Cameras for UV: Pros & Cons of Various Choices

[Andrea B.]

But back to the original post.....

 

Does anyone have any additions, corrections or improvements to the original post?

 

I want to link to the first post in the first Sticky - or perhaps incorporate the first post into the Sticky so that the Sticky will be more useful.

[Timber]

Canon and UV:

2:40. Looks EOS 7D to me B)

[Alaun]

I'd like to keep up the flag for the newer Lumixes :D

 

(as there is a GH4, you could consider the GH3 as old B) )

 

 

 

 

 

ISO 6400 :

 

 

 

(may be I should show more pictures)

 

Werner

 

PS: UV-Nikkor 105 with Baader U,

first one at ISO 200 and 1.3s (f=?11 or even 16)

second ISO 6400 and 1/50s f5.6

The ball in the back is some kind of plastic (I assume something like HDPE, what seems by the way to give a niceWB (tested with other bottles made from HDPE) As the ball shows, the WB might also depend on the "grayscale" the reflecting reference material is seeing?)

[JCDowdy]

But back to the original post.....

Does anyone have any additions, corrections or improvements to the original post?

I might suggest combining Panny and Oly µ4/3 as they use the same mount standard and FFD. I do not recall ever seeing of any of the older 4/3 cameras modified for full spectrum. The term "discontinued" is accurate but might confuse 4/3 with µ4/3. The Oly µ4/3 is used by Boon TANG confirming that the Oly IBIS can be converted, something I was uncertain about.

[Andrea B.]

Timber, thanks for the Youtube link.

 

Werner, thanks for the nice GH3 photos. Great detail in the first one!!

 

John, I added µ4/3 by Panasonic Lumix title so there would be less confusion.

 

I modified the last section to say the following because I didn't want to be promoting one "brand" over another. I merely wanted to give 3 examples of recent cameras with excellent sensors, that's all.

 

The Pentax K5, Nikon D610 or Sony A7 are just three examples of ... cameras with excellent image capability. There are many more.

[enricosavazzi]

I have recently written some thoughts on my web site that may be relevant to conversion choices, especially with mirrorless cameras. The page is not yet linked to the rest of the site, but can be reached at http://savazzi.net/photography/filterpack.html

 

All UV pictures I posted so far on ultravioletphotography.com are taken with a modified Lumix G3.

[Andrea B.]

Thanks, Enrico for the heads-up. That is great information. I've added a link to it in Sticky #1.

[nfoto]

A very nice and thorough write-up, Enrico. I got theoretical confirmation with regard to the practical solution of filter positioning I always have used with my mirrorless cameras as well :D.

[ahrneely]

Okay, for the original post: "Cameras for UV: Pros and Cons of Various Choices" I can only add the following based on my experience with my Canon setup and my personal experience.

 

1. I had a Canon 20D converted by LifePixel. I bought my 20D used on eBay for approximately $50 (US) and the conversion ran $250 at the time (though now it appears the conversion is $325) for a Full-Spectrum (or Broad-Spectrum) conversion. I bought Canon so I could continue to use my set of Canon lenses with the camera for IR shooting, just in cast the full spectrum or UV stuff turned out to not be very appealing.

 

2. For UV, I now own a 50mm El-Nikkor f/2.8 and an 80mm El-Nikkor f/5.6 plus conversion equipment for both of them. All told, I'm probably somewhere around $120 for both lenses, M42 to EOS mount conversion, focus helicoids, and setup-up filter rings for each lens (the step-up ring for the 85mm was difficult to find, but thanks to someone in this forum for recommending RafCamera, I was able to track something down quite easily and with little expense). That total doesn't include additional accessories like the Xume Quick Release Filter Adapters (which I absolutely love).

 

3. I've purchased a couple of UV filters, but none of them have the quality of the Baader-U, so I won't even mention them here.

 

That kind of gives you an idea of the cost (pros vs. cons) of my Canon setup.

 

I don't have a lot of interest in the scientific aspects of the camera builds themselves (nor do I have access to the equipment to test these things), as I'm much more interested in the final output. However, I will say that my output isn't nearly what I would hope it could be.

 

However, I can quite easily suggest that the results I see are not at all what I would hope, and it is possible that the Nikon folks have better image quality results.

