Olympus Internal blocking filters

[dabateman]

I had a chance to finally do some quick scans today of the dust shakers and UVIR blocking filters that I received back from my Olympus full spectrum camera conversions from Kolari.

 

Here is the internal dust shaker from the EM1mk1:

 

Dust shaker from EM5mk2:

 

The dust shaker that I pulled out of my old Olympus E510:

 

This is the UVIR blocking filter from the EM1mk1:

 

This is the UVIR blocking filter from the EM5mk2:

 

 

For fun this is the UVIR blocking filter from my Sigma SD14, which I broke, but I keep the pieces around:

 

 

These are just quick scans so be cautious with any odd conclusions you might draw out. What I might or might not investigate is that 1 percent tail for the EM1mk1, that smooths out to the 350nm glass cut off. That extremely tinny amount of extra UV seems to account for the added UV sensitivity of the EM1 over the EM5mk2. The EM5mk2 drops hard at 375nm, nothing below it.

[colinbm]

Thanks Dave

[Andrea B.]

Interesting right shoulder on the EM UVIR blockers.

And very steep on the left. That's not BG glass alone.

Those must have coatings or are "sandwiches" ?? Do you see anything like that?

 

It appears as though the Sigma blocks violet light. Strange.

[Andy Perrin]

It appears as though the Sigma blocks violet light. Strange.

I think that the Sigma blocking a bit of the violet makes more sense to me (given our computers cannot even display violet - the blue LEDs of our screens are 450nm I think?) than that the EM1 etc. try to record the violet which cannot be displayed!

[dabateman]

Andrea,

They look like this:

https://www.ultravioletphotography.com/content/index.php/topic/3150-enhanced-uvb-sensitivity-in-monochrome-converted-eos-5dsr/page__view__findpost__p__25970

 

A sandwich

[Stefano]

I think that the Sigma blocking a bit of the violet makes more sense to me (given our computers cannot even display violet - the blue LEDs of our screens are 450nm I think?) than that the EM1 etc. try to record the violet which cannot be displayed!

Well, violet light cannot be displayed as well as yellow light, cyan light (that's the hardest) and magenta doesn't even exist.

 

To me instead it makes sense that a camera can see violet (which is a band of visible light often forgotten, like you jump directly from blue to UV), and you can give the illusion of violet with blue + some red. That looks almost exactly the same as real violet (to me, at least). I don't know how you would see violet, your 405 nm laser is definitely violet (I had the same laser), you may look at the diffused light on a PTFE disk or a non-fluorescing surface, of course being careful.

 

Older cameras displayed violet (like from your laser) as blue, which wasn't correct, while more recent cameras with better sensors see colors more accurately and see violet as violet.

 

This is an interesting video about this:

(go to 8:45 to see a comparison between a bad and a good rendition).

 

I actually wondered how this improved CFA (with more red response in the violet range) affects false UV colors. There must be some difference.

[Andy Perrin]

I don't see any red in violet in the first place? I take it other people do, though. To me, in person, the 405nm laser does look like a more intense blue that my computer screen cannot reproduce. But I routinely chop it out of my visible photos with the 420nm cutoff on the Tiffen 2E filter because it makes no sense to me to record it on a computer than can't show it. Your other examples seem totally different to me - yellow is displayed as a mix of red and green, and yellow's wavelength is right between red and green, so the fact that the computer can't display yellow directly makes no difference because it can approximate it. Same with cyan. Magenta is weird because it's a color that's non-spectral, so who cares.

[Stefano]

That’s interesting, you see violet as a stronger blue instead of violet, probably because of your colorblindness (do note that intense violet light does look bluish to me too). As I understand it, our red cones are slightly sensitive to violet, and so you have blue + some red as a signal. “Fake” yellow looks practically the same as actual yellow to me (I have a yellow LED that peaks around 585-590 nm, and so appears as a “sodium” yellow, a bit orangeish. Actual “pure” yellow, visible with a diffraction grating for example, is more around 580 nm to me, and looks just like “screen” yellow). Cyan is very different. “Screen” cyan (0, 255, 255), although being a very nice color to me, is not what I see looking in the cyan region with a diffraction grating, that hue doesn’t exist. My ~500 nm LED looks greenish, a color not so easy to reproduce on a screen, and a ~480 nm LED looks a bit like the sky, a soft blue.

 

Violet, on the other hand, is a color you can easily simulate on a screen, at low brightnesses. My 405 nm LED is a very nice, saturated, deep violet, a really nice color. My 425 nm LED still looks violet, surprisingly similar to the 405 nm LED, but a bit bluer and significantly brighter. My ~450 nm LED, instead, is just solid blue. Our color vision is very complex.

[JMC]

David,

 

Your dust shakers have a similar transmission to the sensor coverglasses on my Canons. In Olympus, what's the layout of these - is it sensor - dust shaker - UVIR 'sandwich'?

 

Andrea, to answer your question, looking at the transmission the 'UVIR filters' are 2 filters together - a UV/IR cut dichroic one, and an IR blocker (like BG39 or similar). The Sigma one is just a UV/IR blocking dichroic filter.

[dabateman]

The layout is sensor, UV/IR blocking sandwich (3 to 4mm thick), rubber gasket thingy, then hypersonic dust shaker thing (0.5mm to 1mm thick), then outside air.

 

Kolari did return my exact stuff. As I had market it with micro scratch. I was curious.

 

The Sigma is just a pop in plastic frame with 0.5mm glass that breaks when you look at it. Dust can enter actually from around it and you do need to remove it to properly clean the sensor. My SD15 dust blocking filter replacements are slightly thicker. Look to be about 1mm, as was the one in the SDQ I rented from Lens Rentals.

 

If anyone is ever curious about a Sigma Quattro camera, I fully recommend renting the SDQ with 30mm from Lens Rentals. That camera had an excellent grip that fit my hand perfectly, but the meter couldn't save its live to properly expose anything. Manual exposure was almost mandatory. When the UV/IR blocking filter was out, it had some of the best black and white renditions I can remember. Almost sold me on a SDQH, but that meter and high ISO was just pointless. So in the end went with a Pen-f for excellent black and white photography.

 

[JMC]

Cheers for explaining the layout David.