Granite Optical Bench Construction

[g4lab]

I thought they had electronic shutters. That is the chip is electronically flushed and then the exposure takes place. I have not yet had the opportunity to fondle one in person.

[AndrewC]

This forum just keeps getting better and better.

On this thread the thought occurs to me, once again, that maybe it's time to
eschew the mirror, and switch to one of the so-called hybrid cameras. (micro 4/3, Samsung NX10 , there is a Sony too I think)
Has anybody done that yet?

Yes, I'm toying with a Panasonic G. The downside is that I can't tether it to my satisfaction - P. did not "properly" implement the protocol.

Don't confuse EVIL cameras with their P&S cousins.. Electronic Viewfinder Interchangeable Lens camera typically do have a shutter, and a big sensor and can take pretty impressive images.

[AndrewC]

But the mirrorless cameras all have mechanical shutters. You would regain the 3 seconds scheduled for the mirror lockup timeout (if you're using it).

But if you suffered from vibration paranoia, the mass of the shutter is probably less than that of the mirror. For those who want to be hyper accurate - the kinetic energy changes are probably more significant for the mirror movement.

Anyway, with M-up mode on most cameras the mechanical shutter movement isn't until the second trigger so isn't a problem you need to delay for anyway ?

[Mitch640]

While I was still in the market for a camera to use with the Fluophot, I ran across several sites that sold digital imaging sensors on a board. Just the sensor and a couple wires you could connect yourself to record images. No shutters needed, as they worked in a frame grabber style using software. Some of you here might be interested in this link.

[mouldy]

I had asked Canon it they knew the travel time of the second shutter curtain so I could figure what the percentage of the exposure time had a mechanical part in motion.

I might be off track here but could this not be determined by the flash sync speed? On the Canons this is usually 1/250. If you try to take a flash picture at 1/320 you will usually be OK (speedlights rather than studio lights) but at 1/400 you will get a dark underexposed bar at the bottom of the picture as the second curtain starts to rise before the flash has fired.
An ambient exposure sequence where the shutter time is longer than the time it takes the curtain to travel is:

• 1. first curtain starts to open
  2. first curtain fully open
  3. wait
  4. second curtain starts to close
  5. second curtain fully closed

Ambient exposure and therefore shutter speed is the time from step 1 to step 4

An ambient exposure sequence where the shutter time is shorter than the time it takes the curtain to travel is:

• 1. first curtain starts to open
  2. wait
  3. second curtain starts to close
  4. first curtain fully open
  5. second curtain fully closed

In this case ambient exposure and therefore shutter speed is the time from step 1 to step 3

When the exposure is such that the second curtain has to start moving at the exact same time that the first curtain reaches fully open, that is the time it takes for the curtain to travel

We know from the flash sync that at 1/320 (usually) we get a full exposure and therefore must be using the first scenario. We also know that at 1/400 the second the second curtain is already moving and is therefore the second scenario. So the point at which the condition above is met must be between the two.

If we assume the time from the first curtain being fully open to the flash firing is negligible (as it is electronically controlled there is no reason for it to be more than a few microseconds) then the curtains must take between 2.5 and 3.125 milliseconds (1/400 and 1/320) to move.

So to answer your question, if you have a shutter speed of 1/250 ( 4 milliseconds) the curtain could be moving for anything between 62.5% and 78.1% of your exposure.
this would obviously only be a factor if you were using ambient light. If your subject is flash lit with very little or no ambient, the curtain closing all takes place in the dark

I stand to be corrected on this so if I have made an error, please do let me know.

[Bob^3]

I accept that Sorbothane is capable of transmitting one particular frequency quite well, it attenuates all the higher ones, to different degrees. I don't think that graph indicates the gain at what is perhaps here misleadingly called the "resonant" frequency, but from the spread of the curve it looks a lot less than unity.

Not "just one particular frequency" but all frequencies below the "natural frequency" and a significant percentage of the frequencies above the natural frequency, near the peak. The graph is a classic normalized resonance plot (the specific resonance frequency and "gain" depends on the system mass, which isn’t indicated). The “natural frequency” (Fn) is the frequency at resonance. The peak in the curve is called the “resonance peak”. It just wasn’t labeled as such in the diagram (see link below).

