UV camera project build question

The question is, will the lens let UV pass?

You need to remove the UV/IR blocking filter in front of the sensor.

Then you need an UV pass filter in front of the lens like this:

https://www.baader-planetarium.com/en/filters/baader-u-filter-(venus-and-uv---350nm).html

You will need to be careful selecting your UV pass filter, all the old film era ones allow NIR through as well (which wasn't seen by film) Digital sensors are sensitive to NIR so the NIR will often swamp the detector. Only a few very expensive filters like the Baader U2, or carefully selected stacks of multiple filters will work acceptably.

The filter stacks may well prove to be too thick for placing on the sensor -the option I'm considering for UV needs at least 3mm thickness of BG38 glass together with a U330. That's a 5mm total filter thickness probably significantly more than the existing hot mirror (usually ~1.5mm thick).

petrochemist wrote:

You will need to be careful selecting your UV pass filter, all the old film era ones allow NIR through as well (which wasn't seen by film) Digital sensors are sensitive to NIR so the NIR will often swamp the detector. Only a few very expensive filters like the Baader U2, or carefully selected stacks of multiple filters will work acceptably.

The filter stacks may well prove to be too thick for placing on the sensor -the option I'm considering for UV needs at least 3mm thickness of BG38 glass together with a U330. That's a 5mm total filter thickness probably significantly more than the existing hot mirror (usually ~1.5mm thick).

If I'm understanding this correctly, the OP is talking about a mod to a Canon IXUS 105, which is a little old PowerShot supported by CHDK .

That means the lens is extremely hard to replace. I'd also be more than a little concerned that the sensor stack itself might block enough UV to make this problematic -- MaxMax claims the CFA and microlenses actually block UV . Incidentally, it is a CCD sensor in most PowerShots, not a CMOS one -- does that help or hurt?

Anyway, it really all depends on the wavelengths of interest, doesn't it? My understanding is that to about 365nm is doable with conventional optics, but shorter wavelengths require quartz optics and other extravagant things. It depends on what you want....

It's not really intended for UV nor for use with cameras, but you can get a theatrical gel filter color swatch book quite cheaply and each filter comes with its spectral profile. Take a look at Rosco's web site to see some profiles -- for example:

Combining two or three filters from a Rosco pack might get you close enough to what you want in UV sensitivity. For example, the above + #375 Cerulean Blue might not be a bad starting point, although you'll be down more than 2 stops around 400nm (and lower elsewhere). Of course, you'll probably need longish exposures and a bright UV light -- and bright UV lights are seriously dangerous.

In any case, good luck.

ProfHankD wrote:

petrochemist wrote:

You will need to be careful selecting your UV pass filter, all the old film era ones allow NIR through as well (which wasn't seen by film) Digital sensors are sensitive to NIR so the NIR will often swamp the detector. Only a few very expensive filters like the Baader U2, or carefully selected stacks of multiple filters will work acceptably.

The filter stacks may well prove to be too thick for placing on the sensor -the option I'm considering for UV needs at least 3mm thickness of BG38 glass together with a U330. That's a 5mm total filter thickness probably significantly more than the existing hot mirror (usually ~1.5mm thick).

If I'm understanding this correctly, the OP is talking about a mod to a Canon IXUS 105, which is a little old PowerShot supported by CHDK .

That means the lens is extremely hard to replace. I'd also be more than a little concerned that the sensor stack itself might block enough UV to make this problematic -- MaxMax claims the CFA and microlenses actually block UV . Incidentally, it is a CCD sensor in most PowerShots, not a CMOS one -- does that help or hurt?

Anyway, it really all depends on the wavelengths of interest, doesn't it? My understanding is that to about 365nm is doable with conventional optics, but shorter wavelengths require quartz optics and other extravagant things. It depends on what you want....