 

I do not have a UV flash, though I do have a UV "Blacklight" flashlight or torch that I have used, but it is an LED and doesn't offer the same "feel" as a strobe or the sun. With that said, most of my images are shot in bright daylight. I typically find myself shooting around ISO 3200 wide open (on the 85mm that's f/5.6) around 1/25 or 1/30 second. That seems exceptionally slow to me.

 

This image:

 

 

This is an image of me with the 85mm f/5.6 using a standard studio strobe monoblock. The light was bare bulb and 20-inches from my face, and set to full power (which I think is actually only 160 w/s - it's not terribly powerful, but for visible light it's enough for most of my work), however it is not UV flash from what I can tell. The exposure was ISO 400, f/8, 1/200" shutter speed (though we know that doesn't really matter for flash photography). It isn't horrible. Definitely under exposed. The only work I have done in post is use Capture One to white balance for my skin tones. I am clearly out of focus (which is my bad for moving and using having no depth of field). It also lacks the contrast I would normally expect from something I shoot, but what can you do? I believe that has to do with the lens more than the camera.

 

This image:

 

 

Is a 100% crop of the previous image, and as you can see there is still quite a bit of noise. More than I'd care for, even at ISO 400. However, that's probably what I should expect for such an old camera body. What I find interesting is how much the noise lacks in uniformity. I guess that makes sense, but there does seem to be quite a few "blotches" and blocky areas that I don't normally see in the visible spectrum (when shooting with a filter that blocks the UV and IR).

 

Finally, here is another image of an outdoor scene that I was able to handhold:

 

 

and

 

. The exposure here was 1/25-second shutter speed, ISO 3200, f/5.6 with the 85mm El-Nikkor. Still not as much contrast as I would have liked, but the exposure difference is huge to me.

 

As I'm sure most of you have seen from other posts I have made here, I am capable of shooting UV with my standard Canon EOS lenses as well, and I typically only see an additional stop added to the exposure when compared to the El-Nikkor lenses. That may be significant for most applications, but I do find it pretty handy at times, so that could be considered a "pro" (maybe).

 

That's probably more information than any of you wanted. Sorry about that. I do like to make sure I can give as much information as possible. All said, I would suggest that, based on what I've seen on this forum, the Nikon users are getting better image quality than I am seeing with my 20D. Perhaps when I have the funds I'll have a 1Ds converted, just to see if the improved sensor makes that much of a difference, or if Canon's sensors simply aren't as up to par with the Nikon sensors for UV photography.

[Andrea B.]

This is useful info, Ahr. So thanks !!

 

I typically find myself shooting around ISO 3200 wide open (on the 85mm that's f/5.6) around 1/25 or 1/30 second. That seems exceptionally slow to me.

 

The Canon 20D is 10 years old, so I'm thinking that this is part of the difficulty you experience in getting a better UV photo. By the time you reach an ISO 3200 signal boost for an older sensor, you're just going to get a lot of noise (both kinds) and loss of colour/contrast. Then you have to add in the effects on image quality of either in-camera or in-editor noise reduction.

 

But UV is a slow, dark art even with newer sensors. Tripods are mandatory. Especially so you can get down off that very high ISO setting!!

 

I just shot a spider web in waning afternoon sunlight at f/11 and ISO800. The exposure time was 1.6 seconds for the overview shot and 1/2" for a closer shot in slightly better light. That was with a Noflexar 35/3.5 on the Nikon D600-broadband. So even with newer sensors, the speeds can be still be slow in actual field work even as we are able to stop down more and use a lower ISO.

 

I would continue to suggest you simply practice and learn all you can on the 20D so that you can determine where you want to go with UV - or if you want to continue further. Then you can upgrade to an improved sensor (brand irrelevant).

 

BTW, were you to look closely at some of my old Nikon D200 shots from 6 or 7 years ago, you would find the image quality not so great in quite a few of them. I had a lot of trouble with blocked shadows and noise back them. Some of those old UV shots make me grit my teeth when I look at them now. :P :D

[ahrneely]

Thanks for the advice Andrea.

 

You have actually spawned another burning question in me that I am going to have to test and experiment with; that is: will I receive less noise with my older sensor by reducing my ISO and increasing my exposure time, or will the noise level from a long exposure be comparable to what I'm seeing now. :D

[nfoto]

All depends on contrast of the scene. In most cases, using a low ISO and longer exposure time will be the better approach. However, if inherent scene contrast is low, all digital captures will be more noisy after they are processed.