You mentioned ( I can't see it while editing!!) about injecting with lead - to introduce some damping. That would be in recognition of the "ringing" quality of the support -…

Don’t get me wrong---Richard’s column will definitely tend to “ring” at some natural frequency if it is excited by vibration which contains the correct high-frequency components. So my reason for suggesting a possible addition of damping material was in case the delay between mirror, shutter or StackShot movement was not long enough to allow any ringing to decay.

The important concept here is that, even if there are no vibrations being produced by the components on the stage (mirror, etc.)---the camera, lens and subject will always be connected to any lower frequency (environmental) energy that can couple through the (Sorbothane) base isolation from “earth”. So the goal in designing the “ideal” isolation table is to break the system into two groups: the “base isolation” group (table, Sorbothane, granite block) and the camera/lens/subject group (I’ll call the stage group). As you point out, Sorbothane will attenuate all vibration above its natural frequency, so only low-frequency energy will reach the stage group.

But another important concept to realize is the low-frequency vibration cannot couple vibrational energy to a system which has a much higher natural frequency. So if components of the stage group are rigidly connected together to form a system with a high natural frequency and high-damping factor (granite), they will all vibrate in unison at the lower frequency of the base, but not relative to each other.

Now, due to its wide-spectrum properties, it might seem logical to want to also use Sorbothane for isolating all the stage components from each other and from the base platform. The problem is that this will also lower the natural frequency of the stage group components. If lowered too far, the low-frequency vibrations propagating through the base isolation will cause the stage components to wobble at different rates resulting in motion blur. See the link below for the typical range of environmental earth frequencies that are present at various locations and why Sorbothane cannot completely remove them.

Wiki didn’t provide any help on these topics, but here is a Melles Griot white paper that covers the it well. If you take the time to review this, especially the graphs, I think you’ll understand my points. This is also a good summary for any forum members who want to better understand the principals of vibration isolation:

http://www.cvimellesgriot.com/Products/ ... lation.pdf

So with regard to your diagram, if you were to replace the upper squishy Sorbothane supports (the little “c” parts) with something stiff like granite and remove the added masses on top of them (which also tend lower the natural frequency), we would be in complete agreement.

As with many of the optical principles discussed in this forum, intuition can fail to provide a correct answer. A full understanding can only be gained by studying the mathematical relationships.

I’m sorry I’m not as good as Rik at concisely explaining these complex concepts. But I hope this makes better sense?

Meanwhile, I'm eyeing up the prices of used Canons!

Me too! Nikon has really missed the boat on this one. It's my biggest pet peeve with all their DSLRs. And it's not like the double mirror slap in Live View mode only effects macro photography. I've seen motion blur on my telephoto landscape shots as well, at shutter speeds below 1/500 second. And the 1 second delay setting from mirror to shutter just isn't long enough. I had high hopes for the D7000 (quiet mode?). But from what I've read so far, it falls far behind Canon's "electronic" shutter implementation.

[Bob^3]

Also Chris, I'd like to say that my intent is not to be argumentative in this discussion; I would just like to discover the source of our disagreement. I'm not interested in who's right, only what's right. If anyone can see flaws in my train of thought, please point them out.

[ChrisR]

Bob I don't want this to go on but perhaps I should explain a little of where I'm "coming from".
The philosophy behind the lecturing I got, and what was used where I was involved in practice, was to keep everything separate, Each element considered for its mass and sensitivity, each source considered separately, but also each connection between them as well. I'm sure I could put that better if there weren't so many years in between.
The Lab I alluded to was involved with sensitive equipment, but not optical - "I could tell you where but I'd have to shoot you.."
The approach was very mathematical, not intuitive at all, in fact more transforms and matrices and integrals than I would ever have enough life to handle.

That is where the desire to separate the camera vibrations from the rigid slab come from. Better to know what they are and knobble them with, if necessary, tuned damping than let them into a rigid element with unspecified behaviour.
In those days Sorbothane wasn't in use so perhaps I'm quoting it too freely. "Squishy" is rather imprecise - perhaps you inferred that the C under the camera would allow it to jiggle about - no.

Richard’s column will definitely tend to “ring” at some natural frequency

That's the point of the "C" under it. Fill it with lead shot, special gel, or whatever. It needn't ring at all if it's damped.

The rigs I have in mind had very many mechanical elements for coupling and damping at different rates with different corner frequencies and different rates (dB/octave) of damping. I wasn't involved but other folk used active (powered) hydraulic dampers.