It's not really intended for UV nor for use with cameras, but you can get a theatrical gel filter color swatch book quite cheaply and each filter comes with its spectral profile. Take a look at Rosco's web site to see some profiles -- for example:

Combining two or three filters from a Rosco pack might get you close enough to what you want in UV sensitivity. For example, the above + #375 Cerulean Blue might not be a bad starting point, although you'll be down more than 2 stops around 400nm (and lower elsewhere). Of course, you'll probably need longish exposures and a bright UV light -- and bright UV lights are seriously dangerous.

In any case, good luck.

IIRC there's no inherent preference for CCD or CMOS, but the thickness of silicone in the sensor can have a significant effect on it's UV sensitivity. On average shorter wavelengths (such as UV) penetrate further than visual light before being detected - this principle is the way Foveon sensors get their colour information.

Typical camera sensors being designed to capture visible light are not particularly sensitive to UV, but are sensitive to NIR - the reason a hot mirror is installed. The rise at the right of the spectra You've supplied shows considerable NIR is transmitted by this gel - FWIW I've not found any gels (other than diffusion ones) that don't transmit NIR well though I've only run about 50 of them on the spectrometer at work. Nearly all transmit more throughout the NIR (700-1100nm) than they do at any visible wavelength. They're definitely good for IR photography!

Schott U360 transmits more (4x?) UV than NIR (still enough IR to cause issues). Small circular examples are available for about £30 - 25mm ones should be big enough for a compact camera.

The wavelength of interest certainly affects things, If you're after anything below 190nm just the air itself will block UV pretty rapidly. Smaller lenses will typically transmit more UV so IF it's not coated the lens in a P&S has a good chance of being usable down to nearly 300nm. Most UV photography is done in this range by the time you've got down to 250nm you're into germicidal UV considerably more dangerous. General UV light sources for Labs. are usually around 365nm, with 254nm sources used for sterilization.

petrochemist wrote:

ProfHankD wrote:

It's not really intended for UV nor for use with cameras, but you can get a theatrical gel filter color swatch book quite cheaply and each filter comes with its spectral profile. Take a look at Rosco's web site to see some profiles -- for example:

Combining two or three filters from a Rosco pack might get you close enough to what you want in UV sensitivity. For example, the above + #375 Cerulean Blue might not be a bad starting point, although you'll be down more than 2 stops around 400nm (and lower elsewhere). Of course, you'll probably need longish exposures and a bright UV light -- and bright UV lights are seriously dangerous.

...

Typical camera sensors being designed to capture visible light are not particularly sensitive to UV, but are sensitive to NIR - the reason a hot mirror is installed. The rise at the right of the spectra You've supplied shows considerable NIR is transmitted by this gel - FWIW I've not found any gels (other than diffusion ones) that don't transmit NIR well though I've only run about 50 of them on the spectrometer at work. Nearly all transmit more throughout the NIR (700-1100nm) than they do at any visible wavelength. They're definitely good for IR photography!

Like I said, you stack 2-3 gels to get the desired profile. According to Rosco, there are definitely gels that kill at least to 740nm. Look at the spectral profiles -- for example, of the #375 Cerulean Blue that I mentioned as a potentially good pairing:

Notice that at least out to 740nm it's flat at 0% transmission. That plus the usual NIR blocking should take care of NIR pretty well... assuming this doesn't take care of it by itself, which it might.  Have you tried any of the gels that claim to kill this range of wavelengths?

Stack that with the other one and you'll get a pretty nice peak around 400nm. Actually, there are probably better pairs in the $2 sampler book, but I just picked two that were easy to find and reasonable rather than paging through all to find the best pair.

Not saying it's perfect, but we're talking about the OP modifying a camera worth less than $25 -- this is obviously a cheap and dirty project.

ProfHankD wrote:

petrochemist wrote:

ProfHankD wrote:

It's not really intended for UV nor for use with cameras, but you can get a theatrical gel filter color swatch book quite cheaply and each filter comes with its spectral profile. Take a look at Rosco's web site to see some profiles -- for example:

Combining two or three filters from a Rosco pack might get you close enough to what you want in UV sensitivity. For example, the above + #375 Cerulean Blue might not be a bad starting point, although you'll be down more than 2 stops around 400nm (and lower elsewhere). Of course, you'll probably need longish exposures and a bright UV light -- and bright UV lights are seriously dangerous.

