User avatar
ab1jx
Posts: 868
Joined: Thu Sep 26, 2013 1:54 pm
Location: Heath, MA USA
Contact: Website

Re: setting lens focus

Thu Nov 30, 2017 12:43 am

This page has a goldmine of ideas: http://truetex.com/raspberrypi.htm I found it searching because my question was whether the original lens is actually glass or some kind of plastic. It seems perpetually dirty, which would imply plastic, which could also have adverse effects from aging.

This guy takes the sensor seriously and retrofits better lenses to it.
Image
The prices for his mounts seem high unless you consider that they're individually custom machined. They could probably be mass-produced more cheaply using CNC techniques if there were demand enough. These are essentially prototypes. I wish he provided a gallery of images taken with his modifications.

I also found a discussion of the spectral response of the OV5647 sensor at https://publiclab.org/notes/khufkens/11 ... efficiency

Something like this
Image

My concern would be the intersection areas from blue-green and green-red where the response is down 50% or so. I don't see a way to correct for it though.

Fer
Posts: 35
Joined: Tue Feb 21, 2017 10:27 am

Re: setting lens focus

Thu Nov 30, 2017 11:31 am

The page you linked seems quite old.
The author says there's no way to get a raw image from the camera module; actually it is possible since a while.
C/CS-Mount camera modules are available without having to resort to expensive hand-manifactured custom mounts (search for Arducam).
Spectral response of OV5647 is correct; individual curves have to dip at intersections (as it's shown on the graph) because it's the combination of the three curves that depicts the actual RGB response.
Without the dip, you would get anomalous peaks at say 500nm and 600nm.

maheshkarajol
Posts: 1
Joined: Mon Dec 16, 2019 8:02 am

Re: setting lens focus

Mon Dec 16, 2019 8:06 am

can anyone send the calculations for NoIR camera V2?

Return to “Camera board”