Capture One Film Curves (FCRV)?

[Marcus Christopher]

Hi all,

I have been a long-time lurker in this forum, so first of all many thanks to all the competent people here.

Anyway, I have used quite a few RAW converters in the past and, like a few of my clients, have come to appreciate the color rendition of Capture One. Not so much the interface, though, but that's a personal thing. So, I wanted to better understand what kind of "film curves" (you know, that drop down that says "film standard", "film high contrast", "linear response" and so on.) CO applies to the RAW images...

Unfortunately, the curve files come in some binary format (ending ".fcrv"), which makes it difficult to decipher them. Has anyone tried to do that? So far, I found only one website (a programmer's forum), where someone simply listed the entries from two of these files, as if it was easy to just open and read them. So, I'm thinking there has to be a way. But the thread in that forum is quite old and I had no luck contacting that guy.

Any help would be much appreciated. Thanks in advance!

[AlterEgo]

Hi all,

I have been a long-time lurker in this forum, so first of all many thanks to all the competent people here.

Anyway, I have used quite a few RAW converters in the past and, like a few of my clients, have come to appreciate the color rendition of Capture One. Not so much the interface, though, but that's a personal thing. So, I wanted to better understand what kind of "film curves" (you know, that drop down that says "film standard", "film high contrast", "linear response" and so on.) CO applies to the RAW images...

Unfortunately, the curve files come in some binary format (ending ".fcrv"), which makes it difficult to decipher them. Has anyone tried to do that? So far, I found only one website (a programmer's forum), where someone simply listed the entries from two of these files, as if it was easy to just open and read them. So, I'm thinking there has to be a way. But the thread in that forum is quite old and I had no luck contacting that guy.

Any help would be much appreciated. Thanks in advance!

quote from a different forum

vgrin ::

    but Curve in Color in C1Pro too can be linear that means that the gamma 1.8 is applied, and the acoustical curve was not applied.
    Curves are stored in "c:\Program Files\Phase One\Capture One 8\Film Curves \" in a binary format fcrv and contain, for example, the following:
    CanonEOS1D-Linear Response.fcrv
    0.000000, 0.000000
    1.000000, 1.000000
    "Linear response"
    CanonEOS1D-Film Standard.fcrv
    0.000000, 0.000000
    0.039474, 0.058387
    0.120614, 0.241626
    0.252193, 0.538556
    0.348684, 0.690488
    0.405702, 0.758989
    0.475877, 0.826794
    0.546053, 0.878867
    0.633772, 0.925495
    0.774123, 0.971589
    1.000000, 1.000000
    "Response similar to film"

x/y coordinates [0.0...1.0] pairs are coded as 4 bytes x 2 containing an UINT (4 bytes, 32bit, little endian) which shall be divided by 4294967295 ( 0xFFFFFFFF ), for example 4 bytes = "0x86" "0xF2" "0x1A" "0x0A" = 169538182 and / 4294967295 = 0.03947368 = 0.039474 (see above)



hex dump, showing different parts in different colors, note that 3 curves are inside

film extra shadow curve

http://1.bp.blogspot.com/-OrghB5FurJI/VRsm9EJuoxI/AAAAAAAAHtA/LZNa-W3HNV0/s1600/1.jpg



linear curve

http://3.bp.blogspot.com/-HofjW71Jmzc/VRs4CsghyQI/AAAAAAAAHtg/H9gPIr4L9W0/s1600/3.jpg



standard curve

http://3.bp.blogspot.com/-AF97oeshlqI/VRswMIR0GTI/AAAAAAAAHtQ/SBIXi4sdXG4/s1600/2.jpg



I think that .fcrv also has instructions whether to apply or not clipped highlight recovery... for exampe "linear scientific" curves do not do this vs simple "linear"

[Marcus Christopher]

Oh, wow, thank your VERY much for this detailed reply. May I ask you, if you had already analyzed these curves before? I'm asking, because I had done a hexdump myself, but it didn't occur to me to search for divisions of UINTs... Did you know what to search for?

Anyway, I will take a more detailed look at what you found and see, if I can work out a method of converting these to simple text files. Will report back. Thank you for your help! Much appreciated. 

[AlterEgo]

May I ask you, if you had already analyzed these curves before?

I did not spend much time on it (other pressing matters) - may be you can...

the file starts with some header (6 bytes in yellow), each curve starts with how many points in it (4 bytes in red) followed by points themselves... you can also see that curve name in stored twice inside the file (with how many bytes name takes - 4 bytes in green, followed by the text itself) - so pick up here and continue decoding the remaining bits

[torger]

What do you intend to do with the curves? If you intend to replicate the look in other software it may not be so easy.

The key thing that also the ICC profile itself applies a curve. I haven't analyzed all C1's ICC profiles, but my guess most do it if not all.

That is the user-selectable curve is doing the major part, but then there's a residual curve applied on top via the profile's LUT. You can see that when you select "linear curve", then the image is still not rendered linear, because the ICC profile applies a curve.

The user-selectable curve is applied as a simple RGB curve, with all the issues of saturation increase and hue shift. However C1 curves have been carefully designed without "the shadow dip" in the S-shape which otherwise causes the worst hue shifts. Instead that "shadow dip" is made by the ICC profile, and that curve is not a RGB curve, it's more similar to a Lab Lightness curve. I haven't analyzed though so I don't know if it really is a Lab curve or not.

Using DCamProf you can extract the film curve by using the TIFF-embedded transfer functions rather than reading the fcurv format directly. This is described here:
http://www.ludd.ltu.se/~torger/photography/camera-profiling.html#the_easy_way_c1