Attempt at stitching using a FF "telecentric" lens

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

Moderators: rjlittlefield, ChrisR, Chris S., Pau

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Attempt at stitching using a FF "telecentric" lens

Post by austrokiwi1 »

This is very much a work in progress post, Rik has been giving me some advice and Lou Jost some supporting information. Lou commented to me that he hoped I would post what I have been doing...so here is an early and very initial post. This is very much a work in progress post I have to purchase a focusing helicoid to improve the adjustment of the aperture ( at the end of the month) and I anticipate I will have to improve both my technique and understanding to get this working well. So far the results have shown promise.

By way of background: Some years back, in ignorant exuberance, I purchased a Schnieder Kreuznach 210/5.6 large format lens. I had, for photographing coins found that a 135mm lens was great so I guessed a 210mm lens would be even more effective. I quickly found out that the required extension and working distance made a 210mm lens unworkable on my vertical rig. I kept the lens with plans that one day I might get a Mitutoyo 10X Objective, and once in a while I have tried it out as a tube lens for reversed enlarger lenses (not an overly successful experience ). I collect coins and on occasion give presentations on coins. I am naturally reluctant to transport rare coins so have resorted to producing posters of rare coins so that participants ( of my presentations) can see the fine detail that isn't obvious in MS PowerPoint.

My main focus is on a series of coins that are all around 40mm in diameter. So to produce posters of those coins I either had to enlarge low magnification shots( .6X) or stitch a series of images taken at 1-1. With the desire to produce high detailed posters (and simply to see if I could do it) I have, over the last few years, tried different methods of creating large 1-1 mages of coins, never with satisfactory results.
Last year I finally started taking coin pictures using Capture-one (pro ), which is the only software ( that I am aware of) that tethers the A7rII. Late last year I realized that tethering might allow me to use that SK 210/5.6 effectively. From my previous experience I knew a telecentric lens would provide me with the best opportunity to succeed ( where I had failed in the past).

To make the lens telecentric I simply followed Riks instructions( which he has repeated in this forum a number of times) as best as I could. I purchased a couple of M42 extension tube sets, and a M42 aperture. I don't have binoculars or a small telescope so to begin with I just looked through the front element of the SK210 and moved the aperture until I could see that the blades were in focus. I then measured that distance and assembled the correct extension from the two extension ring sets, so I could mount the aperture in that position. Years ago I had had an adapter made that allowed an M52 reversing ring to the SK210, so it was a simple matter of getting a M52-M42 stepping ring for mounting the extension rings to the lens.
IN a discussion with Rik it is quite clear that this is not a well "tuned" telecentric rig. At the moment it is as close to telecentric as it is possible for me to make it.( I haven't yet tested it). The next step will be to replace 1-2 extension ring elements with a focusing helicoid. This would make it possible to adjust the position of the aperture with much greater accuracy(the helicoid is in next months budget). Once I have the helicoid I will then test it.. and when satisfied will use epoxy resin to fix the helicoid in the correct position
This is what it looks like set up on my rig( without the lap top connected to the camera)

Image

Its not obvious in the photo but the camera is mounted via two Swiss arca standard rails that allow the camera to be moved in the X-Y axis while being kept static in the Z Axis. Coin collectors will note that the lights are not in the ideal position however the honey comb, that I used to diffuse the light, does a reasonable job of making the light directionless. The light reflection on the wall behind the rig was caused by the flash used to take this photograph. the following photo is of an America Eagle silver coin that is encapsulated( in a Ultra-sonically sealed plastic enclosure). With considerable trial and error ( 6-7 attempts over the last week)I was able to produce the following picture. this picture is a stitch of 10 separate shots( at 1-1 magnification. I took the individual shots in 14 bit Raw, those pictures were converted to uncompressed Tiff. The final stitched picture(Tiff) is a horrifically huge 635mb. For posting here I converted it to Jpeg (103MB) and then resized it down to a much more manageable 25MB. Its not what I would regard as a final photo. if I were aiming to produce a poster I would have stitched and then stacked, the lighting would be very different to what was used here and the editing on the corners would have been much more accurate. This picture was constructed just to show me I was on the right track. What was personally satisfying is that even with pixel peaking I can not find where the pictures were stitched together ( even though I have a good idea where the "joins" are). Note: picture is in-bedded from Photobucket.

