Object movie rig

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

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Graham Stabler
Posts: 209
Joined: Thu Dec 20, 2007 11:22 am
Location: Swindon, UK

Object movie rig

Post by Graham Stabler »

Following on from my previous thread:

http://www.photomacrography.net/forum/v ... php?t=3987

I'm now trying to get everything together to make my object movie rig.

So far I have a heavy base plate, a 630X400X100 granite surface plate:

http://www.cromwell.co.uk/OXD3063840K

To which I intend to bolt (with resin embedded stubs) a kaiser enlarger column:

http://www.kaiser-fototechnik.de/en/pro ... sp?nr=4408

I'm hoping this will give me the basis for a more than sturdy stand with plenty of vertical course adjustablity.

On the surface plate will sit the three axis stepper controlled rig, one vertical axis for stacking and two rotary. The rotary axes will have at least a 400 steps per revolution resolution and the linear axis 2.5um travel per step (1mm lead ballscrew).

I now need to buy a camera, I don't know if yo go Nikon or Canon and I'm not sure once the camera is bought if to go for bellows or a macro lens and some extension tubes if I want to go beyond 1:1.

Then there is the issue of lighting.

I'd appreciate any input.

Thanks

Graham

Harold Gough
Posts: 5786
Joined: Sun Mar 09, 2008 2:17 am
Location: Reading, Berkshire, England

Re: Object movie rig

Post by Harold Gough »

Graham Stabler wrote: I now need to buy a camera, I don't know if yo go Nikon or Canon and I'm not sure once the camera is bought if to go for bellows or a macro lens and some extension tubes if I want to go beyond 1:1.

Then there is the issue of lighting.

I'd appreciate any input.

Thanks

Graham

"Beyond 1:1" has an enormous scope. How far beyond?

The lighting requirement may depend are you answers to the first question.

Harold
My images are a medium for sharing some of my experiences: they are not me.

Graham Stabler
Posts: 209
Joined: Thu Dec 20, 2007 11:22 am
Location: Swindon, UK

Post by Graham Stabler »

I'm wanting to look at the thorax and components of the honey bee and blow-fly. The smallest part I would want to look at would be about 0.3mm in size but that is not to say it must fill the frame.

And I'm not adverse to multiple lens systems and lighting options.

Graham

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

Post by rjlittlefield »

Graham,

It sounds like you want to work in the size range that I've explored for the last few years.

After trying out a lot of combinations, I've become very fond of bellows and a fairly small collection of macro lenses and microscope objectives.

In my case, the bellows and macro lenses are from Olympus (ref). The Nikon CF N 10X NA 0.30 objective is superb -- search the forums for Nikon CF objective (all terms) for recent info about that. High quality enlarging lenses, reversed, are good substitutes for the macro lenses.

For illumination, I'm fond of dual fiber coming off a 150w halogen illuminator, lighting up a hemispherical diffuser (pingpong or whiffle ball, cut in half), as introduced by Charlie Krebs. This allows great freedom in positioning the light sources (fiber heads) to highlight whatever structures you want. Continuous illumination makes it easy to see lighting effects by eye while manipulating the heads.

However, this system also has the drawback that camera vibration becomes an issue at higher magnifications. I use mirror lockup, of course, but shutter vibration is still troublesome. I work around this with shutter times around 1 second, which with my mechanics are enough to make most of the exposure occur after the vibrations have died out. It's hard to predict exactly how camera-induced vibration will affect your setup.

Vibration becomes much less of an issue if you use flash illumination, particularly with rear-curtain sync and a fairly long shutter time. It's harder with flash to specialize the illumination for a particular subject, unless your flashes have good modeling lights, but a lot of good macro work has been done with basic two-flash setups and heavy diffusion. There are lots of tradeoffs -- I can't say which scheme would be better for your purposes. Both are viable.

The choice of best camera seems to change from week to week, and I'm not sure what's on top right now. As general guidelines, I think you'd prefer one that provides live view and full computerized control over the camera. I believe that Nikon, Canon, and Olympus all offer viable models with software available. I suspect you're interested in SDKs (software development kits) that will let you integrate with your own motor controllers. That's a separate issue that you'll need to look at specifically.

Be aware that there can be devils in the details. For example I vaguely recall reading in the Yahoo Microscope group that some early DSLRs that provided live view with magnified focusing and mirror lockup also had strange sequences like dropping and re-raising the mirror when an exposure was taken. This made the vibration problem worse than the authors had expected based on their reading of the camera specs and imagining how the camera would work. There can even be really basic issues like whether the camera's metering works properly for lenses on bellows.

