K
klaus harms
Guest
Zuerst einmal möchte ich Harry S. danken, der es mir gestattet hat, diesen Beitrag aus dem amerikanischen Nikoncafe hier im Nikon Fotografie-Forum zu posten!
Sorry, aber nachfolgend ist der Text aus gegebenem Anlass in englischer Sprache gehalten; ich habe auch auf eine Übersetzung verzichtet - vielleicht eine Aufgabe für jemanden aus dem Forum? -, um den Content ursprünglich zu erhalten.
So, here we go:
This post is very long, so I suggest printing it and sit down somwhere with a real coffee for reading it.
Martin ("kramp") asked me to describe my technique of combining several focal layers for shooting very tiny subjects.
In former days I did this with just one photo and a tremendously stopped-down lens – as expected, diffraction never allowed me to produce such sharp and detail-rich pics as the one shown in this original post.
The digital era changed this dramatically. Many of my colleagues tried to combine focal layers but whenever I looked at the results there was something wrong with the proportions. So I started to experiment until I reached my current state of skill (which by far is not the possible top end).
I have described this technique in DPR a while ago but have decided to create a more detailed tutorial this time because others might be interested as well.
Important: This technique works only with static subjects!
Prerequisites:
- Good macro optics to achieve a much more than life size magnification ratio.
- A focussing rail that allows precise or near precise vertical movements in hair's breadth increments
- A good lighting setup.
- Photoshop and a computer with sufficient RAM
- A graphic tablet for drawing masks might also be very convenient
- A somewhat more than very basic knowledge in PS.
The problem with very tiny (in this case) beetles is that the available DOF never completely covers the vertical extension of their often convex or even highly vaulted bodies. Extremely small apertures cause diffraction and actually undo the gain in DOF. The solution is to capture several images at different focal planes and combine them PS (or even use a software that does it automatically).
The Setup:
I use a wide range of macro optics depending on the subject size. For very small subjects (1-3mm body length) I use a Leitz Photar 25mm/2.5 attached to a Novoflex bellows (Balpro). The bellows is mounted to a repro-stand via a focussing rail. The focussing rail has a vertical focussing screw which allows more or less precise movements in hair's-breadth increments.
The camera is a D1X tethered to the PC with all settings usual for macro still work with long exposures (mirror lock-up, etc.).
Very important: You have to avoid anything that carries the prefix "Auto-", particularly White balance. I usually shoot in manual mode, with preset WB. You don't want to end up with different color casts on each individual layer.
In case of larger beetles I attach a 60mm Leitz to the bellows or even use the naked 60mm Micronikkor.
I could even make use of a 12.5mm Photar, but the working distance would become too short for efficient use of light.
Lighting:
This is one of the most crucial factors for good macro photography. I use two halogen light sources with fibre optics (two goose necks each). Two achieve the most even lighting I filter the light through two layers of diffusion foil (used by the movie people for lamps). This is vitally important to reproduce metallic colors. A blue filter in front of the light source gives it almost the exact color temp of flashlight.
Picture of lens and lighting:
One downside of this is that scattered light might hit the front element of the lens and cause unwanted flare. I have not tried it yet, but a set of small "gobos" might help.
Another problem is that in very glossy specimens the lens and camera are reflected on the specimen which results in a block blotch, which seems impossible to get rid of. A solution might be to increase the working distance, but that requires speciel optics (see links below). I have also seen setups with polarizers in front of the light source, but this might also eliminate desirable reflections.
Bottom: Many insect specimens in collections are mounted on pins or on pinned cards. In former times I just stuck the pins in plastazote or styro foam (see picture above). The downside of this is that this material is elastic and takes some time to settle. I often had some inexplicable blur because I did not recognize that the pin is still slightly moving. Now I use plastilin (bees wax works as well).
The procedure:
That's the easier part. Since DOF is not important you may set the lens to the peak performance. I don't go beyond f/8 with the Photar and f/11 with the 60 Micro. Just focus on the top of the specimen and shoot. When the image appears on the monitor, check if the top layer is really sharp (control through the view finder may be tricky).
Then turn the focussing screw for the next layer – this is a matter of experience, the larger the magnification, the shorter the turn. Check the next layer on the monitor. Toggle back and forth between the pics. There should be a tiny overlap in the focussed areas.
Important: Avoid to touch the camera when turning the screw. The straight x/y misalignments may be easily corrected in PS, but touching the camera often causes slight rotations which are much harder or even impossible to correct.
Combining the layers:
Depending on the vertical extension of your subject you may end up with 4 – 15 layers.
Open the top layer in PS. Open the next layer > Ctrl A > Ctrl C > close. Then just copy this file into the previous one. Layer appears as a separate layer above the background layer. Continue with this procedure until all layers are in one document.
* If you shoot in .tif, the document might become very large, so if you don't have a RAM monster, it's better to split the work flow and combine 4 layers, after flattening import the next 3 or 4.