Image


Just in case any one is curious the full sized image's dimensions are 1.3m by 0.9m
Last edited by austrokiwi1 on Sat Feb 11, 2017 11:09 am, edited 7 times in total.
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

rockycarter
Posts: 80
Joined: Fri Jan 27, 2017 5:21 pm
Location: Canada

Post by rockycarter »

nicely done AK . great photo very good
Rocky Carter

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Glad you posted this! What software are you using to stitch? Nice result.

Do you have (or can you borrow) a second camera? That's all you need to check the aperture for proper placement to get telecentricity. Looking with your eyes is probably not too useful unless you are very far away from the front of the lens. With a camera, just set the lens to infinity and then aim it at the aperture and move the aperture until it is in focus in the viewfinder of the second camera.

For an approach using an added front close-up lens instead of an added aperture, see here (scroll down past the early entries):
http://www.photomacrography.net/forum/v ... highlight=

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Post by austrokiwi1 »

Lou Jost wrote:Glad you posted this! What software are you using to stitch? Nice result.

Do you have (or can you borrow) a second camera? That's all you need to check the aperture for proper placement to get telecentricity. Looking with your eyes is probably not too useful unless you are very far away from the front of the lens. With a camera, just set the lens to infinity and then aim it at the aperture and move the aperture until it is in focus in the viewfinder of the second camera.

For an approach using an added front close-up lens instead of an added aperture, see here (scroll down past the early entries):
http://www.photomacrography.net/forum/v ... highlight=
I stitched in Photoshop elements 14 using the photomerge panorama option.

Once I have the focusing helicoid, as you have suggested, I will use my MFT camera to set the aperture in the correct position.
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

keks
Posts: 44
Joined: Sun Mar 13, 2011 10:41 am
Location: Austria

Post by keks »

I'm definitely not an expert with optics, so I'm wondering why a telecentric setup would be necessary in case of a relatively flat coin?

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

keks wrote:I'm wondering why a telecentric setup would be necessary in case of a relatively flat coin?
I haven't seen any of austrokiwi's images, but based on the nature of his subject I'm thinking that it has to do with sheen.

Using an ordinary lens, the line of sight tips outward from the entrance pupil toward the edges of each frame. This means that as you cross a stitch line, there's a sudden change in the angle of the line of sight. With a shiny subject, that can mean a sudden change in the reflection and thus in the apparent brightness of the coin across the stitch line.

But with a telecentric lens, the lines of sight are always parallel to the optical axis. In that case, as you cross a stitch line there's no change at all in the line of sight, so little or no change in the reflection.

It will be no change at all if the coin does not move with respect to its environment, and nothing moves in the environment that gets reflected. If either of those moves does happen, then there will be some change in the reflection, but with a bit of luck, less than if the line of sight changed.

I will be interested to hear if this explains austrokiwi's difficulties, or if something else is going on.

--Rik

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Post by austrokiwi1 »

But with a telecentric lens, the lines of sight are always parallel to the optical axis. In that case, as you cross a stitch line there's no change at all in the line of sight, so little or no change in the reflection.

It will be no change at all if the coin does not move with respect to its environment, and nothing moves in the environment that gets reflected. If either of those moves does happen, then there will be some change in the reflection, but with a bit of luck, less than if the line of sight changed.

I will be interested to hear if this explains austrokiwi's difficulties, or if something else is going on.
First time I have been given this explanation, which really does enlighten me as to my previous problems. To be clear: yes it does match and explain my previous difficulties. I recall in previous attempts( some years back) I kept the optical system still and moved the subject. That was a huge fail! I later tried different methods where by the subject remained still and the optical system was shifted in the X'Y axis.. however there was only a marginal improvement. I have long ago deleted the pictures I produced but I recall that the worst attempts showed light and dark tiling, which clearly matches Riks hypothesis. I also recall I had to always apply the photoshop "Geometric distortion correction" function when stitching( with sometimes very bizarre results) With the current set up I found the stitching ( in all attempts over the last week) worked best without that function applied.

I earlier posted that with pixel peeping I couldn't see the stitch points. I have subsequently found that in one point in the stitched photo you can see the join ( bottom left section of the coin around the date. I went back to the original two images that were used for that section of the image and discovered further confirmation of what Rik has suggested: As I noted earlier, the lighting set up was not ideal and certainly not usual for imaging such a coin. The coin its self being highly reflective an issue and is compounded by the plastic acrylic case. Around the "2016" the sttch line can be seen because there is some"milkiness" due to light reflection from the coins plastic covering( it looks soft but in the original images it is clear it is unwanted reflection)


I have a better light set up that works best for encapsulated coins... and I believe I can eliminate that problem relatively easily. I picked the slabbed coin as a test target because I knew it was a difficult subject for stitching.