Sorry I can't offer more definite information right now, but I'm really reluctant to recommend anything I haven't actually tried. That's because I get surprised by some devil every time I work with new equipment.

BTW, it would be a good idea to check the archives of the Yahoo Microscope Group with queries like live view.

Hope this is helpful. Let us know what info you find, as you go along. Most likely it will trigger more input from other people too.

--Rik

Graham Stabler
Posts: 209
Joined: Thu Dec 20, 2007 11:22 am
Location: Swindon, UK

Post by Graham Stabler »

Thanks Rik some very useful points.

I think bellows were what I was leaning towards because of the flexibility and the fact you can get some high magnifications, strangely finding suppliers is harder than I expected. Ideally I want bellows without the focusing rack as I will have course focusing on the column and fine on the stepper z-axis.

I searched for the Nikon objective as you suggested and am very impressed, in particular I found a great shot from Charles showing a single image from a stack of a fly's eye, it looked amazing. I'll definitely try and get hold of one.

Shutter vibration is something that Charles mentioned to me in a PM, I was asking him about DLSRs. My first question is why they don't do the shuttering electronically or provide the option? I'm sure there are very good reasons (dust perhaps one good one). I looked at getting a firewire type industrial camera as these have electronic shutters and I can get, say 8M pixels but the sensor is only 8.9X6.7mm, I need to read your thread on sensor size again and think through it. They are quite pricey too:

http://www.edmundoptics.com/onlinecatal ... uctID=2871

I had a thought in terms of reducing the vibration. I know that with a compound microscope you may decouple the camera fully from the microscope (as per Charles' website) and I was wondering if instead of mounting the camera to the bellows I could leave it detached with a super light weight gator or some sort of labyrinth system so it was light tight but not physically connected. The camera or bellows would have to be isolated from the enlarger platform by some sorbothane or something so not to couple the vibration through their mountings.

I have seen the dual fibre and hemisphere illumination a few times and it obviously gives good results. I had been wondering if it might be worth trying some of the high brightness LEDs now available (such as the lumiled K2), I suspect they would be bright enough but I'm clueless on aspects such as colour. FYI I have used red lumiled K2's for illumination of flies in flight when filming at 4000fps.

As far as computer control goes it is not too much of an issue, my experiments are normally controlled by a stand alone microcontroller. I'll probably have to hack into a remote unit to trigger the camera, Canon seems better for this as at least it is wired. It would be nice to be able to transfer the images straight to the PC as there could be rather a lot of them.

I'm looking at the brand new EOS450D at the moment, looks good from what I can tell and does have live view.

Graham

Graham Stabler
Posts: 209
Joined: Thu Dec 20, 2007 11:22 am
Location: Swindon, UK

Post by Graham Stabler »

I found a pdf of the manual for the Canon 450D or Rebel XSi.

http://gdlp01.c-wss.com/gds/0300000933/ ... 50D_EN.pdf

Interestingly it seems mirror lock up is not available when in live view mode however the site I got the link from says:

http://www.northlight-images.co.uk/came ... _450D.html
Enhanced Live View
Introduced in 2007 on the EOS-1D Mark III DSLR, this function makes its debut in an entry level Canon DSLR with the EOS Rebel XSi camera. Live View allows users to frame shots through the LCD screen rather than the viewfinder. Going beyond the manual focus and phase-detection AF capabilities of earlier EOS models with Live View, the EOS Rebel XSi camera adds a new "Live Mode" contrast-detection AF function that allows the camera to focus automatically during Live View without lowering the reflex mirror.
Perhaps mirror lock up is not available because it is not needed, i.e. there would be no 2 second delay.

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

Post by rjlittlefield »

Graham,

Thanks for the links in your last two posts. All were very interesting.

I don't know the answer to your question about why DSLRs don't use electronic shuttering. It's an obvious path to pursue, and since it would eliminate complex mechanics that both wear out and cause vibration problems, I can only presume there are other bad aspects that I just don't happen to know about.

The 450D looks nice, but I wonder how Live View actually works. I have this vision of the sensor being continuously exposed during live view, then at the instant of exposure having the shutter slam closed, then open/close to do the exposure, and finally open up again. I know this sounds a little crazy, but I vividly recall reading a description years ago of what happened at that time inside a medium format film SLR. It read something like this: "When the shutter release is pressed, the in-lens shutter closes, then the mirror flips up while the focal plane shutter opens. The in-lens shutter opens and closes to make the actual exposure. The focal plane shutter closes while the mirror drops back down, and finally the in-lens shutter opens again. And all this happens without a trace of vibration in the camera body." (Yeah, right. I think that last part about the vibration was a bit of hyperbole.)

