http://boku.smugmug.com/photos/27626333-M.jpg)
The Gretag-Macbeth reference file (from Jonathan's website) for comparison:
(http://boku.smugmug.com/photos/27626332-M.jpg)
The script gave me the following results. I ignored everything BEFORE Shadow Tint, per Jonathans recommendation.
(http://boku.smugmug.com/photos/27626336-M.jpg)
-
Is it possible to custom profile your camera? I have a Rebel.
-
I can wait...
But I've tried the mask if that's where you're going... It works, but is not an ideal solution when the corners are significantly off -- at least IMO.
However FWIW, I have found a capture method that generates a nearly perfect capture -- right now I am within 2 points (rgb 8-bit) across all of the corners and center and I think I have the solution to knock that down to 1 or 0. And even if I can't knock that back to 0, I feel 2 points is probably close enough for the mask
-
But what I don't understand is, in the course of landscape photography, are we objective to that extent? I don't know about you folks, but I rather enjoy interpretive color and tone adjustment to bring out a mood or my impression of the original rather than a rigid reading of the original. I feel more like creating than recording. I contemplate this from the very start, even before I set up the shot. For me, it is about what can I do with this raw scene to render it that way I imagine.
Given that subjective approach, what would I get out of camera profiling?
Simple. Your color adjustments will be "creative" tweaks done to accommodate your creative vision, rather than frustrating and usually time-consuming attempts to "fix" what the camera failed to get right in the first place. If the color is dead-on accurate to the original subject, then bumping up saturation to taste is simple and easy. But if the blues have a magenta cast and the greens are too yellow out of camera, you have to waste time dealing with that before increasing saturation will yield the desired result. And if you ever photograph people, you'll appreciate the value of good skin tones straight out-of-camera, even if they're just family snapshots.
-
Bob, I'd like to encourage you to at least give ACR calibration a try. A Color Checker doesn't cost all that much, and I think you'll be surprised at the difference it makes.
Yeah - I bought Fraser's books and never read them. I really need to.
-
A worthwhile read by Bruce Fraser, (find in PhotoshopNews.com)
raw 3.1 professional custom profiles for highend digital cameras (http://photoshopnews.com/2005/06/29/from-the-adobe-camera-raw-user-to-user-forum/)
-
Just an aside
Is it possible to set the tonal range for a specific image manually by doing a test shot of your subject with a colour checker in it, and taking the dropper and sampling the neutral grey square to give you a reading of exactly where your midtone should lie on the histogram? I wasn't sure if this would give you an accurate reading as its relationship with the light source would affect its tone, am I right in thinking this or would it be close enough if the colour checker was facing the camera directly?
I guess that calibration would eliminate the need for this but I'm just curious; when photographing my drawings I have to take a guess where to put the grey point on the histogram so I'd like to make it a more consistant process.
-
That's more or less what I did, but I put the Color Checker on an easel. Put the easel in the middle of a reasonably large sheet of neutral cloth or paper facing direct sun on a clear day and you'll get an emininently useful calibration shot.
-
I have one in preparation, but it might be a month or two till it's done. Other priorities I'm afraid.
Michael
-
Yes it is, but the software and hardware needed to do so properly costs as much as your camera. Less expensive systems simply don't do the job well enough.
I'll have a review and a look at what this is all about next month.
Michael
-
Hi Jack
I hope you'll share your methods once you're happy with them. I use my camera to photograph my artwork, which includes drawings. They are irritating things to light and without fail end up with a subtle shadow passing across the drawing. Though not that noticeable on screen it's very clear in the print. I've developed a method in photoshop to remove it and I can produce an extremely good reproduction with a uniform tone to the paper of the drawing, it would always be nice though to get the picture perfect in the first place.
-
Bob, I'd like to encourage you to at least give ACR calibration a try. A Color Checker doesn't cost all that much, and I think you'll be surprised at the difference it makes.
-
But what I don't understand is, in the course of landscape photography, are we objective to that extent? I don't know about you folks, but I rather enjoy interpretive color and tone adjustment to bring out a mood or my impression of the original rather than a rigid reading of the original. I feel more like creating than recording. I contemplate this from the very start, even before I set up the shot. For me, it is about what can I do with this raw scene to render it that way I imagine.