Once I have managed to tune the telecentric adapter properly I will up date on progress ( late next month I hope). This technique , if it works well, would only be used for just a few rare coins.. the work flow involved in long and complicated so it , in my view, is only worth using for the rarest subjects. I would be intrigued to see how it would work with a butterfly(may be in April)
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Nice work and very interesting topic. Curious about your lighting setup with the "diffuser" panels on each side, what are these?

Thanks,

Mike

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Post by austrokiwi1 »

Thanks for the positive encouragement. I understand you are asking about the diffusers shown in the picture. They are two pieces of 180mm square aluminum Honeycomb. Previous to finding those the only examples of honey comb I had seen were painted black. I discovered these unpainted sheets ideal for coins. Despite the size they are extremely light weight. I used DAS (a form of modeling clay) to make supports.

I found them on Ebay USA, I just tried to find them again... but with no joy.. i recall it took me some time to find this type of honeycomb.

Edit: Found it http://www.ebay.com/itm/60-Aluminum-Hon ... SwImRYLKPc
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

austrokiwi1,

Thank you. Do you think the smaller grid size, or the 60 degree type is better for your coin use?

Best,

Mike

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Post by austrokiwi1 »

I think the smaller grid size is better, however I should clarify how I use the honey comb. I position the lights and the grid so that the light does not have a direct path through the holes. The diffusion comes from the light being reflected through the grid. I use the 28 degree I have no idea if it is better or worse than the 60 degree
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

I am involved with imaging Si chips, some have many solder balls (40~100um dia) that act like tiny spherical mirrors. Lighting has proved to be very difficult and must be absolutely uniform. My solution is a double light tent, one inside the other and multiple (up to 12) highly double diffused strobes. Needless to said this is physically large and cumbersome! Thus my interest in your lighting approach and grids.

Best,

Mike

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Post by austrokiwi1 »

an update: I have consulted with rik a couple of times.. purchased first one then another extension ring set. Rik had asked me to send him privately copies of test shots for him to look at and advise me on what I was doing.. but I am not there yet. To be honest It is now only stubbornness that keeps me going. I now have over 200 mm extension between the rear element of the SK 210 and the aperture. With extension to get just under one to one the lens is only just manageable on my rig. My main issue is trying to position the aperture correctly ( and I suspect) If any one can be so kind as to post pictures of how they do it may be I will see what I am doing wrong) I may need just a bit more extension( its getting a little ridiculous). MY latest try is clearly much better but its not as accurate as int could be. from today's try here is the relevant line from console log . (I am using a full frame camera)

197.127: Merging, finalFit = RegistrationParameters: xoffset = -0.002166826942346378, yoffset = 0.010158024063645799, scale = 1.0011513202036495, rotate = 7.382962775190904E-4.
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

I'm curious, what made you decide to use pure extension on such a long lens, instead of coupled lenses and a telecentric stop between them?

I can place the iris with great accuracy by mounting step-down and step-up rings on the inside threads of other step-down rings:

http://www.photomacrography.net/forum/v ... highlight=

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Post by austrokiwi1 »

Lou Now its just pure stubbornness to get this method to work. The 210 is living upp to its previous frustration... its just too long a focal length. It would work well if I had a 2m high column. I have a very good Minolta lens (MD 135 F2) that also has a 72mm diameter front element. I will hunt round to see if I have the adapters to mount the M42 extension rings to it ( I think I might be able to do something with the front section of an old Minolta Sr extension ring set( has a 45mm thread( I might even have the right M45 -m42 stepping ring adapter) I will later try it your way with lens coupling( I want to try both methods in the end)

Edit: May be I shouldn't give up quite yet. Adjusting the aperture in increments to the furthest possible distance I performed another test:

191.672: Merging, finalFit = RegistrationParameters: xoffset = 0.0030993917129252258, yoffset = -0.006871362792917224, scale = 1.0010824616832428, rotate = -4.2331333772938073E-4

The result is not good but it shows me I need just slightly more extension. I am going to get one more extension ring set before giving up
Last edited by austrokiwi1 on Fri Apr 21, 2017 11:15 am, edited 1 time in total.
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

Post Reply Previous topicNext topic