• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

The many roles of look up tables

Doug Kerr

Well-known member
Early in Asher's serial tutorial on color management and such, he mentioned a role of look up tables that was unfamiliar to me.

I did a little poking around, and find that I am now uncertain on a much broader basis.

Let me start by telling of my understanding of look up tables (LUTs) (in the context of interest here) prior to today.

Firstly, LUT is not the IATA airport code for Luton.

• In a display adapter (often called a "graphics card), there is provision for mapping between the incoming R, G, and B values and the three "drive signals" to the display proper. This is typically done by a "look up table" (LUT), in simple adapters usually just three one-dimensional tables.

When we "calibrate" a display chain (the intent being to make it conform, as closely as is practical, to some standard color space), the calibration system develops a set of values for this LUT, which is saved in a file (often in a "caboose" to the associated profile file, about which more soon) and implanted in the LUT of the adapter each time on system startup.

• When we "profile" a display system, the profiling system determines its color behavior (it already having been calibrated) and develops a mapping between the RGB of a certain standard color space (called the Profile Color Space PCS) and the color space that the display system (as calibrated) evidently follows.

This mapping is defined in the profile file, in one of (at least) two ways:

•• In terms of a linear transform, defined by a matrix, between the incoming and output color space coordinates.

•• In terms of a 3-dimensional look up table (3D LUT). For each of many discrete combinations of the incoming RGB values, this table delivers (arbitrarily) a set of output RGB values. As the table does not have a zillion entries, interpolation is used to deal with the precision of the color space coordinates.

This latter is considered superior.

That's what I knew about when I went to bed last night.

Today I have learned that there are one or more other ways in which the LUT concept plays a role in image color management. Evidently, these practices largely emerged in the world of digital cinema (where color management is a very serious business), but have flowed back into "our" world of still photography.

As I understand this, after fully 45 minutes of study, two of those uses of an LUT is:

A. To do a remapping of the image colors on their way to the fancy display in the suite where color adjustment ("color grading" in the digital cinema world, the successor to "color timing" in the film world) is done so that the display will emulate the "look" that will result in the final film print (when film prints are involved).

B. [I am less sure about this.] Perhaps to implement changes in color chosen by the colorist on the image itself (which of course may well vary between scenes or subscenes).

Perhaps the "script" of these changes, described by different sets of LUT values, are preserved, and only actually applied to the final output after all the "color grading" is decided upon.

Now, back to the world of still image processing.

Asher spoke of, in more sophisticated image processing, capturing color "adjustments" made by the editor in the form of settings for an LUT. Then, when the same image is addressed again, the file carrying those LUT settings would be recalled, presumably so that we would now see the image color the way we left it after the last session.

Now, one question I have is this. In this scenario, is the impact of the LUT settings that characterize the current state of the color adjustment applied to the image, as we expect to see it on the screen (and presumably as it would be applied to the final "output" file):

• Applied to the image on its way to the display system, so it receives the "adjusted" image..

• Implanted into the LUT of the display adapter, so it will apply the adjustments while handling the unadjusted image fed to it by the editing application..

In other words, I am missing a block diagram of this scenario.

Who can help me out here?

"Color timing", by the way, is called that because the "program" for the color adjustments through the film, which may vary throughout the film, are captured in a "script" in terms of time. Then, in the earlier days, a technician attending the optical printer which made the release prints would follow that script, watching a clock, and make adjustments of the color filters to suit.

Also included would be a program of changes in exposure (to adjust scene brightness), something that had been done by a "timer", before the days of color. This was usually executed by changing the speed of the printer.

In later days, the script would be transformed into a program in a mechanical or electronic programming system, again time-based, which would make the filter changes (and printer speed changes) automatically.

Best regards,

Doug
 

Doug Kerr

Well-known member
Well, I had lunch, and spent a while cussing out an a insurance adjuster who was supposed to visit today but didn't, and now I'm ready to take on another notch of my ignorance about LUTs.

I have looked a little into the use of LUT's in Photoshop CS6 (I only have CS5, so I never heard of that).

As I get it, in PS CS6, one can establish a color adjustment layer whose behavior is defined by a color lookup table (LUT) defined in a .3DL file. Apparently, quite a library of these files is provided with PS CS6.

As with all PS adjustment layers, the "strength" of the adjustment can be controlled by setting the opacity of the layer, and as usual we have a choice of various "blend modes" that control just how the adjustment is applied.

Now, as I understand it, in the other direction, the collective effect of one or more adjustment layers, created using the various familiar Photoshop tools, can also be consolidated and exported as an LUT file. Thus, the same regimen of adjustment can be applied to another image.

Now, for those of you who really know how this works, am I at all on the right track?

Thanks.

Best regards,

Doug
 
Top