On Monday 07 August 2006 05:27, Cory Papenfuss wrote: > > (lit by a 5000k light), and then maybe converting it > > to TIFF and use SCARSE (www.scarse.org) or whatever. > > Would that be already sufficient to get good results? > > Or at least beeter than sRGB? Not sure about it... > > > > It seems that it is still not too common for manufacturers > > of cameras to simply provide ICC profiles? > > Although I'm still a bit ignorant of all the intricies of color > management, as I understand it the problem is that it's impossible to get > an accurate ICC profile for a camera unless the lighting is held constant. > ICC profiles translate into (and out of) an *absolute* colorspace (like > Lab or XYZ)... that isn't possible unless you know the white point of the > lighting of the shot. > > I wish that they would at least provide the profiles for their > built-in settings... "Sunny," "Shady," "Tungsten," etc.
I know that Nikon uses only one profile for all of these. The normal raw conversion process applies the white point BEFORE it does Bayer integration and then applies the color profile. I know this is how UFRAW works. I have found that getting a good direct sun light profile for my DSLR is all I need for light sources that have a smooth spectrum since it is always applied to the image after it has been corrected to something close to D50. This covers all of your natural lighting plus flash and incandescent lighting. Lighting that does not have a smooth spectrum such as florescent(or mixed florescent and sun/incandescent/flash) lighting is a different matter since these lighting sources will have significant spikes and drop outs in the spectrum and it also varies for different brands and models of bulbs. Because of this I have profiles for locations with this type of lighting where I shoot with some regularity. I will also shoot an IT8 target at locations where I don't know what the lighting is and where the shots are important. So an IT8 target is part of my camera kit as is a standard gray card (which I have had in my camera kit for decades) for setting the white point of the camera. It only takes a few minutes and it can result is a huge improvement in the quality of the photos. Using LProf I can create a profile from the Target images in about 2 minutes and applying it in UFRAW only takes seconds to setup. > > Oh, and until Gimp does >8 bits, it's pretty useless for accurate > photographic work. With only 8-bits you must read in a photo in a > gamma-corrected space (like sRGB or AdobeRGB). If you use a gamut wide > enough to capture everything (like ProRGB or its ilk), you'll have way too > much posterization to be useful for photos. I mostly agree and that is why I am currently using CinePaint. Wide gamut color spaces should be used only with higher bit depth images. Also ProPhotoRGB, which is recommended by some sources but not me, has a gamut that is much wider than is needed by any existing photographic process and actually makes this issue more of a concern because of this overly wide gamut. I use BetaRGB because it has been specifically designed to have enough gamut to not lose information from any existing photographic process but only just big enough to do this. Typical photos only use 50% to 70% of the space of ProPhotoRGB and at least 20% of ProPhotoRGB is for colors that humans can not even see. So an additional issue with ProPhotoRGB is that you can adjust some colors into parts of the colors space that are not even visible. For more information see the Bruce Lindboom's site (http://www.brucelindbloom.com/), under info, Information About RGB Working Spaces. The thing you need to do is keep your images in a higher bit depth format until conversion for output (the very last processing stage) such as printing or going to the web (or emailing to someone) and treat the 8 bit/channel images as output and not as the real image. I always leave the original high bit depth version alone (other than basic corrections) for archival purposes. I also archive the RAW files since these are my "digital" negatives and I have found that over time the RAW processing software and profile generation software have improved and this allows for the possibility of going back and getting even better results. This is sort of like keeping (real) negatives or slides because you might some day get a better lens for your enlarger or Kodak, Fiji or Ilford might, as they have over the years, make a better photographic paper and/or improve the processing of photographic paper. As an example Wolf Faust as told me that he is now experimenting with new photo paper and new processes that result in the DMAX of his reflective IT8.7 targets going from around 2.0 to around 2.8. This is more than a two stop increase in the dynamic range of the patches on these targets which results in the possibility of increased profile quality and in particular improved modeling of the darker regions of the tone curves. I am not sure when or if these new targets will be made widely available. Hal ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Lcms-user mailing list Lcms-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lcms-user