Given that subjective approach, what would I get out of camera profiling?
Bob, remember that thread of mine awhile back where I was asking for a technical critique and you said that the colors looked "muddy" compared with my earlier film work? Well, the single biggest factor (by far) in getting rid of that muddiness was doing the camera calibration thing in ACR that Jonathan suggested. It's hard to color correct (either to what's accurate or to what you want) when things just don't look quite right, but you're not sure why.
Bob, I'd like to encourage you to at least give ACR calibration a try. A Color Checker doesn't cost all that much, and I think you'll be surprised at the difference it makes.
I second the motion!
Lisa
Lisa,
You folks have a valid point. Here's my game plan. Next week I have vacation. I am going to read Fraser ("Camera Raw with PSCS" or "Color Management"? I have both.) and absorb this technique. I'll order the color checker and when it arrives have a calibration event. I am quite open to improvement. Everything I learn here has helped.
I figure I'll have this all accomplished in a few weeks. I'll ask questions if I need to, OK?
-
Bob, I'll be happy to answer any questions about ACR calibration that don't involve focusing "beyond infinty".
-
there is something i dont get. if we profile our cameras, then what's the use of the WhiBal? i mean, after all, isn't that WhiBal that gives correct color? what's the point of using a WhiBal when you'll leave the White Balance as As Shot anyways?
WhiBal sets the color balance for a particular scene. A camera profile normalizes the spectral and tonal response of the sensor and circuitry in the camera. Two different things.
-
Is it possible to set the tonal range for a specific image manually by doing a test shot of your subject with a colour checker in it, and taking the dropper and sampling the neutral grey square to give you a reading of exactly where your midtone should lie on the histogram? I wasn't sure if this would give you an accurate reading as its relationship with the light source would affect its tone, am I right in thinking this or would it be close enough if the colour checker was facing the camera directly?
I guess that calibration would eliminate the need for this but I'm just curious; when photographing my drawings I have to take a guess where to put the grey point on the histogram so I'd like to make it a more consistant process.
Some people actually do this (taking an image with the color checker in it before the "real" image), but it has less to do with calibration than with helping to set the white balance afterward in the RAW conversion, to get rid of any color casts associated with the local lighting conditions and find an appropriate white balance setting.
I think that it would not replace calibration, though, as it would be difficult to figure out how to adjust the colors of your image to match the chart without a lot of trial and error.
Lisa
-
It seems from what I've read that the main problem with profiling your camera is producing a uniformly lit test subject.
Would it not be possible to create (either by the manufacturers of profiling software or at home using a technique like the one Jonathan W. outlined in a different thread on tinted gels) a transparent colour chart which could be placed on a light box? This would give uniform lighting without much hassle, wouldn't it?
-
And even if you have the proper software and hardware, capturing a perfectly (evenly) lit image of the McBeth chart is supremely difficult.
-
I'd like to propose a philosophical question regarding this...
I see the point of monitor and printer color management - getting results that you expect consistently. I feel that camera profiling is more about getting authentic colors that record the scene in a way that the final image can closely match the original scene. I also see camera profiling as a way of getting exposures from mutiple cameras to match.
But what I don't understand is, in the course of landscape photography, are we objective to that extent? I don't know about you folks, but I rather enjoy interpretive color and tone adjustment to bring out a mood or my impression of the original rather than a rigid reading of the original. I feel more like creating than recording. I contemplate this from the very start, even before I set up the shot. For me, it is about what can I do with this raw scene to render it that way I imagine.
Given that subjective approach, what would I get out of camera profiling?
-
But what I don't understand is, in the course of landscape photography, are we objective to that extent? I don't know about you folks, but I rather enjoy interpretive color and tone adjustment to bring out a mood or my impression of the original rather than a rigid reading of the original. I feel more like creating than recording. I contemplate this from the very start, even before I set up the shot. For me, it is about what can I do with this raw scene to render it that way I imagine.
