Hi,
Lots of food for thoughts here. Marie-Noelle, isn't the IIC file necessary for our settings in Digikam to turn the RAW into the right Jpeg (providing the screen has been calibrated). Then the colours obtained, whether for display on screen or printing would be delivered at least in compliance with the standards. For exemple, if you want to bring your USB key to the best shop in town to make a large prints for your wall with some pictures that you have tweaked as raw-turned-to-jpeg, it would have the right material at his/her end. And having a presumably calibrated printer, your prints would come out fine. Pierre _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
2012/7/20 Pierre Edelman <[hidden email]> Hi, Yes, it would be better to have an ICC profile, but it is your computer screen you need a profile for, so your post-treatment won't be impaired by a not-accurate enough rendering on your screen. A camera profile wouldn't be useful for that: each profile is used by the corresponding device to 'interpret' the image and render it faithfully; if you're not going to look at your picture on your camera (other than just for checking composition, histogram, and such), you don't need a profile for it. Marie-Noëlle PS : By the way, if you can read french, I might have some interesting links that explain this much better than I do. Then the colours obtained, whether for display on screen or printing -- Mes dernières photos sont dans ma galerie. Retrouvez-moi aussi sur mon blog. Et parcourez les Cévennes à ma façon avec Cévennes Plurielles, _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
On Fri, 20 Jul 2012 18:38:35 +0200
Marie-Noëlle Augendre <[hidden email]> wrote: > A camera profile wouldn't > be useful for that: each profile is used by the corresponding device > to 'interpret' the image and render it faithfully; if you're not > going to look at your picture on your camera (other than just for > checking composition, histogram, and such), you don't need a profile > for it. That is not the case, a camera profile is used to modify the colours saved in the RAW file when they are turned into what is displayed on the computer or stored in a colour-corrected TIFF, JPEG etc. The profile is produced by looking at the colours the camera shows for a known set of tiles in an image of a colour target, it will be specific to the lighting so you need more than one profile depending on the colour temperature you are shooting under. You will never see the corrected image on the camera's own screen unless you re-upload a colour-corrected JPEG to its storage. -- Brian Morrison _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Am 20.07.2012 19:35, schrieb Brian Morrison:
> On Fri, 20 Jul 2012 18:38:35 +0200 > Marie-Noëlle Augendre <[hidden email]> wrote: > >> A camera profile wouldn't >> be useful for that: each profile is used by the corresponding device >> to 'interpret' the image and render it faithfully; if you're not >> going to look at your picture on your camera (other than just for >> checking composition, histogram, and such), you don't need a profile >> for it. > > That is not the case, a camera profile is used to modify the colours > saved in the RAW file when they are turned into what is displayed on the > computer or stored in a colour-corrected TIFF, JPEG etc. Not quite right. The camera profile is used to transfer the data from raw processor (after demosaicing ...) into the working colour space (that are usually sRGB, prophotoRGB, eciRGB) which are transfered into the display (or storage or print) colour space (sRGB, AdobeRGB ...). The extra step with the working colour space is used to get a defined base for further transformation. > The profile is > produced by looking at the colours the camera shows for a known set of > tiles in an image of a colour target, it will be specific to the > lighting so you need more than one profile depending on the colour > temperature you are shooting under. It does depend on the spectral colour distribution of your light source and the filters and lenses you have in front of your sensor. > > You will never see the corrected image on the camera's own screen > unless you re-upload a colour-corrected JPEG to its storage. Even then you will not see the correct colour as no camera I know of has a calibrated display. Even if a camera would have a calibrated display it is not exact as the environmental light differs. Martin _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
On Fri, 20 Jul 2012 19:58:48 +0200
"Martin (KDE)" <[hidden email]> wrote: > > That is not the case, a camera profile is used to modify the colours > > saved in the RAW file when they are turned into what is displayed > > on the computer or stored in a colour-corrected TIFF, JPEG etc. > > Not quite right. The camera profile is used to transfer the data from > raw processor (after demosaicing ...) into the working colour space > (that are usually sRGB, prophotoRGB, eciRGB) which are transfered into > the display (or storage or print) colour space (sRGB, AdobeRGB ...). > The extra step with the working colour space is used to get a defined > base for further transformation. Yes indeed, you added the detail that I decided I would take a long time to type ;-) Colour spaces can get very confusing... -- Brian Morrison _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
In reply to this post by tosca
On 20/07/12 17:38, Marie-Noëlle Augendre wrote:
> each profile is used by the corresponding > device to 'interpret' the image and render it faithfully; A camera profile is an input profile whereas a monitor profile is an output profile. So a camera profile is used to adjust the raw data to allow for the colour response of the lens and sensor. This means that in the memory of the computer you have an accurate representation of what was seen. In order to see that accurate representation you need a monitor profile and calibrated monitor. A camera profile is not for calibrating the display on the camera LCD, that would be utterly pointless. So you see the use of tethering or not is completely irrelevant to the need to use a camera profile or not. Tethering is only about transferring the data directly to a computer or storing it on the camera memory card. Andrew _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
That's very clearly explained, thank you.
