Making Pentax SMC 105 Telecentric Using Raynox 250

[mjkzz]

This is more or less a demonstration of a how-to in optical playground. Others have done it before with Canon 100mm macro and I was bored so I did it with Pentax SMC 105mm f/2.8 lens. In terms of magnification, it is about 17/21 = 0.81x, image quality is not as great as the hot "line scan/copy" lenses discussed here lately, but not too bad.

Ignore the camera and adapter to mount M42 to MFT adapter, if you use other cameras, it fine as long as you have the right adapter to mount a M42 lens. It is important to stress that the hood is critical, and since the WD is very long (about 110mm), it does not pose any problem at all.

All of these tubes there are to make sure the focal plane (of Raynox) coincides with where entrance pupil of the Pentax lens is located. Aperture is adjusted on the Pentax lens.

A - Pentax SMC 105mm f/2.8 M42 mount
B - 49 male to 49 male adapter ~3.5mm thick
C - 42 to 49 (not 49 to 42) ~4mm thick
D - 28mm segment of 42 extension tube
E - 14mm segment of 42 extension tube
F - 7mm segment of 42 extension tube
G - 20mm long M42 with through female thread.
H - 42-43 adapter ring ~4mm thick
J - Raynox 250
K - lens hood came with Pentax SMC 105mm lens.



Test image done with a Sony A7III, stacked with Zerene, scale factor is around 0.9983 between two images acquired at 200um spacing, I think it is close enough.

[Beatsy]

That's really interesting! Couple of questions...

1. Why the male to male adapter on the lens filter thread? Is that just a case of being able to use the parts you had? I have a 49 to 42 filter adapter, so I presume that would work just fine too.

2. If the rig were extended to include an infinity objective on the end (with the Raynox acting as its tube lens), would that remain a near-telecentric setup too?

Thanks for sharing. Please get bored often

[mjkzz]

That's really interesting! Couple of questions...

1. Why the male to male adapter on the lens filter thread? Is that just a case of being able to use the parts you had? I have a 49 to 42 filter adapter, so I presume that would work just fine too.

2. If the rig were extended to include an infinity objective on the end (with the Raynox acting as its tube lens), would that remain a near-telecentric setup too?

Thanks for sharing. Please get bored often

1. I could not find a 49 to 42 step down ring in my pile of adapter rings, so I have to improvise. The actual configuration is not important as long as we can place the Raynox at the right displacement in front of the Pentax, in this case about 80-81mm (not very sure about the thickness of B). At this displacement, it seems the focal plane of Raynox coincides with the location of entrance pupil of the Pentax. That is what makes it telecentric.

2. I am not sure that would work, maybe for some objective (like high powered Mitties), but not all.

With this method, we can use a camera with a lens focused to infinity and look through from the subject side of the setup, if we can get a sharp image of the aperture, then we get a near-telecentric optical system.



Anyways, there are 3 ways (I can think of) to make a telecentric optical system, this is just one of them. If I can find some of my old telecentric setups, I will take some pics and post them here.

[Beatsy]

Thanks very much for this. I didn't know the test for telecentricity was so easy (imaging the entrance pupil at infinity focus). Although I mentioned infinity objectives in my original question, I think I'll try it with my Linoscan scanner lens (that I ended up mounting on a Raynox as tube lens). I suspect it won't work on a "sounds too good to be true" basis, but nothing ventured etc...

Be interested to see the other telecentric setups too. Thanks again.

[mjkzz]

Thanks very much for this. I didn't know the test for telecentricity was so easy (imaging the entrance pupil at infinity focus). Although I mentioned infinity objectives in my original question, I think I'll try it with my Linoscan scanner lens (that I ended up mounting on a Raynox as tube lens). I suspect it won't work on a "sounds too good to be true" basis, but nothing ventured etc...

Be interested to see the other telecentric setups too. Thanks again.

You can try the second method -- putting your own aperture (paper one is OK) at the focal plane of your lens on the image side and use the same method to check telecentricity. Be warned that you can not image something larger than your front element of a telecentric system. I will shoot some pics of this setup I have used.

[Lou Jost]

It would be good to test the tetecentricity using a more 3-dimensionally-complex subject. Note that the main lens shouldn't necessarily be focused at infinity; ;but the telecentric distance for the Raynox can depend on the focus distance.