Given that subjective approach, what would I get out of camera profiling?
Simple. Your color adjustments will be "creative" tweaks done to accommodate your creative vision, rather than frustrating and usually time-consuming attempts to "fix" what the camera failed to get right in the first place. If the color is dead-on accurate to the original subject, then bumping up saturation to taste is simple and easy. But if the blues have a magenta cast and the greens are too yellow out of camera, you have to waste time dealing with that before increasing saturation will yield the desired result. And if you ever photograph people, you'll appreciate the value of good skin tones straight out-of-camera, even if they're just family snapshots.
I guess I just haven't seen odd color shifts from my cameras that I haven't attributed to reflected surfaces in the scene or the lighting color itself. I know there is nothing I detect that is constant regardless of the scene conditions.
I get the point about skin tones. If a camera has native color shift, profiling would at least eliminate the need to normailize that for every shot. Skin tones are much more sensitive to fine color adjustments. Our eyes are trained to expect certain renditions. I no longer shoot people (except for downsizing), so I don't give it much thought.
-
But what I don't understand is, in the course of landscape photography, are we objective to that extent? I don't know about you folks, but I rather enjoy interpretive color and tone adjustment to bring out a mood or my impression of the original rather than a rigid reading of the original. I feel more like creating than recording. I contemplate this from the very start, even before I set up the shot. For me, it is about what can I do with this raw scene to render it that way I imagine.
Given that subjective approach, what would I get out of camera profiling?
Bob, remember that thread of mine awhile back where I was asking for a technical critique and you said that the colors looked "muddy" compared with my earlier film work? Well, the single biggest factor (by far) in getting rid of that muddiness was doing the camera calibration thing in ACR that Jonathan suggested. It's hard to color correct (either to what's accurate or to what you want) when things just don't look quite right, but you're not sure why.
Bob, I'd like to encourage you to at least give ACR calibration a try. A Color Checker doesn't cost all that much, and I think you'll be surprised at the difference it makes.
I second the motion!
Lisa
-
there is something i dont get. if we profile our cameras, then what's the use of the WhiBal? i mean, after all, isn't that WhiBal that gives correct color? what's the point of using a WhiBal when you'll leave the White Balance as As Shot anyways?
-
Bob, I'll be happy to answer any questions about ACR calibration that don't involve focusing "beyond infinty".
Thanks. I've been this side of infinity for nearly 18 years now.
"Purple haze runnin through my brain."
"scuse me while I kiss the sky"
-
OK, I ordered the color checker, read Jonathan's article, stopped by the website of the guy with the script. I'll be doing this.
Where in the world will I find a neutral location to take the test shots? Don't you need a predictable light source?
What if I set this up in my studio using photogrey seamless for the background and studio strobes for illumination? Will that work OK? I just can't imaging getting anything accomplished in daylight outdoors. The light varies and there is no neutral environment.
-
But what I don't understand is, in the course of landscape photography, are we objective to that extent? I don't know about you folks, but I rather enjoy interpretive color and tone adjustment to bring out a mood or my impression of the original rather than a rigid reading of the original. I feel more like creating than recording. I contemplate this from the very start, even before I set up the shot. For me, it is about what can I do with this raw scene to render it that way I imagine.
Given that subjective approach, what would I get out of camera profiling?
Bob, remember that thread of mine awhile back where I was asking for a technical critique and you said that the colors looked "muddy" compared with my earlier film work? Well, the single biggest factor (by far) in getting rid of that muddiness was doing the camera calibration thing in ACR that Jonathan suggested. It's hard to color correct (either to what's accurate or to what you want) when things just don't look quite right, but you're not sure why.
Bob, I'd like to encourage you to at least give ACR calibration a try. A Color Checker doesn't cost all that much, and I think you'll be surprised at the difference it makes.
I second the motion!
Lisa
Jon & Lisa,
I did it. Things have improved enough to make me a believer. The script took about 40 minutes to run. I set up RAW as explained at Jonathan's website. Tried a few files - great results. I won't be doing as much "doctoring" any more.
Thank you both.
I love this place.