[Andrea B.]

My guess is that you would have somewhat less noise with the low ISO, longer exposure over the high ISO, shorter exposure. But it may depend on how the camera handles noise reduction.

 

It's always good to shoot a test series at each ISO stop (with a fixed aperture). Then you can pixel peep and determine where you want to "draw the line" on your ISO setting. Make two such series - one with and one without in-camera noise reduction. Sometimes noise reduction can cause detail loss - which fact I'm sure you know! - but the point is that you might find a better noise reducing tool in editing software than you have in the camera.

 

UV seems to always have more noise than the corresponding visible shots.

[Timber]

It also matters what material they used for the conversion I believe... sometimes they only put optically clear glass, which is not the best solution for UV if I am not mistaken. The conversion service I used (eBay) offered the Spectrosil 2000 silica, which has a very good transmission rate, so that's the best for Full Spectrum conversion if you are planning to do UV as well. I am not 100% sure if the EOS 20D was using CMOS or CCD? That could also make a huge difference if I am not mistaken.

[nfoto]

I think most companies use Schott glass with a cut-off around 280 nm. Should suffice with the current state of digital cameras.

 

As to CCD vs CMOS I don't think the spectral response of each type is much different, but my CCD-based DSLRs consistently had an edge over CMOS in terms of sensitivity. At least 1 stop advantage for the D40X I recall.

[Adrian]

Does anyone have any comments re. a full spectrum conversion, rather than a UV only conversion? The full spectrum version is much cheaper! I am particularly interested in using it for UV images of flowers and carnivorous plants .I am thinking of getting a Nikon D300 converted by Life Pixel, or ACS in the UK

Adrian Davies

[nfoto]

The conversion might be cheaper, but the filter for proper UV photography still is expensive ....

 

In my experience, "full spectrum" cameras entail a lot of filter swapping and you are best served with lenses that won't have significant focus shifts, or you have to rely on LiveView feature. The latter is neither easy (nor always practical) when you move away from an outdoor situation to overcast with low UV, or go indoors in a studio setting. So, a UV 'torch' is required as well, and did I mention UV-protective goggles?

 

The point here is UV shooting easily runs up the associated costs even when the camera itself might not be so expensive.

[Adrian]

Thank you Bjorn,

I already have the Baader U fiter, though can't yet justify a UV lens! I am currently using an El Nikkor lens, which works well, though, as you say, with a significant focus shift!

Adrian Davies

[Bill De Jager]

Adrian, I have both IR-converted and full-spectrum converted cameras. The latter are used for UV. Other than flexibility in use for both UV and IR, the advantage among Lifepixel conversions is that the broad-spectrum conversions go down to a much lower range in the UV spectrum that the UV-only conversion. Go to the FAQs on the Lifepixel site and click on the "What kind of filters do you use for conversions?" line to see both graphs. On the other hand, unless you have a dedicated UV lens (UV-Nikkor or the like) it may not make much of a difference which filter you get, due to limitations in how short a wavelength the lens glass will transmit. Perhaps others here can step in and clarify this issue.

 

Another consideration is the practicality of filter use. If you're using a SLR or enlarging lenses on a mirrorless camera then it's possible to mount the filter between the lens and the camera due to the amount of backspacing the lens needs. However, if you're using SLR lenses on an SLR (especially the case for Nikon with its large backspacing) then this is likely not an option. Meanwhile, it may not always be easy or practical to put a filter on the front of the lens; for instance, the lens may require filters much larger than the largest Baader Venus filter. That might still work in some cases with step-down rings when the lens is placed on a crop-sensor camera, but there is a possibility of vignetting. Some lenses won't take filters, may not have filter threads (many Leica screwmount lenses, for instance), or may have internal filters that would require custom fabrication of UV-only replacements.

 

This is why I just sent a Canon T3i (600D) off to Lifepixel today for a UV-only conversion. While I've been too busy (and somewhat unwell) for quite some time to attend to this long-promised task, or any significant UV work, I still want to test a number of my lenses for UV performance. Having camera with an internal filter will make this not only easier but also much more practical. While I've been using Nikon for over 30 years, one shortcoming of Nikon cameras is the relatively large backspacing which prevents use of most other makes of SLR lenses without losing infinity focus. That's why I selected the Canon EOS mount this time since it'll allow use of a number of other mounts such as M42 screwmount.

