[filter test] 300Bp10

[dabateman]

I purchased a cheap 300bp10 12.5mm filter from ebay and a filter wheel with seven 12.5 mm filters, one of which also being a 300bp10 filter. These filters surprisingly cover the m43rds sensor without vignetting. I mount these filters in a 25mm filter frame using a custom plastic washer. Basically I drilled a 1/2 inch hole in an old black DVD case, then used the 25mm glass from the frame to mark the outer circle and then cut that out with a razor blade.

 

I think this is the limit for my lights. The filters were very different, with the first one having lots of IR leakage making exposures very difficult. The one I pulled out from the filter wheel is much better with a lot less IR leak.

 

First 300bp10 fitler stacked with 330WB80 improved filter to cut the IR.

Image using UAT 85mm f4.5 lens, at F8 ISO 200 Using 125W lucky herb mercury light, exposure time was 17 minutes 15 seconds.

 

Filter pulled from wheel only, can see some IR leakage in image.

Image using UAT 85mm F4.5 lens, At F8 ISO 200 Using 125w lucky herb mercury light. The exposure time on this shot was 9 minutes.

 

I will have to do a better comparison and make sure to rotate out my PFTE block, so it doesn't reflect out so much light.

 

Comparison image with 330WB80Improved ISO 200 F8, 2 seconds with UAT:

[Cadmium]

What do you think? Conclusions?

I might be looking for a lot of cerulean and turquoise down in that range, theoretically.

[dabateman]

After looking as far down as I can. I think, at least for my sensor, that its all green down there. All the filters I have tested leak some IR, possibly in the over 850nm range. This gives some times a gentle or strong blue glow. Websites that I have read, that speak of blue or blue white at the UVb range to UVc range are really looking at IR leakage.

The green channel, at least for me, is very strong.

I will have to test a trial version of software to really analyze the Raw data, but that is what I see.

The UAT also holds focus point down at 300nm. Which I think is amazing. I can get signal with 313bp25 filter with igoriginal 35mm and Steinheil 50mm f2.8 lenses. But the focus shifts a lot from previous Baader adjusted focus point. Also images not so usable, from those lenses.

I am seeing some slight different coloration of the flower with different specific filters. I may post a composite photo to show this at some point.

[dabateman]

In smmary its a long road to UVb or UVc, you end up spending lots of gold. But you get to the Emerald City, where every thing is green.

 

[JMC]

Just been outside and done a Sparticle photo, using my converted d810 and UV Nikkor lens (ISO3200, times between 1s and 30s, f4.5). These were recorded as RAWs and then white balanced in darktable using PTFE tile.

 

I've selected the 4s one here (the filter center wavelengths are above the circles);

 

Everything gets nice and green at 321nm.

 

The 303nm one is barely visible even at 30s exposure. But has a green tinge to it.

[Andrea B.]

What does the raw file look like, please?

[JMC]

This is the RAW file put into RawDigger and displayed as a RAW composite. I have then cropped it to the same size as above;

 

And as the RGB render;

 

A lot more yellow and gold now than green....

[Andrea B.]

Thanks!

I like to avoid the side effects of white balancing. Too many variables in that process.

 

Curious what the 303 looks like with a longer exposure? Anything there?

 

At high ISOs and long exposures I wonder how much the effects of noise are going to skew colors (whether false or real)? Although that D810 is a champ at high ISOs. This range is where it would be very cool to have a 330nm UV-Led flashlight/torch to shorten exposure times as much as possible so that ISO can be kept low. :D

[JMC]

Andrea, there will be pretty much nothing at 303nm with my ACS converted d810, as their filter looks to be similar to the Baader U from what I have seen so far. I'll need to use my multispectral Canon 5DSR, and run long exposures to see what 303nm looks like. Can't do it today though, its raining here.

 

EDIT - the rain went away and the sun came out. This test is with my EOS 5DSR multispectral conversion (with a WG280 window internally). Sunlight, but the Sparticle was not pointed at the sun, I always have the setup itself in shadow, but pointed at the open sky. UV Nikkor at f4.5, ISO6400 and 30s. I blocked off all but the top 3 filters in my Sparticle - 303nm, 321nm and 341nm. The camera was so sensitive to the other wavelengths that I was getting flare without them blocked off. I also added a U340 2mm filter to try and cut down on non UV wavelengths, and still let light in the 300-350nm region through.

 

Firstly the RAW composite (from RAW Digger);

 

Now the RGB render (again from RAW Digger);

 

And finally white balanced (using PTFE tile in Darktable);

 

In the non white balanced images the 321nm one looks nicely golden again (despite this being a different camera). The 303nm is barely visible even here. It's clearest (and I use that term loosely) in the RAW composite image, and to me looks like the 321nm one, just a lot darker.

