Telecentric combo at 0.8X to 1.69X

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

killes wrote:Any suggestions for a test setting?
If you just want to test whether the modified lens is truly telecentric, then one good test is the tipped grid shown at http://www.photomacrography.net/forum/v ... 0750#50750.

If you're intending to use the lens for stack-and-stitch with no parallax or other alignment problems, then the relevant test is whether you can get a stack with no computational alignment that's as clean as what you get with ordinary alignment. Using Zerene Stacker, that's done by clearing all the checkboxes at Options > Preferences > Alignment.

A lens does not have to be perfectly telecentric to pass the latter test. It just has to be close enough to telecentric that the stacking process can impose the right geometry without introducing image artifacts. There's some discussion and an example of that aspect HERE.

I'm curious -- what is your application for telecentrics?

--Rik

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

killes wrote:I've found a friendly photo shop owner who was able to make me a spacer with an embedded diaphragm from an old teleconverter. He made an adapter on top of it so that I can attach any lens.

I've attached my Sigma 105mm lens and - as predicted - the part of the sensor that has a sensible image is rather small. I've then added a stack of spacer rings and now the whole sensor area gets used. Total magnfication is roughly between 1:2 to 2:1.
By the way, I assume that when you say "Total magnfication is roughly between 1:2 to 2:1", you're referring to adjusting extension behind the added aperture. If you're talking about turning the focus ring on the lens in front of the aperture, then I'd be very surprised if the setup remains telecentric when you do that.

--Rik

killes
Posts: 7
Joined: Thu Feb 14, 2013 7:36 am

Post by killes »

rjlittlefield wrote:
killes wrote:Any suggestions for a test setting?
If you just want to test whether the modified lens is truly telecentric, then one good test is the tipped grid shown at http://www.photomacrography.net/forum/v ... 0750#50750.
Thanks! I'll see if I can set up such a test.
If you're intending to use the lens for stack-and-stitch with no parallax or other alignment problems, then the relevant test is whether you can get a stack with no computational alignment that's as clean as what you get with ordinary alignment. Using Zerene Stacker, that's done by clearing all the checkboxes at Options > Preferences > Alignment.

A lens does not have to be perfectly telecentric to pass the latter test. It just has to be close enough to telecentric that the stacking process can impose the right geometry without introducing image artifacts. There's some discussion and an example of that aspect HERE.
Thanks again!
I'm curious -- what is your application for telecentrics?
I've in the past tied to make stacks of macro shots of plants but I found it difficult to get proper stacks when the plant did have hairs or spines. I would always get artefacts at the edges of the frame, making the usable area smaller.
rjlittlefield wrote: By the way, I assume that when you say "Total magnfication is roughly between 1:2 to 2:1", you're referring to adjusting extension behind the added aperture. If you're talking about turning the focus ring on the lens in front of the aperture, then I'd be very surprised if the setup remains telecentric when you do that.
I was actually referring to turning the focus ring of the lens in front. I will need to see for which positions it is actually telecentric, but wasn't the point of this setup with the extra diaphragm at the flange distance that it would make any lens telecentric? Maybe I misunderstood this.

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

killes wrote:I was actually referring to turning the focus ring of the lens in front. I will need to see for which positions it is actually telecentric, but wasn't the point of this setup with the extra diaphragm at the flange distance that it would make any lens telecentric? Maybe I misunderstood this.
Probably you misunderstood.

Adding a diaphragm at the flange distance will make any lens telecentric, but only when the lens itself is set to infinity focus. That combination places the diaphragm at the rear focal point of the lens, which is one of the criteria to be telecentric. The problem is that when you refocus the lens in front of the diaphragm, then the rear focal point moves, the diaphragm is no longer located at it, and the combo goes non-telecentric.
I've in the past tied to make stacks of macro shots of plants but I found it difficult to get proper stacks when the plant did have hairs or spines. I would always get artefacts at the edges of the frame, making the usable area smaller.

Yes, telecentric will solve this problem, assuming you can retain good image quality clear to the edge of each source frame.

But frequently this problem can be solved in a simpler way by changing the way you focus. See discussion and illustrations at http://zerenesystems.com/cms/stacker/do ... versusrail. The key question for this issue is how much perspective changes between front and back focuses of the stack. With many lenses, particularly internal focusing ones, the perspective change is much less if you change focus by ring than if you change focus by distance.

It's probably worth pointing out that our galleries are full of high quality images and almost all of them were shot with non-telecentric optics. Personally I only bother with telecentric when I want to stack-and-stitch, because in that case the other ways to avoid parallax are even more painful.