...

Typical camera sensors being designed to capture visible light are not particularly sensitive to UV, but are sensitive to NIR - the reason a hot mirror is installed. The rise at the right of the spectra You've supplied shows considerable NIR is transmitted by this gel - FWIW I've not found any gels (other than diffusion ones) that don't transmit NIR well though I've only run about 50 of them on the spectrometer at work. Nearly all transmit more throughout the NIR (700-1100nm) than they do at any visible wavelength. They're definitely good for IR photography!

Like I said, you stack 2-3 gels to get the desired profile. According to Rosco, there are definitely gels that kill at least to 740nm. Look at the spectral profiles -- for example, of the #375 Cerulean Blue that I mentioned as a potentially good pairing:

Notice that at least out to 740nm it's flat at 0% transmission. That plus the usual NIR blocking should take care of NIR pretty well... assuming this doesn't take care of it by itself, which it might. Have you tried any of the gels that claim to kill this range of wavelengths?

Stack that with the other one and you'll get a pretty nice peak around 400nm. Actually, there are probably better pairs in the $2 sampler book, but I just picked two that were easy to find and reasonable rather than paging through all to find the best pair.

Not saying it's perfect, but we're talking about the OP modifying a camera worth less than $25 -- this is obviously a cheap and dirty project.

To get UV response you have to remove the hot mirror, which is what stops the sensor picking up 700-1100nm 740 is no good. Of the 50 I tested a handful were low transmission in the 700region but these all transmitted over 30% at 750nm (I've only got tabulated data readily to hand).

My other gel booklet which happens to be on my desk has #375 Cerulean Blue. It's visible light transmission is considerable - I'll check the full spectra in a bit.

To record UV you're going to need all wavelengths from 400-1100nm blocked to under 1% I don't see it helping much!

Transmission data according to a Perkin Elmer Lambda 35 research spectrometer:

900-1100nm 87 to 93% transmission (interference between surface reflections has made this area less precise)

800nm   82% transmitted

750nm   41.7% transmitted

700nm   0.23% transmitted

600nm   4.1% transmitted

500nm   66% transmitted

400nm   53% transmitted

350nm   18% transmitted

300nm to 200nm less than 0.02% transmitted

From the areas under the spectra around 10% of the normalised transmitted light would be UV, visible light mainly blues & green s account for about 30% & NIR about 60%

Taking into account camera sensor response will boost the visible portion at the cost of both UV & IR.

The UV transmission is better than most filters, but both IR & visible are HUGE.

petrochemist wrote:

Transmission data according to a Perkin Elmer Lambda 35 research spectrometer:

900-1100nm 87 to 93% transmission (interference between surface reflections has made this area less precise)

800nm 82% transmitted

750nm 41.7% transmitted

700nm 0.23% transmitted

600nm 4.1% transmitted

500nm 66% transmitted

400nm 53% transmitted

350nm 18% transmitted

300nm to 200nm less than 0.02% transmitted

Nothing surprising -- that matches the published spec well -- except wow, it sure climbs fast after 740nm! They say 0% at 740nm, and judging by 700nm, I believe them.

I'm quite surprised by this, although it creeping up wouldn't have been a big surprise. Then again, you're certainly right that organic dyes rarely block NIR -- that's what got Sony's NightShot into trouble. 

From the areas under the spectra around 10% of the normalised transmitted light would be UV, visible light mainly blues & green s account for about 30% & NIR about 60%

Taking into account camera sensor response will boost the visible portion at the cost of both UV & IR.

The UV transmission is better than most filters, but both IR & visible are HUGE.

Fair enough.

It might still be possible for the OP to computationally separate UV out using RGB CFA sensitivity differences, but it's pretty clear noise level will be HUGE.