 

The big issue I see here is the camera is so sensitive to longer wavelengths, that you would need a monochrome light source at 300nm to work down there, or a 300nm band pass filter. Trying to see 300nm + other wavelengths, the longer wavelengths would just swamp the 300nm signal.

[Cadmium]

Jonathan, What filter are you using on the lens, other than the added U-340 you mention?

[JMC]

Nothing in addition to the U340 on the lens Steve. On those final images it's the U340 on the lens, and filters from my Sparticle, and that's it.

[Cadmium]

Jonathan, thanks. Cool. Interesting. Thus probably the slight color shift downward from the WB Sparticle shot, you think perhaps?

So conclusion? Sensor limit to about 300 then? What is your conclusion here?

[ulf]

I would expect that the peak transmission of the filters are different.

Often for the ones I have seen atOmegabob, the short wavelength variants has less transmission.

Jonathan, have you measured the transmission of the filters and can share the results?

[JMC]

Steve. Its difficult to draw a conclusion about the limit of the sensor from my images here. There's a few reasons for that. There's very little light down there in sunlight anyway. I have a WG280 window over sensor as part of the conversion. However other work I have done measuring the spectral response curves would suggest that on my cameras at least the sensitivity drops down to virtually zero around 300nm. My monochrome conversion has more sensitivity than the one with the CFA still attached, but even the monochrome struggles down there. If I had a different window (instead of the WG280) then maybe it might go a little lower, but with the CFA in place there is virtually nothing down there.

 

Ulf, the transmission curves for my filters in the Sparticle are here;

http://www.ultravioletphotography.com/content/index.php/topic/2580-build-thread-at-home-measurement-of-camera-uv-spectral-response/page__view__findpost__p__22555

[OlDoinyo]

When a signal is as faint in relative terms as those shown here, one must wonder if the color swatch at 300 nm is really due to light at that wavelength as opposed to out-of-band leakage from longer wavelengths. This is a difficult thing to prove one way or the other, but perhaps one could devise some way to test this.

[dabateman]

Steve and Jonathan,

My gut feeling is that our 550nm optimized sensors have a limit at 290nm. This is based on my exposures, tests with bandpass filters and Andrea's images with her 293bp10 filter.

In the high quality filter wheel I won, I have 260bp10, 280bp10, 300bp10, 340bp10, 350bp10, 380bp10 and 400bp10 filters which are small at 12.5mm but are good at IR blocking. Although from first series of tests their need exposure values is long, possibly indicating only around 15 to 20 % transmission. As I have compared with my 335bp10 and 340bp10 25mm filters where I know the transmission is 80%.

 

Jonathan you could try to remove your 304nm filter and mount it in an isolated filter holder as I have to attempt images with your monochrome camera. Would be interesting to see your exposure values. I would guess 3 minutes would be needed with a mercury vapour lamp.

 

We also have to remember that down there, I think most organic thinks will be absorbing. Protein will be absorbing at 280, as will glycoprotein and aromatic linked polymers. DNA absorbs at 260nm and everything absorbs at 230nm. So it will a dark world, like trying to fly the Daster areas ship from Hitchhiker guide to the Galaxy. Black light iluminating black buttons with black letters.

 

I will have to see if there are fruits or vegetables with known polymers. They maybe reflective. I will have to try a spider web. That might be interesting as it forms a know amyloid shape and might work down there.

[JMC]

David, building the Sparticle was a one way trip for my filters. The only way they are coming out is in pieces.

 

Thing is I've done these measurements with my monochromator/integrating sphere set up that I use for measuring spectral sensitivity. There is still a tiny amount of sensitivity at 300nm. Even less at 280nm, basically just above the noise. I can't go any lower, and with my WG280 window on the sensor I wouldn't expect to see anything anyway.

[dabateman]

Well I must say if it wasn't free the Sun is really a horrible lighting source. The UV was not very strong today. I tried to do a 300nm image in sunlight and got the following image:

 

I have no idea why the 300nm filter is moving through a time tunnel like Dr. who. I have no obvious light leaks, It was shaded and I had a hood on. There seems to be some scatter though. The pattern reminds me of the aperture blades, so there maybe some internal scattering. I may try this reverse mounted.

 

The same flower using a 313nm filter with ISO 200, F8, 60 second exposure.

 

Here is the same flower using Baader venus filer ISO 200 F8, 6 seconds.

 

These are all full image No cropping.