The problem is that this will also lower the natural frequency of the stage group components. If lowered too far, the low-frequency vibrations propagating through the base isolation will cause the stage components to wobble at different rates resulting in motion blur.

But you would never let that happen. The maths would show the base isolation to be inadequate.

Granite is conveniently massive, and cheap, but combines mass with (unspecified) damping properties, which is a conflict. Being a composite it has helpful complex time and frequency domain properties, which traditionally have been found useful on optical structures.
A quick look at the net http://en.wikipedia.org/wiki/Optical_table
found that modern tables use steel, - which would have special structure and predictable behaviour, rather than simply

"high natural frequency and high-damping factor (granite), "

However, granite is cheap and readily available, and I indeed have a slab ready for the day when I have time and space to use it in a rig. I will definitely NOT be doing anything but basic maths, nor smothering it in accelerometers! (Though there is at least one designed to work with a PIC which is interesting). At my scale a traditional approach with some educated intuition will be fine I'm sure.


Richard - if you're still there...
I'd suggest sealing the base end of your column, perhaps with epoxy, ready for thhe day when you may wish to pour something in there.

[Bob^3]

Yes Chris, I don’t want to keep this going forever either. But I think it has been an interesting discussion. And perhaps someone besides us has learned something from the exercise.

The philosophy behind the lecturing I got, and what was used where I was involved in practice, was to keep everything separate, Each element considered for its mass and sensitivity, each source considered separately, but also each connection between them as well. I'm sure I could put that better if there weren't so many years in between.

I think you put that very well, indeed (that must have been wuite a lecture). I fully agree point for point, (depending on the definition of the term “separate”).

That is where the desire to separate the camera vibrations from the rigid slab come from. Better to know what they are and knobble them with, if necessary, tuned damping than let them into a rigid element with unspecified behaviour.

Agreed again, especially about the “tuned damping”. Only for optical systems, I would say rigid but specified high-frequency damping behavior.

In those days Sorbothane wasn't in use so perhaps I'm quoting it too freely. "Squishy" is rather imprecise - perhaps you inferred that the C under the camera would allow it to jiggle about - no.

True, “squishy” was an imprecise term. It was meant to convey any range of relatively soft rubber-like materials used for low-frequency base isolation and damping. If Sorbothane were made in a very hard durometer that would not couple significant low-frequency vibrations to the camera (say below 100 Hz vibration or “wobble”), but would still decouple high-frequencies, that should work for the C connecting the stage components; so I think we also need to define precisely what the term “rigid” means. On the other hand, the “C” used for base isolation should as conformal as possible (choose your term, squishy, springy, soft rubber, low durometer); but it must also be very highly damped to avoid oscillation (resonance) at the system’s natural frequency.

Quote:
Richard’s column will definitely tend to “ring” at some natural frequency

That's the point of the "C" under it. Fill it with lead shot, special gel, or whatever. It needn't ring at all if it's damped.

I think you may still be stuck on the need to add some relatively rubbery-like material like Sorbothane between Richards’s granite block and the aluminum column (However, if the damping material your referring to is a rigid materail like a platform made from the laminated steel material you mentioned, I would agree).
. The column already is damped (decoupled) for high frequencies by the granite block
What I’m trying to say is that “best practice” would be to construct the column out of light-weight, rigid material designed to have high-frequency, self-damping qualities and rigidly attach it to a high-frequency, self-damping base isolation plate. The columns used to mount optical components on interferometers are made like this, special laminated structures like the steel tables you mention but in the form of a rod or post---move the base isolation table around and all the components stay locked to the table and stationary relative to each other within a distance less that a wavelength of light. If you were to place anything like Sorbothane (even a very thin layer of very hard grade Sorbothane) between the base and the post, the system would fail due to (even very minuet) vibrations coming from the local environment. The isolation requirements in this case are orders of magnitude higher than is needed for macro photography, but the principles are exactly the same.

Again, in Richard’s case, since the granite base is providing decoupling for high-frequency vibration and his Canon camera should not be adding any shock impulse to the column, I would guess that he won’t have a significant vibration issues---as long as his Sorbothane base isolation is adequate for the environmental ground vibrations at his location!

