Did you see the thread describing a simple calibration device based on that same principle?( a bit of card with two slots cut in it)
Build a shutter tester for Focal Plane shutters - Cheap, Easy & it Works
-
A
- Thread starter Niglyn
- Start date
Recent Classifieds
-
For Sale Classic Hasselblad 500 C/M 6x6 all manual Medium Format film camera system
- Started by pdccamerqs
-
For Sale Horseman SW612 w/Rodenstock 45mm f4.5 Apo Grandagon, Ground Glass
- Started by djdister
-
For Sale Canon 28mm f/3.5 LTM plus finder in case
- Started by Steve Goldstein
-
Want to Buy Durst Printo / or any Table Top Processor
- Started by juniejun
-
Want to Buy LF/WTB: Ilex No. 5 Shutter
- Started by Jurgen Estanislao
Forum statistics
When I have tested the next iteration, will attach to another post.
Github maybe it's final resting place, but I don't know how github works, so need to look into it, sometime.
Nope, but the piccy looks like they using Audacity to measure a pulse length. Nice hack if they do not have access to an oscilloscope.
It is from a thread I had posted on calibrating my existing Audacity based shutter tester. My interest in Arduino is so I don't have to boot up my workbench computer and attach the A/D converter every time I want to test a camera.
It is from a thread I had posted on calibrating my existing Audacity based shutter tester. My interest in Arduino is so I don't have to boot up my workbench computer and attach the A/D converter every time I want to test a camera.
Not sure how this would calibrate an arduino. The internal arduino clock will be far more accurate than a drill, paper with slots in it, running into a computer where the operating system can go off and do something, the computer with its own clock that will not be 100% accurate into a program designed for audio.
I really do think you are chasing your tail with this.
The notion of calibration in this context seems to stem from a misunderstanding of what a microcontroller circuit is capable of. A typical Nano will be equipped with a +/-20ppm (typical max. boundary) quartz oscillator. Provided that the code doesn't mess things up (and this is easy to do!), this allows the thing to be accurate in its timings within 0.002% accuracy.
Of course, there will be some delays inherent to the laser gate circuitry. However:
1: These will be in the low microsecond range
2: They tend to be pretty symmetrical for both laser gates
3: They tend to be fairly stable over time and operating conditions
In short, they tend to cancel out and/or are insignificant in this application, provided somewhat sensible component choices.
The Arduino might be nice to calibrate that electric drill contraption, though
With a 1mm sensor the effective slit will be 2.25mm + the extra 1mm, so 3.25mm. When adjusting the closing curtain cam, this error will direct one to have the closing curtain cam release the closing curtain at 1.25mm to obtain 1/1000 of a second from the shutter tester.
This sensor width error is irrespective of how the sensor output is being read, Arduino, oscilloscope, etc. The wider the sensor the more error. If the sensor were 36mm across the sensor output would read 1/60 at all the faster speeds.
The calibration factor, likely will be the same across the board for all speeds for each sensor. It can be calculated a priori, but an empiric determination is better.
A slit of known velocity and width can be passed between the laser and the sensor using a shutter of known velocity and slit width. One can construct a gold-standard shutter with known slit width and velocity easily. An example is shown in post #80. But there can be other methods of passing a known slit width across the sensor if one does not have a motor or know how to determine RPM.
This sensor width error is irrespective of how the sensor output is being read, Arduino, oscilloscope, etc. The wider the sensor the more error. If the sensor were 36mm across the sensor output would read 1/60 at all the faster speeds.
The calibration factor, likely will be the same across the board for all speeds for each sensor. It can be calculated a priori, but an empiric determination is better.
A slit of known velocity and width can be passed between the laser and the sensor using a shutter of known velocity and slit width. One can construct a gold-standard shutter with known slit width and velocity easily. An example is shown in post #80. But there can be other methods of passing a known slit width across the sensor if one does not have a motor or know how to determine RPM.
Last edited:
So what needs to be determined, is the width on & off of the sensor?
A rotating disc would introduce error as the sensor has both height and width, would a rotating drum be better?
Would a better way to measure sensor width, to use a card with a slot & move this across the sensor with a precise measuring aid, maybe a vernier or set it up on a milling machine with digital readout? We would know the actual measurement of the slot and the length travelled by the card.
The speed of each curtain and the distance over which this was measured is known. This is measured by two sensors, so there should be no width of sensor error.
Assuming the 'on-off' distance of the sensor is known, by measurement as described above, it seems just a maths problem to calculate the actual slot width.
