i went for another long-ish drive with the raspberry pi noir camera v2 a couple days ago. instead of west to the beach (like the day before), i went east to the country (and through the city). this drive wasn’t as relaxing, perhaps as there were more people on the road and as i was paying more attention to driving a particular route (with a number of specific turns) slowly so i could make it back home (with the small battery) without stopping at a public charger (and without running the battery too low). i’m not sure if i went any faster than 45 mph the whole way, and most of the time was slower than that (as i tried to take roads with slower speed limits, multiple lanes, or rural roads with little to no traffic). i was able to average 6.8 mi/kWh the whole 102.3 miles (the equivalent of 147 Wh/mi or 229 MPGe, about twice as good as my car’s EPA rating).. looking quickly at my notes (which are missing lots of data), that appears to be the highest efficiency i’ve gotten on a long drive. when i made it to some of the rural parts that i had never been to before, i discovered a very nice, quiet, peaceful area, which was a pleasant surprise.

this was shot with the same settings as the one the day before: 1/128 shutter speed @ iso 60 with custom white balance (and lee 87c visible light blocking / infrared pass filter), at an interval of one shot every 3 seconds (and the video rendered at 10 fps). to try to compensate for the strange glare i was getting from the filter, i added a long extension rod to the suction cup mount that was on the windshield.. the camera was able to sit further back into the car to reduce its direct exposure to sunlight. this helped some, though it still had some sun hitting it (depending on the direction the car was going) and the whole camera was bouncing up and down the whole time (due to the additional leverage of the long rod it was mounted on). i was wondering if i was going to get anything useful out of it or if the suction cup was going to fall off during the drive.. fortunately the camera movement isn’t too noticeable in the time-lapse (perhaps the rate of movement was somewhat aligned with the interval the shots were taken, and/or the time-lapse playing back at a slow stop-motion-like frame rate of 10 fps effectively helps reduce the visibility of the constant bouncing of the camera), and it held the whole time (just under 4 hours). i was also wondering if there would be motion blur as the shutter speed wasn’t super fast.. looking closely at the individual images, they aren’t as sharp as they could be, but again, being played back in a time-lapse video like this hides the imperfections.

again, this is by no means a perfect video.. i think next time i’m going to see if i can mount the camera behind the tint at the top of the windshield (to see if that works to block the glare from sunlight hitting the filter directly). i’ll also be doing a future time-lapse with an interval of 1 second (and more jpeg compression or lower resolution so the (non high speed) sd card can keep up) to make the time-lapse video appear more smooth. i forgot to mention in the previous video notes that there is a part of the image area that is not in focus like the majority of the image (seemingly due to imperfect alignment of the lens to the sensor), and i’ve attempted to hide this by placing the camera upside down (so the lower + left area(s) of the image are the ones that are less sharp).