Hi Chris,
Thank you for the tips.
This represents a very different approach, trying to calibrate the gopro cameras.
I am the novice here… was advised that post production color matching from “flat color” video would be the easiest / most satisfactory results.
I will try the other route first and see how easy it really is. 
Hi Austin,
I have tried to follow your recipe… no luck at all.
I have taken still images from 8 places in my flight and put them into a project.
https://www.dropbox.com/s/iy8l9o376s1jflb/ColorGrade%20001.zip?dl=0
My idea would be to match the images over the “scenic” part say set 05, 06
I hope that these setting would work for the less important traveling there and back bits of the flight.
You help with how to match these images would be very much appreciated.
It’s how the pro’s do it 
I think you’ll find it a lot easier if you have a reference chart to use.
Hi Austin,
I have made some progress with my project.
I would like if possible to place the results in the “clipboard” the shotcut uses for its filters… is this possible?
Here is a download link to the finished app.
Workflow is quite simple.
Thank you! I’ll try it out!
Update.
Added range limit override.
I’m a bit (a lot!) lost trying to follow and understand the conversation above. @Derek could you give us a brief presentation of your app ?
Hi MusicalBox,
The app calculates the required input parameters that the Color Grade filter can use to match 2 images.
You might to call it a color matching tool as there is already something in Shotcut called “Color Grading.” Then, make a new post in the #resources category with the download link.
I don’t think this is possible. The Shotcut clipboard is internal to the app. It isn’t the operating system clipboard.
Your app looks great!
I finally got a chance to try out your MLT file and apply some filters to it. I was able to get sets 5 and 6 looking much closer, but it was not a fast process.
From what little reading I’ve been doing, it seems that GoPro in flat mode has different gamma curves per color channel, which effectively multiplies the effort using scopes by 3x. I can’t verify that’s true, but it sure felt like it. I had to consistently pull red down in the highlights and green down in the midtones.
Once I got sets 5 and 6 working, those values did not apply uniformly to any other sets. I’m guessing one or more cameras was in Auto ISO mode, which is usually necessary when outdoors. So, the work kinda has to start all over again for the next clip.
For your use case, your app is probably the fastest way to get a match.
Thanks for the effort.
Here I am in dire need of a haircut, hiding behind a Kodak Gray Card Plus.
I got a fairly decent grade with Shotcut. YouTube didn’t mess up the levels too badly.
White: Y = apx. 235
Gray: Y = apx. 111
Black: Y = apx. 16
https://studio.youtube.com/video/mI2shfS94os
The camera was a Logitech Brio webcam.
Your link is broken.
I just clicked on it and the video comes right up.
Thanks Chris. colors look great.
I didn’t tweak any of the colors. All I tweaked were lift, gamma and gain, with all three primaries set to the same values, so they are effectively affecting only the Y (luma/luminance). The webcam did a pretty good job of auto white and black balancing — no adjustment needed. Again,
white: Y = 235 (90% reflectance on the chart)
gray: Y = 111 (gamma 2.2) (18% reflectance on the chart)
black: Y = 16
These are basically the BT.709 Y values. I used an eyedropper/color picker tool calibrated in Y using the BT.709 coefficients.
This topic was automatically closed after 90 days. New replies are no longer allowed.
