That's cool. I'm going to experiment with seeing how useful it is with video color matching .

Gary

Any one tried this utility with the color corrector?
I tried it yesterday to match 2 cameras without success.
I would be very interested to know how I would translate the RGB value obtained by this utility in to angle/magnitude in vegas corrector.

Frankly spoken, this tool sounds nice.

But there are NLEs which adjust two different videos by different camcorders automatically. THAT would be an improvement in Vegas, everything else is a workaround.

But there are NLEs which adjust two different videos by different camcorders automatically. THAT would be an improvement in Vegas, everything else is a workaround.
You are right. Vegas colour correction is excellent but is lacking the simple step of transferring a set colour value. I have not yet worked out what the 'pick color' does.

Mahesh, what is, "a set colour value" ? Is this a value you have "set" within Vegas OR a colour balance done in camera? If you are setting a colour USING the CC then when you use Copy > Paste Attributes and then you/I/we can copy OVER the settings to this other event, which WOULD copy a setting. BUT but, but this would/could be most inappropriate to what you WANT for that "other" event.

I use "Copy To Clipboard" and the Scopes to share and compare one event to another. This way I can "iron-out" CC differentials - if I wish. I really can not see how this could be done automatically. I have been experimenting recently with the remarkable Paint Shop Pro settings for CC and Luminance and Chroma and creating my own templates. But all of this will ultimately depend on the "look" of the original and me wanting to use these "set-in-stone" templates - see? . .. .

In summary: That which is set/filmed to a 4k template will appear as ZEROS in Vegas CC - something set to 6k will also appear as ZEROS in Vegas CC. I cannot see how Vegas can do anything else? The only thing I can see Vegas or another s/w package doing is to allow me to share and compare Scopes and Visual look? BUT, there maybe s/w out there that does this? Yes? A type of dynamic Colour Correcting process that reads what it "sees" and compares this to a set algorithm of templates? Maybe?

. . . or maybe I haven't understood correctly . . .

Excellent thread! This is something that I've been exploring - for other reasons - over the last 4 months.

Mahesh experiment with "Colour Pick", get Gary Kleiner's DVD on V5! His section on Colour Correction was worth to me JUST for this! The shades just dropped from my eyes when he explained this to me.

Grazie

Grazie, thanks for that insight. As you say, this could be an excellent thread if individual problems are aired. I am dreadful at explaining but let me try.
I have footage from 2 cameras, looking at the same scene from different angles. Both set on auto-white-balance because of local lighting conditions.
During filming, WB on one camera drifted because camera man panned to audience. When returning to main subject WB had changed.

I took some readings on a white table cloth on the scene using the TrueColor utility.
Cam 1 was R=227 G=227 B=231
Cam 2 was R=222 G=219 B=231

Being a simple minded person, I thought if I could make Cam 2 to be the same value as cam 1 at that point then the rest should just drop in to the place.
I played with the colour wheel but could not corallate these values.

So I did the next best thing.
Copied cam1 image to clip board. Drew rectangle in the preview window so that I could see the white table cloth from both cams. Twiddled the color wheel until cam2 whites were simillar to cam1.

Okay, job done. but not very scientific.
How would others approach this.
Before anyone says this, I never white blance my cameras when filming live events.

Before anyone says this, I never white blance my cameras when filming live events.

Then don't ever expect to have footage that matches. Why don't you? There is virtually always something you can balance on, even if it's a tux shirt, large number on a jersey, whatever. Or, put the cam in auto-white balance, and some cams do a good job with this while others don't. But...If you can, you should. If you can't, then you deal w/it, but rendering color correction when you can avoid it on the shoot is a much better option, IMO.

I guess what one could have that'd automate it to some extent is being able to select a reference area in one shot and then select another area in another and have something automatically match the two.
All that'd be fine if the two scenes had an aweful lot in common to start with and the things you were matching were the same, say a reference white card but I can see this being far from goof proof.
I used to dable a little in still photography and so the lab could correctly color match I'd always take a shot of a 90% reference card but you needed to do that everytime you moved anything.

The tool that started this discussion I think highlights the limitations of DV, the color resolution is so low that trying to sample a color based on one pixel is probably going to be inaccurate.

Bob.

"but rendering color correction when you can avoid it on the shoot is a much better option, IMO. "

Absolutely, and I'd add getting all your light source at the same CT to boot (if possible). I think the reason is that the camera is applying the CC in either the analogue domain or in a higher res color space, then it gets compressed to DV and a LOT of the color information is ditched in the process. Trying to correct gross color errors after that is impossible.
Bob.

