Lens for increasing magnification on Photron Fastcam SA-X2
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Lens for increasing magnification on Photron Fastcam SA-X2
Hi,
I am new to this forum and the reason I am here is that I need some advice on an experimental setup.
The setup is as follows:
I am using a Photron Fastcam SA-X2 right now in combination with a Sigma APO Macro DG HSM 150mm. With this setup I can achieve a 1:1 magnification. This is not good enough, though. Extension tubes will not work, since I cannot move the camera closer to the object I am filming (working distance around 200mm).
My question is: What would be a way to go to let's say 2x-10x magnification without loosing too much light. I am ordering the Sigma 2x Teleconverter, but am afraid, that the picture will be too dark, since I loose about two stops of light.
The budget for this is flexible, so even expensive solutions will be considered.
I am new to this forum and the reason I am here is that I need some advice on an experimental setup.
The setup is as follows:
I am using a Photron Fastcam SA-X2 right now in combination with a Sigma APO Macro DG HSM 150mm. With this setup I can achieve a 1:1 magnification. This is not good enough, though. Extension tubes will not work, since I cannot move the camera closer to the object I am filming (working distance around 200mm).
My question is: What would be a way to go to let's say 2x-10x magnification without loosing too much light. I am ordering the Sigma 2x Teleconverter, but am afraid, that the picture will be too dark, since I loose about two stops of light.
The budget for this is flexible, so even expensive solutions will be considered.
- enricosavazzi
- Posts: 1479
- Joined: Sat Nov 21, 2009 2:41 pm
- Location: Västerås, Sweden
- Contact:
Re: Lens for increasing magnification on Photron Fastcam SA-
This is correct, you will lose two stops of exposure with a 2x teleconverter. In practice you can compensate for this by increasing fourfold the intensity of the subject illumination, or (if using a continuous illumination) with a fourfold increase in exposure time.Sebs wrote:[...]I am ordering the Sigma 2x Teleconverter, but am afraid, that the picture will be too dark, since I loose about two stops of light.[...]
Whether the image quality remains acceptable depends on the lens quality and your demands on image resolution. A 2x teleconverter works by taking the center area of the image and magnifying it by 2x. Lens aberrations are likewise magnified, so a lens that will give you e.g. a resolution of 100 line-pairs per mm on the image, after adding the teleconverter, will give you at most 50 line-pairs per mm. This assuming that the teleconverter adds only a neglectable amount of aberrations of its own compared to the lens' aberrations, which is generally true for good teleconverters.
The Sigma 150 mm macro is a good lens, so unless you need literally a pixel-peeping level of detail, image quality may remain sufficiently high even with the teleconverter. If you do get to pixel-peeping on e.g. a 32 Mpixel APS-C sensor, then you will see a slight image degradation with the teleconverter.
--ES
The problem is that I illuminate the subject with a high speed laser (and expanding optics) and although it is already high power $100k laser, I am struggling with image intensity. That's why I am wondering, if there is a better way to achieve higher magnification levels without sacrificing light.
I am also thinking about buying the Nikon 200mm f/4 AF-D Macro lens and combining that with some other optics to achieve some more magnification.
I am also thinking about buying the Nikon 200mm f/4 AF-D Macro lens and combining that with some other optics to achieve some more magnification.
- rjlittlefield
- Site Admin
- Posts: 23626
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Let's quickly review the physical limitations: To a good approximation, if your lens aperture is wide open then all the light that enters the lens does get transmitted through to the sensor. Adding the teleconverter does not actually lose light, it just spreads the same light over an area that is 2 times larger on each axis. This does make the image less bright. Unfortunately, any other method of increasing magnification will have the same problem of spreading the available light over a larger area of the sensor.
So, to get more magnification while still maintaining the same image brightness, you have to either direct more light onto the subject, or collect the light through a lens that has a wider entrance cone, for example, a lens that provides a larger entrance pupil at the same distance.
I assume you're running the Sigma 150 f/2.8 wide open. As a rough guestimate, that's giving you a 150/2.8 = 54 mm aperture, at your minimum working distance of 200 mm. It's not immediately apparent that switching to a 200 mm f/4 = 50 mm aperture is going to give much improvement.
If you're not running the 150/2.8 wide open, say because you need some specific depth of field, then your problem gets even harder. That's because depth of field tracks the angular width of the entrance cone, so if you're limited by depth of field then that also limits the width of the entrance cone, which limits how much light you can collect. You might be able to recover more DOF by combining a wide aperture with fancy wavefront coding techniques, but I have no idea what equipment is currently available to do that, or whether the reconstruction process could tolerate the noise produced by a high speed camera.
Another possibility is to add an electronic image intensifier relay system between the lens and the camera. Something like https://stanfordcomputeroptics.com/prod ... odule.html . I don't know anything about the speed or resolution of those systems, other than what's stated in the sales literature.
