New camera, first tests and a long story to tell

[Stefano]

My Panasonic DMC-F3 broke some time ago. The lens became stuck for unknown reasons and never worked again. I took about 15,000 photos with that camera, a lot of photos. The camera worked quite well, it delivered nice images, but had some big limitations:

- Fixed lens. The lens was not interchangeable, I had to work with the lens I had. It was good to about 360 nm, maybe 355, but not lower than that. It was enough to produce violets (close to a blue) and yellows, but I couldn't see my 340 nm LED;

- Limited ISO (1600), and quite nolisy;

- I couldn't set the aperture, nor the shutter speed, and I couldn't set an in-camera white balance in "starry sky" mode (long exposure mode)

 

After I announced my camera broke, Ulf offered me one of his Canon EOS M. Long story short, I bought a white one that was still stock, and I had to convert. It was a very nice camera, and I converted it with my dad, being careful not to trap dust between the sensor and the full-spectrum filter (Astronomik MC-Klarglas filter). After assembling it together, and disabling the dust shaker, the camera worked. Then discovered there was a lot of dust trapped inside, took it apart again, removed the filter, closed it again, but it rattled, opened it again, closed it, it didn't work, opened again, tried to adjust a connector, closed again... The camera never worked again. It turns on, but gives an error. I didn't follow Ulf's instructions to work calmly, with "baby steps", and that was a great life lesson. He has been very helpful the whole time, and I want to thank him again. He suggested me to buy an adaptor with helicoid, and a Soligor 35 mm f/3.5 lens (no KA serial number), which performs very well in UV.

 

Before breaking that camera, I took some photos with a pinhole and a magnifying glass (about 300 mm in focal length), stopped down to about f/10 (I didn't have the adaptor yet). I didn't increase the saturation in any image. For all UV images, I used a ZWB2 (2 mm) + Chinese BG39 (2 mm).

 

Pinhole:

 

Magnifying glass

 

I then bought a black Canon EOS M from Alan Burch, it took a while to arrive because of brexit things, but works well. It has some dead pixels, but I assume it is normal? Maybe it was irradiated with gamma rays? Anyway, not a big deal. It is a very nice camera. I adapted my filters to fit the Soligor (will show them in the future), and started taking some pictures:

 

Handheld camera, handheld filter (only first image)

 

 

Sunscreen:

 

 

 

Image brightened in post:

 

Half banana:

 

The Soligor is quite sharp at f/3.5 in UV, but is much sharper at f/22.

 

f/3.5:

 

f/22:

(also notice the interesting pattern)

 

ZWB1 (3 mm thick):

 

Under a 340 nm LED... (the black thing in the image is a plastic lens which is completely opaque at 340 nm):

 

One of the old vintage lenses I have, owned by my grandfather and maybe by my great-grandfather. it appears slightly purple, which indicates it is good for UV but not as good as the Soligor:

You can also see some residues of sunscreen on my finger.

 

And now some IR (Hoya R72):

 

 

 

The lens seems to be hotspot-free in the NIR. At f/22, you can literally see the Airy's disks, so it must be used at wider apertures.

 

Having both cameras, I have two batteries, two chargers, two straps... I have a spare for everything. In particular, the double battery is very useful.

 

Colors

My old camera had a slightly different UV palette. The purple/lavenders were closer to a blue, and the yellows were closer to a pure yellow. Here I have true lavenders, and greenish yellows. I guess it is due to the better UV-reach.

 

In IR, instead, the colors are not as strong. The skies were more yellow with the old camera, but I can still play with the saturation after taking the photos.

 

All the images above were taken as .jpg and were white balanced in-camera. I will have to learn how to work with RAW images.

 

I am still learning a lot. I hope to do a lot of things with this camera. It is very promising.

[bvf]

Glad to see you are equipped and ready to go, Stefano.

