Question about grey cards

[genf]

I have recently purchased a grey card package... one actually grey, one white and one black.

Much to my surprise, the white card looks a little bluish. So I photographed the set, imported that in Lightroom and updated WB based on the grey card.
These are the approximate RGB values, measured in Photoshop:

- grey: 132,132,132 (neutral)
- white: 218,218,230 (blue cast)
- black: 40,35,32 (red cast)

So indeed, neither the white nor the black are neutral. Is this normal? If so, can someone explain the deviations? Can I trust the grey card (which I will be using most of the time) being neutral?
Thanks for your help.

[Tony W]

Re: Question about grey cards

As your grey card does appear to be neutral then it can be relied on to set accurate white balance both in camera and as a reference source if included in one of your images.

With the required exposure correction it can give a good guide to the 'correct' exposure for midtones - bear in mind that this may not always be helpful if the scene dynamic range exceeds that of the camera sensor.

Some photographers may prefer a white card for WB but imo it rarely matters and if it does then you should find a neutral source. My guess is that if the less than neutral white is by design then it is probably there to give a slightly warmer tone that many may prefer for skin tones or highlights.

Unless you are undertaking scientific or forensic work then the small variations may not be of great importance as a perfect WB does not necessarily consititute an aesthetically pleasing image - sometimes departure from a true and accurate representation may be desirable?

[klev]

Re: Question about grey cards

You aren't doing this in a very scientific way. Were they photographed under reasonably even lighting laying parallel to the camera plane? You should try that. Setting them as the white balance target is fine if it produces usable results. If they aren't neutral come photoshop, that's a matter of how Lightroom processes the images based on the provided data. White balance is just one thing that can be applied. It's also important to note that saying something looks bluish is irrelevant to its use as a reference. The manufacturer should specify their intended use and other details. Anyway I hate the way lightroom processes data, but if you're interested in making a science project out of it, look into video work and scene referred workflows.

[andrewrodney]

Re: Question about grey cards

The only way to know if the card is neutral is to measure it with a Spectrophotometer. Then examine the spectral plot or the aStar and bStar axis. Forget what it looks like visually. Photographing it then examining the resulting values in Photoshop isn't providing accurate readings of that card.

[Tony W]

Re: Question about grey cards

I agree that measuring with a Spectrophotometer will give you accurate objective figures to work from but not everyone has access to accurate devices and exact measurement and confirmation may be overkill, but it depends on needs.

If you really want something that should be accurate there are grey cards around that are supposedly certified to fall within spec. WhiBal cards come to mind as they are said to have both *a and *b Lab channel values of 0 (+/- 0.5) each card measured to certify that they conform to this spec.
http://michaeltapesdesign.com/whibal.html or the Robin Myers grey cards seem to be well regarded but not sure if they are certified in any way

Never used either product so cannot comment on quality

[andrewrodney]

Re: Question about grey cards

Got WhiBal cards (they are good), those from X-rite and the best alas no longer being sold (babelcolor.com). I'd go X-Rite Passport for $99 as it's a multi tasker.

[Tony W]

Re: Question about grey cards

Thanks for the confirmation on the Whibal cards quality. I have to say I have been toying with the idea of the X-Rite Passport or the larger Color Checker cards think the Passport probably best all round. At least I should bin my Kodak 18% greys as even without a measurement device my eye tells me they have actually yellowed

[andrewrodney]

Re: Question about grey cards

I measured the WhiBal (off white) and the off white of the Passport (the large target), I get this:

WhiBal: 94.3/-0.97/1.41

Passport: 80.8/-0.11/1.05

Macbeth is a bit more neutral.

And the sad news, here are the measurements of the BableColor tile no longer offered:

100.3/-011/0.44 Awesome!

Used an EyeOne Pro-2 and BableColor's Patch tool for measurements.

[genf]

Re: Question about grey cards

Everyone, thanks for your reactions.
Yes, I am fully aware that I can't trust my eyes to judge color neutrality. The same for a photograph taken in a non-controlled environment.

The purpose of my small experiment was to find out if the three cards have equal neutrality. The photograph was taken on a desk, lit by a northern window in overcast conditions. No artificial light. Even though it's not scientific, I would not expect the values I measured. The AB values of my white card measure 2/-6. If that's by design, I'm (unpleasantly) surprised. If it's a production error, then I lose confidence in all three cards.

Yes, maybe I should look for a more reliable brand. Thanks for all tips.

[Tony W]

Re: Question about grey cards

I am not sure I understand why you would loose confidence in your cards for a number of reasons:

Your grey card appears to be neutral, although as Andrew points out unless you have equipment to accurately measure then you are taking a wysiwyg stance for accuracy.

Unless you have bought a certified set of cards, those mentioned or even These then you are at the mercy of the manufacturer and their production and quality controls. If you do buy certified how do you confirm that your set conform within the manufacturers tolerance? And if they do not then how important is this accuracy to your image making?

At least one company Warmcards.com manufacture white cards designed to warm and produce a 'more pleasing' skin tone than neutral white cards and also cool cards for the opposite effect. I suspect that your white card manufactured with similar aims - of course it could be just poor or no qc at the factory!

[andrewrodney]

Re: Question about grey cards

Couple points. If you're working with raw, you're better off with an off white (non spectral white) than gray. That's due to the linear data encoding of this data. You can use gray, it's not optimal for raw but is for a gamma corrected image. Using gray could produce some color shifts, kind of defeating the purpose.

Does the off white card need to be totally neutral? Be nice. But I find that 99 times out of 100, a WB on a neutral card doesn't appear as I desire and at least in Adobe raw converters, I have to season to taste using Tint/Temp sliders. Apparently X-rite and Warmcards agree as at least with the Passport, you have a group of off-white (cool to warm) white's to use for WB. Clever and it works real well.

So I wouldn't agonize over the neutral values unless you have a very specific kind of work to do where that would be rather critical. I've heard of people using coffee filters for WB cards with great success! But again, if starting from scratch, I'd go Passport in a second. You get a Macbeth color checker which is useful. You get the differing whites plus one big target that is rather neutral. You get a protective case. You can build DNG camera profiles with it. All for $99. Screaming deal.