Reading these color charts..?

[Advil Dremali]

This is going to come off as extremely stupid and noobish, but. How do you read these charts? The ones with the different colors on it and stuff... And.. well, whats it for?

I know their used to rate something, I'm just...dumb.

Here's an example of these charts, I don't know what they're called.

http://www.camcorderinfo.com/images/...20autoweb1.jpg

http://www.camcorderinfo.com/images/...20autoweb1.jpg

...yeah.

[Boyd Ostroff]

This might provide a few clues...

http://www.dsclabs.com/product_reviews.htm

[Rokta Bija]

Displayed properly, the charts are a valuable information tool, unfortunately Camcorderinfo doesn't do this.

It's pointless to publishing a color\resolution chart if your going to display it at a reduced size and in a lossy compression format. You don't know if the noise in the color is from the camera or the compression format and the resolution part of the chart is totally useless. About the only thing it's good for is comparing the color saturation between cams.

In some of their reviews they have started using the excellent ISO 12233 resolution chart, but at the same reduced size and compressed format and you have to wonder why they even include it since it doesn't convey any useful information displayed that way.

[Glenn Chan]

The one thing those pictures are good for is to look for false colors. The black and white converging lines (normally used to test resolution) are a good test for how well the camera avoids false colors. The lines are black and white in real life, and should be black and white in the captured image.

An example is the Sony TRV19/22 (both models have the same CCD/optics):
http://www.camcorderinfo.com/content..._500_group.htm

The charts would be a good indicator of resolution if they were the original resolution (which they're not). However, you can get an ok idea of a camera's resolution (i.e. the Sony TRV22 is very noticeably less sharp than other cameras). However, the lower resolution might hide edge enhancement. On a full resolution picture you would see halos around high contrast edges.

The chart also tests color rendition. You need a vectorscope to check color accuracy. Many editing programs have software versions of vectorscopes, although you may need to configure them correctly.

The four patches are supposed to be flesh tone colors I believe. These tones usually should be reproduced accurately because people know what they look like.

The middle chart has bars for white and black balance. If you eyedropper them in Photoshop, you should be able to see how accurate the camera's autoWB is. In the link above, you'll see that the JVC's autoWB is closest.

The black/grey/white bars of varying brightness (chip chart is what I call it) show the camera's gamma response. You can see the apparent contrast you will get from the cameras. I don't think you can figure out a camera's exposure latitude from the chip chart unfortunately. However you can see if white balance is consistent across the bars.

I believe the chart is normally shot so different cameras can be more easily (color) matched in post. I'm not sure if it's that good of a test chart, although as a test chart you can see the things I talk about above.

[Glenn Chan]

I just did some messing around with the images from the camcorderinfo site and the perfect version of the chart at the ChromaDuMonde site (http://www.dsclabs.com/chromadumonde.htm). If you pull up vectorscopes in a program such as Sony Vegas (which I used; Premiere, Final Cut, and Avid I think also have scopes), then you can see how color accurate the cameras are.

The ChromaDuMonde is your base reference. Bring it up on the vectorscope and look at where the dots on the hexagon are. On the image capture from the camcorders, that's where the dots should land.

In the camcorder shoot-out link I gave (http://www.camcorderinfo.com/content...500_group.htm), I'm beginning to believe the reviewer is totally off in her comments. All the cameras:
A- (minor criticism) all the cameras exhibit false colors, which is not mentioned.
B- The color accuracy comments are all wrong. Bring the images up on a vectorscope and you see which cameras are accurate. All but the JVC are slightly mis-white balanced. All lack the color saturation that's supposed to be happening (I might be wrong here, but they have like half the color saturation they're supposed to; however, in real world conditions it's rare that you'll encounter colors like the ones on the test chart).
A lot of the colors have are of the wrong hue 'slightly'. Some of them do miss the targets on a vectorscope even if you boost saturation.

Of all the cameras, the Panasonic and the JVC have the most even color reproduction. You need to raise the saturation in post (which you need to do for all those cameras anyways).

2- Flourescent lighting (and other low CRI lighting) might screw everything up? I don't know.

3- I don't think it's obvious how those charts should be interpreted. Personally, I'm just guessing that the proper way to do things is to bring the images up on a vectorscope.

[Boyd Ostroff]

He hasn't updated it for a long time, but Scott Billups has a lot of hi-res images of this chart taken by a variety of SD and HD cameras on his site here: http://www.pixelmonger.com/hg_cam.html

[Dennis Vogel]

The charts on Billups page are also in a lossy format (jpg). Wouldn't they suffer the same problems outlined above?

Good luck.

Dennis