 

I started my full-spectrum carer with an EOS M as well. It gave great results, but there was one slight problem: if you look VERY closely, you can sometimes see a row of dots in UV images, which I believe is to do with the autofocus mechanism. These will not cause any problem unless you start doing focus stacking - on a few shots this will cause an effect like this:

 

 

Regarding the Soligor lens, I suspect that is the same as the Igororiginal lenses I covered in my post today about lens resolution. They perform well in UV, especially at the centre. In UV you will get optimum image quality by stopping down to f/11 or smaller, but in IR you will need to use a much larger aperture - perhaps f/5.6 or even larger - because of the greater impact of diffraction.

[Stefano]

I too noticed those dots. You really have to zoom a lot to see them.

 

I don't know how deep the Soligor goes, but maybe it reaches 320 nm or so. At f/3.5, in UV, the images look OK, but there is a slight halo around bright objects. At f/22, the images are sharp to the pixel.

 

I am thinking a Focotar-2 is one of the best lenses for UV. Goes below 310 nm and is sharp. But I would like focal lengths shorter than 35 mm. The Soligor, although 35 mm is "medium" still feels too telephoto sometimes.

 

I would like to try your Tri-color UV images one day. I don't know if the Soligor will allow that.

[colinbm]

Good to see you are back in business Stefano.

[Stefano]

Thanks Colin.

[Andy Perrin]

I’m glad you are back to normal. I am not surprised by your camera conversion problems because it’s actually rather difficult between the dust and every camera being a bit different to take apart. I did my first camera, the Coolpix, but I didn’t dare convert my Sonys.

 

I believe that Zenit has a very good reputation— one of the Russian photographers I follow on Instagram likes that lens.

[dabateman]

Congratulations on getting a working camera. Canon M-mount is an interesting choice. The 18mm flange back distance should make adapting lenses, not too hard.

Also the smaller 1.6x sensor (22.3mm x 14.9mm) should allow for 25mm filters to be used like four thirds system. You should be able to mount C-mount lenses without too much issue.

[bvf]

 

I don't know how deep the Soligor goes, but maybe it reaches 320 nm or so. At f/3.5, in UV, the images look OK, but there is a slight halo around bright objects. At f/22, the images are sharp to the pixel.

 

I am thinking a Focotar-2 is one of the best lenses for UV. Goes below 310 nm and is sharp. But I would like focal lengths shorter than 35 mm. The Soligor, although 35 mm is "medium" still feels too telephoto sometimes.

 

I would like to try your Tri-color UV images one day. I don't know if the Soligor will allow that.

 

Assuming the Soligor is the same as an Igororiginal, it will have good UV reach. It works with the 315BP20 filter, although there is vignetting with the small (25mm) filter I have. You may have problems finding a wider angle lens. I have a 28mm f/3.5 Lithagon which has been rated as UV-friendly, but I don't think the UV reach is very good. The other problems you have as you go to wider angles are that you will need larger filters (I have not been able to try the Lithagon with my tri-colour UV set) and you may experience the problems with dichroic filters that have been discused on the forum.

[bvf]

 

Also the smaller 1.6x sensor (22.3mm x 14.9mm) should allow for 25mm filters to be used like four thirds system.

 

 

If this Soligor is the same lens as my original, the problem is that the front element is deeply recessed in the lens housing and so the filter is some way in front of the lens. So I do experience vignetting with 25mm filters.

[bvf]

 

I don't know how deep the Soligor goes, but maybe it reaches 320 nm or so

 

 

By the way, there is a good discussion of the Soligor 35mm in UV here: http://www.savazzi.net/photography/35mmuv.html

 

Be aware that there are 2 versions of these 35mm f/3.5 lenses, and one of them is not so good for UV. The way to check is to measure the diameter of the rear element. The "good" lens has a diameter of about 12mm (the glass, not including the metal around it) and the "bad" lens is about 15mm across. The front elements are also different, although more difficult to measure because they are recessed: the "good" lens is about 20mm diameter and the "bad" lens about 30mm. (There is also a post about this somewhere on UVP.)

[bvf]

 

I believe that Zenit has a very good reputation— one of the Russian photographers I follow on Instagram likes that lens.