Now - the question is - what will Michael be writing an article about? Originally he mentioned he had something forthcoming re: camera profiling.
-
Browsing around for camera profiling I found http://www.qpcard.se (http://www.qpcard.se) .
They offer the grey card QPcard 101, the color checker QPcard 201 and the free program QPcolorsoft 501 (product support > downloads) to create a profile.
So basically all that I need are the two cards. The color checker will be taken under controlled light conditions in the studio to create a profile, and shooting outside I can use the grey card to adjust the WBT settings to a custom setting right in the camera or I can correct the white balance later converting the raw file.
Anybody uses the solution of QPcard with success?
-
Anybody uses the solution of QPcard with success?
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=130987\")
Haven't tried them but I don't see anything at all unique here. More white and gray cards.
The best white card I've seen so far in terms of the neutrality and brightness which I've measured with an EyeOne (among other targets for an upcoming article in DPP) is this one:
[a href=\"http://www.babelcolor.com/main_level/White_Target.htm]http://www.babelcolor.com/main_level/White_Target.htm[/url]
The LAB values are nearly prefect.
-
Thanks for the link, Andrew. The White Target is a smart solution.
Browsing through the BabelColor website i noticed the document "RGB coordinates of the Macbeth ColorChecker".
My dumb question: If I reproduce the ColorChecker in my studio isnt't it possible to create a camera profile having the given RGB coordinates for the ColorChecker?
-
My dumb question: If I reproduce the ColorChecker in my studio isnt't it possible to create a camera profile having the given RGB coordinates for the ColorChecker?
[a href=\"index.php?act=findpost&pid=131155\"][{POST_SNAPBACK}][/a]
Using Camera Raw or Lightroom, you can tweak the existing profiles to produce a numeric match of the various patches of the Macbeth.
Some older profile packages used the Macbeth target to build profiles. Not what it was designed for, not enough patches and like ALL camera profile packages that use a target, the idea is to define the gamut of the capture device based on the target. That's a poor approach, treating a camera like a scanner. A scanner profile can be successfully built from a target made of film, that target exhibits the same gamut as what you hope to match. Not so with a camera. The day we have targets that can define all scene gamut, I suspect we'll also be using transporters to get about as well.
-
Anybody has tested or uses inCamera from http://www.pictocolor.com/incamera.htm (http://www.pictocolor.com/incamera.htm) ?
Review from 2002: http://ftp.luminous-landscape.com/reviews/...amera-pro.shtml (http://ftp.luminous-landscape.com/reviews/software/incamera-pro.shtml)
-
The ACR script uses the normal colour checker card. I've got a version of the script for the ColorChecker SG card that I developed when I was reviewing the camera profiling in Eye One Match.
The data you need (and a copy of the script) is at:
www.northlight-images.co.uk/article_pages/eye_one_photo_SG.html (http://www.northlight-images.co.uk/article_pages/eye_one_photo_SG.html)
For most of -my- commercial work I find camera profiles more hassle than they're worth, but I do have some for some locations where I've regularly worked that have particularly nasty mixes of 'energy saving' lightbulbs everywhere...
I also noticed that the current version of DxO Optics Pro (review) (http://www.northlight-images.co.uk/reviews/dxo_optics_pro4.html) has a raw processing mode specifically to produce images that you can then use for profile creation, and allows you to use the profile for conversion later.
-
Like ALL camera profile packages that use a target, the idea is to define the gamut of the capture device based on the target. That's a poor approach, treating a camera like a scanner. A scanner profile can be successfully built from a target made of film, that target exhibits the same gamut as what you hope to match. Not so with a camera. The day we have targets that can define all scene gamut, I suspect we'll also be using transporters to get about as well.
[a href=\"index.php?act=findpost&pid=131166\"][{POST_SNAPBACK}][/a]
I'm not so sure it's all about the gamut.
My T-shirt says "It's all about Me!"
I've distributed profiles made with Xrite and Basiccolor profiling packages. All have their uses. A decent profile for controlled lighting is someting a buyer of these packages can reasonably expect. A profile for general use for a camera needs skill to create.
Edmund