The rigs I have in mind had very many mechanical elements for coupling and damping at different rates with different corner frequencies and different rates (dB/octave) of damping. I wasn't involved but other folk used active (powered) hydraulic dampers.

What you are describing sounds exactly like the first stage “base isolation” half of the Melles Griot publication “Fundamentals of Vibration Isolation” I linked to. And for non-optical applications where micron-amplitude movement doesn’t matter, base isolation is all that is needed (I won’t ask you to explain any more details of your activities, because I’d kinda like to live a little longer ) In critical optical applications, “tuned” base isolation systems, which are controlled by active electronic damping drives can almost completely eliminate most high and low frequency environmental sources of vibration.

Quote:
The problem is that this will also lower the natural frequency of the stage group components. If lowered too far, the low-frequency vibrations propagating through the base isolation will cause the stage components to wobble at different rates resulting in motion blur.

But you would never let that happen. The maths would show the base isolation to be inadequate.

The fact is for sensitive optical systems mathematics may show that you simply can’t isolate all of the environmentally generated ground vibration noise below a certain frequency. Unfortunately, I can’t link directly to the section in the Melles Griot document that covers this, but it does a better job than I can at explaining the importance of this concept. For the magnifications we are dealing with in this forum, it may or may not be an issue in capturing sharp macro or micro images.

And this is a very difficult potential problem to show or prove in a practical real-world sense because it is inherently intermittent and completely governed by the “noise” at a particular location. If Rick (no relation to the admin ) happens to live in the suburbs far from heavy road traffic, he might well be able to successfully capture sharp macro images with his a Canon camera sitting on top of a cardboard box. While Bob, who lives next to a train station or freeway (or dump-truck driven alley), might have problems that even long-Bungee-cord isolation or Lefkowitz’ 175 lb-mass-on-top-of-inner-tubes may not solve at 1x magnification (but, apparently it did solve his refrigerator vibration problem at >5x). The top of base isolation table might only “wobble” at 1 Hz with an amplitude of 5mm. But would you grab your table and move it back and forth 5mm during image capture?

Granite is conveniently massive, and cheap, but combines mass with (unspecified) damping properties, which is a conflict. Being a composite it has helpful complex time and frequency domain properties, which traditionally have been found useful on optical structures.
A quick look at the net http://en.wikipedia.org/wiki/Optical_table
found that modern tables use steel, - which would have special structure and predictable behaviour, rather than simply

Quote:
"high natural frequency and high-damping factor (granite), "

All very true, not all granite is created equal. And there are certainly better, lighter, more rigid and more “predictable” materials to mount the optical stage elements to (but granite can be graded for these properties to have more “predictable behavior”). I just used granite as an example in this discussion because, as you say, it’s cheap and available and it’s what Richard, I and maybe you(?) choose to use. And even grades of granite with “unpredictable behavior”, are probably more than good enough for the current application.

I will definitely NOT be doing anything but basic maths…

Fine, and depending on the factors discussed above, that’s probably all you will need. Obviously, most if not all of the members of this forum do not need to attempt to implement every extreme design measure called out in the Melles Griot document and even simple designs can be perfectly functional, as is proven by the many fine designs and resulting images in this forum.

… nor smothering it in accelerometers!

Yeah well Chris that’s probably just me. It’s what I “do” S. S. I. I. = he Smothers Stuff In Instrumentation (and my brother does, too ).

(Though there is at least one designed to work with a PIC which is interesting.)

Truth is, several times I have considered starting a thread titled something like “ Using accelerometers to optimize your stage”. But I haven’t because I’m not sure there would be an audience for my blabbering. I did a quick forum search for “accelerometer” but only had a couple hits. There was a short discussion HERE brought up by elf (my stack and stitch mentor). But it but it never seemed to go anywhere. Also, I detect a certain tendency to view any concepts that might be new and poorly understood to be deemed overly complicated, too high-tech, total overkill and completely unnecessary to get good results from a macro stage. And to some extent I would have to agree with this assessment, in the same sense that automated stacking systems were considered totally unnecessary a couple years ago (and I would guess still are to some members). This is drifting way off subject for this thread now. So I’ll just offer a short summary of the potential benefits of using accelerometers to measure and analyze stage vibrations.