 

 

The Helios-44 is a much discussed lens. It is based on the Biotar design - another piece of technology the USSR acquired as war reparations. It was the main kit lens on the Zenith cameras from the 1960s on. There were a number of iterations of the lens, each bringing improved performance and/or mounts. The -5 model in Stefano's is quite a late version and well respected. The pinnacle of the family was the -7 version which is rated as particularly good (although Russian quality control was a bit hit and miss so that there were good copies and poor ones), and the -7 version attracts high prices. The problem is that many, if not most, -7 versions on the market are fakes - it is easy to replace the front ring on an older version of the Helios with a ring claiming to be the -7 version. A useful check is to look at the serial numbers: Soviet/Russian cameras and lenses identified the year of manufacture in the first two digits of the serial number, and the -7 version came out in '92.

[ulf]

I started my full-spectrum carer with an EOS M as well. It gave great results, but there was one slight problem: if you look VERY closely, you can sometimes see a row of dots in UV images, which I believe is to do with the autofocus mechanism.

I do not think so. AFAIK the EOS M is focussing purely by contrast and you can chose focus point freely on the touch screen.

It is too old to have any dedicated focus pixels.

 

Some dead or hot pixels can occur and sometimes the camera is not able to mask them after it has been converted.

My EOS 60D had some hot pixels that started to brighten during long exposures. They were either red blue or green.

Some RAW-processors can take care of that problem in the post process of the RAW-file.

As I only use RAW-files I have no idea about how Canon is doing when converting to jpg.

[ulf]

Stefano!

I am glad that you finally is up and running with your new camera from Alan.

 

Don't forget BABY STEPS and much research before you do something physically with the camera like cleaning the sensor.

The sensor will be dirty and will need to be cleaned eventually.

Then use proper materials and strict routines.

 

The best way to keep it clean is to expose it as little as possible to the environment and be very careful when changing lenses.

[dabateman]

Bernard read this about reesesing filters deeper into the front of lenses here:

http://www.savazzi.net/photography/mounting_filters.html

 

And bottom of page here:

http://www.savazzi.net/photography/uvpassfilters.html

 

However, since the flange back distance is 18mm, and the sensor is about the size of a micro four thirds sensor, my point was a 25mm filters can actually be positioned behind the lens using a dual C-mount to M42 mount adapter. I must have forgotten to explicitly say that.

 

But just a quick Ebay search and I see them for Nikon 1 mount, Sony E-mount, Fuji X-mount, and MFT, but not Canon. You can make one with an other adapter fitting either a C-mount to M42 or 1.25" to M42 no lip adapter inside the center of a thicker M42 adapter.

[ulf]

When changing filters on te same lens you have to expose the sensor for dust every time That is not good.

There is a risk of vignetting if the filter gets too close to the sensor. The diagonal is 29mm and the 25mm filter ring has a clear passage smaller than 25mm, like 21 or 22mm

 

The diagonal of a FourThirds sensor is much smaller than an APS-C sensor.

https://www.google.c...QAAAAAdAAAAABAD

[Stefano]

I will leave the lens mounted on the camera as much as possible in order to seal the sensor inside.

[dabateman]

All depends on the print size. I like 8x10 or a 1:1 square. So I almost always throw the sides out. So even on four thirds the side gets cropped off, how I like to look at things. So that is why height is more important to me than width or diagonal. The EF-M is nearly identical, 13mm vs 14.8mm.

 

But I have know idea what Stefano prints at, so shouldn't assume similar.

[ulf]

Rather few people limit themselves to 1:1 all the time, if they have a sensor with the dimension-ratio of 2:3.

Sometimes I also crop to 1:1 but I do not like to be forced to do that by vignetting if it is possible to avoid that.

Small fiters, rear mounted, are best to mounted close to the lens rear element to avoid vignetting as much as possible.

[bvf]

 

I do not think so. AFAIK the EOS M is focussing purely by contrast and you can chose focus point freely on the touch screen.

It is too old to have any dedicated focus pixels.

 

Some dead or hot pixels can occur and sometimes the camera is not able to mask them after it has been converted.

 

 

I may be wrong about these pixels being for autofocus, but they are not dead or hot pixels. There are at least two rows of them, in a dead straight line and at even spacing. You can see them in the image I posted above where the effect is made prominent by focus stacking.

