Hello Guys and Girls,
I am wondering if anyone has experience of using a microscope camera and how it compares to the DSLR?
Due to the unsharp photos above 25X magnification with my Leica Z16 I am wondering if this can be improved by using a microscope camera.
With the smaller camera chip the camera and prime lens would be close resulting in an increase of mag. How would the quality be though?
Regards,
Bob
Canon EOS D50 and Cognisys Stackshot. Z16 APo with 2x prime lens
DSLR v microscope camera
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Sorry but I can't answer your question. BUT I use a Nikon D90 and can get sharp images way above 25x.
http://www.photomacrography.net/forum/v ... hp?t=16707
The left image was at 25x and the right image at 150x with the D90.
http://www.photomacrography.net/forum/v ... hp?t=16707
The left image was at 25x and the right image at 150x with the D90.
NU.
student of entomology
Quote – Holmes on ‘Entomology’
” I suppose you are an entomologist ? “
” Not quite so ambitious as that, sir. I should like to put my eyes on the individual entitled to that name.
No man can be truly called an entomologist,
sir; the subject is too vast for any single human intelligence to grasp.”
Oliver Wendell Holmes, Sr
The Poet at the Breakfast Table.
Nikon camera, lenses and objectives
Olympus microscope and objectives
student of entomology
Quote – Holmes on ‘Entomology’
” I suppose you are an entomologist ? “
” Not quite so ambitious as that, sir. I should like to put my eyes on the individual entitled to that name.
No man can be truly called an entomologist,
sir; the subject is too vast for any single human intelligence to grasp.”
Oliver Wendell Holmes, Sr
The Poet at the Breakfast Table.
Nikon camera, lenses and objectives
Olympus microscope and objectives
-
rjlittlefield
- Site Admin
- Posts: 23621
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
The sensor in a DSLR is higher resolution and has less pixel noise than any correspondingly priced microscope camera. Perfectly coupled, the DSLR will give better pictures. But it's very easy to mess up the coupling so that the image is degraded either by optical aberrations or by camera vibration being coupled into the scope.
Have you explored why you're getting unsharp photos above 25X? How do you have the camera coupled to the scope?
--Rik
Have you explored why you're getting unsharp photos above 25X? How do you have the camera coupled to the scope?
--Rik
-
Charles Krebs
- Posts: 5865
- Joined: Tue Aug 01, 2006 8:02 pm
- Location: Issaquah, WA USA
- Contact:
Bob,
Have you changed anything from your set-up discussed here?:
http://www.photomacrography.net/forum/v ... hp?t=15550
I contacted Leica after this earlier discussion and was able to obtain some good Z16 information that I believe might at least partially explain your dilemma.
The key things to look at here are the NA for the provided magnification. Since this chart represents using 10X eyepieces, you can divide the "Total Magnification" column by 10 to determine the actual magnification provided by the image forming optics, and the NA at that setting. (The "Y tube" adds 1.25X but that is unavoidable when using a viewing head).
If this were looked at in terms used for individual microscope objectives you have the equivalent of the following with the Z16 and 2X Apo (from the Leica chart):
2/0.046 (Resolution: 7.25micron)
2.5/0.054 (Resolution: 6.17micron)
4/0.082 (Resolution: 4.07micron)
10/0.158 (Resolution: 2.11micron)
20/0.214 (Resolution: 1.56micron)
These next values are fairly typical of a single microscope objective as used by many here:
2/0.055 (Resolution: 6.10micron)
2.5/0.07 (Resolution: 4.79micron)
4/0.16 (Resolution: 2.10micron)
10/0.28 (Resolution: 1.20micron)
20/0.42 (Resolution: 0.42micron)
If any of the 10X or 20X single objectives just listed were used for "direct projection" (1X) onto a sensor, as is often done here, the results would be clearly "sharper".
Much will depend on any "relay" optics used and thus the actual "on sensor" magnification. If you are using the 50D in "silent-mode live-view" you should not be incurring any camera induced vibration if continuous light were used. I would not expect to see a "sharpness" improvement by going to a dedicated microscope camera. If you are getting an actual 25X on camera sensor, at the highest possible NA specified above (0.224) it means that the smallest detail that can be resolved (1.49 micron as per Leica chart) will measure 37.25 micron on the sensor surface. This will not look extremely sharp.
Another way to look at this could be in terms of "effective aperture". If you are getting 25X on sensor with a 0.224 NA, you are working at an effective aperture of about f/56. Lots of diffraction resolution loss there.
