Camera sensor UVIF

[aphalo]

I am puzzled. I have been scratching my head for ages trying to work out why UV photographs taken with the full-spectrum-converted Olympus E-M1 (Mk1) camera I have been using for some years show diffuse yellow halos around strong highlights... The answer I seem to have found... seemed illogical at first, but I am now quite convinced the camera sensor itself fluoresces. It fluoresces enough to be visible to the eye. Has anybody seen something like this before?

 

It seems to be the sensor itself that fluoresces, but it could also be related to the "proprietary coated glass" used as filter replacement by DSLR AstroTEC (http://www.dslr-astr.../index-eng.html) for the conversion.

 

Fluorescent coatings are used on some sensor arrays of spectrometers to extend the sensitivity deeper into the UV, but this makes little sense for a normal photographic camera.

 

What is the evidence:

1) For some time I have noticed that the yellow halos did not differ much in sunlight, using a "full-spectrum" flash or using a 365 nm flashlight.

2) I started suspecting that the cause could be fluorescence of the lens itself as I was using front mounted filters (both Baader U, and StraightEdgeU.

3) I could still see the halos with SWB1 and SWB2 3mm filters mounted behind the lens in a DEO-Tech OWL adapter with a filter drawer.

4) Yesterday I slightly "shortened" a spare Baader filter ring, so that the Baader filter would easily fit in the OWL adapter, the halos were still there.

5) I went into a darkroom shined my 365 nm LED "flood" flashlight into the camera, and the sensor glowed rather faintly clearly yellow.

6) Today I took photographs of the glowing sensor (of course also showing the blue fluorescence from dust and lint).

 

 

 

Both photographs were taken with an off-the-shelf E-M1 mk II, with a M.Zuiko 60 mm f:2.8 macro with a Zeiss T* UV filter stacked with Tiffen Haze 2A filter.

 

In addition to the fluorescence of the sensor, the strongly fluorescent red index dot can be seen. The badly done reassembly and chipped glass edges can be also seen. The M.Zuiko 60 mm f:2.8 transmits very little UV radiation itself.

[dabateman]

Yes I see the same.

Never thought to look. I just shawn my convoy 365nm light on a stock Em1mk1, Pen-f, full spectrum converted Em1mk1 and full spectrum converted Em5mk2.

They all have some fluorescence. The stock cameras have it but minor due to sensor glass filtration. The full spectrum converted cameras its quite strong for both Em5mk2 and Em1mk1.

I also looked at my PiHQ sensor and it too is fluorescent. My monochrome pi HQ sensor is not fluorescent, but has some extremely bright fluorescence around it at small points.

My lodestar camera with GMYC color array is possibly twice as fluorescent as the other cameras.

 

So my conclusion is that the RGB dyes used on the sensor for color are fluorescent. I don't have an other camera brand that is full spectrum to test. But this might be Universal.

 

Message Jonathan as he has a Canon full spectrum and monochrome converted cameras and could say if its universal or just with Sony/Panasonic sensors.

[Cadmium]

Pretty cool. :-)

[JMC]

I see a similar effect with one of mine with a glass window on the sensor, I don't believe I see it with the quartz window, so I'm not so sure about it being the Bayer filters. I'll try and get some comparative shots when I can.

 

EDIT - I stand corrected. I have a Eos 5DSR with a WG280 sensor window, and a Sony A7III with a quartz window (both with the Bayer filter attached). Both give the orange appearance 365nm UV. So it would appear that something about the sensor is fluorescing.

 

I also have a monochrome Canon 5DSR and Nikon d800 with Schott WG280 coverglass and a Nikon d850 with quartz coverglass, so will check those too.

 

EDIT 2 - Ok, I still see the fluorescence on the monochrome converted cameras. One last thing to check - I have an old EOS with a talf the Bayer filter layer removed. I'll charge up the batter for that, lock the mirror up and have a look at it. If someone has a Sigma with a foveon sensor it may well be worth checking that too.

[colinbm]

I know of a guy that purchased a mono converted from Max Max & it still had significant colour on some pixels.

[colinbm]

I think the significance of this should be viewed in the way we use the UV lights, we are not photographing the light source, just the reflected light, so how much UV is actually getting down their to the sensor ?

[dabateman]

Oh good points Jonathan. My piHQ monochrome has nothing on it. Its a bare sensor. Since it doesn't fluorescence, it may not just be the color filter array, but could be the glass used on the surface of the sensors.

[JMC]

Ok, so this is interesting. I have a spare sensor (not sure what from) but if has part of the Bayer filter removed to reveal the bar sensor underneath.

 

This is the sensor in visible light. On the right hand side of the sensor part of the Bayer Filer and microlenses has been removed.

 

And now under 365nm (Nemo torch and it's own built in filter).

 

From this is looks as though the bare sensor itself is fluorescing more than the Bayer filter layer. This would suggest that the Bayer filter layer absorbs some of the UV and prevents the bare sensor from fluorescing as much as the area where it has been removed. Not sure what to make of that.

[colinbm]

Do you have access to a silicon wafer with whatever they dope them with to make make a photographic sensor.

[JMC]

Do you have access to a silicon wafer with whatever they dope them with to make make a photographic sensor.

 

Me, no, sorry I don't.

[dabateman]

Jonathan,

I am confused by that sensor. I see the left half clearly has the microlenses and CFA. The right half looks to have them removed. Then a central portion in that right half seems to have some kind of color smearing in the visible.

Then looking at the 365nm only the smeared area has fluorescence.

 

Is that correct, there is pixels on the right absent of CFA and microlenses that is not fluorescent?

Only some damaged sensor section or mixed up CFA region is fluorescent?

 

Also on my Em1 and Pen-f cameras the shutter curtain is wickedly fluorescent.

[JMC]

David, on the right side, most of the right hand side has had the microlenses either fully or partially removed (that's the bit that looks darker in the visible image). The central portion of the right hand side has had the Microlenses and underlying Bayer filter removed. I think all this was done by polishing the sensor. The bright are in the UV fluorescent area is the part which has had the microlenses and Bayer filter removed.

 

Interestingly on my EOS 450d which has had the half the Bayer filter removed (and does not have any cover glass at all), I get a deep red fluorescence across the entire sensor, not an orange glow.

[aphalo]

I found some extreme examples of yellow halo from last year. I have either discarded or cannot find less extreme cases. These are taken with an M.Zuiko Macro 30mm f3.5 MFT objective and either Baader U or StraightEdgeU. UV from a filtered Convoy 2+ with 365nm Nichia LED. Same full-spectrum converted E-M1 camera whose sensor's fluorescence can be seen in the first message of this thread. White slab is virgin PTFE 6mm thick. Black slab is 6mm black Nylon 6 painted with Tetenal black "camera varnish".

 

 

 

Visible:

[Stefano]

Seems like a ring, so it is localized. That's odd, I would have expected something more uniform. I don't think it is around bright areas (at least it doesn't appear so to me judging from this image only).

 

Also, is that a UV-yellow rock/mineral? That's cool.

[KhanhDam]

On 3/9/2021 at 7:53 AM, JMC said:

From this is looks as though the bare sensor itself is fluorescing more than the Bayer filter layer. This would suggest that the Bayer filter layer absorbs some of the UV and prevents the bare sensor from fluorescing as much as the area where it has been removed. Not sure what to make of that.

 

 

Thanks for those photos. definitely looks like the bayer layer is darker in UV. Doesn't look like it's fluorescing to me though, just looks darker. 

[KhanhDam]

found this youtube video and did 2 frame grabs showing UV light used to bond sensor cover glass over the sensor.