ProfHankD wrote:

petrochemist wrote:

Transmission data according to a Perkin Elmer Lambda 35 research spectrometer:

900-1100nm 87 to 93% transmission (interference between surface reflections has made this area less precise)

800nm 82% transmitted

750nm 41.7% transmitted

700nm 0.23% transmitted

600nm 4.1% transmitted

500nm 66% transmitted

400nm 53% transmitted

350nm 18% transmitted

300nm to 200nm less than 0.02% transmitted

Nothing surprising -- that matches the published spec well -- except wow, it sure climbs fast after 740nm! They say 0% at 740nm, and judging by 700nm, I believe them.

Actually it climbs rapidly after 705nm. IIRC it was up to more than 30% by 740nm

I'm quite surprised by this, although it creeping up wouldn't have been a big surprise. Then again, you're certainly right that organic dyes rarely block NIR -- that's what got Sony's NightShot into trouble.

From the areas under the spectra around 10% of the normalised transmitted light would be UV, visible light mainly blues & green s account for about 30% & NIR about 60%

Taking into account camera sensor response will boost the visible portion at the cost of both UV & IR.

The UV transmission is better than most filters, but both IR & visible are HUGE.

Fair enough.

It might still be possible for the OP to computationally separate UV out using RGB CFA sensitivity differences, but it's pretty clear noise level will be HUGE.

I don't think theres enough data available for that. The ratios of the RGB channels are already used for getting colour information. This is how the eye works & cameras too. All the visible spectrum gives some signal in each of the channels.

Here's one of the on-line spectra for the Bayer filters https://static5.olympus-lifescience.com/data/olympusmicro/primer/digitalimaging/images/cmos/cmoschipsfigure4.jpg?rev=88D6

Blinkie wrote:

Wow thank you everyone for your advice. There is a lot I need to understand. The reason I thought about this project was seeing a converted camera for sale on ebay. It wasn't a lot of money and the seller explains that converting the cameras was a hobby of theirs. I thought it might be fun to try myself. The seller doesn't explain the sort of filter he incorporates into the cameras. Apart from giving the nm figure. Have a look for yourselves at an old listing.

https://www.ebay.co.uk/itm/Canon-PowerShot-A480-Infrared-Converted-IR-670nm-Digital-Camera-Custom-Firmware-/352465874416?nordt=true&orig_cvip=true&rt=nc&_trksid=p2047675.m43663.l10137

The seller doesn't have anything currently for sale but but every now and then will list a similar item.

B

Compact cameras are often very easy to convert, I've even done one myself (though I need to revisit it to improve the focus range).

The nm figure quoted for filters is roughly the point where 50% is transmitted. Unless stated otherwise it will be a long pass filter, which effectively only transmits above the listed cut off. In most cases long pass filters will have a rapid transition, only small traces will be seen below the quoted nm but there are a few filters that have much more gradual changes.

670nm (as your example) is in the deep red so will give NIR with a trace of the visible red. You won't see any UV with that conversion. (Visible is roughly 400-700nm, UV is below 400 & NIR above 700)

Some sellers use 'full spectrum' for conversions where glass or a UV filter has been used instead of the hot mirror. IMO these should be called 'two spectrum' as UV is reduced well below the sensors native sensitivity.

At least compact cameras have small lenses, so the glass in them will not absorb too much UV. If the lens coatings don't block UV they should be a fairly good option.

Blinkie wrote:

Wow thank you everyone for your advice. There is a lot I need to understand. The reason I thought about this project was seeing a converted camera for sale on ebay. It wasn't a lot of money and the seller explains that converting the cameras was a hobby of theirs. I thought it might be fun to try myself. The seller doesn't explain the sort of filter he incorporates into the cameras. Apart from giving the nm figure. Have a look for yourselves at an old listing.

https://www.ebay.co.uk/itm/Canon-PowerShot-A480-Infrared-Converted-IR-670nm-Digital-Camera-Custom-Firmware-/352465874416?nordt=true&orig_cvip=true&rt=nc&_trksid=p2047675.m43663.l10137

The seller doesn't have anything currently for sale but but every now and then will list a similar item.