 

Rangefinder lenses like the Leica screwmount lenses will still need to be used on a mirrorless camera. Unfortunately, Lifepixel does not offer the UV-only conversion on any of these cameras yet. MaxMax offers extremely limited UV conversions, at least the last time I checked a couple of years ago. I'm not familiar with ACS.

 

Best of success!

Bill

[OlDoinyo]

At the risk of being thought behind the times, I would like to draw attention to one more option--the humble film camera, loaded with b&w film, and equipped with an appropriate filter. It is true that one cannot hope to do false-color work this way, and processing and scanning film is an extra task which must be dealt with; on the other hand, the sensitivity of film is definitely not to be sneezed at--I have seen working ISO numbers of 10 through a Baader filter for Shanghai GP3 processed at label speed (not pushed) and 50 with Ilford Delta 3200. This last may not seem impressive compared to these films' visible performance, but it edges into the territory where shutter speeds such as 1/125 second can be employed in bright light, meaning hand-held shooting is a reality (I have taken pictures while riding up a platter-pull ski lift, and tried a couple shots of airplanes taking off.) To approach this sensitivity even with very good digital gear requires insanely high sensor gain settings which can result in poor image quality. Infrared contamination is also a virtual non-issue with film, meaning that some cheaper Wood's-glass filters such as the 403 become usable. This is actually how I got into UV, and you can do it without breaking the bank; the equipment is relatively inexpensive, and no specialized films are needed (in stark contrast to film IR.)

 

My favorite type of camera for film UV is the twin-lens reflex, which provides an old-fashioned equivalent of "live view" and can be purchased used for under $200 these days. Rangefinder cameras are also a good compact alternative, though framing is a bit less precise. I do not recommend SLR cameras for this purpose.

[enricosavazzi]

Another potential advantage of film (or at least some film, especially if the gelatin layer is thin) is that it can record shorter wavelengths than ordinary solid-state sensors, which are rarely sensitive enough to allow the recording of 300 nm and shorter wavelengths. Only special sensors without microlenses and Bayer filters can compete with film in this respect. This of course requires the use of a lens that transmits these wavelengths, like the UV Nikkor 105 mm. Even the Coastalopt 60 mm Apo is not really usable below about 290 nm.

[nfoto]

B/w films can have coatings that prevent ultradeep UV response, even if the lens allows this by itself. All colour films are worse in this respect, of course. It is probably true that you can record useful information as low as 300 nm, though. Whether or not that region of the UV registers significantly different from the 'normal' range covered by digital cameras remains to be demonstrated. The few times I've tried results were inconclusive.

[Bill De Jager]

I've only tried color film for UV once. Fujicolor Pro 160 had been touted in one photo.net post as having a useful UV response. I tried it with a Nikkor 105mm bellows lens and got nothing despite massive bracketing. Later I read that it had meanwhile been reformulated (new coating) to better block UV. That was my first UV effort, long ago.

[OlDoinyo]

Some color films are overcoated, giving extremely weak UV response (Fujichrome RTP is an example of this, in my experience: exposures of 30 seconds or more are required in broad daylight, and the final result is clearly hybrid, containing leaked visible light as well as UV.) Other color films have sometimes not been including some of the Kodak products. Aerochrome, unsurprisingly (who needs overcoating when you are shooting through a yellow filter, anyway?) and some of the Portra negative emulsions turn out to be examples; working ISO of these films ends up in the double digits in daylight. However, I have never gotten any color film to give me much beyond tinted monochrome results, as almost all the UV response in a laminar sensing medium is confined to the top layer, leaving the lower layers to record mainly leakage. The one reseau-type color film that existed in the modern era, the short-lived Polachrome, could in theory have been different, but in reality it had about the worst UV sensitivity I have ever seen in a film, I would guess 14+ stops below visible. Thus, the only point of using color film is if you want to mix UV and regular color in one roll (and the UV quality is not as good as that of B&W film.) Otherwise, B&W film is better; and the ones I have used (Shanghai GP3, Ilford Delta 3200, and Kodak Tmax 100 and 400) do not act as if they are overcoated. Recall that a lot of B&W photography is traditionally done through colored filters that themselves are efficient UV blockers, so overcoating would seem to serve little purpose.

[colinbm]

I have just run into this innovative site, someone must find something interesting here ??

 

http://www.eigenimaging.com/

 

Col