Now comes the tricky part:
Alignment and resizing:
(As a preparation to speed your work, you should write an action set containing actions with transform commands for resizing. I have actions for all 0.1% increments between 99.9 and 99.0).
Deactivate all layers excpet the background layer and the next one. Lower opacity of layer 2 to about 50% to be able to see both. Correct wrong alignment with the move tool. By repeatedly deactivating and activating layer 2 you will immediately recognize that the specimen in layer 2 is somewhat larger (which is clear because you moved the lens closer to the subject). Now you have to transform layer 2 until it perfectly matches the background layer.
When this is does to your satisfaction, return to 100% opacity, add a layer mask and invert it. Then choose a white, moderately soft-edged brush and paint over the areas of layer 2 which are in focus. Actually I paint everything white that is not in focus on the background layer, because everything else becoming a bit sharper, makes subsequent alignment much easier.
All you have to do now is repeat this step one by one with each subsequent layer. You will recognize that you have to choose increasing amounts of resizing with each step.
When everything is done, flatten the layers and you have a pic that is sharp throughout the entire Z-axis.
Now comes the usual enhancement work flow, which I use to do after replacing the back-ground. This I usually do by a color channel mask. Only when I have not enough contrast between fore and background I use different methods.
I do levels adjustments with the final mask activated because then the histogram shows me only the information of the subject.
All insect pics (except the out door pics) in my gallery ahve been done using this technique:
http://www.pbase.com/rovebeetle/mostly_beetles
Additional information:
As mentioned above, it is sometimes useful to have longer working distance which is difficult to achieve at high magnification ratios.
There is a US company that offers custom solutions:
http://www.microptics-usa.com.
Among their products are some very useful items, like a long distance microscope, a fibreopic lamp that can produce flashes, an accentuator that allows vertical movements of the focussing rail controlled from the PC, a video camera attache to the viewfinder which transfers a magnified viewfinder image to the monitor, etc.
Also there are different kinds of software to automatize the combining process. There are freewares (like Z-combine; sorry, don't have a link ready) or a very expensive one (the best, however) like Automontage by syncroscopy:
http://www.syncroscopy.com/syncroscopy
Anyway, I hope this was informative. I am sure there are many issues which I forgot to mention, but some of them are difficult to explain by words and may be figured out by practise and eyperiment.
Any additional questions will be gladly answered and suggestions for improving this technique or the setup are highly appreciated.
Cheers
Harry S.
----
Einige sehr imposante Beispiele könnt Ihr auf DIESER SEITE bewundern.
Sorry, aber nachfolgend ist der Text aus gegebenem Anlass in englischer Sprache gehalten; ich habe auch auf eine Übersetzung verzichtet - vielleicht eine Aufgabe für jemanden aus dem Forum? -, um den Content ursprünglich zu erhalten.
So, here we go:
This post is very long, so I suggest printing it and sit down somwhere with a real coffee for reading it.
Martin ("kramp") asked me to describe my technique of combining several focal layers for shooting very tiny subjects.
In former days I did this with just one photo and a tremendously stopped-down lens – as expected, diffraction never allowed me to produce such sharp and detail-rich pics as the one shown in this original post.
The digital era changed this dramatically. Many of my colleagues tried to combine focal layers but whenever I looked at the results there was something wrong with the proportions. So I started to experiment until I reached my current state of skill (which by far is not the possible top end).
I have described this technique in DPR a while ago but have decided to create a more detailed tutorial this time because others might be interested as well.
Important: This technique works only with static subjects!
Prerequisites:
- Good macro optics to achieve a much more than life size magnification ratio.
- A focussing rail that allows precise or near precise vertical movements in hair's breadth increments
- A good lighting setup.
- Photoshop and a computer with sufficient RAM
- A graphic tablet for drawing masks might also be very convenient
- A somewhat more than very basic knowledge in PS.
The problem with very tiny (in this case) beetles is that the available DOF never completely covers the vertical extension of their often convex or even highly vaulted bodies. Extremely small apertures cause diffraction and actually undo the gain in DOF. The solution is to capture several images at different focal planes and combine them PS (or even use a software that does it automatically).
The Setup:
I use a wide range of macro optics depending on the subject size. For very small subjects (1-3mm body length) I use a Leitz Photar 25mm/2.5 attached to a Novoflex bellows (Balpro). The bellows is mounted to a repro-stand via a focussing rail. The focussing rail has a vertical focussing screw which allows more or less precise movements in hair's-breadth increments.
The camera is a D1X tethered to the PC with all settings usual for macro still work with long exposures (mirror lock-up, etc.).
Very important: You have to avoid anything that carries the prefix "Auto-", particularly White balance. I usually shoot in manual mode, with preset WB. You don't want to end up with different color casts on each individual layer.
In case of larger beetles I attach a 60mm Leitz to the bellows or even use the naked 60mm Micronikkor.