I wish I could be more optimistic. If anybody else has a known-good solution, I will be interested to hear it.
--Rik
So, to get more magnification while still maintaining the same image brightness, you have to either direct more light onto the subject, or collect the light through a lens that has a wider entrance cone, for example, a lens that provides a larger entrance pupil at the same distance.
I assume you're running the Sigma 150 f/2.8 wide open. As a rough guestimate, that's giving you a 150/2.8 = 54 mm aperture, at your minimum working distance of 200 mm. It's not immediately apparent that switching to a 200 mm f/4 = 50 mm aperture is going to give much improvement.
If you're not running the 150/2.8 wide open, say because you need some specific depth of field, then your problem gets even harder. That's because depth of field tracks the angular width of the entrance cone, so if you're limited by depth of field then that also limits the width of the entrance cone, which limits how much light you can collect. You might be able to recover more DOF by combining a wide aperture with fancy wavefront coding techniques, but I have no idea what equipment is currently available to do that, or whether the reconstruction process could tolerate the noise produced by a high speed camera.
Another possibility is to add an electronic image intensifier relay system between the lens and the camera. Something like https://stanfordcomputeroptics.com/prod ... odule.html . I don't know anything about the speed or resolution of those systems, other than what's stated in the sales literature.
I wish I could be more optimistic. If anybody else has a known-good solution, I will be interested to hear it.
--Rik
- enricosavazzi
- Posts: 1479
- Joined: Sat Nov 21, 2009 2:41 pm
- Location: Västerås, Sweden
- Contact:
Looking more in detail at the camera specifications, the camera you selected has a resolution of only 1 Mpixel as far as I can see. This resolution is so low that the limitations on image resolution normally discussed in this forum (especially diffraction) are not, or only marginally, relevant to your equipment.
Sensor size is somewhat intermediate between APS-C and Micro 4/3, albeit in a square format. Pixel size is 20 micron, so about 4 to 6 times larger linear size (16 to 36 times the area) than in commonly used system cameras. This explains the high native ISO sensitivity, necessary for high-speed sequences. Global electronic shutter is also a must for this type of camera, and something many users of ordinary cameras are sorely missing.
It would help to know more of the subjects you need to image (size in mm of desired field of view, any constraints on DoF and distance between lens and subject). Supplementing the 150 mm Sigma lens with a second lens of shorter focal length mounted reversed at the front of the 150 mm might be a way to increase magnification without loosing too much effective aperture. This has the effect of reducing the combined focal length of the optical system.
I agree that using a 200 mm f/4 instead of the 150 mm f/2.8 would do nothing useful in your case except increasing the working distance to the subject. Electronic image intensifiers may not perform well for high-speed imaging, so probably not the way to go.
Sensor size is somewhat intermediate between APS-C and Micro 4/3, albeit in a square format. Pixel size is 20 micron, so about 4 to 6 times larger linear size (16 to 36 times the area) than in commonly used system cameras. This explains the high native ISO sensitivity, necessary for high-speed sequences. Global electronic shutter is also a must for this type of camera, and something many users of ordinary cameras are sorely missing.
It would help to know more of the subjects you need to image (size in mm of desired field of view, any constraints on DoF and distance between lens and subject). Supplementing the 150 mm Sigma lens with a second lens of shorter focal length mounted reversed at the front of the 150 mm might be a way to increase magnification without loosing too much effective aperture. This has the effect of reducing the combined focal length of the optical system.
I agree that using a 200 mm f/4 instead of the 150 mm f/2.8 would do nothing useful in your case except increasing the working distance to the subject. Electronic image intensifiers may not perform well for high-speed imaging, so probably not the way to go.
--ES
Reversed lenses mounted on the Sigma won't work, since they always have flange distances << 200mm, and that's where the subject should be.
Is your camera the monochromatic version?
You quote a fixed working distance of 200mm, but you say the CAMERA cannot be moved closer. So just to clarify, is 200mm the distance from the front of the lens to the subject, or is that the distance from the camera to the subject?
If your subject is monochromatic and the lens (but not the camera) can be closer than 200mm to the subject, then photolithography lenses become relevant, and they are very very fast. I am currently selling two of them. They are 5x lenses:
https://www.photomacrography.net/forum/ ... highlight=
Is your camera the monochromatic version?
You quote a fixed working distance of 200mm, but you say the CAMERA cannot be moved closer. So just to clarify, is 200mm the distance from the front of the lens to the subject, or is that the distance from the camera to the subject?
If your subject is monochromatic and the lens (but not the camera) can be closer than 200mm to the subject, then photolithography lenses become relevant, and they are very very fast. I am currently selling two of them. They are 5x lenses:
https://www.photomacrography.net/forum/ ... highlight=
Ok, so I am imaging microfluidic flows. My illumination source is a high speed laser and the camera I am using is the greyscale version (No color information needed - basically just how thick is the flow field).
