Nikon bodies good for UV/insect-vision macro z-stacking?

[Daniel Geiger]

Hi Andrea, thanks for the pointer re file extension.

 

Ulf: Tried MPE65 and there was a bit of UV transmission but back when I tried I had a pretty weak UV light source. I may try again with the stronger studio strobe. Could be worth-while.

 

Also got some pointers from Klaus Schmidt (most helpful as always) re additional and quite a bit more sensibly priced UV microscope objectives.

 

Below is a comparison shot of the same species of orchid with visible and UV light plus filters. Scale bar is on first image. That's the kind of thing I want to do, possibly get single flower full frame, and this species is moderately large for the genus. What is interesting to me from a pollination biology perspective is that the rachis has different color from flower in pure UV (Baader U) and that lip is darker in the insect vision images (Bug U3/5).

 

[ulf]

The MPE65 has not a good UV-transmission.

I did a very quick test on mine a while ago and the UV transition (-50%) was around 385nm.

At 365nm there was nothing left.

 

This morning I found the box with my stacking lenses including the Canon 20/3.5.

Amazing how the brain is processing questions during sleep, when you have begun thinking about something.

It is almost ten years since I had any reason to use or think about those lenses or anything related to focus stacking.

 

I will check and see if there are any promising UV-alternatives. I hope to be able to return with result within a few days.

Is it correct that you are looking for a magnification-range of 1:2 - 1:10?

[Daniel Geiger]

Ulf: thanks for the info on the MPE 65 and the Canon 20. You are correct re magnification range.

[nfoto]

I would have thought 2:1 to 10:1, i.e. magnification?

[ulf]

I would have thought 2:1 to 10:1, i.e. magnification?

 

Ops, naturally, that is what I ment to write.

I think Daniel read what I ment and not what I wrote too.

 

Sorry for the confusion and thank you for the correction, Birna.

[Daniel Geiger]

I also automatically read it as 2x to 10x mag, pretty clear in context.

[dabateman]

@ulf W,

I disagree with saying the Canon 65MPE is not good for UV. 50% at 385nm sounds great. Remember, sunlight is shifted to these wavelengths and full spectrum cameras are really only good from 370nm and above.

I was working with the 65MPE and it seemed quite good. Using better lights will help. I now have some 365nm led bulbs, which work well. If your using flash, I also think it should still work for you. That is if you own the Canon 65MPE, I would run out and buy it just for UV work.

 

Also a cheap alternative that I too may test is the UKA optics 25mm c-mount lens on extension. You can buy a c-mount to M42 adapter, then use a M42 mount adapter to your camera. Using the Wollensak 25mm f1.5 and these adapters on my olympus E510 I was able to focus 52mm away from a subject, about 2 inches. On my Sigma SD14, I was able to focus about 30mm away. I am not sure the magnification off the top of my head. But I think it would be in your 2x to 10x range. Canon has the same flange back distance as Sigma.

[ulf]

@ulf W,

I disagree with saying the Canon 65MPE is not good for UV. 50% at 385nm sounds great.

 

Well, then lets agree to disagree. :)

 

This might be a case of semantics, but also something caused by that I wasn't specific enough about what I stated.

 

I wrote "the MPE65 has not a good UV-transmission".

I ment compared to other lenses suitable for UV, discussed on the forum.

 

Depending on what you are investigating and what filters, cameras and lightsources are used the importance of a short wavelength transition differs.

However I would expect most members on the forum would consider a lens with a transition wavelength of 385nm or worse as only barely suited for UV-work.

 

Normally the wavelength transitions in a transmission measurement are specified by where the relative transmission has fallen to 50% of the peak transmission.

That is the estimated value I gave for the MPE65. The 50% is not the absolute transmission at that point. It is certainly worse for this lens.

 

The MPE65 is a very nice macro-lens, but not that well suited for UV-work.

 

Have you really tried the "365nm led bulbs" together with the MPE65?

At 365nm there is very little transmission left in a MPE65.

[dabateman]

Ulf W,

Yes I agree with what you wrote above. More clear. Also I ment to say I would Not buy the 65mm just for UV work. I really need to watch that.

From what you previously wrote I thought you actually measured the transmission. Doing that vs feelings is quite different.

In most cases we give much more credit to something by feeling than by true measurement.

I haven't had the chance to test it with the 365, but I will and let you know.

David

[ulf]

David, I really did measure the transmission, but only a relative transmission normalised to the level at 410-420nm.

Not a feeling, but a good indication about where the cutoff wavelength is for this MPE-65 lens.

 

I do not have a system with an integrating sphere, monochromator and high sensitivity detectors.

 

I use an UV-Vis array spectrometer, UV-rated collimators and fibers and a deuterium light source.

The system covers ca 200-1000nm.

The sensitivity and light source are not strong enough to work well with an integrating sphere, big enough to capture all the emerging light from a lens.

 

Without that a proper absolute transmission-measurement of a lens cannot be done.

However if you are only looking for detecting the cutoff wavelength it is doable as long as you can see the flat passband.

Good UV-capable lenses have a distinct visible passband and are easy to define the cutoff for.

 

The MPE65 is a borderline case where the transmission level still is rising at 410nm.

 

I hope I have clarified things a bit now.

[Andrea B.]

Dabateman: ...full spectrum cameras are really only good from 370nm and above.

 

I don't quite understand this?

I don't quite understand why you are saying this?