Stereos and macroscopes are great for many uses, but for photography they are generally best used at a pretty modest "on sensor" magnification if the utmost resolution is desired. The Z16 with 2X objective, and 0.225 max NA is far better than the vast majority of this type of microscope in this regard. Many good stereos max out at an NA of 0.09. If it is possible to direct project onto the 50D sensor then it would be quite competitive with a compound microscope that were set up with, say, a 2X addition relay magnification to the cameras as is often the case.
Even after this lengthy answer I still wonder about your initial set-up shown in the earlier discussion. Sometimes it is difficult to determine if the image "softness" that is seen is attributable diffraction or vibration (or a combination of both). The horizontal set-up you showed appeared to be quite susceptible to vibration. Even though you were using flash it would probably be a good idea to provide multiple support locations for such a lengthy apparatus. Might make a difference.
Have you changed anything from your set-up discussed here?:
http://www.photomacrography.net/forum/v ... hp?t=15550
I contacted Leica after this earlier discussion and was able to obtain some good Z16 information that I believe might at least partially explain your dilemma.
The key things to look at here are the NA for the provided magnification. Since this chart represents using 10X eyepieces, you can divide the "Total Magnification" column by 10 to determine the actual magnification provided by the image forming optics, and the NA at that setting. (The "Y tube" adds 1.25X but that is unavoidable when using a viewing head).
If this were looked at in terms used for individual microscope objectives you have the equivalent of the following with the Z16 and 2X Apo (from the Leica chart):
2/0.046 (Resolution: 7.25micron)
2.5/0.054 (Resolution: 6.17micron)
4/0.082 (Resolution: 4.07micron)
10/0.158 (Resolution: 2.11micron)
20/0.214 (Resolution: 1.56micron)
These next values are fairly typical of a single microscope objective as used by many here:
2/0.055 (Resolution: 6.10micron)
2.5/0.07 (Resolution: 4.79micron)
4/0.16 (Resolution: 2.10micron)
10/0.28 (Resolution: 1.20micron)
20/0.42 (Resolution: 0.42micron)
If any of the 10X or 20X single objectives just listed were used for "direct projection" (1X) onto a sensor, as is often done here, the results would be clearly "sharper".
Much will depend on any "relay" optics used and thus the actual "on sensor" magnification. If you are using the 50D in "silent-mode live-view" you should not be incurring any camera induced vibration if continuous light were used. I would not expect to see a "sharpness" improvement by going to a dedicated microscope camera. If you are getting an actual 25X on camera sensor, at the highest possible NA specified above (0.224) it means that the smallest detail that can be resolved (1.49 micron as per Leica chart) will measure 37.25 micron on the sensor surface. This will not look extremely sharp.
Another way to look at this could be in terms of "effective aperture". If you are getting 25X on sensor with a 0.224 NA, you are working at an effective aperture of about f/56. Lots of diffraction resolution loss there.
Stereos and macroscopes are great for many uses, but for photography they are generally best used at a pretty modest "on sensor" magnification if the utmost resolution is desired. The Z16 with 2X objective, and 0.225 max NA is far better than the vast majority of this type of microscope in this regard. Many good stereos max out at an NA of 0.09. If it is possible to direct project onto the 50D sensor then it would be quite competitive with a compound microscope that were set up with, say, a 2X addition relay magnification to the cameras as is often the case.
Even after this lengthy answer I still wonder about your initial set-up shown in the earlier discussion. Sometimes it is difficult to determine if the image "softness" that is seen is attributable diffraction or vibration (or a combination of both). The horizontal set-up you showed appeared to be quite susceptible to vibration. Even though you were using flash it would probably be a good idea to provide multiple support locations for such a lengthy apparatus. Might make a difference.
DSLR v microscope camera
Hello NU, Wow, just been looking at your pictures and article on the Horse Fly head. They are an inspiration and speak volumes for the Nikon CF Plan objectives. I will continue my search for a 40x ELWD.
Rik. Thanks for the info about less noise and higher resolution on the DSRL.
Charles, What can I say. Thank you is so inadequate. You come up with the info allowing us to progress and achieve new heights, plus you provide photos to inspire us.
Many thanks for that useful chart with the NA and Resolution. That will be valuable in working out the necessary size of steps - below 25x!
Clearly, Nikon CF M Plan objectives have the edge every time. If only they were more readly available, however, will keep looking for a 40x ELWD.
Yes, I did alter my rig after your recent advice. Photos herewith.
Regards, Bob
Rik. Thanks for the info about less noise and higher resolution on the DSRL.