--Rik

killes
Posts: 7
Joined: Thu Feb 14, 2013 7:36 am

Post by killes »

rjlittlefield wrote: Probably you misunderstood.
Quite likely! 8)
Adding a diaphragm at the flange distance will make any lens telecentric, but only when the lens itself is set to infinity focus. That combination places the diaphragm at the rear focal point of the lens, which is one of the criteria to be telecentric. The problem is that when you refocus the lens in front of the diaphragm, then the rear focal point moves, the diaphragm is no longer located at it, and the combo goes non-telecentric.
I see. That also means I don't need to bother with the heavy macro lens and can use a leightweight 50mm one.
I've in the past tied to make stacks of macro shots of plants but I found it difficult to get proper stacks when the plant did have hairs or spines. I would always get artefacts at the edges of the frame, making the usable area smaller.

Yes, telecentric will solve this problem, assuming you can retain good image quality clear to the edge of each source frame.
I believe I can solve that by adding extension tubes in the back.
But frequently this problem can be solved in a simpler way by changing the way you focus. See discussion and illustrations at http://zerenesystems.com/cms/stacker/do ... versusrail. The key question for this issue is how much perspective changes between front and back focuses of the stack. With many lenses, particularly internal focusing ones, the perspective change is much less if you change focus by ring than if you change focus by distance.
I don't think this will work in my case. I usually use the macro lense near its 1:1 setting and when I focus by ring, I'll also change magnification. It is also not focussing internally, so the position of the front element changes as well.

I also try to minimize problems with the background by using a flash. But of course this is not always feasible.
It's probably worth pointing out that our galleries are full of high quality images and almost all of them were shot with non-telecentric optics. Personally I only bother with telecentric when I want to stack-and-stitch, because in that case the other ways to avoid parallax are even more painful.
I don't think the current setup is very painful. Unfortunately, the teleconverter (ab)used is of the manual kind, so I am fully manual now. But I still got TTL and the autofocus indicator even works.

It's sunny outside, I think I'll try to obtain some samples. :)

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

killes wrote:I don't think this will work in my case. I usually use the macro lense near its 1:1 setting and when I focus by ring, I'll also change magnification. It is also not focussing internally, so the position of the front element changes as well.
Change of magnification by itself is no problem -- any of the stacking programs can compensate perfectly for that.

Movement of the front element is definitely an issue. I have one macro lens with similar characteristics (a Sigma 105, old model). Around 1:1 with that lens, both methods of focusing are far from perfect.

However, that lens has one very nice feature: the entrance pupil maintains a constant distance from the front of the lens as focus changes. This simplifies the added-lens approach because no change in separation is required as the lens ring is turned to change magnification/focus.

Using the Raynox DCR-250 in combination with the Sigma 105, the combo becomes telecentric with about 45 mm of rings between the Sigma and the Raynox. The useful magnification on APS-C without added extension behind the lens varies from about 0.83:1 to 1.59:1, field width 27 to 14 mm. [Numbers corrected from initial posting!] Image quality is good corner to corner, with some correctable red/cyan CA.

Fully automatic operation is retained, including auto diaphragm, auto focus and shooting stacks under software control using the A/F motor. The setup can be reproduced by simply screwing together new manufacture parts easily purchased on eBay at any time. I mention these points because they are some of the reasons why I have come to prefer the added lens approach, and I think it's important for other readers to understand the issues.

Thank you for prompting me to check the Sigma+Raynox combination.

--Rik

Edit: to correct magnifications as described below.
Last edited by rjlittlefield on Sun Mar 03, 2013 5:23 pm, edited 2 times in total.

killes
Posts: 7
Joined: Thu Feb 14, 2013 7:36 am

Post by killes »

rjlittlefield wrote:
killes wrote:I don't think this will work in my case. I usually use the macro lense near its 1:1 setting and when I focus by ring, I'll also change magnification. It is also not focussing internally, so the position of the front element changes as well.
Change of magnification by itself is no problem -- any of the stacking programs can compensate perfectly for that.
I have to admit: I didn't know that.
Movement of the front element is definitely an issue. I have one macro lens with similar characteristics (a Sigma 105, old model). Around 1:1 with that lens, both methods of focusing are far from perfect.
That's just the lense I have, too!

http://www.sigma-imaging-uk.com/lenses/macro/105mm.htm
However, that lens has one very nice feature: the entrance pupil maintains a constant distance from the front of the lens as focus changes. This simplifies the added-lens approach because no change in separation is required as the lens ring is turned to change magnification/focus.
This is exactly the reason why I did not want the "lens in front" approach, I thought the changing distance would be a problem. :D
Using the Raynox DCR-250 in combination with the Sigma 105, the combo becomes telecentric with about 45 mm of rings between the Sigma and the Raynox. The useful magnification on APS-C varies from about 1.05:1 to 1.85:1 (field width 21 mm down to 12 mm). Image quality is good corner to corner, with some correctable red/cyan CA.

