Page 1 of 1

18 percent gray card

PostPosted: Mon Jan 31, 2005 9:42 am
by Rusty W. Griswald
Hello

I downloaded an 18% gray card from someone in a topic on DPreview.

First of all, if I want to be really cheap and not pay $11 for a KODAK gray card, could I just get this printed at BIGW for 33c or would that pose some sort of colour matching problem?

Also, if I was to get it printed there, should I get it on gloss or matt?

Thanks
Rusty.

Image

PostPosted: Mon Jan 31, 2005 10:08 am
by AlistairF
Hi Rusty,

The problem you face is in the color of the paper that you're printing to. Even white papers vary in colour density. Hopefully the white paper you use is "colour neutral" if you're using it for white balance and the 18% factor is not as important, just that it's a middle grey.

A friend of mine did a similar thing and did a printout just using a laser printer on stiff white paper. He just filled the page with grey using Windows Paint with the color (R128,G128,B128). As toner is jet-black you'll get a fairly neutral result, but not a perfect one. Give it a go. There's a reason why these babys are $11 each and Expodiscs are $150.

Have fun,

Alistair

PostPosted: Sat Feb 05, 2005 12:19 am
by Onyx
Rusty, I tested this theory so you wouldn't have to. ;)

I downloaded the grey patch in your post and opened it in PS. When converted to RGB, it was 102 on all channels. I corrected this to be 128 (theoretical middle grey) and printed it out with my inkjet using glossy paper (Canon Photo Paper Pro) and Canon provided (default) profile for the paper type & quality setting. I did set all printer settings to bypass auto corrections and reproduce faithfully as dictated by PS.

I then used the printout to preset WB on the D70, hoping to compare it to Auto WB. I took a series of images under both modes with everything in-camera set to neutral. It seems neither gave optimum results. The camera under Auto WB showed significantly higher readings in the red channel than the green and blue - both of which were consistent.

Using the grey card with preset WB, I had expected fairly consistent RGB channels. The results I observed were red deficient - ie. when green and blue channels (both seemed consistent) reporting around 90 with PS eyedropper tool, the red channel typically had a reading of low 80's. The overall images had a slight visible blue cast too.

Looking at the physical grey card printout, I can see perhaps a tinge of magenta - however I may be perceiving this only because I'm expecting to.

At this stage, I haven't tested a store bought Kodak Grey Card yet, so I can't say whether it would be better. All I can conclude is, a home printed grey card (with the correct printer profiles) doesn't seem to work for preset WB.

PostPosted: Sat Feb 05, 2005 8:36 am
by gstark
Too hard by half.

It really doesn't matter too much how the sample "grey" image prints out at all.

Just take an extra image or two as calibration images before you begin your real shoot, and include the grey card that you've printed within the area of those first couple of test images. Remember to reshoot these calibration images if your shooting conditions change.

Set your monitor and printer up to reproduce those calibration images with the grey card appearing exactly as it does in real life, and you'll have the correct callibration set for everything else you've shot under those same condtions.

PostPosted: Sat Feb 05, 2005 9:16 am
by MattC
I played with this one overnight.

I printed to plain paper, using my calibrated monitor for comparison. The version I did on Epson premium semi gloss did not work at all - magenta colour cast and too reflective.

I did not really get very good pre-set WB (mixed bag actually) values from my plain paper grey card, although I did find it useful for correcting WB and exposure in NC - within the limitations of uncalibrated grey card. For setting exposure it was spot on.

One thing that I did get out of all of my experimenting was that a grey card will make a significant improvement to my shots and make PP a lot easier, and therefore, make an important addition to my kit. Now, where can I get a Kodak R-27 grey card and a Kodak Q-13 colour seperation chart from in Oz without paying horrendous prices?

Cheers

Matt

PostPosted: Sat Feb 05, 2005 11:10 am
by lejazzcat
gstark wrote:Too hard by half.

It really doesn't matter too much how the sample "grey" image prints out at all.

Just take an extra image or two as calibration images before you begin your real shoot, and include the grey card that you've printed within the area of those first couple of test images. Remember to reshoot these calibration images if your shooting conditions change.

Set your monitor and printer up to reproduce those calibration images with the grey card appearing exactly as it does in real life, and you'll have the correct callibration set for everything else you've shot under those same condtions.


Yep - just remember to shoot RAW when you do this. Then youll be able to adjust the Wb afterwards in PS, the grey card in your calib shot to 18% grey and all the other colours then 'should be spot on' .

PostPosted: Sat Feb 05, 2005 11:12 am
by gstark
lejazzcat wrote:Yep - just remember to shoot RAW when you do this.


You mean that there are people out there who do not :!: :?: :!:

:)

PostPosted: Sat Feb 05, 2005 11:29 am
by Rusty W. Griswald
Onyx, thanks for doing the detailed test. I took the file into the Kmart DIY machine yesterday afternoon just to see what happens, and nothing happened! It wont recognise the file I think as it is only 2kb (unless it sees that there is nothing on the print and ignores it).

I had a look through a friends Thom Hogan D70 ebook after that and the chapter on white balance was very long indeed.

He was talking about field tests using the preset wb mode showing a neutral grey card works better than a white card.

I think he said a grey card should have no color cast and a white card is often underexposed. So if you were to print one yourself, should you take the colour cartridges out and only use the black cartridge so there are no other colours leaking in?

Matt said he would like a card without paying horrendous prices, I had a quick look at a couple of sites and it looks like I might give a card from Vanbar a go, price are below.

The other thing I thought about with your post Onyx was there was lots of noise on the other forum about people wanting a firmware fix for the D70 because of the way it handles colour temperatures or something.
http://forums.dpreview.com/forums/read. ... e=12009478
But I think that only relates to the Flash WB, would that be right?

http://madsens.com.au/
Madsens
JJ013 Jessop Grey Card $22.80

http://www.vanbar.com.au/
Vanbar
"Colorama Grey Card 140gsm 8.5"x11"` (Ref No: 33-901)
Order:
The new Colorama Crey Card.
8.5" x 11" 140gsm.
The gray card reflects approximatly 18% of the incident light. These cards are double sided grey.
AUD Inc. Tax: $3.30"

And they also had the Kodak
Kodak R-27M Mini grey card $5.72
Kodak R-27 Grey card ( 1 only) $22.00

PostPosted: Sat Feb 05, 2005 5:37 pm
by Matt. K
The KODAK manufactured grey card is 100% colour neutral and that is why it costs a bit...that is also it's value. Any printed copy is bound to be inferior.....but you may get it close enough for your use.

PostPosted: Sat Feb 05, 2005 6:18 pm
by lejazzcat

PostPosted: Sat Feb 05, 2005 7:02 pm
by Onyx
Teds have the Lastolites too - I saw them in stock and enquired when I purchased a monopod there few weeks back ($50 seemed a bit much so I gave it a miss).

Rusty, the orig file (as linked in first post) is a monochrome JPG. It may have confused the poor minimum wage Kmart technician as it's not sRGB or in the form of a popular consumer format (would they even recognise TIFFs?).

Gary, what you've described is another way to skin the cat - one based upon matching monitor and printer colour profiles and using one's eye to discern between onscreen and print colour differences. IMHO it's too subjective, and I'd rather have one element in all that's objective and acts as the "standard" - ie. a known neutral grey (ala Kodak R27) as you could potentially end up with results that are consistent but wrong.