B

Well, I don't know what a "670nm filter" means, but most of us think of NIR as starting around 750nm+ wavelength. The fun stuff happens mostly between 850nm and 1200nm, with most camera sensors falling off by 1100-1200nm. Then again, the price quoted was really cheap (ok, the base A480 is worth about $10) and it's really no fun to do that to a little PowerShot (I've taken 'em apart myself before).

Perhaps I am mis-reading your post, but you don't want an internal filter. You do want a IR-VIS-UV pass broadband window, e.g. quartz with broadband AR coatings, that make up the optical path length of having the IR cut filter. Without this, lenses will not focus to infinity and / or you will have ghosting.   Then you use an external UV short pass filter, or whatever filters you want.  .   The camera will also work for IR and general multi-spectral work.  t.

I and some of our researchers here, have had good experiences with this eBay seller, including his converted cameras and filters.    I have one of their larger UV cut filters   It is great and rejects IR and VIS > 250:1.

http://www.ebay.com/usr/image-laboratory

As for a fun DIY project, if you want something to play with, web cams are a great way to do this, as pretty much they are a bare sensor, so easily modified.  And given how bad some people's UV photography turns out, you may actually be better off.

Also, be in the lookout for old spectrophotometers ( the IR-VIS-UV type or anything with UV.   Some of them will have a lens assemblies, and this will necessarily pass UV.   Small UV lenses ( suitable for the webcam ) can be had from optics places.   Some times colleges / universities chem and bio department have heaps of this stuff hidden away and a polite request will liberate one of the broken ones.

BobORama wrote:

Perhaps I am mis-reading your post, but you don't want an internal filter. You do want a IR-VIS-UV pass broadband window, e.g. quartz with broadband AR coatings, that make up the optical path length of having the IR cut filter. Without this, lenses will not focus to infinity and / or you will have ghosting. Then you use an external UV short pass filter, or whatever filters you want. . The camera will also work for IR and general multi-spectral work. t.

I and some of our researchers here, have had good experiences with this eBay seller, including his converted cameras and filters. I have one of their larger UV cut filters It is great and rejects IR and VIS > 250:1.

http://www.ebay.com/usr/image-laboratory

Wow.  Great reference!  I never stumbled on this before....

As for a fun DIY project, if you want something to play with, web cams are a great way to do this, as pretty much they are a bare sensor, so easily modified. And given how bad some people's UV photography turns out, you may actually be better off.

Yeah, UV just isn't as much fun in a variety of ways... and then there's also the fact that bright UV light is bad for living things (like us humans).

Also, be in the lookout for old spectrophotometers ( the IR-VIS-UV type or anything with UV. Some of them will have a lens assemblies, and this will necessarily pass UV. Small UV lenses ( suitable for the webcam ) can be had from optics places. Some times colleges / universities chem and bio department have heaps of this stuff hidden away and a polite request will liberate one of the broken ones.

Another interesting idea....

ProfHankD wrote:

Blinkie wrote:

Wow thank you everyone for your advice. There is a lot I need to understand. The reason I thought about this project was seeing a converted camera for sale on ebay. It wasn't a lot of money and the seller explains that converting the cameras was a hobby of theirs. I thought it might be fun to try myself. The seller doesn't explain the sort of filter he incorporates into the cameras. Apart from giving the nm figure. Have a look for yourselves at an old listing.

https://www.ebay.co.uk/itm/Canon-PowerShot-A480-Infrared-Converted-IR-670nm-Digital-Camera-Custom-Firmware-/352465874416?nordt=true&orig_cvip=true&rt=nc&_trksid=p2047675.m43663.l10137

The seller doesn't have anything currently for sale but but every now and then will list a similar item.