As an 'artistic' photographer, I had never seen to point to investigate this further. My guess is this kind of input profile will mainly be useful for photographers that want/need a totally accurate rendering of the 'real' colours; those specialized in fine arts, and may be a few other domains I don't think about at the moment. Marie-Noëlle 2012/7/20 Andrew Goodbody <[hidden email]>
-- Mes dernières photos sont dans ma galerie. Retrouvez-moi aussi sur mon blog. Et parcourez les Cévennes à ma façon avec Cévennes Plurielles, _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
In reply to this post by Martin (KDE)
On Friday 20 July 2012 19:58:48 Martin wrote: > > That is not the case, a camera profile is used to modify the colours > > saved in the RAW file when they are turned into what is displayed on the > > computer or stored in a colour-corrected TIFF, JPEG etc. > > Not quite right. The camera profile is used to transfer the data from > raw processor (after demosaicing ...) into the working colour space > (that are usually sRGB, prophotoRGB, eciRGB) which are transfered into > the display (or storage or print) colour space (sRGB, AdobeRGB ...). The > extra step with the working colour space is used to get a defined base > for further transformation.
Just to dot all the i's and cross all the t's: That's not quite rigth (or quite complete) either.
Let's first define what we are talking about: - for me, the camera *profile* is a correction curve to compensate the differences between what a colour should be and what it is in practice (strictly the same as what a display or printer profile does); - in addition, there's a camera *working space*, which defines how the camera encodes the colours it sees. The working space is essential, unchangeable, and present in the RAW decoders (implicit or explicit).
So, a camera profile is not strictly required (then we assume that the camera has no deviations in the colour representation),which is where the advice "don't worry about it unless you have special needs" came from.
So now we have our colours in our photo editor, in a well defined colour space. If your editing working space corresponds to the device's working space, no further transfer is necessary (device profiles still have to be applied though).
And the 'extra step' with the working colour space is not extra at all, it is a essential step in the process, as that is the first step where your colours are well defined across devices/systems.
And for quite a few uses, there's no need to translate from one colour space to another. For screen display the (de facto) standard is sRGB (with some high-end displays having the ability to use Adobe RGB), and most print shops also accept only sRGB. (some printers can handle Adobe RGB).
And indeed, colour management involved a lot of steps, and is complex...
Remco
_______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
In reply to this post by tosca
On Friday 20 July 2012 20:31:09 Marie-Noëlle Augendre wrote: > That's very clearly explained, thank you. > As an 'artistic' photographer, I had never seen to point to investigate > this further. My guess is this kind of input profile will mainly be useful > for photographers that want/need a totally accurate rendering of the 'real' > colours; those specialized in fine arts, and may be a few other domains I > don't think about at the moment.
Some portrait photographers also use input profiles, to make sure they get the skin tones right (and that's commercial portrait work). After that, they might play with the white balance to get slightly warmer tones (for exemple), but they want to have a well defined starting point.
Remco _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Free forum by Nabble | Edit this page |