Alternately, what about mounting the shutter tester vertically & dropping a weighted card with a slot cut, though it? The two lasers would measure speed & the slot width is known.
Latest code.
Now has shutter bounce detection. To use this, I would suggest moving the sensors/camera, so the second laser is right at the edge of the gate.
Has anybody built one yet?
Now has shutter bounce detection. To use this, I would suggest moving the sensors/camera, so the second laser is right at the edge of the gate.
Has anybody built one yet?
Not yet, just waiting on the lasers and detectors to come, they're on a slow boat from China, should come this week.
Seems my laser units (ordered from the link you posted) are coming from China, so it will be a couple of weeks yet! But they have set off.
Apologies, did not notice the item location when I found the item on Ebay.
At least it will give time to get the software set up & communicate with the Arduino board
)Not yet, I do have all the parts, but haven't got the time yet to put it all together. Also have to learn a bit about the Arduino, this is my first project
Apologies, did not notice the item location when I found the item on Ebay.
At least it will give time to get the software set up & communicate with the Arduino board
In case there are any other Arduino newbies following along, when I did a trial upload of the software on the Arduino IDE to my Nano it gave me a compile error as the LiquidCrystal_I2C library was missing but this was easily solved with Tools > Manage Libraries in the IDE and searching for LiquidCrystal_I2C.
All errors would be grouped into a single calibration factor by comparing a known shutter to the device output.
One does not really need to determine the sensor size, unless calculating , rather than measuring, a calibration factor.
It probably is a good idea to have the laser beam totally covering the sensor.
The nice thing about a rotating test shutter is it’s constant velocity.
I also found using an essentially new Nikon f6 as a standard didn’t work because the shutter won’t fire at high speed with the back open.
Last edited:
With a disc, how do you know the sensor is triggering on a circumference of 200mm, not 201mm or 199mm?
I doubt that the rotation speed of a drill is calibrated or that it's rotational velocity is constant.
I thought the whole point of your posts was that the sensor size would have an impact on the accuracy of measurement?
When I tested my tester against my electronically timed Canon T70, the results were spot on, so that is close enough for me.
The ISO standard for 1/500s is between 1/374s & 1/699s and for 1/30s it is between 1/25s and 1/39s
If you like maths, read this.
The Way to Modern Shutter Speed Measurement Methods: A Historical Overview - PMC
Exposure time is a fundamental parameter for the photographer when the photo is composed, and the exact length of the exposure may be an essential determinant of performance in certain camera-based applications, e.g., optical camera communication ...
www.ncbi.nlm.nih.gov
Good question. You know that the closer to the center the angular velocity is less, so, for the same shutter time, the gate has to be narrower. Likewise, the angular velocity farther out is greater, so for the same shutter time the gate needs to be larger.
This is accounted for by the shape of the openings. See how they are not rectangles but trapezoid openings. When laying the thin, slide masking tape on the edges of the openings, the tape is aligned with the spokes radiating from the center.
The same principle applies to rotating shutters in movie cameras. The opening is trapezoid and the entire frame gets even exposure.
I have many electric motors, but by good fortune the Makita drill had a speed that allowed the disk to be not too big or not too small to travel at the same velocity as a camera's curtains.
Also, as you get into the finer points of the math, my wheel actually gives speeds of 1/120, 1/240, 1/480 and 1/960 based on the progression of halving the 36mm shutter opening with each increasing speed.
Below, some movie camera rotating focal plane shutters that provide even illumination to the frame in spite of the outer portion moving faster than the inner portion:
The 'ticks' on an oscilloscope (Audacity screen) from the openings were measured over ten revolutions during the calibration process to get the exact speed (which btw was identical to the speed I measured before constructing the wheel as I did not run the drill at all between measuring its initial speed and sizing the wheel in software and printing it). The speed was very constant over the measuring period of ten revolutions with the interval between the 'ticks' being essentially constant. The battery would not drain in that short period, ten revolutions in 1.652 seconds ; there was no battery drain in 1.6 seconds.
The ten second measuring period gave ten samples of each of the shutter speeds to average, but they were essentially all the same +/- 1 or 2/4410.
Last edited:
That is good news!! Using Audacity, the photodiode turns ON instantly and stays on until no more light hits it. Thus the errors from the sensor size. Perhaps the sensors in the Arduino world register ON after 1/2 of the sensor is exposed and register OFF when the other 1/2 is exposed, effectively making it like a point sensing source.
Good question. You know that the closer to the center the angular velocity is less, so, for the same shutter time, the gate has to be narrower. Likewise, the angular velocity farther out is greater, so for the same shutter time the gate needs to be larger.
