Jump to content
UltravioletPhotography

UV Sparticle BandPass Filter Array


Recommended Posts

It seems to me having the sun in the filter will produce unreliable results, simply due to the intensity of light across the UV and IR, and the filter will leak more than usual.

Sky/clouds away from the sun seem to me to be much more reliable test. It seems as sensitivity drops considerably between 370 and 350 nm, consistent with the drop of UV-transmission in many modern multi-element multicoated lenses.

Link to comment

Thanks very much for looking Alex.

Yes I realise that shooting the Sun was cheating, I just did it to see if there was any different colour response from the sensor.

I was pleasantly surprised at the 370nm response at 180 deg away from the Sun, that shows the camera is able to achieve that much.

Now to find a 360nm filter ?

Cheers

Col

Link to comment

What are the specs, bandpass, max transmittance and most important out of band OD, on your filters? I ask because they look rather different from one another.

Are you using a filter on the lens and what is the lens. Did you white balance the camera on a UV neutral target - if possible for your camera (I don't know).

I have been wanting to build one also. I have several filter arrays, I just need to order filters that fit.

- John

Link to comment

The filters are from Omega.....

http://www.ebay.com.au/itm/151151248197?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

http://www.ebay.com.au/itm/350222327560?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

http://www.ebay.com.au/itm/350334204972?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

http://www.ebay.com.au/itm/160623995186?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

Despite what is in these tables the cans are marked....

340BP10

350BP10

370BP20

386.5NB2

When I was testing these filters, with an Amici Prism Spectroscope on the camera, I noticed some IR contamination, so the UV Sparticle test was with a Schott S8612 on the lens.

The camera is a full spectrum, Sigma DP2, APS-C Foveon sensor Compact, with a fixed lens & without the internal IRC filter.

I used the Sunlight White Balance for these.

I will be looking for some more bandpass filters between 340nm & 360nm.

Cheers

Col

Link to comment
igoriginal

Now to find a 360nm filter ?

 

Not necessary, and probably somewhat redundant, since you have a "370BP20" within your array.

 

Why? Because the "BP20" part of the filter naming convention indicates a bandpass transmission width of as much as 20 nanometers across (although estimated, probably, within a margin of error of + or - 3 nanometers, given that filter production results can lead to variations). Which means, that even though the filter is rated as "peaking" at 370nm (again, estimated), it will also continue to transmit some UV down to as low as 360nm, and also up to as high as 380nm. (380 - 360 = 20 nanometers of transmission bandwidth, before a complete cut-off).

 

Thus, you'll likely get some transmission "overlap" if you add a "360BP10" alongside your existing "370BP20." Especially so, since the "360BP10" also has an estimated bandpass width tolerance to as high as 365nm (and as low as 355nm), hence the "BP10" within the filter naming convention.

 

Also of note: The only true narrow-band filter within your list, above, is the "386.5NB2", since it has a rated transmission bandwidth tolerance of only 2 nanometers across.

 

I noticed some IR contamination, so the UV Sparticle test was with a Schott S8612 on the lens.

 

Speaking of the "IR contamination" that you are supposedly getting with these narrowband filters, that is highly unusual ... unless, you are unaware that these filters are direction-specific. Meaning, unlike absorption-type UV-pass filters which can be used with either side, these are interruption-type filters involving one side coated with a mirror-like surface for IR interruption, and so, must be orientated with the correct side facing out (you'll usually find an arrow along the filter rim, indicating proper orientation of the filter).

 

Of course, I could be wrong, but I've also read that these interruption-type, metallic-like, mirror coatings for IR interruption can become unreliable, due to various environmental factors which could lead to them becoming compromised in some way (Oxidation? Coming into contact with oils from fingertips? Other contaminants? Other factors?)

 

Anyway, maybe you are already aware of the fact that these filters are designed to only work "correctly" with a one-directional orientation, so please disregard that, if so.

 

(And, I think it is wise to add a Schott S8612 to the lens, regardless. So, good call. Maybe there are greater risks for IR leaking than these manufacturers would like to admit within their listed filter specifications. Shrug.)

 

I, myself, have a whole batch of them laying around, because I have been procrastinating to build my own sparticle bandpass test (for testing various lenses). So, I guess I should play around with my own stuff, before any further input. :D

Link to comment

Col,

 

Omega makes good filters with standard out of band attenuation OD ≥5. That said you are buying eBay seconds or overruns that for whatever reason aren't inventory standard. Looking directly into the sun even OD≥5 filters will likely still pass enough out of band energy to detect, which you clearly show in my estimation.

