I am new to microphotography and this site has been very helpful. I’ve just been putting together what I think is a pretty good kit but am questioning my initial results – they seem much less sharp than I would have expected. I would be interested in feedback as to whether my results are normal, and if not, suggestions to improve them.
My gear is as follows:
Olympus BHTU microscope with DPlan objectives
Trinocular head (BH2-TR30)
Olympus photomicro adapter L
Fotodiox Pro lens mount adaptor (Olympus to Canon lens adaptor)
Olympus 1.67X and 2.5X photo projection eyepieces
Canon 7D Mark II and 50-D cameras
Various botanical slides from Carolina Biological Supply
I believe I have centered my condenser, set my field iris diaphragm and aperture iris properly for each objective according to the BHTU manual instructions. I have even tried stopping down the aperture iris a bit to try and improve contrast.
I am shooting tethered and using live view on Canon EOS Utility; cameras are set to Silent Mode to use the electronic first shutter curtain. (The resulting images are exactly as they appear on my laptop prior to shooting, so I am quite sure there are no vibrations affecting my results.)
I have tried shooting with various professionally prepared slides, both photo projection eyepieces (1.67X and 2.5X), both cameras (50D and 7D2), and using 10X and 40X objectives. Results are all similar. My goal is to capture a stack of images to combine in Zerene Stacker. However, I don’t feel I can hit any sharp images/image parts as I move through the range using fine focus. My Live view image (and resulting photos) just move through a range of unfocused to “almost-focused” as I adjust the fine focus. I can’t seem to come up with anything sharp that would be useful in the focus stack. (When I look through the eyepieces everything looks fine/focused, but of course things are not zoomed in as much as they are on EOS Utility Live View.)
The attached images are typical of my results; this is the best general focus I can get. One image is the full sized image (downsized for this forum) and the other is cropped from the exact center of the image (with less downsizing needed). The downsizing did not change the quality of the image. This was shot on my 18 megapixel 7D2, using the 1.67X projection eyepiece and the 10X objective. This image was shot at ISO 100, shutter speed 200.
I have scoured this website for tips but don’t see that there any other variables to play with to try and improve my results. Any input and advice would be very much appreciated!
Full size image:
Central crop of first picture:
My focus seems inadequate - can it be improved?
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Hi fastbilly, welcome aboard!
I see nothing really wrong with your images, they are pretty typical
I've reprocessed them in PS adjusting levels, color and adding some sharpening and IMO they improve quite a bit (I can post them here if you want)
I see nothing really wrong with your images, they are pretty typical
I've reprocessed them in PS adjusting levels, color and adding some sharpening and IMO they improve quite a bit (I can post them here if you want)
Likely this is the key: microscope images typically are diffraction limited and except with the best optics at low magnification they doesn't look sharp at 100% pixel view. It's the typical frustration of photographers when first become microscopists. Microscopes are designed for resolution (detail) at the subject side, not to deliver the sharper image at the camera sensor.(When I look through the eyepieces everything looks fine/focused, but of course things are not zoomed in as much as they are on EOS Utility Live View.)
Pau
That image doesn't look too bad to me. This may help explain what you are seeing.
I'll assume your 10x lens has a numerical aperture of 0.25 (it will be marked 10/0.25 if that is the case). The smallest area that this objective can resolve (in microns) is roughly 0.55/(2*NA) = 0.55/0.5 = 1.1 microns. This is magnified 10x by the objective and a further 1.67x by the projection eyepiece. So the size of the smallest resolvable (sharp) spot projected onto the camera sensor is 1.1 * 10 * 1.67 = 18.37 microns. The pixels on your 7D MkII sensor are 4.1 microns. This means the smallest resolvable spot covers 18.37/4.1 = ~4.5 pixels. That is why it looks soft at 100% crop.
The situation will be worse for a 40/0.75 objective. There the resolveable spot is smaller (0.36 microns) but magnified more on sensor (to 24.5 microns) covering almost 6 pixels.
In short, your camera sensor is out-resolving the objectives. This is normal. Just resample the image down to half size or even quarter, then it will be sharp.
There could be some setup issues that will get you a *bit* more resolution if corrected, (e.g. distance from projection lens to sensor, or NA of the condenser), but I'll leave more experienced members to chime in on that.