B

Well, I don't know what a "670nm filter" means, but most of us think of NIR as starting around 750nm+ wavelength. The fun stuff happens mostly between 850nm and 1200nm, with most camera sensors falling off by 1100-1200nm. Then again, the price quoted was really cheap (ok, the base A480 is worth about $10) and it's really no fun to do that to a little PowerShot (I've taken 'em apart myself before).

Most websites on IR photography will include quite a selection of 'High color' 'super color' etc conversions. The capture NIR along with some visible, the Bayer filter shows much more variation for the visible portion, so interesting colour effects can be introduced using the small portion of the visible that's captured. See https://www.lifepixel.com/ for example.

My current favorite filter is a 590nm. It still looks pretty much like a 720nm image if totally desaturated, but also opens up false color options. A selection of my IR & IR+ shots can be found athttps://www.flickr.com/gp/petrochemist/g7BR91 most of the latest are with a 590nm, the 670 should be similar but give a little less color.

BobORama wrote:

Perhaps I am mis-reading your post, but you don't want an internal filter. You do want a IR-VIS-UV pass broadband window, e.g. quartz with broadband AR coatings, that make up the optical path length of having the IR cut filter. Without this, lenses will not focus to infinity and / or you will have ghosting. Then you use an external UV short pass filter, or whatever filters you want. . The camera will also work for IR and general multi-spectral work. t.

I and some of our researchers here, have had good experiences with this eBay seller, including his converted cameras and filters. I have one of their larger UV cut filters It is great and rejects IR and VIS > 250:1.

http://www.ebay.com/usr/image-laboratory

Interesting my searches hadn't turned them up... Now I've ordered a couple more filters - Properly mounted BG40 & UG11 not cheap but hopefully the AR coatings will make the postage/import duties worthwhile, and the mountings will be more convenient than my current bare 25mm discs!

As for a fun DIY project, if you want something to play with, web cams are a great way to do this, as pretty much they are a bare sensor, so easily modified. And given how bad some people's UV photography turns out, you may actually be better off.

Also, be in the lookout for old spectrophotometers ( the IR-VIS-UV type or anything with UV. Some of them will have a lens assemblies, and this will necessarily pass UV. Small UV lenses ( suitable for the webcam ) can be had from optics places. Some times colleges / universities chem and bio department have heaps of this stuff hidden away and a polite request will liberate one of the broken ones.

All the spectrometers here I've used (10 different models IIRC, including UV/Vis, IR, atomic & Fluorescence) use mirrors to direct/focus the light. The focal length of lenses varies with wavelength so would not be very good.

There might still be an interesting source of components to play with potentially having half silvered mirrors, lasers, diffraction gratings, light sources... Unfortunately by the time works finished with them the remaining life hasn't made it worthwhile trying to sort out portable power supplies, which only leaves mirrors & gratings. I might try stripping out current UV when it gets replaced, but I'd rather keep it running as long as possible - the proposed replacement will not take lenses or filters in the sample compartment!

Maybe some practical information about how I did my conversions may give the OP more ideas. I used Nikon N1 bodys which are quite cheap. The advantage is that you can change lenses.

For one body I replaced the filter stack in front of the sensor with a Schott B270 1,35mm thickness. You can cut the glas to the needed size with a glas cutter and cutting oil. The other N1 body I shimmed the sensor assy and brought it forward be 0,5mm.

I also converted a Sony A7 by shimming the sensor.

The results in IR are the same for me. You can see some results on my Flickr account.

I am also interested in UV photography. But so far I have not got some useful results.

I use a UG-11 and a IR blocking filter in front of a Staeble Lineogon 35 mm f/3.5 Lens which should be good for UV Photography according to these sites:

http://photographyoftheinvisibleworld.blogspot.com/2011/01/uv-lenses-tested-good-for-uv.html

http://www.savazzi.net/photography/35noflexar.html

https://kolarivision.com/uv-photography-lens-compatibility/