Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.
Category: Science
If you measure something often enough, it becomes science
According to the Arducam doc, their Motorized Focus Camera has a 54°×41° field of view, (roughly) equivalent to an old-school wide(-ish) angle 35 mm lens on a 35 mm still camera. For my simple purposes, the camera will be focused on objects within maybe 200 mm:
Arducam Motorized Focus Camera – desktop test range
The numeric keys are 6.36 mm = ¼ inch tall, the function keys are 5.3 mm tall, and the rows are 10 to 11 mm apart.
The focusing equation converting distance to lens DAC values depends critically on my crude measurements, so the focus distance accuracy isn’t spot on. Bonus: there’s plenty of room for discussion about where the zero origin should be, but given the tune-for-best-picture nature of focusing, it’s good enough.
I set the CANCEL legend at 50 mm and it’s in good focus with the lens set to that distance:
Arducam Motorized Focus Camera – 50 mm
Focusing at 55 mm sharpens the ON key legend, while the CANCEL legend remains reasonably crisp:
Arducam Motorized Focus Camera – 55 mm
Adding another 5 mm to focus at 60 mm near the front of the second row shows the DoF is maybe 15 mm total:
Arducam Motorized Focus Camera – 60 mm
Focusing at 65 mm, near the middle of the second row, softens the first and fourth rows. Both of the middle two rows seem OK, making the DoF about 20 mm overall:
Arducam Motorized Focus Camera – 65 mm
Jumping to 100 mm, near the top of the first function row:
Arducam Motorized Focus Camera – 100 mm
At 150 mm, about the top of the far row just under the display:
Arducam Motorized Focus Camera – 150 mm
I think 200 mm may be the far limit of useful detail for a 5 MP camera:
Arducam Motorized Focus Camera – 200 mm
At 300 mm the DoF includes the mug at 600 mm, but the calculator keyboard is uselessly fuzzy:
Arducam Motorized Focus Camera – 300 mm
At 500 mm, the mug becomes as crisp as it’ll get and the text on the box at 750 mm is entirely legible:
Arducam Motorized Focus Camera – 500 mm
At 1000 mm, which is basically the edge of the desk all this junk sits atop, the mug and text become slightly fuzzy, so the DoF doesn’t quite reach them:
Arducam Motorized Focus Camera – 1000 mm
I limited the focus range to 1500 mm, which doesn’t much change the results:
Arducam Motorized Focus Camera – 1500 mm
I could focus-stack a set of still images along the entire range to get one of those unnatural everything-in-focus pictures.
I spotted this little gadget chugging steadily across a table in the living room:
Chestnut parasite larva – detail
Nearby, two of its friends / siblings / clones remained near their landing craft:
Chestnut parasite larvae – overview
They’re about 5 mm long and, although there are no larva-size holes visible in the chestnuts tucked inside the burr, that’s definitely where they started their journey.
A few hours later, the rest of the crew bailed out:
Chestnut parasite larvae – irruption
The exit hole must be on a nut under the curve of the husk, but they’re sufficiently squishy to wriggle their way out. The little brown dot over on the left belongs to the top larva of a pair queued in the exit corridor:
Chestnut parasite larvae – exiting husk
I lost count at 18. There’s surely more where they came from, so I replaced the plate with a bowl to reduce the quantum tunneling probability.
In an ideal world, they’d grow up to be chestnut weevils, but I put them out near the suet feeder and, a few hours later, my offering was accepted.
The values written to the I²C register controlling the Arducam Motorized Focus Camera lens position are strongly nonlinear with distance, so a simple linear increment / decrement isn’t particularly useful. If one had an equation for the focus value given the distance, one could step linearly by distance.
So, we begin.
Set up a lens focus test range amid the benchtop clutter with found objects marking distances:
The camera defaults to a focus at infinity (or, perhaps, a bit beyond), corresponding to 0 in its I²C DAC (or whatever). The blue-green scenery visible through the window over on the right is as crisp as it’ll get through a 5 MP camera, the HP spectrum analyzer is slightly defocused at 80 cm, and everything closer is fuzzy.