Spot
I think you misunderstood what I was trying to say.
When filming a live event, you are presented with a mixture of daylight/tungsten/other lighting in the same room.
I always use cameras from same manufacturer, Sony. The cameras are left in AWB and they do a pretty good job, 95% of the time. In manual white balance, there is a good chance you would get a mismatch if the camera moves to an area with different lighting whilst still recording.
Occassionally, I have to color correct one camera so that swithching between the 2 during edit looks acceptable.
So moving on to my original question, is there a scientific way of changing colour balance by using the colour values of a group of pixels?

I understand what you mean now, apologies for missing the point.
You can identify the color value of a particular pixel, yes. Then you can identify the same location in the other camera shot, and using scopes, match them up. This is exactly what scopes are for.
One book I highly recommend to everyone is Steve Hullfish' book on color correction. It's very, very well written.
I also have a vid on color correction I've just tried to upload, but this hotel's connection keeps timing out before it uploads completely. I'll try it from KL tomorrow evening.

Thanks Spot. It is my fault. I did not explain very well.
You mention using scopes. I presume you mean vectorscope. I am not quite sure how one would use this in a situation where the overall chroma information is quite different. I mean if 1 camera is in CU and one is wide, the chroma information displayed would be quite different, especially if one is trying to match whites or grays.
Without sounding flippant, I have used vectorscopes for last 30 years in broadcast environment. I am sure you may be right when you say that one can identify a colour phase even if the actual U and V values are different.
Spot, please do not this as criticism of what you said. I am just trying to get my head around software colour correction.
I really look forward to looking at your video on colour correction.

1- Some color correction tools have a match color tool that'll do what the name suggests. They'll get you pretty close but not all the way there... there will still be minor differences between different sources unless you shoot with the same camera.

2- Video noise: A big problem with consumer cameras is that they have a lot of chroma noise. If you take just one pixel sample on an even area of color, it'll differ a bit from another pixel. This could really throw off any color correction tool.

In Vegas, you can get around this by using the eyedropped and drag-selecting an area. Vegas will average out the values (and the noise).

I don't DV compression is a big problem.

You get around this by applying a simple mask and then looking at the values in the scopes. The book SPOT mentions covers this but once you get the concept easy enough to do.
One could for example use the event pan/crop tool to enlarge the same object in two shots to full frame, then the scopes will read just that object.
Bob.

Don't forget the scopes will read split screens, too, and while this doesn't help with the vector scope or histograms very much, the waveform monitor can help you see this easily.

And do all of this while you have the piece is LOOPING, so you can see the scopes coming closer and closer to what you want.

Having the Clip Board working gives the split screen, "steal" the look directly from the 1st event, and you have then the REAL events playing and the scopes AND you can adjust the levels until you have it on the money - as it were!

Grazie

Farss
Using mask and pan/crop is a brilliant idea.
Basically vectorscope is only good if you are matching a calibration chart or colour bars. With 2 cameras looking at the same scene and different picture information, scope would not be able to pinpoint the area. Idea of using Mask is good. I shall try that.

Spot
Waveform scope is also a good idea. It works well with color curves. If you have the brightest part in the clip board, curve can be adjusted to match that, keeping an eye on the black level.

Grazie
I had never thought about looping. I just looked at a frame, at best quality, for match using split screen.

There are some wonderful gems flowing from all you experts. Please let them roll.

Here's another cockup we did recently.
Filming a stage event with 7 cameras but 3 operators. We knew we would not have time to do manual white balance because the curtains would not be opened with lights-on for us.
Lighting was tungsten with no filter gels. So we decided to switch all cams to tungsten setting. One moron mistook the little sun symbol for little bulb symbol in semidarkened auditorum. (hang-on, that moron was me.) I thought fine, no prob. Correct it in post.
Try it. It's definitely not a five minute job.
Now my Paint Shop Pro has color temperature corrector. A slider which moves from cold to hot colours.
Second best is a template in color corrector. Say increasing/decreasing colour temperature Daylight to/from tungsten.

I think that's about the biggest ask of all, the camera is already at the bottom end of its capabilities and then you have to dial in more gain on the blue in post. At least if you shoot daylight as tungsten you've probably got a lot more light than there's usually on a stage so the noise isn't likely to be such an issue.
Bob.

Filming a stage event with 7 cameras but 3 operators. We knew we would not have time to do manual white balance because the curtains would not be opened with lights-on for us.
Lighting was tungsten with no filter gels. So we decided to switch all cams to tungsten setting.
=======
I shoot predominantly stage events. I stopped using manual white balance long ago. I get much better results and find it much easier to color correct even the funkiest stage lighting using the tungsten setting. If you are using Sony cameras you CAN modify the setting using the CP.

John