Chromatic abberation play (combined UV and IR in photos with uncorrected UV5035BK lens)

[lukaszgryglicki]

Hi, I've found this interesting.

 

Universe Kogaku Optics UV5035BK isn't corrected for different wavelengths, so I was thinking that I will easily reproduce affect I once had with UV filter with IR leak in Phoenix, AZ.

 

I've took just Hoya U-340 4mm which passes a bit deeper UV and then has IR leak. That IR leak is small but IR amount in sunlight is bigger than UV, they about cancel out creating image with IR and UV. The lens has huge focusing difference between UV and IR, even at f=11 or f=16 which gives two kinda-focused images when adjusting focus. One with UV sharp and IR blur and another with IR sharp and UV-blur. See them (NEF images from Nikon D600 full spectrum - white balanced to concrete somewhere in images).

 

Two last images are just:

- SEU - upper UV-A so blue dominates a lot.

- SEU+Hoya U-340 - this is to go deeper in UV while blocking IR - no idea what is the transmission of that stack.

 

There is some light leak/reflection that was not visible while shooting - Nikon only produces either just RED images or mono - it is never able to WB in-body.

[Stefano]

This nicely shows the focus shift of uncorrected lenses. Corrected lenses like the UV-Nikkor should not do this, or do it much less.

[lukaszgryglicki]

UV-Nikkor also has focus shift between IR and the rest (but it doesn't have between UV & Visible). But I need to test for this specific effect and compare...

 

[Wayne Harridge]

Any idea how focus stacking software would cope with this situation?

 

[Doug A]

Neat effect. The difference in focus is easy to see. Looking forward to your UV Nikkor test results.

Thanks for sharing,

Doug A

[Andy Perrin]

Wayne, focus stacking probably wouldn’t help because this is caused by having multiple depths-of-field in the SAME PHOTO, rather than a single DOF that shifts location. I would imagine that having two DOF at a time would badly confuse stacking software!

I believe the software works by building depth maps, but each point in the image would have two different apparent depths here.

[nfoto]

"Confusion of depth" replaces "depth of confusion" .... Or "confusion of circle" substitutes for "circle of confusion". I'm confused and no wonder the software responds in a similar manner.

 

I agree. Focus stacking is not the solution here. Do not ask the lens to deliver the impossible, or use it with a narrow bandwidth filter.

[lukaszgryglicki]

I know it is designed to be used with narrow band filters - I will be using this in my UV-B stack.

I just wanted to shouw "two focus" effect that is interesting imho - I can use a filter which transmits say 70% of UV and leaks 2% or IR and then because sunlight output is kinda reverse - I can do two focus photos - one with sharp IR and UV blur and another with sharp UV and IR blur... and because focus differs so much - it looks "interesting" - I  don't think any software will be able to fix this - how software could create a sharp image from blur? Impossible IMHO.

 

[dabateman]

@lukaszgryglicki

"how software could create a sharp image from blur? Impossible IMHO"

 

These days, Just type the text into Ai program and it spits the image out. Soon working from a blur template to get a simulated sharp image, will be a click away. 

[Stefano]

I have experimented with DALL-E 2 and it knows what a NIR image looks like (Wood effect), but only knows UVIVF images. This is a reflection of how much less known UV photography is compared to IR photography (AIs mirror us, they have learned from us).

[lukaszgryglicki]

As an almost 15+ year software developer I don't really like AI.