Experimentally, the low byte of the I²C word written to the DAC doesn’t change the focus much at all, so what you see below comes from writing a focus value to the high byte and zero to the low byte.
For example, to write 18 (decimal) to the camera:
i2cset -y 0 0x0c 18 0
That’s I²C bus 0 (through the RPi camera ribbon cable), camera lens controller address 0x0c (you could use 12 decimal), focus value 18 * 256 + 0 = 0x12 + 0x00 = 4608 decimal.
Which yanks the focus inward to 30 cm, near the end of the ruler:
Arducam Motorized Focus test – focus 30 cm
The window is now blurry, the analyzer becomes better focused, and the screws at the far end of the yellow ruler look good. Obviously, the depth of field spans quite a range at that distance, but iterating a few values at each distance gives a good idea of the center point.
A Bash one-liner steps the focus inward from infinity while you arrange those doodads on the ruler:
for i in {0..31} ; do let h=i*2 ; echo "high: " $h ; let rc=1 ; until (( rc < 1 )) ; do i2cset -y 0 0x0c $h 0 ; let rc=$? ; echo "rc: " $rc ; done ; sleep 1 ; done
Write 33 to set the focus at 10 cm:
Arducam Motorized Focus test – focus 10 cm
Then write 55 for 5 cm:
Arducam Motorized Focus test – focus 5 cm
The tick marks show the depth of field might be 10 mm.
Although the camera doesn’t have a “thin lens” in the optical sense, for my simple purposes the ideal thin lens equation gives some idea of what’s happening. I think the DAC value moves the lens more-or-less linearly with respect to the sensor, so it should be more-or-less inversely related to the focus distance.
Take a few data points, reciprocate & scale, plot on a doodle pad:
Arducam Motorized Focus RPi Camera – focus equation doodles
Dang, I loves me some good straight-as-a-ruler plotting action!
The hook at the upper right covers the last few millimeters of lens travel where the object distance is comparable to the sensor distance, so I’ll give the curve a pass.
DAC MSB = 10.8 + 218 / (distance in cm) = 10.8 + 2180 / distance in mm)
Given the rather casual test setup, the straight-line section definitely doesn’t support three significant figures for the slope and we could quibble about exactly where the focus origin sits with respect to the camera.
So this seems close enough:
DAC MSB = 11 + 2200 / (distance in mm)
Anyhow, I can now tweak a “distance” value in a linear-ish manner (perhaps with a knob, but through evdev), run the equation, send the corresponding DAC value to the camera lens controller, and have the focus come out pretty close to where it should be.
Much to our utter astonishment, this appeared on the driveway:
Chestnut burr
We’ve since found half a dozen chestnut burrs in the yard, which means at least two trees (it takes two to cross-fertilize) are growing in the immediate area.
We originally thought they were American Chestnuts, but Mary (being a Master Gardener) found enough references including comparative burr pictures to convince us they’re Chinese Chestnuts.
We’ve seen squirrels carrying the burrs in their mouths from the trees to wherever they bury their food supply, as shown by this gnawed spot on the other side of the burr:
Chestnut burr – gnawed section
I regard this as conclusive proof that squirrels either have no sense of pain or no lips, because I can’t imagine carrying that thing in my hand, let alone gnawing through it to extract the nuts inside.
Each burr contains three nuts, although this empty husk shows some nuts can fail to fill out:
Chestnut burr – interior with failed nut
We don’t know where the trees are, but the squirrels seem to carry the burrs across our yard from north to south, so they can’t be too far from us or each other.
Despite our conclusion, it’s faintly possible they’re American Chestnuts, in which case they’re definitely survivors!