Here are my experiments with a passionflower, which has lots of flower parts crossing each other in the air, with crossing points that are very dependent on perspective:
http://www.photomacrography.net/forum/v ... elecentric

[mjkzz]

It would be good to test the tetecentricity using a more 3-dimensionally-complex subject. Note that the main lens shouldn't necessarily be focused at infinity; ;but the telecentric distance for the Raynox can depend on the focus distance.

Here are my experiments with a passionflower, which has lots of flower parts crossing each other in the air, with crossing points that are very dependent on perspective:
http://www.photomacrography.net/forum/v ... elecentric

Thanks Lou.

Since I do not know too much about optical stuff, I always try to keep thing simple, so I always set the rear lens (in this case, the Pentax 105mm) to focus to infinite because, as you mentioned, the apparent location of its aperture might change depending on where it is focused to. So by keeping it focused to infinity, I have one less variable to think about when constructing a (near) telecentric system. But it is good to know that it does not have to focus to infinity, I will take your word for it.

As for testing, I am leaning towards the scale factor from Zerene. But good to know a more 3D structure is a better.

For a single image test, without stacking, photographing a (relatively) long tube head on should result two clearly defined concentric circles.

[Lou Jost]

The Zerene scale factor is definitive, but it is hard to know how close it has to be to 1.0000 for a given setup to behave as if it were telecentric. The ultimate test is whether or not it works in your application. The flower I showed in the link I gave you was extremely complex, and my set-up worked perfectly, so I am sure it can handle anything less complex, whatever the scale value was.

My other test is to compare a stack with scaling turned on versus the same stack with scaling turned off. If they look the same, I am happy to conclude that for this set-up, the lens is close enough to telecentric.

[mjkzz]

Silly me, I converted a Canon EF 50mm f/1.4 lens to (near) telecentric, but could not find the lens (it is on the camera I am holding to photograph the following). So I played with that Nikon El-Nikkor 50mm f/2.8 to make it (near) telecentric.

A - aperture assembly with M42 on each side
B, C, D - M42 to M42 adapter ring 4mm thick
E - M40.5 to M42 adapter ring 4mm thick

Total displacement from Nikon El-Nikkor 50mm to A is about 16mm, but there some gap between actual blades and edge of aperture assembly. It still needs some more fiddling to make it better, but the idea is there. Once this is constructed, it can be mounted on different tube lenses (or just use it as is) and stay (near) telecentric.

[mjkzz]

The Zerene scale factor is definitive, but it is hard to know how close it has to be to 1.0000 for a given setup to behave as if it were telecentric. The ultimate test is whether or not it works in your application. The flower I showed in the link I gave you was extremely complex, and my set-up worked perfectly, so I am sure it can handle anything less complex, whatever the scale value was.

My other test is to compare a stack with scaling turned on versus the same stack with scaling turned off. If they look the same, I am happy to conclude that for this set-up, the lens is close enough to telecentric.

I think scale factor of 0.9998 or less is good enough and the idea of turning on and off the scale in Zerene is brilliant.

[rjlittlefield]

I didn't know the test for telecentricity was so easy (imaging the entrance pupil at infinity focus)

I guess it's time for a FAQ on telecentric lenses, because some basic stuff is getting overlooked.

Until the FAQ can get itself written, I suggest to read the Jan 2007 topic "Telecentric optics, third round", and maybe the references included therein. You will find the test for telecentricity, and why it works, and why the telecentricity does not have to be perfect, and some hints about how far off it can be and how to test for that. That particular thread predates Zerene Stacker, so of course you won't find anything about its Scale number in there.

--Rik

[mjkzz]

Bored again . . . I think I did this before but just forgot this configuration. This is to validate that a lens does not have to focus to infinity to be modified for telecentricity (thanks Lou).

Lens is an old Canon EF-100mm macro lens, the adapter is 58 to 43, and finally the Raynox DCR-250. I have focus peaking turned on on the GX85 and the Leica 25mm focused to infinity (MF), so you will see some blue edges due to focus peaking.

[mawyatt]

Peter,

Very interesting indeed!!

Best,

[Lou Jost]

Check out the test images of the passionflower on the link I gave in an earlier comment; this method works astoundingly well.

[mawyatt]

Lou,

Thanks, that's very interesting and useful. Now you and Peter have got me thinking about conjuring up something with some of the loose ends I have hanging around

This will have to wait until some other things get accomplished first

Best,