 

Actually I've just found this at https://www.cameralabs.com/canon_eos_m/ :

 

... the EOS M features a hybrid AF system which employs a combination of contrast-based and phase-detect AF systems, the latter integrated into the sensor itself. As I understand it, the 31 phase-detect AF points are concentrated in an area towards the centre of the frame and are used to perform initial calculations and determine the direction the lens should focus, before leaving the contrast-based system to nail the final focus. Obviously should the subject fall outside of the phase-detect AF points towards the edges of the frame, the system becomes contrast-based only.

[ulf]

Bernhard you are better informed about this than I am and I have today learned something new.

Thank you for the correction.

 

My hot pixels in the 60D were always at the same place, but placed quite random.

There were less than ten of them and easily cloned away manually, but my Raw processor did that automatically for me.

[Andy Perrin]

I change lenses pretty frequently and also deal with dust a lot. I've gotten used to editing it out in photoshop with a mask of the dust locations and Content Aware Fill. When it gets dirty enough, I have a loup and light combination and a sensor pen (it was sold as one kit) that I use to clean it. But dust is a fact of life with interchangeable lenses, at least if you want to have that advantage.

[Stefano]

I had some fun cranking the ISO all the way up as well as doing some very long exposures. I also discovered how to enable extended ISO (H, "High"), which for my camera is 25600 equivalent for photos and 12800 equivalent for videos. I like having it in the ISO bar at the bottom, even if I will not always use it.

 

For all images, the lens was the Soligor (shown below).

UV filter: ZWB2 (2 mm) + Chinese BG39 (2 mm).

 

I tried some very long exposures in BULB mode:

 

f/3.5, ISO 12800, 5592.5 s exposure. My longest single exposure yet (more than a hour and a half). UV, nighttime. The little light I captured is probably of artificial origin. Also, there's a TON of noise.

 

f/3.5, ISO 12800, 1800.4 s exposure (half an hour). The bottom purple light is a fluorescent tube, the yellow is probably the 365 nm I-line of nearby mercury-vapor streetlights.

 

f/3.5, ISO 25600, 1/125 s exposure.

 

f/3.5, ISO 25600, 1/40 s exposure. Photo taken in the shade if I remember right. Not very focused, but that's a pretty nice image for being UV in the shade, handheld at ISO 25600. It may be nothing special for your cameras, but something like this was unthinkable with my previous camera, limited to ISO 1600 and 1/8 s.

 

At ISO 12800 I can easily record UV videos in the shade. The noise is there, but not too bad. I can stop the lens down to f/8 under direct afternoon sunlight.

 

And now, my lens. A Soligor 35 mm f/3.5 (is this its name?) I measured the rear lens diameter, and it is 12 mm, so it looks like I got a UV-capable lens.

 

 

As you can see, the serial number doesn't begin with "KA".

 

 

 

The filters (here the UV stack) are attached to a paper/cardboard/tape "ring" that fits perfectly the focus ring of the Soligor.

 

 

 

Looks "bad", but it works perfectly.

[Andy Perrin]

Looks like you're in business again! I wouldn't necessarily trust that you are seeing UV only in a night photo unless you have done an additional test stacking with a long pass filter to check for out of band stuff. I don't know about your filters in particular but there was a post a bunch of years ago showing that the Baader leaks under night conditions sometimes.

[dabateman]

Awesome, that first and second image show you exactly were the PDAF points ate on the sensor in the center.

 

Now you can have some real UV fun.

[Stefano]

Looks like you're in business again! I wouldn't necessarily trust that you are seeing UV only in a night photo unless you have done an additional test stacking with a long pass filter to check for out of band stuff. I don't know about your filters in particular but there was a post a bunch of years ago showing that the Baader leaks under night conditions sometimes.

Yes, I probably have to check that. I don’t know the blocking of my filters, if they were the Schott/Hoya equivalents at that thickness they would be over OD 5 (like Hoya U-360 2 mm + BG39/S8612 2 mm), but being “unknown” filters I can’t know for sure. I am able to force a leak stacking them with a Hoya R72 and looking at the tungsten filament of an incandescent/halogen bulb, but this doesn’t tell me much.