The appendages at the tip of her abdomen were spread to the sides and her whole body moved in small circles, although I couldn’t get a good view of the proceedings. Building an ootheca apparently requires concerted effort, as she was still hard at work when dusk fell.
The NYS DOT has been improving the pedestrian crossings at the Burnett – Rt 55 intersection. I expect this will be a bullet item in their Complete Streets compliance document, with favorable job reviews for all parties. The situation for bicyclists using the intersection, which provides the only access from Poughkeepsie to the Dutchess Rail Trail, hasn’t changed in the slightest. No signal timing adjustments, no bike-capable sensor loops, no lane markings, no shoulders, no nothing.
Here’s what NYS DOT’s Complete Streets program looks like from our perspective, with the four-digit frame numbers ticking along at 60 frame/sec.
We’re waiting on Overocker Rd for Burnett traffic to clear enough to cross three lanes from a cold start:
Burnett Signal – 2020-09-25 – front 0006
That building over there across Burnett is the NYS DOT Region 8 Headquarters, so we’re not in the hinterlands where nobody ever goes.
About 1.5 seconds later, the vehicles have started moving and we’re lining up for the left side of the right-hand lane:
Burnett Signal – 2020-09-25 – front 0752
There’s no traffic behind us, so we can ride a little more to the right than we usually do, in the hopes of triggering the signal’s unmarked sensor loop:
Burnett Signal – 2020-09-25 – front 1178
We didn’t expect anything different:
Burnett Signal – 2020-09-25 – front 1333
We’re rolling at about 12 mph and it’s unreasonable to expect us to jam to a stop whenever the signal turns yellow. Oh, did you notice the truck parked in the sidewalk over on the left?
As usual, 4.3 seconds later, the Burnett signals turn red, so we’re now riding in the “intersection clearing” delay:
One second later, we’re still proceeding through the intersection, clearing the lethally smooth manhole cover by a few inches, and approaching the far side:
Burnett Signal – 2020-09-25 – front 1771
Here’s what the intersection looks like behind me:
Burnett Signal – 2020-09-25 – rear 1
Another second goes by and we’re pretty much into the far right lane , with the westbound traffic beginning to move:
Burnett Signal – 2020-09-25 – front 1831
The pedestrian crossing ladder has fresh new paint. They milled off the old paint while reconstructing the crossing, so the scarred asphalt will deteriorate into potholes after a few freeze-thaw cycles. Not their problem, it seems.
Although it’s been three seconds since Rt 55 got a green signal, the eastbound drivers remain stunned by our presence:
Burnett Signal – 2020-09-25 – rear 2
After another second, we’re almost where we need to be:
Burnett Signal – 2020-09-25 – front 1891
There’s a new concrete sidewalk on the right, with a wheelchair-accessible signal button I can now hit with my elbow when we’re headed in the other direction. It’s worth noting there is no way to reach Overocker by bicycle, other than riding the sidewalk; there’s only one “complete” direction for vehicular cyclists.
One second later puts us as far to the right as we can get, given all the gravel / debris / deteriorated asphalt along the fog line near the curb:
Burnett Signal – 2020-09-25 – front 1957
Which is good, because four seconds after the green signal for Rt 55, the pack has overtaken us:
Burnett Signal – 2020-09-25 – rear 3
If you were the driver of the grayish car in the middle lane, directly behind the black one giving us plenty of room, you might be surprised at the abrupt lane change in front of you. Maybe not, because you had a front-row seat while we went through the intersection.
Elapsed time from the green signal on Burnett: 25 seconds. My point is that another few seconds of all-red intersection clearing time wouldn’t materially affect anybody’s day and would go a long way toward improving bicycle safety.
Unlike the pedestrian crossing upgrade, NYS DOT could fix this with zero capital expenditure: one engineer with keys to the control box, a screwdriver or keyboard (depending on the age of the controls), and the ability to do the right thing could fix it before lunch tomorrow.
The Smell of Molten Projects in the Morning 