 

That said, think there is an underlying issue with what you are attempting to do. I am unconvinced that the Foveon sensor will detect any UV in the deeper green and red layers, if blue won't penetrate UV certainly won't. At best you may have a monochromatic image in the blue channel depending on the sensitivity of that layer. Your struggle is compounded by the fact that the fixed lens of a point and shoot camera, while probably serviceable for NIR, unlikely passes more than meager levels of UV.

 

One easy test you might consider is to place a strong UV blocking filter over the lens, something like a Schott GG400 or GG420. Take another direct sun shot, any image then recorded must be due to out of band radiation at longer wavelengths than the cut off of these filters.

 

Also, the Schott S8612 also cuts off on the short wavelength UV side, ~70%T @350, dropping to ~50%T @ 330nm and ~25%T @ 320nm.

 

I salute your diligence,

 

John

Link to comment

Thanks John

It is a short journey when you go around & around in circles :D

I have one each of L370nm & L390nm UV cut filters, I'll try these. I would like some other longpass filters, but they are harder to find & companies that sell them want two & three times the filters value for specialist freight & won't send by regular post :D

I will make a new holder for the four filters & make them have equal apertures & cut the light leaks down, then photograph again with & without the S8612.

I will do some more with these filters, with the Amici Prism Spectroscope too.

Cheers

Col

Link to comment

Dumb question: I wonder if it isn't just easier and (perhaps better?) just to put the narrowpass UV-filter onto a given UV-lens and modded camera and then try to make a UV photograph of a fixed target under some kind of fixed UV illumination?

 

Needs must to have the test set-up and results repeatable whether using filter board or simply direct photography.

So the UV illumination part is more than just a little bit tricky (as Shane has often pointed out).

If using Sunlight, then you need to always have the board at the same angle to the sun at the same time of day on a very clear day at the same altitude in summer to ensure the same amount of UV for each test????

Hmmmm.......just trying to think it through. Likely overthinking it as usual. :D :D :D

Link to comment
Question for John: Any filter manufacturer/vendor you particularly like? I've been looking into getting some narrowpass UV-filters. Although I may not have any $$$ left after this Norway trip. "-)
Link to comment

Thanks Andrea, no dumb question.....all food for thought...........

The immediate problem is these 'Sparticle' filters are, they are only 10 to 15mm in diameter, not much to get an image into.

Second, I am just trying to get a UV mono response for each band (filter) & compare the results with the others, in one shot, to cut down the variability.

I am using Sunlight as it is what we are trying to photograph under, with the limited, UV supply that arrives on Earth :D

Cheers

Col

Link to comment

Question for John: Any filter manufacturer/vendor you particularly like? I've been looking into getting some narrowpass UV-filters. Although I may not have any $$$ left after this Norway trip. "-)

 

Andrea,

 

I hope y'all are enjoying your excursion!

 

As for filters, I buy most often from Edmund. Not because they are particularly any better but they have a very wide selection and I have an account with EO - :D

 

Omega and Newport, ThorLabs are also reputable suppliers. I recently have been looking at some interesting specs from Chroma Technology Corp and Asahi Spectra but have yet to sample any of their wares.

 

..and Semrock too has an interesting filter, see: http://www.semrock.com/FilterDetails.aspx?id=FF01-300/80-25

 

Edmund and ThorLabs and Newport also carry nice hardware, mounts and such!

 

- John

Link to comment

I have one each of L370nm & L390nm UV cut filters, I'll try these. I would like some other longpass filters, but they are harder to find & companies that sell them want two & three times the filters value for specialist freight & won't send by regular post :D

 

Yes, scientific grade filters are not cheap! Here is a very simple alternative which may not cost you anything.

 

Get a pair of yellow/amber tinted polycarbonate UV safety glasses. If you are playing around in the UV you likely already have a pair and if not you need them anyway. I prefer the fit of UVEX brand, Fisher or LAbSafety are also to be found around my lab.

 

Much like the first layer of your Foveon sensor the yellow or amber tinted lenses also block well into the blue, knocking down pretty much all the UV to get there.

 

So, screw on the L390 and also put the amber glasses in the light path and shoot a shot of the sun through your filter array. If you see anything it won't be UV. I predict if you image anything it will be dim solar disc.

 

- John

Link to comment
John, thanks. Those seem to be the same filter suppliers whose online pages I've been prowling. But it's good to have a confirmation. "-)
Link to comment

John, thanks. Those seem to be the same filter suppliers whose online pages I've been prowling. But it's good to have a confirmation. "-)

You are welcome, as I am sure you know, the thing to be sure of it the wavelength range and OD of the out of band blocking.

That 300/80nm Brightline that Semrock has seems interesting but it is thick and pricey.

Link to comment

Yes, scientific grade filters are not cheap! Here is a very simple alternative which may not cost you anything.