[dabateman]

Andrea,

Sorry, I guess I am being too biased based on Jonathan's results. Yes full spectrum converted cameras will see 340nm and even 320nm as both you and Cadmium have clearly shown many times.

I am basing my 370nm and above, due to the cameras higher sensitivity at 385nm, and the tendency for solar light to be higher wavelengths. But I should not make such broad incorrect statements. I need to proof read what I write and place things in context. Getting only 2 hour sleep per night for the last 3 years has really caught up with me and I find I make too many mistakes lately.

 

UlfW,

I tested the Canon 65MPE this morning with my lights and will up date my 65mpe post about it. Full disclosure though, I don't have a full spectrum Canon camera to test with. I can only use it on my Stock Olympus EM1, so that will significantly affect my range. The quick result using the convoy S2+ flash light 2 inces away from subject and the Canon 65MPE at 2.2x mag f8 iso 800 8sec was acceptable to me. Adding the speed booster had an impact of just less than 1 stop. The 365nm led bulb I have was acceptable at iso 3200 30sec and the light was 4 inches away from subject. My 405nm led was acceptable at ISO 800 30sec, and my compact fluorescent black light was acceptable at ISO 3200 60sec. All tests are with the Baader venus filter.

I tried to compare the EL 80mm f5.6 lens head to head, but I only had 50mm of extension beyond the 70mm back focus distance. So the full flower was in view and not much magnification was obtained. With the convoy ISO 200, f8, 4sec was acceptable.

But it reminded me that in general using high magnification is dark unless you use twin flashes. I didn't test my Canon 199As, but that may help. Being close to the subject really makes the lighting hard.

[JMC]

I did a quick test today, using a 50mm Canon Macro lens, 65mm MPE lens, and the 85mm Asahi UAT on my Eos 7D modified for UV imaging by ACS with one of their proprietary filters. The sensitivity of this camera as a function of wavelength is similar to a multispectral camera with Baader U filter. The images were taken under hazy sunlight within a couple of minutes of each other. Whitebalance images were taken using a PTFE tile, and whitebalance as done in Dark Table afterwards. They were all at f11 and ISO400. However the time for exposure varied (given they have different magnifications that not unexpected anyway).

 

Firstly the control - 85mm UAT lens;

 

Not the best focussing, but this shows the typical dark, almost black centre to the flower and yellow petals when imaged with this lens. The cardboard the flower was on went almost grey in whitebalancing. White balance temp was about 2700K, with a tint of about 1.7 in Darktable when checked using Spot whitebalance on the tile.

 

Secondly the 50mm macro EOS lens;

 

The yellow of the petals is there, but it has the distinct pink tinge even after white balance. The black centre of the flower is now much less well defined. The white balance was around 4700K and a tint of 2.3. The 50mm macro lens is not a good UV lens. It gives about 1 less stop transmission than my 40mm EOS pancake lens, which is itself not as good as some of the enlarger lenses and older 35mm lenses.

 

Thirdly the 65mm MPE lens (at about 2x magnification);

 

Ignore the dept of field issue, as I didn't spend a lot of time with it to try and get it right. As with the 50mm macro lens, the petals look yellow, but with a pink tinge. Also the black centre is again much less well defined. The white balance was similar to the 50mm macro lens - about 4700K and tint of 2.3.

 

While there is some UV getting through with the 50mm macro and 65mm MPE lenses, otherwise you wouldn't see a slight darkening of the centre of the flower, it's far from ideal and doesn't go too far into the UV. The white balance is significantly different to the UAT lens, suggesting a big difference in the range of wavelengths of light getting through to the sensor.

 

I did some calculated transmission measurements of the 50mm macro lens, while I was playing about with the camera sensitivity measurements. The link to that is here;

http://www.ultraviol...dpost__p__19599

 

Of the lenses I'd tested to that point it has the worst UV transmission in terms of how far into the UV it went, so not surprising it doesn't give a distinct colour difference across the image here.

[ulf]

I think I have a solution that will work well for UV up to 5:1 - 6:1.

The correct version and vintage Leitz Foccotar 50/4.5 has a very good UV-transmission.

See my findings here: post #1 and #22 http://www.ultraviol...__fromsearch__1

 

The Focotar 50/4.5 are rather common on eBay often with a reasonable price.

The Focotar-2 50/4.5 is not seen that often and more expensive.

Both are very good, but the Focotar-2 is slightly better.

 

I quickly tried one of the old Focotar-lenses for magnification.

I used a long extension-tube, helicoid combination, set to 25cm.

 

That gave a magnification of ca 5x.

The distance from the front of the lens body to the focal plane became around 40mm.

 

That leaves enough space for illumination, but I would aim for placing the UV-pass filter somewhere behind the lens, in the tube structure.

This could be done with suitable step rings.

 

A Baader U in original m48 filter cell will need four rings, if a M42-system is used: M42-M52, M52-M48, up and down versions.

For mounting the Focotar a M39-M42 ring is needed.

 

Multiple helicoids is not really needed and give worse stability. I just grabbed what I had available.

Rigid extension tubes are superior to both helicoids and bellows for stability.

[Andy Perrin]

I laughed at all those extension tubes.

[ulf]

In the beginning I too thought it looked crazy, but I got used to the view. :) :) :)

 

The setup I used for the stacked images of the fly and hair in post #15 in this thread, was just a bit shorter.

 

It will work and is much more stable than any bellows setup.

The risk for leakage is also smaller.

[Cadmium]

:) Ulf, Wow! That is some setup there!