Fully automatic operation is retained, including auto diaphragm, auto focus and shooting stacks under software control using the A/F motor. The setup can be reproduced by simply screwing together new manufacture parts easily purchased on eBay at any time. I mention these points because they are some of the reasons why I have come to prefer the added lens approach, and I think it's important for other readers to understand the issues.
Definitly!
Thank you for prompting me to check the Sigma+Raynox combination.
Thank you for doing that!


Some findings from my setup:

- when using a 50mm normal lens, I do not need to add extra extensions, the remaining part of the teleconverter acts as a 15mm extension tube. The image does not have vignetting.

- when I use it like that, the magnification is about 1:1

- When stacking, I can't change the magnification at the front lense, since it needs to be at infinity, but I can change the extension of the rear ex-teleconverter. Only 8mm, though.

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

killes wrote: That's just the lense I have, too!

http://www.sigma-imaging-uk.com/lenses/macro/105mm.htm
Mine is even older, not even a DG.
Some findings from my setup:

- when using a 50mm normal lens, I do not need to add extra extensions, the remaining part of the teleconverter acts as a 15mm extension tube. The image does not have vignetting.

- when I use it like that, the magnification is about 1:1
This makes sense. It sounds like you have extended the lens from its normal distance by an amount equal to the flange distance plus 15 mm, so about 60 mm total.
- When stacking, I can't change the magnification at the front lense, since it needs to be at infinity, but I can change the extension of the rear ex-teleconverter. Only 8mm, though.
Why only 8 mm? You can add ordinary extension tubes behind the ex-teleconverter.
rjlittlefield wrote:However, that lens has one very nice feature: the entrance pupil maintains a constant distance from the front of the lens as focus changes. This simplifies the added-lens approach because no change in separation is required as the lens ring is turned to change magnification/focus.
This is exactly the reason why I did not want the "lens in front" approach, I thought the changing distance would be a problem.
Well, it's a tradeoff. With the added aperture you have to change extension at the camera, where with the added lens you have to change extension between lenses. It seems the same amount of trouble either way, and with the other advantages of added-lens versus added-aperture, I see the scales tipping toward added-lens most of the time.

--Rik

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

rjlittlefield wrote:Using the Raynox DCR-250 in combination with the Sigma 105, the combo becomes telecentric with about 45 mm of rings between the Sigma and the Raynox. The useful magnification on APS-C varies from about 1.05:1 to 1.85:1 (field width 21 mm down to 12 mm).
Sorry, those numbers are not quite right. I had accidentally left a 20mm tube on the camera when switching lenses. Without the extra tube, the magnifications range from about 0.83 to 1.59 (field widths 27 to 14 mm). I have corrected the earlier post in hopes of reducing confusion later.

--Rik

Photo-DIY
Posts: 40
Joined: Mon Apr 04, 2011 7:56 am
Location: Germany
Contact:

Post by Photo-DIY »

killes wrote:Wouldn't it be possible to take an old manual lens and remove the actual lenses from it to get a tube with a diaphragm inside it? That diaphragm would still be operated by the camera. You just need to figure out a way to attach the actual lens to this and to keep the proper distance. Possibly use a reversal ring.
Should work - but as Rik wrote you need the front lens in constant distance to the iris.
I have build automatic Canon EF iris assemblys in manual lenses. It is a complicated process to get the correct distances between front and back lens assembly. But for a telecentric adapter this should be easy.

I had the same problem with stacking of images with ~3.5mm depth and ~10:1 magnification. For example with a leg in front of the insects eye. But my used Mitutoyo lens is a infinity lens, and it seems focal point is very close to the back side of the lens, or even inside the lens.
Photo Tinkerers Heaven and Hell:
http://www.4photos.de/index-en.html
Some own work - and a bigger list of DIY links

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Photo-DIY wrote:I had the same problem with stacking of images with ~3.5mm depth and ~10:1 magnification. For example with a leg in front of the insects eye. But my used Mitutoyo lens is a infinity lens, and it seems focal point is very close to the back side of the lens, or even inside the lens.
Do you mean the Mitutoyo Plan Apo 10X NA 0.28 WD 33.5mm, like the one HERE?

If so, then I think you have mis-identified the cause of your problem. That lens is very close to telecentric by itself. I have just now measured the change in magnification as approximately 1 part in 10,000 per DOF step. It would be a very unusual situation where that amount of change would cause a problem.