I have no experience with the firewire cameras at Edmund Optics. I do have experience with one IP network camera, in particular the IQeye 705 unit described here. At work, we integrated it with a power tilt/pan head and power zoom/focus lens, mounted the whole unit on a small crawler for a tricky remote inspection task. The camera itself was insanely simple to integrate -- it runs a small Linux kernel that provides an interactive command shell and ftp/http interfaces for image transfer. We controlled the whole crawler from a customized UI running on a laptop, with the crawler tethered on a hundred feet or so of two-wire power plus a network cable.

The image quality from that particular camera was notably inferior to what comes out of my Canon 300D and A710IS cameras, more like what came out of the early consumer digitals. I wouldn't dream of trying to sell them on the basis of photographic quality. But given decent lighting, they were pretty good in terms of information content. I didn't try stacking any, but I did stitch several hundred of them together into a full spherical panorama that ended up looking surprisingly good considering the ghastly lighting conditions we had to work under.

I gather your interests are more scientific than aesthetic. Given the advantages of the firewire camera in terms of vibration and interfacing, I'd recommend taking a very hard look at it. Yes, the purchase price is quite a bit higher, but you have to factor in the cost of your time as well. If it takes you an extra week or two to make the DSLR actually do the job, then you've lost money overall.

But back to the DSLR, your ideas for handling vibration by isolation sound correct to me -- they're exactly what I would do, but I haven't done them yet so I can't say for sure if anything goes wrong. It seems to me that Charlie's setup provides a pretty good existence proof.

About the bellows, with my Olympus units it would be child's play to remove the lower rack and hard-mount the rail to something else. I don't think you'd even have to drill and tap it -- the rack is fastened on by two screws, just remove those and reuse the holes for mounting.

--Rik

DaveW
Posts: 1702
Joined: Fri Aug 04, 2006 4:29 am
Location: Nottingham, UK

Post by DaveW »

For the problems with electronic shutters see here:-

http://www.photographyblog.com/index.ph ... s_15_10_06

Some DSLR's have a combination of mechanical and electronic shutter.

DaveW

Graham Stabler
Posts: 209
Joined: Thu Dec 20, 2007 11:22 am
Location: Swindon, UK

Post by Graham Stabler »

I think you might well be right about the shutter, if I was braver I would dismantle and disable it, then add one isolated from the lens/camera. In practise that could be very tricky depending on how intellegent the camera is. Having hacked a few printers I know what a job it can be to do this sort of thing.

Perhaps I'll email Canon, who knows I might even get a reply.

In terms of the firewire cameras they are a piece of cake, at least my Sony is. I set it up to add captured images to an AVI file on the PC then my experimental set up sends a TTL trigger pulse to the camera every time an image is required. Definately easier and definately solves some problems but....
I gather your interests are more scientific than aesthetic.
This is the issue. While I am not as fussy (in a good way) as you guys the point of creating this rig is the recording of three dimensional structures in as much detail as possible. The purpose is to allow analyse without having to repeat a dissection or dig out a sample. In a way a comprehensive set of all-in-focus images will actually be better than the real thing and much easier to manipulate. It will also allow comparison between structures (or parts of the same structure) and dissemination of information to other researchers.

I am sure you have seen some of R.E. Snodgrass' wonderful drawings of insect anatomy but I can tell you that they miss out so many important details, written explanations are almost impossible too even if you know the terminology (and there is a lot of it).

So I worry if I spend the cash on the firewire and the images are poor it will be money down the drain followed by the development time with the DSLR.

I take it the olympus OM system is no longer in production? The nikon bellows look pretty good and there is even an extension for super long setups which has a very sturdy looking mount.

I saw mention of a 2.8/50 nikon enlarger lens in one thread which looked pretty good from Charles' test shots.

Graham

Graham Stabler
Posts: 209
Joined: Thu Dec 20, 2007 11:22 am
Location: Swindon, UK

Post by Graham Stabler »

Thanks for that Dave, confirmed my suspicions and an interesting last line:
Even so, with the current state of CMOS Image Sensor technology, mechanical shutters for digital cameras are just a ‘click’ away from extinction
Also makes me wonder if a simple way to improve the performance of an electronic shuttered system would be to turn off the lights just before and after the exposure. If I use LED illumination and choose the right trigger type on the firewire camera this would be easy and would still allow for continous lighting for focussing etc.

