Multi-camera color balance

robert-bogue wrote on 4/17/2012, 6:08 AM
I've got two different cameras which, obviously, shoot different color curves. I've shot a set of color calibration targets, taken the sample values, and know (more or less) how to adjust the two color curves so that the cameras will look the same (again, more or less). The challenge is getting this plugged into Vegas. The Color Curves FX is the right control -- but it doesn't allow me to do mathmatical placement of the points.

Does anyone know of more sophisticated color correction/alignment FX that I might purchase to get the two cameras color to sync up better?

Comments

megabit wrote on 4/17/2012, 6:29 AM
I do a lot of MC myself, and I had high hopes for Frederic Baumann's ColorMatch FX (http://www.fbmn-software.com/en/color-match.html).

However, while it does great job when having pure & saturated R G B areas in both materials to match, in real life situation it proved to be of little use to me.

YMMV

Piotr

AMD TR 2990WX CPU | MSI X399 CARBON AC | 64GB RAM@XMP2933  | 2x RTX 2080Ti GPU | 4x 3TB WD Black RAID0 media drive | 3x 1TB NVMe RAID0 cache drive | SSD SATA system drive | AX1600i PSU | Decklink 12G Extreme | Samsung UHD reference monitor (calibrated)

Baron Oz wrote on 4/17/2012, 7:18 AM
I always thought there should be an algorithmic way to match color, provided that you shot the same color chart in the same light.
robert-bogue wrote on 4/17/2012, 10:25 AM
Yea, that's what I'm testing. The problem is how do I feed the individual biases for each channel at each point in the output curve into the software. So for instance at a nominal output of 40 (out of 255) one of the cameras is biased +12 of the other in red. So I need to for each camera do a shift of 6 to get them to meet. The Sony color curves will do the processing -- if I can plug in all the data points but that seems to be a problem.
robwood wrote on 4/17/2012, 12:08 PM
sounds like the Grade Node in Nuke.

there's a tutorial from Peachpit that describes what Grade Node does here:
http://www.peachpit.com/articles/article.aspx?p=1707325&seqNum=5
fausseplanete wrote on 4/17/2012, 4:41 PM
Algorithmic? Like this one a colleague made for Prem/AE: http://www.mattroberts.org/MBR_Color_Corrector/ ? The free version is not very crippled, quite usable.

Or in Vegas I've done it manually, just reconstruct an XRite/Munsell/Gretag/Macbeth card in RGB (eg as a composite of different tracks/layers for different color squares each as a PiP) and see where the points should be on Vegas's vectorscope. You'll notice for instance the "yellow" square is in fact slightly orangey. Then apply color correction to footage (featuring same chart) from each camera. For example to make a Z1 look (more) like an EX3 I had to apply Secondary CC first, to increase cyan saturation (only), kind of normalizng the footage, then Primary (3-way) or whatever downstream of that. Likewise for a handycam in low light I found Secondary CC needed to remove excess magenta. Same principle.