In general, the DOF of microscope objectives at 10X and above is so shallow that telecentricity is not a significant issue. You can just pretend that the lens is telecentric, force telecentric geometry by telling the software to ignore scale changes, and the stack will be fine. In fact it's usually better to do that than to let the stacking software try to correct for perspective, because the software is very likely to be led astray by changes in appearance with focus that have nothing to do with perspective. See detailed discussion at http://www.photomacrography.net/forum/v ... 9878#79878 and in the surrounding thread.

I would have to study your stack myself to be sure, but based on other stacks I've studied I would bet heavily that your problems were actually with bogus alignment due to blooming of out-of-focus features, and not with the perspective of the lens.

--Rik

Photo-DIY
Posts: 40
Joined: Mon Apr 04, 2011 7:56 am
Location: Germany
Contact:

Post by Photo-DIY »

Thank you Rik for your profund explanation!
It is the Mitutoye 20x Plan Apo SL lens with a 90mm Sigma macro lens as tube lens, used on a Canon Crop DSLR.
Need to upload the image - but probably better in another thread?
I have used the Helicon software :-/
Photo Tinkerers Heaven and Hell:
http://www.4photos.de/index-en.html
Some own work - and a bigger list of DIY links

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Photo-DIY wrote:Need to upload the image - but probably better in another thread?
I have used the Helicon software :-/
Yes, definitely better in another thread.

The basic issues are the same for Helicon Focus and Zerene Stacker. The default methods and parameter settings are different, and Zerene Stacker tends to be more aggressive about adjusting scale.

--Rik

TheLostVertex
Posts: 317
Joined: Thu Sep 22, 2011 9:55 am
Location: Florida

Post by TheLostVertex »

rjlittlefield wrote: Adding a diaphragm at the flange distance will make any lens telecentric, but only when the lens itself is set to infinity focus. That combination places the diaphragm at the rear focal point of the lens, which is one of the criteria to be telecentric. The problem is that when you refocus the lens in front of the diaphragm, then the rear focal point moves, the diaphragm is no longer located at it, and the combo goes non-telecentric.
I am a little bit unclear. When you say flange distance you are referring to the flange focal distance of the lens? Or do you mean something else? I ask because when I use a lens on bellows it would be useful to be able to make the lens telecentric in such a trivial manner. Setting the lens to infinity on bellows hardly makes much of a difference.

Also, when using a reversed lens, would that mean you can set an aperture at where the lens's focus point would be and it will be telecentric? When using a lens combo figuring out how to make it telecentric wouldnt be as simple as having the rear lens focus at the front lens's iris would it? I am guessing using things like the Raynox to make something telecentric is more of a guessing game.

rjlittlefield
Site Admin
Posts: 23563
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

TheLostVertex wrote:I am guessing using things like the Raynox to make something telecentric is more of a guessing game.
Not much guessing. A very sensitive test can be made by focusing binoculars at infinity, then using those to look into the front of the optics to see if the limiting aperture is sharply focused. If it is, then the aperture is virtually located at infinity, and that's one of the two tests for telecentricity. If it isn't, then you need to either move the aperture or change the separation of the lens combo. Either way, the binoculars can tell you which direction by observing when the aperture appears too close or too far away ("beyond infinity"). The only thing left to guesswork is how much things need to be moved to reach the telecentric condition. (You could measure that too, but it's more trouble than it's worth.)

Choosing an added lens to make a telecentric combo is also pretty cut-and-dried in concept. You look into the front of the lens to see where the aperture currently appears to be, decide where you would like to place the added lens, and specify the focal length of the added lens to be just the distance between where it goes and where the aperture appears to be without it. See http://www.photomacrography.net/forum/v ... php?t=1472.
Also, when using a reversed lens, would that mean you can set an aperture at where the lens's focus point would be and it will be telecentric?
Yes, it would! :D
I am a little bit unclear. When you say flange distance you are referring to the flange focal distance of the lens?
Yes, flange focal distance. The point is that when you focus a lens at infinity and place a plate with a small hole in the sensor plane, you have placed the hole where it needs to be to make a telecentric aperture. But in that configuration, it's not acting as an aperture; it's just acting as a mask. To make it act like an aperture, you have to then add extension, moving the lens and the hole as a unit away from the sensor, until the hole becomes the limiting aperture for light cones all over the sensor. Not by coincidence, that occurs at the same time that the setup ceases to show "mechanical" and "optical" vignetting, in the sense of http://wikipedia.org/wiki/Vignetting.
Or do you mean something else? I ask because when I use a lens on bellows it would be useful to be able to make the lens telecentric in such a trivial manner.
See http://www.photomacrography.net/forum/v ... =5609#5609 and the surrounding thread.

--Rik

Post Reply Previous topicNext topic