Charles, What can I say. Thank you is so inadequate. You come up with the info allowing us to progress and achieve new heights, plus you provide photos to inspire us.
Many thanks for that useful chart with the NA and Resolution. That will be valuable in working out the necessary size of steps - below 25x!
Clearly, Nikon CF M Plan objectives have the edge every time. If only they were more readly available, however, will keep looking for a 40x ELWD.
Yes, I did alter my rig after your recent advice. Photos herewith.
Regards, Bob
Hi Bob, I use a Leica MZ16 with a Leica DFC420C camera. There is a photo of it near the end of this thread: http://www.photomacrography.net/forum/v ... highlight=
The camera was ~$6,000 and in my opinion was way over priced. Hot or dead pixels seem to be a common problem with the dedicated microscope cameras. The camera sensor is on the whole time the camera is plugged in to the computer. With a DSLR the sensor is on only when the camera is activated.
I have used this system up to 160x, but even with APO objectives the resolution problematic above 80x.
My 2cents,
Steve
The camera was ~$6,000 and in my opinion was way over priced. Hot or dead pixels seem to be a common problem with the dedicated microscope cameras. The camera sensor is on the whole time the camera is plugged in to the computer. With a DSLR the sensor is on only when the camera is activated.
I have used this system up to 160x, but even with APO objectives the resolution problematic above 80x.
My 2cents,
Steve
"You can't build a time machine without weird optics"
Steve Valley - Albany, Oregon
Steve Valley - Albany, Oregon
DSLR v Microscope Camera
Hello Steve,
Thanks so much for your reply. Good to know that some one else is using a Z16, even though in my case it's only the optical system.
Marvellous set-up you've got there with two macro systems. I see you're using your studio flashes with brollies. Am using an Einstein 650 with just a polystyrene cup diffuser.
Thanks for the tips on the DFC 420C I will bear that in mind. Only reason I'm enquiring about the microscope cameras is that I figure with a smaller chip I could mount it nearer the prime lens and thus increase my magnification.
Realise the best option would be a Nikon M CFI 60x but they're a bit like gold dust.
Best Wishes,
Bob
Thanks so much for your reply. Good to know that some one else is using a Z16, even though in my case it's only the optical system.
Marvellous set-up you've got there with two macro systems. I see you're using your studio flashes with brollies. Am using an Einstein 650 with just a polystyrene cup diffuser.
Thanks for the tips on the DFC 420C I will bear that in mind. Only reason I'm enquiring about the microscope cameras is that I figure with a smaller chip I could mount it nearer the prime lens and thus increase my magnification.
Realise the best option would be a Nikon M CFI 60x but they're a bit like gold dust.
Best Wishes,
Bob
DSLR vs Microscope Camera
Hello,
I'll add my own stone to that discussion since your problematic seems some way similar to mine.
1) Setup & purpose :
Our optical lab is used for taking pictures of electronics and mechanical parts during failure analysis or inspection.
It is composed of :
2) Dead/stuck pixels
We have also this problem, mainly with the MZ16.
We got it with both the DFC295 and the previous camera (a DFC290) that failed due to a mechanical weakness of its connector (now resolved on the DFC295 through a built-in screw fastening)
As I will state it below, I am trying to use a canon EOS 550D in place of the DFC295 or as an alternative : While it worked flawlessly for months being used as a regular DSLR, the same type of dead/stuck pixels appeared on its sensor after I used it with the MZ16.
I suppose that the root cause is tightly linked to the 0/100% optical coupler, which alternatively directs the whole left light path towards either the left eyepiece or the camera:
When the camera is unlit while turned ON, it certainly sets an analogue gain to its maximum; then it is suddenly hit by an high light flux (the cold light sources, be it halogen or LED, are very powerful and the resulting light density is rather high) => I suppose that it is this way that the pixels fail : because of the high light that hits the sensor before the gain is automatically adjusted; I think that some pixels get oversaturated (I think that it must generate an overvoltage in the capacitor of the sensor cell that destroys it).
3) Using a DSLR in place of a dedicated digital camera
The cost of the dedicated cameras, the dead pixels occuring rather quickly and the large pixel count gap made me think about this for a few years; thought, as I stated it above, this acquisition system has simplicity, precise dimensional measurements and integration needs that made an integrated live acquisition (and not image download in a second time) and image calibration mandatory.