I could even make use of a 12.5mm Photar, but the working distance would become too short for efficient use of light.
Lighting:
This is one of the most crucial factors for good macro photography. I use two halogen light sources with fibre optics (two goose necks each). Two achieve the most even lighting I filter the light through two layers of diffusion foil (used by the movie people for lamps). This is vitally important to reproduce metallic colors. A blue filter in front of the light source gives it almost the exact color temp of flashlight.
Picture of lens and lighting:
One downside of this is that scattered light might hit the front element of the lens and cause unwanted flare. I have not tried it yet, but a set of small "gobos" might help.
Another problem is that in very glossy specimens the lens and camera are reflected on the specimen which results in a block blotch, which seems impossible to get rid of. A solution might be to increase the working distance, but that requires speciel optics (see links below). I have also seen setups with polarizers in front of the light source, but this might also eliminate desirable reflections.
Bottom: Many insect specimens in collections are mounted on pins or on pinned cards. In former times I just stuck the pins in plastazote or styro foam (see picture above). The downside of this is that this material is elastic and takes some time to settle. I often had some inexplicable blur because I did not recognize that the pin is still slightly moving. Now I use plastilin (bees wax works as well).
The procedure:
That's the easier part. Since DOF is not important you may set the lens to the peak performance. I don't go beyond f/8 with the Photar and f/11 with the 60 Micro. Just focus on the top of the specimen and shoot. When the image appears on the monitor, check if the top layer is really sharp (control through the view finder may be tricky).
Then turn the focussing screw for the next layer – this is a matter of experience, the larger the magnification, the shorter the turn. Check the next layer on the monitor. Toggle back and forth between the pics. There should be a tiny overlap in the focussed areas.
Important: Avoid to touch the camera when turning the screw. The straight x/y misalignments may be easily corrected in PS, but touching the camera often causes slight rotations which are much harder or even impossible to correct.
Combining the layers:
Depending on the vertical extension of your subject you may end up with 4 – 15 layers.
Open the top layer in PS. Open the next layer > Ctrl A > Ctrl C > close. Then just copy this file into the previous one. Layer appears as a separate layer above the background layer. Continue with this procedure until all layers are in one document.
* If you shoot in .tif, the document might become very large, so if you don't have a RAM monster, it's better to split the work flow and combine 4 layers, after flattening import the next 3 or 4.
Now comes the tricky part:
Alignment and resizing:
(As a preparation to speed your work, you should write an action set containing actions with transform commands for resizing. I have actions for all 0.1% increments between 99.9 and 99.0).
Deactivate all layers excpet the background layer and the next one. Lower opacity of layer 2 to about 50% to be able to see both. Correct wrong alignment with the move tool. By repeatedly deactivating and activating layer 2 you will immediately recognize that the specimen in layer 2 is somewhat larger (which is clear because you moved the lens closer to the subject). Now you have to transform layer 2 until it perfectly matches the background layer.
When this is does to your satisfaction, return to 100% opacity, add a layer mask and invert it. Then choose a white, moderately soft-edged brush and paint over the areas of layer 2 which are in focus. Actually I paint everything white that is not in focus on the background layer, because everything else becoming a bit sharper, makes subsequent alignment much easier.
All you have to do now is repeat this step one by one with each subsequent layer. You will recognize that you have to choose increasing amounts of resizing with each step.
When everything is done, flatten the layers and you have a pic that is sharp throughout the entire Z-axis.
Now comes the usual enhancement work flow, which I use to do after replacing the back-ground. This I usually do by a color channel mask. Only when I have not enough contrast between fore and background I use different methods.
I do levels adjustments with the final mask activated because then the histogram shows me only the information of the subject.
All insect pics (except the out door pics) in my gallery ahve been done using this technique:
http://www.pbase.com/rovebeetle/mostly_beetles
Additional information:
As mentioned above, it is sometimes useful to have longer working distance which is difficult to achieve at high magnification ratios.
There is a US company that offers custom solutions:
http://www.microptics-usa.com.
Among their products are some very useful items, like a long distance microscope, a fibreopic lamp that can produce flashes, an accentuator that allows vertical movements of the focussing rail controlled from the PC, a video camera attache to the viewfinder which transfers a magnified viewfinder image to the monitor, etc.
Also there are different kinds of software to automatize the combining process. There are freewares (like Z-combine; sorry, don't have a link ready) or a very expensive one (the best, however) like Automontage by syncroscopy:
http://www.syncroscopy.com/syncroscopy
Anyway, I hope this was informative. I am sure there are many issues which I forgot to mention, but some of them are difficult to explain by words and may be figured out by practise and eyperiment.
Any additional questions will be gladly answered and suggestions for improving this technique or the setup are highly appreciated.
Cheers
Harry S.
----
Einige sehr imposante Beispiele könnt Ihr auf DIESER SEITE bewundern.
.