I'll assume your 10x lens has a numerical aperture of 0.25 (it will be marked 10/0.25 if that is the case). The smallest area that this objective can resolve (in microns) is roughly 0.55/(2*NA) = 0.55/0.5 = 1.1 microns. This is magnified 10x by the objective and a further 1.67x by the projection eyepiece. So the size of the smallest resolvable (sharp) spot projected onto the camera sensor is 1.1 * 10 * 1.67 = 18.37 microns. The pixels on your 7D MkII sensor are 4.1 microns. This means the smallest resolvable spot covers 18.37/4.1 = ~4.5 pixels. That is why it looks soft at 100% crop.
The situation will be worse for a 40/0.75 objective. There the resolveable spot is smaller (0.36 microns) but magnified more on sensor (to 24.5 microns) covering almost 6 pixels.
In short, your camera sensor is out-resolving the objectives. This is normal. Just resample the image down to half size or even quarter, then it will be sharp.
There could be some setup issues that will get you a *bit* more resolution if corrected, (e.g. distance from projection lens to sensor, or NA of the condenser), but I'll leave more experienced members to chime in on that.
Pau is correct, nothing seriously wrong with your images.
Your equipment is very similar to mine, and your results are similar to mine.
I set white balance in EOS Utility, and use the histogram to make sure that the exposure is correct.
After using Zerene Stacker, I use Photoshop Elements to adjust levels and sharpen. I used Photoshop Elements on your photo, and it looked much better.
I can't see any sign of vibration spoiling your images. My microscope and computer are on a heavy desk. I sometimes have my mouse on a separate shelf so that I don't touch the desk during an exposure.
One thing to watch for with a BHTU - the glass between the nosepiece and the head can attract dust and might reduce contrast, so keep it clean.
You don't need to change the distance between the NFK projection lenses and the sensor. With a 2mm thick OM-EOS adapter the distance will be exactly as Olympus intended.
Alan Wood
Your equipment is very similar to mine, and your results are similar to mine.
I set white balance in EOS Utility, and use the histogram to make sure that the exposure is correct.
After using Zerene Stacker, I use Photoshop Elements to adjust levels and sharpen. I used Photoshop Elements on your photo, and it looked much better.
I can't see any sign of vibration spoiling your images. My microscope and computer are on a heavy desk. I sometimes have my mouse on a separate shelf so that I don't touch the desk during an exposure.
One thing to watch for with a BHTU - the glass between the nosepiece and the head can attract dust and might reduce contrast, so keep it clean.
You don't need to change the distance between the NFK projection lenses and the sensor. With a 2mm thick OM-EOS adapter the distance will be exactly as Olympus intended.
Alan Wood
- rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Another point about the same issue that Beatsy describes...
In addition to limiting your maximum resolution, diffraction causes contrast for fine detail to be cut way down. Even for features that are twice as large as the limit, diffraction will cut the contrast by more than 50% .
As a result, it is completely appropriate to sharpen microscope images quite heavily. You can, and should, use sharpening filters that are so strong they would make garbage of ordinary sharp photographs.
With a 10X objective in my setup (Canon T1i, 15 megapixels on APS-C), a Photoshop unsharp mask of 100% at 4 pixels would not be unusual. The best filter depends on NA, total magnification, and sensor resolution. As a matter of practice you'll have to play with it -- just recognize that the right amount is a lot, so don't be timid.
For some more formal discussion, see my post about sharpening versus deconvolution, HERE.
--Rik
In addition to limiting your maximum resolution, diffraction causes contrast for fine detail to be cut way down. Even for features that are twice as large as the limit, diffraction will cut the contrast by more than 50% .
As a result, it is completely appropriate to sharpen microscope images quite heavily. You can, and should, use sharpening filters that are so strong they would make garbage of ordinary sharp photographs.
With a 10X objective in my setup (Canon T1i, 15 megapixels on APS-C), a Photoshop unsharp mask of 100% at 4 pixels would not be unusual. The best filter depends on NA, total magnification, and sensor resolution. As a matter of practice you'll have to play with it -- just recognize that the right amount is a lot, so don't be timid.
For some more formal discussion, see my post about sharpening versus deconvolution, HERE.
--Rik