Another point that made the live driver mandatory is that though the daily use will certainly not be 18Mpixel images (rather 3-5Mpixel in order to keep decent image byte weight, especially when you take a few dozens a day), but when it is useful, focusing sharply a 18Mpixel image using the integrated display of the camera or even a fullHD monitor used via an HDMI link, is definitely not easy, while these acquisition systems have dedicated focus zoom functions far more efficient.
Here comes the difficulty: Hardly no software available can do this!
After a long research and numerous demonstrations/trials made with all the softwares that I found around the web, I finally found the only one that (as far as I know) proposes it without any custom development : Imagic Image Access (for our need, we are going to buy the Easylab version that has database/server for remote clients capabilities) with its Canon EOS capture module. The demo is not that old : we had it yesterday afternoon and it worked : the driver perfectly connected the EOS 550D and performed the acquisition
The "funny" thing is that it appears to be the successor of the IM1000 we had since 10+ years (it was here before me ) which was a rebranded / customized version of this software!
The second part of using a DSLR in place of a dedicated camera is the optical adaptation, which was not obvious (Leica was no more really pushing towards it even if I recently learned from them that they have the right adapters at rather competitive prices : around 1K€ for the one that plugs into the trinocular of our MZ16), so I turned towards an independent optics manufacturer specialized in DSLR adaptation on microscope, that proposes more or less equivalent products with similar prices; I had an optical coupling issue (spherical image blur), that should be soon resolved thanks to the kind help of members of this Forum
The Third and last point that is not technical is an assets security issue : this lab is intended to be used by several people and hence in free access ... some times you don't find all you left the previous day
... I hope that the kensington-like anti-theft cable will be sufficiently dissuading when people will see a DSLR body, be it without optics 
From a financial point of view, the choice between a 3K€ 3Mpixel dedicated camera with a ~1K€ software vs a ~3K€ software + 1K€ optical adapter + ~500€ 18Mpixel DSLR body is rather equivalent ... but when the sensor gets out of order you only pay a new 500€ body and you get 18Mpixel instead of 3Mpixel!
From a sensitivity / functionality point of view, dedicated cameras were long advertised as far more sensitive than DSLR, but at least in the case of the "low-end" DFC295 (CMOS not cooled), the sensitivity match is largely won by the EOS 550D:
The DFC has a easily noticeable noise at low light levels (though acquiring images of lit specimen) while I replaced it by the EOS and took a look at the sensor of the DFC in a medium lit room, with no additional light turned on => the sensor which is located at the bottom of a shape of cylinder 1cm deep was clearly visible (once again with NO cold light ON), without too munch noise (obviously the exposure time, left in auto mode, was around 1 sec).
A little technical problem we could have if the DSLR was adapted onto the MZ16 is that we have a adjustable observation angle (ergo) module that is disturbed by the weight of the optical adapter/DSLR that plug into the trinocular on its top (the whole tilts forward) ... Fortunately the primary need is on the DMLM microscope that does not have this problem 
So, after some 2-3 years of investigation, I am finally just a few weeks away from replacing the old VGA-like DC180 / IM1000 by the EOS 550D + Imagic Image Access Easylab with the Canon EOS capture module
All the parts were tested separately and seem to work (except the optical adapter issue that should be OK with a 1X C-mount instead of the present 0.5X); if it works well all together, I'll see in a second time if I do the same with the MZ16, but as long as the DFC295 does correctly the job, the need is less critical.
I'll tell & show you the result when this is be bought/installed.
I'll add my own stone to that discussion since your problematic seems some way similar to mine.
1) Setup & purpose :
Our optical lab is used for taking pictures of electronics and mechanical parts during failure analysis or inspection.
It is composed of :
- An MZ16 binocular with various objectives (the most used being a 0.5X PLAN, which I prefered, due to its larger FOV, to the 0.63X PLANAPO which has better optical properties and is the MZ16 "standard" objective) + a Leica DFC295 camera connected through a 0/100% optical coupler + Leica Application Suite.
A DMLM microscope with 5X/10X/20X/50X objectives and an old DC180 + VGA camera connected through a Leica/Matrox proprietary acquisition card to the Leica IM1000 software (10+ years old)
A macro photo stand for uncalibrated wide views using the DSLR & its standard EF-S 18-55 IS
2) Dead/stuck pixels
We have also this problem, mainly with the MZ16.
We got it with both the DFC295 and the previous camera (a DFC290) that failed due to a mechanical weakness of its connector (now resolved on the DFC295 through a built-in screw fastening)
As I will state it below, I am trying to use a canon EOS 550D in place of the DFC295 or as an alternative : While it worked flawlessly for months being used as a regular DSLR, the same type of dead/stuck pixels appeared on its sensor after I used it with the MZ16.
