I agree with AMSP. When I would retouch my own editorial work, I would mix 1ds2, 5D, Aptus 22 files, and make them all look the same. I could get them all to agree, even when processing Canon in C1 and Leaf in ACR/LC10/11. In fact, I could get really, really good skin from a 1ds2 and Cs2/CS3. The defaults for Leaf produce better skin, and teh Dalsa backs have a different response, I think, at least out of the box. The Sinar 54M had the best default skin tones I've ever seen. Too bad it was a 4 seconds a frame tethered back.
Hi,
With ACR calibrator or others there is a very good "compatibility" mode without a doubt.
Still in nuances especially in the extreme values you will always see some differences.
As mentioned with my calibration sample of displays, all cameras and devices have their own "look" and "feel" there is no way you can get all looking exactly the same.
But to deliver to a client, no problem at all, especially when you use different lightsetups with the different cameras there is no way telling which camera was used in print on tones only.
Problems will arrive when doing good printing and using a same setup with 2 different cameras let's say a 1DsIII and a Phase One or Leaf back.
There will be nuance differences that WILL show up.
Calibrations are often regarded as "if I calibrate everything will be 100% accurate" what people seem to forget however that the only thing a calibration does is getting everything as close as possible (and there lies the problem) to the original LAB values. This is done for printing, scaning, cameras and monitors. This way you can work on one system and print on another or display on another and have a similair look/feel. Also for working in different colorspaces and converting calibration is a must, simply put you're working in prophotoRGB on a 97% ARGB monitor and delivering to the customer as sRGB you want it to look the same (as much as possible) on all systems with as little tweaking as possible.
The reason a calibration is not 100% accurate is simple to explain.
There is always the case of what a sensor can capture.
Let's make it very simple.
We take one color in this case RED.
Sensor 1 can capture red on coordinate XYZ
Sensor 2 can capture red on coordinate XYZ
But the coordinate XYZ on sensor 1 can be different than on sensor 2.
to be honest it is 100% sure, it can even happen between two cameras of the same brand.
What the calibration does is getting both sensors as much as possible to the standard XYZ coordinates that are fixed.
Some cameras can capture a much wider colorspace in certain colors than other cameras, meaning that whatever you do the more intense colors of sensor 1 CAN and WILL differ from sensor 2.
The problem that rises from this is that we work in a 10/12/14/16 bits colorspace and that the colorspace of the camera (cameraRAW) has to be mapped to the colorspace you choose, for example prophotorgb.
This mapping is hugely different per raw convertor. The colors I get from Leaf capture for example are different from ACR, you could calibrate both with a profile and in a perfect world both would now be equal, except it is not.
Calibration is a can of worms when you look at it from a viewpoint that it all has to be perfect after the calibration.
In fact calibration however is very simple.
There is a standard tabel with coordinates, on those coordinates the colors should be for a given lightness, hue and saturation.
AFTER the calibration the source (being it printer, camera or monitor) will try to get the image you see/capture/print as close as possible to that standard.
For most people this means that you will see no or hardly no difference between two cameras (again in print form for magazines you won't see a difference in color when done correctly).
HOWEVER it's not true that a Leaf back will look exactly the same as a Phase One back or a Canon.
When working in a good workflow with good display and printers and calibration they come extremily close but there will be a difference.
And that difference is why some people prefer X or Y.
The canned profiles is why most people prefer X or Y.
