Kolari Vision Filter Update

[Nemo Andrea]

I went back to a different flower (this one is:helianthus tuberosus; as indicated by a sign). Here an image in the visible:

 

Now for UV. White balance was set to white teflon sheet. It seems to be thick enough for this application.

Settings: ISO 100, Canon 1000D, Nikon 50mm f/1.8 @f/5.6, Exposure time: 25s. Unforunately this is as close as the lens will focus in my setup.

 

 

What do you think? I personally don't see much brown in this flower, but I don't know if this flower type is familiar to you. I checked if somehow the high ISO made it tricky, but at ISO1600 colours seemed to be unchanged.

[ulf]

This UV-image looks much more typical to me, with a reasonably dark center of the flower. I cannot see any brown either.

 

The reason many of us jumped at the possibility of an IR-leakage is that brown is normally never seen in proper UV-images.

This might have been corrected by a better white-balance now.

 

However the image do not completely rule out the possibility for IR-leakage in the filter, especially if it is just OD 3.5.

The effect of leakage is not always visible. It depends on light situation and motif.

It is most visible if there is much IR and little UV in the light.

The motif best showing the effect is something like a flower being really UV-dark and very IR-bright at the same time.

 

I think the https://en.wikipedia.org/wiki/Lotus_corniculatus is one of them. It might still be in bloom in the Netherlands.

Here where I live it is very common.

[nfoto]

The last image looks much better. Still, I think there is room for improvements in terms of UV colour balance. The vegetation has some additional magenta tint to it and there still is a slight brownish tinge to the basal patch on the ligulate rays.

 

When you do w/b on PTFE material, it is important to have an exposure that still leaves some headroom so as not to blow out the target completely. Overexposure means the final adjustment to UV w/b is fraught with uncertainties.

[Nemo Andrea]

The last image looks much better. Still, I think there is room for improvements in terms of UV colour balance. The vegetation has some additional magenta tint to it and there still is a slight brownish tinge to the basal patch on the ligulate rays.

 

When you do w/b on PTFE material, it is important to have an exposure that still leaves some headroom so as not to blow out the target completely. Overexposure means the final adjustment to UV w/b is fraught with uncertainties.

 

I've done some thinking, and I reckon it's a limitation of my raw editor (DxO Photolab) ability to pick white balance. I will get affinity photo at some point for another project, which can handle UV and IR white balance natively (without requiring perfect setting during the shoot or custom profiles). Here the same image using the white balance set to a random street during daytime through the filter. Of course this image is not set to a standard so it will probably have some cast to it, but its better than the reference PFTE when the software can't properly WB to it.