I suppose that the root cause is tightly linked to the 0/100% optical coupler, which alternatively directs the whole left light path towards either the left eyepiece or the camera:
When the camera is unlit while turned ON, it certainly sets an analogue gain to its maximum; then it is suddenly hit by an high light flux (the cold light sources, be it halogen or LED, are very powerful and the resulting light density is rather high) => I suppose that it is this way that the pixels fail : because of the high light that hits the sensor before the gain is automatically adjusted; I think that some pixels get oversaturated (I think that it must generate an overvoltage in the capacitor of the sensor cell that destroys it).
3) Using a DSLR in place of a dedicated digital camera
The cost of the dedicated cameras, the dead pixels occuring rather quickly and the large pixel count gap made me think about this for a few years; thought, as I stated it above, this acquisition system has simplicity, precise dimensional measurements and integration needs that made an integrated live acquisition (and not image download in a second time) and image calibration mandatory.
Another point that made the live driver mandatory is that though the daily use will certainly not be 18Mpixel images (rather 3-5Mpixel in order to keep decent image byte weight, especially when you take a few dozens a day), but when it is useful, focusing sharply a 18Mpixel image using the integrated display of the camera or even a fullHD monitor used via an HDMI link, is definitely not easy, while these acquisition systems have dedicated focus zoom functions far more efficient.
Here comes the difficulty: Hardly no software available can do this!
After a long research and numerous demonstrations/trials made with all the softwares that I found around the web, I finally found the only one that (as far as I know) proposes it without any custom development : Imagic Image Access (for our need, we are going to buy the Easylab version that has database/server for remote clients capabilities) with its Canon EOS capture module. The demo is not that old : we had it yesterday afternoon and it worked : the driver perfectly connected the EOS 550D and performed the acquisition
The "funny" thing is that it appears to be the successor of the IM1000 we had since 10+ years (it was here before me
) which was a rebranded / customized version of this software!The second part of using a DSLR in place of a dedicated camera is the optical adaptation, which was not obvious (Leica was no more really pushing towards it even if I recently learned from them that they have the right adapters at rather competitive prices : around 1K€ for the one that plugs into the trinocular of our MZ16), so I turned towards an independent optics manufacturer specialized in DSLR adaptation on microscope, that proposes more or less equivalent products with similar prices; I had an optical coupling issue (spherical image blur), that should be soon resolved thanks to the kind help of members of this Forum
The Third and last point that is not technical is an assets security issue : this lab is intended to be used by several people and hence in free access ... some times you don't find all you left the previous day
... I hope that the kensington-like anti-theft cable will be sufficiently dissuading when people will see a DSLR body, be it without optics From a financial point of view, the choice between a 3K€ 3Mpixel dedicated camera with a ~1K€ software vs a ~3K€ software + 1K€ optical adapter + ~500€ 18Mpixel DSLR body is rather equivalent ... but when the sensor gets out of order you only pay a new 500€ body and you get 18Mpixel instead of 3Mpixel!
From a sensitivity / functionality point of view, dedicated cameras were long advertised as far more sensitive than DSLR, but at least in the case of the "low-end" DFC295 (CMOS not cooled), the sensitivity match is largely won by the EOS 550D:
The DFC has a easily noticeable noise at low light levels (though acquiring images of lit specimen) while I replaced it by the EOS and took a look at the sensor of the DFC in a medium lit room, with no additional light turned on => the sensor which is located at the bottom of a shape of cylinder 1cm deep was clearly visible (once again with NO cold light ON), without too munch noise (obviously the exposure time, left in auto mode, was around 1 sec).
A little technical problem we could have if the DSLR was adapted onto the MZ16 is that we have a adjustable observation angle (ergo) module that is disturbed by the weight of the optical adapter/DSLR that plug into the trinocular on its top (the whole tilts forward
) ... Fortunately the primary need is on the DMLM microscope that does not have this problem So, after some 2-3 years of investigation, I am finally just a few weeks away from replacing the old VGA-like DC180 / IM1000 by the EOS 550D + Imagic Image Access Easylab with the Canon EOS capture module
All the parts were tested separately and seem to work (except the optical adapter issue that should be OK with a 1X C-mount instead of the present 0.5X); if it works well all together, I'll see in a second time if I do the same with the MZ16, but as long as the DFC295 does correctly the job, the need is less critical.I'll tell & show you the result when this is be bought/installed.