UV-green with GFX 50R

[lukaszgryglicki]

So I was trying to use my tri-filter UV-B stack on my converted Fuji GFX 50R.

I already used that stack many times with Fuji 50R mono and it has a peak around 306-307nm and passes very very little above 340nm, there were many discussions related to this.

So this stack was usable with GFX live-view with mono-conversion, but the live view is no longer usable with a non-bayered sensor. I was focusing by guessing - many trial-and-errors on ISO 25600, UV-Nikkor at f=4.5, and exposures around 10s-20s in full sunlight. I wanted to see what colors will be recorded by the RGGB matrix that low in UV (UV-B).

 

They are mostly green, and next comes red (a lot less) - blue is basically pure noise.

 

From all shots I made, those with longer exposures were unusable due to light leaks (looks like mount adapter leaks not filters, my custom hood that puts 2 IV 308nms inside and Hoya U-340 nm outside may have some reflections as well - even if it looks black).

 

This is the filter stack transmission info: https://docs.google.com/spreadsheets/d/1zMXCT4vtTR6PGf-yzFgUJgUop3gfkpaRAZ50IqN2lsQ/edit#gid=416424503 (look for data for stack 2 x IV308 + Hoya U-340)

 

Those very very very barely usable files were shoot at ISO 12800, f=5.6, t=20s or more (UV-Nikkor).

 

Here are RAF files: https://teststats.cncf.io/backups/uv-green/

 

JPGs will follow in next post (because I'm a bit afraid that I will see server error again as I saw maybe 5 times already today).

JPGs SOOC (as saved by the camera).

[lukaszgryglicki]

Pasted JPGs in that link too (small and full) - sorry but it is not possible to upload anything to the forum.

 

I constantly see this or uploads just fail.

/cgi-sys/images/e.png

ERROR 500 - INTERNAL SERVER ERROR

Why am I seeing this page?

Is there anything that I can do?

Understanding Filesystem Permissions

How to modify your .htaccess file

How to modify file and directory permissions

[JMC]

When you said "non-bayered sensor", I presume you mean 'non-debayered'?

 

The Bayer filter will absorb a lot of UVB - my experiments have shown about a 5 stop difference between 313nm and 365nm for a camera with the Bayer filter still intact. I presume the live view needs a certain amount of light to work, and there just isn't enough getting through to allow it to work. I even get this occasionally with my mono cameras on my microscope - at 313nm with high magnification objectives, live view will work fine at 365nm but not at 313nm.

 

It will be green at around 308nm with a Bayer filter. But keep in mind that the blocking you will need will be much much more than for the mono camera -  with a Bayer filter it'll be much less sensitive to UVB, while essentially being just as sensitive to Vis and IR. When I force IR leaks with a Bayer filter I often get a green tinge to the image, so it can be tricky to decide if it is leaks or true UVB.

[dabateman]

That looks like what I would expect.  UVB seems to be green. That is what I see at 303nm, especially.  

I can white balance a little at 313nm to get some more yellow. 

But its really mostly green.

Sunlight will be hard, as there is very little UVB there.

Was the coverglass of the sensor removed?  Or just the UV/IR blocking filter for the full spectrum converted 50R with bayer color filter intact? 

If not the coverglass,  than that is good news. A converted 50S or 50R might be one of the few cameras to see UVB then.

[lukaszgryglicki]

This was from Fuji 50R with UV/IR glass removed but Bayer matrix intact.

[JMC]

Ok, with the Bayer filter intact, a rough estimate is 4ish stops less sensitive at about 310nm compared with the monochrome version of the same sensor.  Think 1 or more ideally 2 OD better blocking of the longer wavelengths being needed when using a sensor with a Bayer filter vs a monochrome one for UVB work. UVB in sunlight with a camera with a Bayer filter intact is a challenge.

 

As a simple test, if you want to check for leaks, then just try a decent UV blocker such as a Zeiss T* filter in front of everything - if you still get a similar image with that in front of everything then you have major contributions from the visible and/or IR.

[Stefano]

So this is another example that most sensors see green at and around 310 nm. With most cameras green is what you should expect to see.

[lukaszgryglicki]

I think I have enough blocking (stack of 3 filters, IV308 x 2 and Hoya U-240) that I'm able to see pure UV-B, other "prove" is that I see nothing when I close window or when I use any lens other than UV-Nikkor and UV5035BK. Signal is mosly green, but focusing in LV or EVF is no longer possible, while the same stack with mono camera (same camera type) allows focusing in LV and EVF easily, just a bit slow like 3 fps...