This is a summary of my tests of the RPi HQ Camera and some Canon EF-S lenses.
(all photos made with the HQ camera are untouched and full size, click to view).
The results are not bad, the Canon lenses are usable, easy to change, focus is manually, they enhance the quality. The 6mm RPi is no match to any Canon, the 16mm is not bad at all.
Much has already been written about the new (spring 2020) HQ camera available for the Raspberry Pi.
Now one can argue if this is a HQ camera (12M pixels, no focus facilities), it certainly got the interest of the Pi community, the cameras and lenses seem to sell well.
Not to be positioned as a Pi Camera V2 replacement (much more expensive, larger in dimensions), it can be a quality camera to do projects with.
Certainly not a photo/video camera replacement (even a phone camera is much easier to wok with, depth of view and focussing is too limited) , but will have its place in vision applications demanding higher quality images under computer control.
Since the camera comes without lens, but has a C or CS lens mount adapter, it is possible to choose a lens, like the two available form the Pi Foundation.
I have bought both together with the camera, resulting in reasonable images but a pain to focus, let alone to change the focus often. The small one has a reasonable depth of view, and a wide view and lots of cushion distortion and color aberration. The bigger one is better, more a telelens.
The C mount allows more freedom in the choice of lenses, so I thought about checking out my (really high quality) Canon EF lenses.
This requires a C mount to Canon adapter. A metal one is on its way (May 10) from China, a metal one. In the meantime I set my 3D printer to work on 3D printed adapters.
I found two in Thingiverse:
and printed both.
Some remarks about the two designs:
The STL is missing the sizing information as required by the slicer program, as noted in the comments section. The trick to get a correct STL is to upload the STL in Autodesk Fusion and export as STL (with a new name of course!). It then prints fine in the right scale, once rotated and landed lying on its Canon side with supports and a brim. Reduce layer to 0.1.
It fits the Pi camera without adapter. I suppose I should have used a finer print, screwing on was hard, a gap was still present.
The tripod adapter needs to be removed from the camera, otherwise it will not fit.
The stand makes it a very stable setup on a table. I did not yet glue a tripod screw on it.
Prints fine, rotated and landed lying on the Canon side, with a brim. Reduce layer to 0.1.
It fits fine in the Pi camera (without c to cs adapter). The camera’s tripod support can be used, the result is quite unstable with a small table tripod and a real lens, a real tripod works fine, but the Raspberry Pi computer needs to be close and not dangling on the side. A longer camera cable of 50 cm helped a lot.
With each lens a photo of a test chart, printed on my laser printer, on the wall and a plant in the window.
Distance between camera and object for all lenses is 60 cm.
Focussing and capturing photo has to be done manually by looking at the monitor display running this command:
$ raspistill -fw -md 3 -q 90 -k -v -o filename
The – fw switch requires downloading and building Raspbian userland, it gives a numeric indication of the quality of the center part of the screen. Very helpful to get to the best focus.
The -k -v keep the camera image on screen with a prompt until Return key is pressed. Note that the -k in combinationwith other image formats like PNG do not work.
Unable to focus
A Raspberry Pi or Arduino is a perfect companion for experiments, but you need some hardware support during development...
Over the years I have collected several tiny touch LCD screens for the Raspberry Pi which are based on sending the frame...
A page describing how to get the latest Freepascal and Lazarus development systems on the Raspberry Pi in optimal form...
On this page I will show you how to access the Raspberry Pi in a Windows environment. - Acces the home directory on t...