The front of the lens cannot be closer to the subject than about 200mm.
I don't care too much about DOF, since the things I am imaging are mostly in one plane parallel to the sensor.
The things I am trying to look at is about 12mm x 12 mm big or 6mm x 6mm.
The front of the lens cannot be closer to the subject than about 200mm.
I don't care too much about DOF, since the things I am imaging are mostly in one plane parallel to the sensor.
The things I am trying to look at is about 12mm x 12 mm big or 6mm x 6mm.
The two photolithography lenses I am selling both have unusually large working distances at 5x. If pushed down to 2x, their working distances might reach 200mm. Depending on how quickly these lenses degrade as they are moved away from their design points, they might work for you. I'm in Amsterdam right now but will be back in the US Oct 19 and can check the working distances at 2x.
What color is your laser light wavelength?
Going back to the teleconverter solution: if you increase the magnification from 1x to 2x, the area to be illuminated could be reduced by the same factor. In that case you could use a narrower beam spreader and the gain in illumination would cancel out the light lost by teleconverters. The two effects would always cancel out, no matter how many teleconverters you put in series.
What color is your laser light wavelength?
Going back to the teleconverter solution: if you increase the magnification from 1x to 2x, the area to be illuminated could be reduced by the same factor. In that case you could use a narrower beam spreader and the gain in illumination would cancel out the light lost by teleconverters. The two effects would always cancel out, no matter how many teleconverters you put in series.
-
- Posts: 200
- Joined: Sat Dec 06, 2014 6:46 am
- Location: Allentown, PA, USA, Earth, etc.
- Contact:
Infinity Photo-Optical
Hi,
I think Lou's option sounds good, honestly, if the price is right.
If this is intended for published research or QA where repeatability is a concern, contact Infinity Photo-Optical, they make video lenses which are used specifically for this purpose and are specialists in long distance video microscopy. Their equipment is often cited in the literature for this sort of work. ( I have their K2-Distamax internal focus tube lens and some of their objectives - this platform is very flexible but costly. They have less expensive series which are more purpose built for a particular application. They are a great company to deal with and support what they sell. I can't say enough good things about them. )
I think Lou's option sounds good, honestly, if the price is right.
If this is intended for published research or QA where repeatability is a concern, contact Infinity Photo-Optical, they make video lenses which are used specifically for this purpose and are specialists in long distance video microscopy. Their equipment is often cited in the literature for this sort of work. ( I have their K2-Distamax internal focus tube lens and some of their objectives - this platform is very flexible but costly. They have less expensive series which are more purpose built for a particular application. They are a great company to deal with and support what they sell. I can't say enough good things about them. )
I just bought the Sigma teleconverter since it was cheap and will see how it goes. I also contacted Infinity Photo optical and see, which one of their products they would suggest. Thank you so much for your recommendations.
In fact this work will be published as part of my PhD thesis. That's the only reason I can afford such expensive camera, laser and optics ..
Edit: The laser wavelenght I am using is 532nm, but I am using a longpass filter - so all light above 532nm is recorded by the camera.
In fact this work will be published as part of my PhD thesis. That's the only reason I can afford such expensive camera, laser and optics ..
Edit: The laser wavelenght I am using is 532nm, but I am using a longpass filter - so all light above 532nm is recorded by the camera.
That's a good wavelength. Many older photolithography lenses, including one of the lenses that are for sale at the link I gave you, are optimized for the 546nm mercury line. That's plenty close enough to your wavelength, given the low resolution you'll be getting.
One other thing to consider is lens distorion. Used as directed, the photolithography lenses have essentially zero distortion, but when used far from their design point, that is probably not true. Ordinary lenses plus teleconverters are also likely to intriduce distortion. I assume you will have some kind of calibration program to eliminate that.
One other question: Could a relay lens be inserted into the system closer to the subject? I assume not, but I ask just to cover all bases.
One other thing to consider is lens distorion. Used as directed, the photolithography lenses have essentially zero distortion, but when used far from their design point, that is probably not true. Ordinary lenses plus teleconverters are also likely to intriduce distortion. I assume you will have some kind of calibration program to eliminate that.
One other question: Could a relay lens be inserted into the system closer to the subject? I assume not, but I ask just to cover all bases.
So it boils down to getting a higher magnification at a given working distance? The obvious route then would be getting a longer focal length, right? Either with a teleconverter or a different lens altogether. My first thought is that you should test your 150/2.8 plus 2x teleconverter. This results in a 300mm f/5.6 lens. If you find you need more magnification, you need a longer focal length, if you need/want more light gathering capability you need a faster lens. The 1 Mpixel resolution may give you the option of using lenses less optimised for macro than we are normally considering here, so maybe a manual focus 300/2.8 or even 400/2.8 with extension would work for you.