 

Get a pair of yellow/amber tinted polycarbonate UV safety glasses. If you are playing around in the UV you likely already have a pair and if not you need them anyway. I prefer the fit of UVEX brand, Fisher or LAbSafety are also to be found around my lab.

 

Much like the first layer of your Foveon sensor the yellow or amber tinted lenses also block well into the blue, knocking down pretty much all the UV to get there.

 

So, screw on the L390 and also put the amber glasses in the light path and shoot a shot of the sun through your filter array. If you see anything it won't be UV. I predict if you image anything it will be dim solar disc.

 

- John

Yep, thanks John

I did the L390 test on the 'UV Sparticle' today & it knocked out the three lowest filters, 340, 350 & 370nm, & left the 386.6nm mostly unaffected.

Col

Link to comment

I did the L390 test on the 'UV Sparticle' today & it knocked out the three lowest filters, 340, 350 & 370nm, & left the 386.6nm mostly unaffected.

 

I do not know the transmittance of the L390 but if it is using conventional longpass filter nomenclature it may be 50% transmission at 390nm. Depending on how sharply the L390 is cutting off that's likely close enough to overlap the 386.6 ±2nm filter.

 

Using a sharp cut long pass blocking filter is a standard technique to test for stray & scattered radiation in a monochromator. If you replicated your "sun in the filter" exposure geometry then that is very good indication you were not just imaging longer wavelengths passing under the blocking OD.

 

Question, was this done with or without the S8612? If so, it could have attenuated some the out of band radiation, and you might want to check it with the sun in the 350 or 340 without S8612 but with the L390 blocking filter.

Link to comment

I have made an adaptor to use the ‘UV Sparticle BandPass Filter Array’ with the Sigma DP2 full spectrum camera & photographed the results.

 

Col, I like your adapter very much.

 

I was in fact considering placing one of my small filter arrays inside a tube. Looks like you beat me to the Mark-1 design! :D

 

Looking at the linked photos I offer these observations:

 

Your L390 filter is not cutting off enough UV to clearly delineate long UV-A from blue. I am not certain if L390 is exactly the same as Hoya L39 but here is a link to Hoya showing filter transmittance of L39 ( http://www.hoyacandeo.co.jp/english/products/eo_color_01.html ). As you can see the L39 ideally transmits 50% 390±5nm with a transition from 5%-72% of <35nm (click into transmittance glossary). This means that the L39 is not acting as a barrier to your 370±20nm or 386±2nm filters because L39 could be transmitting as much as 5% @ 367nm & 50% @ 385nm and still be within Hoya's specs.

 

The S8612 looks promising, almost a bright as the first shot without any extra filter, but S8612 will pass all the visible blue so that still has not distinguished UV from blue.

 

The image made with additional R72 filter (http://www.hoyacandeo.co.jp/english/products/eo_color_03.html) looks like it was aimed at the sun. This image clearly shows detectable IR passing the 350±10 & 370±20 filter but not your 386±2 consistent with typical high out of band OD of such very narrower filters. I thought the shorter 340nm filter showed something looking into the sun before?

 

Personally I think you are imaging some UV, primarily because very narrow band pass filters like your 386±2 typically have lower MAX transmission and higher out of band OD and second because your 370 filter signal appeared reduced by L39. However you have not demonstrated how much of the signal is UV-A -vs- blue. The Hoya L42 or WL420 will block all UV and some blue visible as will the yellow/amber UVEX safety glasses.

 

Good job Col, very encouraging results so far.

- John

Link to comment

Thanks John

Yes you have the correct Hoya filters, L39 & R72. The 52mm filter rings fit inside the tube opening in front of the UV Sparticle.

All the UV Sparticle shots were at 90deg from the Sun & aimed at the sky / trees.

The 3 UV shots were at 1 second exposure.

The IR shot was at 15 seconds.

I don't have a L42, but I can try with a K2 (yellow) http://www.hoyafilter.com/hoya/products/coloredfilters/k2yellow/

Cheers

Col

Link to comment

John the L42 is 420nm at 50% & the K2 is about 470nm at 50%.

They should be both the same filters whether photography or technical, just the quality of the surface finish ?

Cheers

Col

Link to comment
Col, yes, the transmittance plot of K2 in the link you posted looks longer than 420nm now that I look at it again. So K2 is more similar to Y-48 then? That seems to long a wavelength, cutting much to much blue. The Hoya link I posted jumps from L-39 to L-42 skipping L-40. That makes me wonder if Hoya quit making L-40. L-40 is ≈ to Schott GG-400 which might be an ideal UV barrier filter, I have some around somewhere I need to check.
Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...