Graham

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

Post by rjlittlefield »

Graham Stabler wrote:
I gather your interests are more scientific than aesthetic.
This is the issue. While I am not as fussy (in a good way) as you guys the point of creating this rig is the recording of three dimensional structures in as much detail as possible.
I understand your concern. From my limited experience, let me offer the following thoughts...

I don't think there's a serious issue about image quality for your application. Achievable image quality is more a matter of optics and pixel count. Noise can be an issue under some circumstances, particularly the relationship between noise level and effective resolution for low contrast detail. Larger sensors are better in this regard if you are limited to a single exposure. However, for a static subject, you always have the option to take multiple exposures and add/average them together. When you do this, the statistics work out just the same for small and large sensors. I notice also that the camera at Edmund lists a "16 Bit High SNR Mode", undoubtedly some sort of internal averaging. In comparison, my 300D is only 12 bits coming off the sensor, and a couple of those are noise.

Given all this, I'd say the odds are very high that the firewire camera would work well for you -- better than a DSLR. It's not a sure thing, of course -- nothing ever is. But if I were in your shoes, the firewire camera would be my first shot. I'd be checking to see where I could get one with some sort of money-back trial period, then I'd get everything I could set up before delivery and test the dang thing into the ground in about the first two days after it arrived.

BTW, I'm assuming that you've used that AVT Oscar camera at Edmund only as a talking point. There are undoubtedly similar cameras available from several manufacturers. Edmund does good work, but it would be only prudent to hunt down and read whatever manuals and experience reports you can find about other units.

--Rik

Graham Stabler
Posts: 209
Joined: Thu Dec 20, 2007 11:22 am
Location: Swindon, UK

Post by Graham Stabler »

I suspect you are probably right, at the end of the day most of the advantages of a DSLR come in when you are not sticking it in a static lab rig and focussing manually.

I think you are right that the AVT camera can bin frames to get better SNR.

I'm still looking for other potential cameras, nothing really stands out yet.

This company:

http://www.diaginc.com/cameras/

Do some interesting cameras, some use pixel shifting technology to get increased resolution, but I have no idea of the costs or how well they actually perform in reality. So far the AVT Oscar is the biggest straight pixel count I can find.

Graham

Charles Krebs
Posts: 5865
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

Graham,

Don't know your budget, but take a look at this camera:
http://www.lumenera.com/scientific/inf411.php

"Full frame" (24x36mm), 11Mp, large pixels, Nikon F mount. For your purposes something like this would be great. DSLR's can do a great job, but you really need to take some extra measures to deal with the vibration issue.
One reason I was willing to do so was my desire to work with live subjects and electronic flash. Many of the "scientific" cameras cannot be synchronized with electronic flash. The cost of these cameras is a big issue as well. :wink:

One thing to consider with the "live view" function on the Nikon and Canon cameras (in addition to the mechanical operations that must still take place) is that as far as I know both manufacturers have limits on the amount of time it can be left "on". Heat will build up and the camera will either advise you to "shut down" or actually shut down automatically when a certain temperature is reached. The new Sony cameras (don't know if they are out yet) have a novel "live view" approach. A mirror inside the pentaprism head "flips" and directs the image to a completely different sensor dedicated to provide the live view only. The regular camera sensor is used for the picture. I have no idea if this will minimize the heat issue or how well it will work.

Harold Gough
Posts: 5786
Joined: Sun Mar 09, 2008 2:17 am
Location: Reading, Berkshire, England

Post by Harold Gough »

Charles Krebs wrote:Many of the "scientific" cameras cannot be synchronized with electronic flash.
Open flash (I think that is the term) may be an option. Open the shutter, fire the flash manually, close the shutter. The shutter could be operated by selecting, say, a one second speed on the dial or via a cable release on B setting. Ambient lighting would have to be sufficiently low to not register in the image.

Harold
My images are a medium for sharing some of my experiences: they are not me.

Graham Stabler
Posts: 209
Joined: Thu Dec 20, 2007 11:22 am
Location: Swindon, UK

Post by Graham Stabler »

Charles,

Thanks for the link, I found these cameras but ruled them out due to the lack of a hardware trigger, I could certainly write some code for capture and interface that to my control system but I just don't have the time in this case and manual clicking is not an option when I might be taking deep stacks from many angles.

Most of the industrial non microscope specific cameras allow external triggering, if you wanted to you could trigger the camera and flash together or have a lag as required. The Sony I use also allows the trigger pulse to determine the exposure if you want. The cameras with global shutters would be especially good but might be a pain in the field even with a tiny laptop in tow.

Graham

Post Reply Previous topicNext topic