First, without being able to place accelerometers directly on a camera sensor, they are unlikely to provide a calibrated, objective measurement of how any vibrations in a given optical system will affect motion blur or the sharpness of the final image. You can only assess that by analyzing the shift of the image as projected onto the camera sensor. What the accelerometer can do is provide a relative indication of how much acceleration, velocity and displacement is occurring at each accessible component in the optical path. And it can then be possible to establish some correlation between those measurements and the image results. Also, the accelerometer data can be extremely valuable in quickly finding and resolving any sources of vibration that are measurably degrading the images. So it can be effectively employed to avoid this dreaded syndrome from a recent thread:

I think collectively we are evolving a monster - I'm waiting for the first view of a stage made out of a 600lb granite slab, suspended from bungee cords, installed in a room carved out of living rock, overhung with an 8 legged multi-jointed lighting "Shelob", hooked to a 20 petaflop computer to enable real time instantaneous stacking, camera moving on a maglev table with laser positioning and all based on a $20 lens picked up on eBay .....

For instance in the current case with Richard’s stage if he was seeing motion blur in his images, it would be possible to measure the setup at various places, the granite block, top of the aluminum column, lens and subject stage to quickly find the vibrations sources in real time. He could then take proper measures to eliminate only the sources of the problem, thus avoiding “system-design overkill”.

I happen to own a several very sensitive, wide bandwidth accelerometers, and an oscilloscope for making these measurements. But even the new MEMs accelerometers (a few $ on eBay) used in everything from Wii controllers to smart phones like the iPhone and many Android phones can have some limited use. Here is an image of the LCD screen on my DroidX cell phone running a free app called Accelogger to record data from the built-in 3-axis MEMs accelerometer. The phone was resting on top of the D700 hotshoe during a normal low and high speed capture sequence (no mirror lock-up)---sorry for the motion blur :

Richard - if you're still there...

Hopefully, Richard will forgive us based on the fact that this discussion is happening at the end of the thread? And Richard if you are still reading this, I think you should get an award for “Longest First Equipment Thread” = 6 pages, 83 responses, 2175 reads so far! So some poor misguided folks must be reading this stuff!!! Or they are only reading the first 2 pages until they see Chris’ diagram and read the word “resonance” and their eyes glaze over.

[richard martel]

Bob,
Yes I'm still here. I have been taking advantage of the cool weather in the Keys to work on my deck. In fact I was reading over the posts that I had missed just today. Very interesting and I'm learning much. Thank you all. I did order the book you recommended. Also my 416 SS plate is on order. I must say that I'm surprised at all the attention this thread has generated. Hopefully it may turn into, with everyones input including any photos related to the topic title, the "go to" place for info of this sort.

Best wishes to all...Richard

[Bob^3]

Glad we didn't loose you Richard.

[richard martel]

There have been so many discussions on vibration that upon completion of this project I'd like to test the system for vibration. So I think I'll try to find a xyz accelerometer, o'scope and maybe a signal generater. I briefly read the white paper from Melles Griot on the subject too. There would be quite a few places one could put the accelerometers and from that see how the total system responds...Stay tuned (pun intended)

Regards, Richard

[rjlittlefield]

Truth is, several times I have considered starting a thread titled something like “ Using accelerometers to optimize your stage”. But I haven’t because I’m not sure there would be an audience for my blabbering.

I guarantee at least a small audience. If you have time to write it, I have time to study it.

move the base isolation table around and all the components stay locked to the table and stationary relative to each other within a distance less that a wavelength of light. ... The isolation requirements in this case are orders of magnitude higher than is needed for macro photography

Not always. Unless I've blown the calculation rather badly, when I stick a 40X objective on my T1i the resolution on subject is 0.117 microns per pixel -- around 1/5 of a wavelength. High mag macro really is pretty demanding.

--Rik

[Bob^3]

Not always. Unless I've blown the calculation rather badly, when I stick a 40X objective on my T1i the resolution on subject is 0.117 microns per pixel -- around 1/5 of a wavelength.

True. But in defense of my oversight, I'll point out that I did not specify the precise wavelength of light nor use the term "visible".
I should have said "much smaller than the wavelength of visible light".

[Bob^3]

And by correcting the former statement to "much smaller than the wavelength of visible light", I'm referring to some sensitive interferometers that can measure phase shifts of less than 1/1000 the wavelength of visible light, in the sub-nanometer range. They require very rigid damped mounting combined with very good base isolation.