This is accounted for by the shape of the openings. See how they are not rectangles but trapezoid openings. When laying the thin, slide masking tape on the edges of the openings, the tape is aligned with the spokes radiating from the center.
The same principle applies to rotating shutters in movie cameras. The opening is trapezoid and the entire frame gets even exposure.
I have many electric motors, but by good fortune the Makita drill had a speed that allowed the disk to be not too big or not too small to travel at the same velocity as a camera's curtains.
Also, as you get into the finer points of the math, my wheel actually gives speeds of 1/120, 1/240, 1/480 and 1/960 based on the progression of halving the 36mm shutter opening with each increasing speed.
Below, some movie camera rotating focal plane shutters that provide even illumination to the frame in spite of the outer portion moving faster than the inner portion:
The 'ticks' on an oscilloscope (Audacity screen) from the openings were measured over ten revolutions during the calibration process to get the exact speed (which btw was identical to the speed I measured before constructing the wheel as I did not run the drill at all between measuring its initial speed and sizing the wheel in software and printing it). The speed was very constant over the measuring period of ten revolutions with the interval between the 'ticks' being essentially constant. The battery would not drain in that short period, ten revolutions in 1.652 seconds ; there was no battery drain in 1.6 seconds.
The ten second measuring period gave ten samples of each of the shutter speeds to average, but they were essentially all the same +/- 1 or 2/4410.
Fascinating stuff. The Arduino can drive a stepper motor, one of the basics in the tutorials, which will give an accurate rotation speed. Aliexpress has steppers & interface boards for silly money. A photo-interrupter can be used to give feedback to confirm rotation speed.
Aliexpess also sell a complete tacho module for silly money. I have one around somewhere. I bought two, one of which I put onto a lathe to confirm rotation speed.
Arduino and Stepper Motor Configurations | Arduino Documentation | Arduino Documentation
Learn how to control a variety of stepper motors using unipolar / bipolar circuits with Arduino.
docs.arduino.cc
The Wokwi site is pretty amazing, I was able to build a complete version of the device with virtual components that actually runs Niglyn's code. I got both sensors to work on the simulation.
The file to create the model is attached below, it is a screen shot, so you'd have to type it in as a ".json" file to make the model.
I think there is a way to share a public link to the actual working virtual model running the code... stay tuned.
The file to create the model is attached below, it is a screen shot, so you'd have to type it in as a ".json" file to make the model.
I think there is a way to share a public link to the actual working virtual model running the code... stay tuned.
This makes me think of the tester I built years ago. The only difference with mine is that I use a single visible light detector. At first I thought it would be great but in the end it turns out that it only works reliably with leaf shutters 
One feature I had implemented is that I store the shutter speeds into memory, calculate the average speed and when I trigger a switch to reset the device, it shows me the lowest and highest speeds achieved.
But all this is not very useful if I can't get reliable results. I'll have to rebuild mine to include the dual laser system you implemented.
One feature I had implemented is that I store the shutter speeds into memory, calculate the average speed and when I trigger a switch to reset the device, it shows me the lowest and highest speeds achieved.
But all this is not very useful if I can't get reliable results. I'll have to rebuild mine to include the dual laser system you implemented.
Here is a public link to Niglyn's sketch running on a virtual Arduino UNO. The sketch is modified somewhat to run under the simulation, so anything odd in the behavior is due to the simulation, not Niglyn's code.
wokwi.com
Public Shutter Tester - Wokwi Arduino and ESP32 Simulator
Run IoT and embedded projects in your browser: ESP32, Arduino, Pi Pico, and more. No installation required!
I've saved it and I'll have to compare them when I have a bit of time.
Here's the sketch I used on mine.
I use a 2 line LCD Keypad shield by D1Robot hooked to an Arduino UNO.
I figured that since the device is not something I want to carry in my pocket, I might as well go for the UNO.
But I made it with a removable sensor like on the commercial machines and I regret doing so a bit. Thing is that the commercial ones used in repair shops have a light that is collimated. This makes the readings much more precise.
They also probably have a slit aperture on the sensor to increase precision, something that mine doesn't have yet.
This thread got my interest back in writing computer code. The last thing I wrote was in Think-C on my Macintosh Quadra 700. When I upgraded to an iMac, I bought Code Warier but it would not run on the iMac I got and never did anything since.
| Photrio.com contains affiliate links to products. We may receive a commission for purchases made through these links. To read our full affiliate disclosure statement please click Here. |
PHOTRIO PARTNERS EQUALLY FUNDING OUR COMMUNITY:  |