Comparison of White Balance Targets for IR and UV Photography or Video

In this post I'll be looking at various different white balance targets and seeing how they compare when doing an in-camera white balance for infrared photography or video. (Only actually looking at stills but if you can get the in-camera white balance set correctly for stills then it will work for video too). I'll be looking at the results when using filters that pass both visible and infrared light, one that passes mostly just infrared, and also one that passes just ultraviolet.

I'll be looking at both reflectance based targets where you take the white balance using the light reflected from the target. And also semi-transparent incident based targets where you place the target over the end of the lens and shoot towards the light to take the white balance reading.

The White balance targets I'll be comparing are:

A4 sheet of white card
X-Rite / Calibrite Color Checker Passport Photo 2
Labsphere Spectralon reflectance standard
3.5mm thick PTFE disc, as both a reflectance based and incident based WB target
3mm thick expanded PTFE disc, as both a reflectance based and incident based WB target
Expodisc v2
Expodisc v3, as both a reflectance based and incident based WB target
0.1mm, 0.25mm, and 0.3mm thick PTFE sheets as incident based WB targets
Clouds
Using white balance colour picker in post on the Clouds

When looking at the comparison images, what I am looking for is neutral gray clouds. So if I hue shift or channel swap the image, the clouds should remain neutral because the red, green, and blue channels all have the same values for the clouds.

To evaluate the results I used color sampler points placed on the clouds in Photoshop and looked at the differences in the R,G,B values. And there is also my subjective opinion on whether the clouds look neutral.

If you'd prefer a video version of this comparison, you can find that here: Comparison of White Balance Targets for IR and UV Photography or Video.

Test with Lee 101 (Visible + IR)

To start off, I'll compare the different white balance targets when shooting a scene with the Lee 101 filter. This filter passes a mix of visible and infrared light. White balance was set in camera using each of the different targets, and a 98° hue shift has been applied in post.

White card

WB on clouds is off, though doesn't look too bad once hue adjusted. You have to adjust exposure when taking the white balance image to avoid over-exposing the target. (If the white balance target is over-exposed you won't be able to use it for setting white balance).

Colour Checker

With hue adjustment, the image is more red than when white balanced with the card. Using the large grey white balance target on the color checker, no exposure adjustment was needed when taking the white balance image.

Labsphere Spectralon

Again, doesn't look too bad, but the Clouds are not neutral. White balance is almost identical to the White card. With a white standard, as used here, exposure adjustment is needed when taking the WB image, but you could just use a grey one.

The main issue is that the small size of the standard means it is difficult to make it large enough in the frame to cover the WB sample area while avoiding shadow (e.g. of the camera) on the target:

PTFE disc reflected

Pretty much the same white balance as the Labsphere and White card, WB is not neutral on the clouds but the resulting image is not too bad. Exposure adjustment was needed for the WB photo to avoid over-exposing the target.

It is better than the Labsphere in terms of having the target big enough in the frame while avoiding shadows, but not as good as an A4 sheet of card:

Expodisc v3 reflected

I asked Expodisc if you could use the Expodisc as a reflective target, and they said that it has not been designed for that, so WB results may not be accurate if used that way. But I decided to try it anyway. With the hue adjustment applied, there is less red and more green than the PTFE disc reflected. I wouldn't say it is better or worse than PTFE used for reflective WB readings, just different.

Expodisc v3 covering lens, camera pointed towards light

Using the Expodisc correctly, you take your WB photo with the Expodisc over the lens and point the camera towards the light source. You use it like an incident light meter. I find this an inconvenience as I set up my framing for the image I wanted, then had to remove the camera from the tripod to take the white balance image, then have to put the camera back on the tripod and adjust the framing slightly again. But this is more of a problem of getting used to having to take the WB image first than it is an actual issue with the Expodisc.

No, or minimal exposure adjustment is needed when taking the WB image, which is nice.

The resulting image isn't that great though, with the hue shift applied, blue tint is fairly strong and red is quite weak. Quite possibly the White balance is technically correct, but I prefer a White balance that gives neutral clouds.

The Expodisc does come with an adaptor to fix the disc onto the end of your lens (image below) but it's much simpler to just hold the Expodisc over the end of the lens manually.

Expodisc v3 covering lens, camera pointed towards subject

If you use the Expodisc incorrectly, and take the white balance photo with the lens pointed towards the subject rather than the light source, this gives a bad result. With the hue shift applied, red is very weak and green is quite strong. Exposure adjustment is needed when taking the WB image to avoid under-exposing it.

PTFE disc covering lens, camera pointed towards light

I wanted to see if using PTFE over the lens would work, in the same way the Expodisc is used. Indeed it does, and gives us a similar result. As with the Expodisc, using 3.5mm thick PTFE means no or only a small adjustment in exposure is needed when taking the WB image.

PTFE disc covering lens, camera pointed towards subject

I also tested the PTFE over the lens and with the lens pointed to the subject. This gives a bad result, very similar to what we got with the Expodisc when pointed in the wrong direction.

ePTFE disc reflected

Using an expanded PTFE target gives much the same result as standard PTFE. This has the same issue of needing to adjust the exposure to avoid over-exposing the target for the WB image.

ePTFE disc covering lens, camera pointed towards light

Using ePTFE over the lens like the Expodisc does not work as well as standard PTFE and we get an image with less red (after hue adjustment applied).

ePTFE disc covering lens, camera pointed towards subject

This required a large amount of positive exposure adjustment (compared to the exposure required for the image) to get the WB image exposed reasonably. And like you'd expect based on the previous tests, the resulting WB is not very good.

Expodisc v2 covering lens, camera pointed towards light

The Expodisc v2 was only designed for visible light. The resulting image, once hue adjusted, has less red than the Expodisc v3. So v3 does seem a bit better for infrared, but not a massive difference here.

0.1mm thick PTFE sheet covering lens, camera pointed towards light

Trying thin PTFE over the lens, when taking the WB image the sun can be clearly seen and a lot of exposure adjustment is needed to avoid over-exposing the WB image. The resulting image once hue adjusted does not have enough green, making for a magenta shift.

WB image using the 0.1mm thick PTFE sheet over the end of the lens. (0.25mm thick and 0.3mm thick PTFE sheet WB images were also similar).

For comparison, the 3.5mm thick PTFE disc over the end of the lens gives a much more evenly lit WB image. (Expodisc and 3mm thick ePTFE disc WB images were also similar).

0.25mm thick PTFE sheet covering lens, camera pointed towards light

A better result, but the problems in taking the WB image are much the same as when using 0.1mm thick PTFE over the lens.

0.3mm thick PTFE sheet covering lens, camera pointed towards light

Basically the same as 0.25mm PTFE, unsurprisingly.

Clouds (WB image taken with 55-250 lens)

Due to the cloud coverage on this day I had to use a longer lens to take the WB photo to white balance on clouds. I'm not sure why, but the resulting white balance is not very good and the image has a strong red tint (once hue adjusted). In all my previous tests WB on clouds has given me the best result, in terms of getting an image where the clouds are neutral in colour. All other results in this test were pretty much consistent with what I saw in previous tests, so I'm not sure what happened here.

Having to swap lenses, which meant also swapping over the filter used, just to take a WB image is very inconvenient. Hence why I generally prefer a WB preset image.

Clouds WB in post

Using the color picker to set white balance on the clouds in post gives the best result in my opinion.