808#16D camera modified for near IR for sale

edited November 2013 in KAP Gear Sources
I found the below page soon after purchasing an 808#16D and successfully followed the directions by John Wells.

Simple and cheap near IR conversion of the wide-angle HD 808#16D camera

I recently bought a mobius actioncam and do not use the 808#16 anymore. It includes an unmodified as well as the modified lens and everything in the photo. You will need to provide your own microsd card.

$70 shipped.




  • Hi, What is the quality of the camera like compared with the 808#16D and how much does it weigh?
  • John, I don't own a Mobius, but I checked out some full-res videos made with one, and they're excellent. Full 1920x1080, and the still frames are a lot cleaner than the 808. It still suffers from barrel distortion, but not as bad as the 808.

    This guy uses a Mobius:


    View it at full 1080 and take a look at an individual frame. It's from a glider, so you'll get a good feel for its aerial performance.

    Same guy has a video showing some color shift behavior when it saturates, though. That's the most recent one in his video stream.

  • Sorry to utterly hijack a thread, but this might actually be a selling point for the IR-converted 808 #16D:

    Has anyone tried mounting two 808 #16D cameras, one IR converted one left as-is, both pointing as close to the same direction as possible? I've brought this up in the past, but it seems like you could pull the R channel out of the stock camera and a monochrome image out of the IR converted camera and use them to make an NDVI movie. It would require a truckload of number crunching, but it could be really really useful for people doing soil studies, archaeology, etc. The number crunching would be considerably less if you were pulling still frames out of the video streams and only crunching numbers on the frames of interest. The software for doing the NDVI processing is free, so this might not be too far off-base from what you're doing with these cameras, John.

  • It's possible but a little cumbersome to do the NDVI video processing using ImageJ/Fiji. The processing is reasonably fast but far from real time. It's a lot easier if you use a single camera with a blue or preferably red filter so you don't have to worry about syncing the the frames of two cameras. If the video frames are not synced well there will be too much parallax (assuming the camera is moving around) to combine the images to make a clear NDVI. If the cameras are high enough that's less of a problem.
  • The mobius actioncam weighs 39grams. I use it to fly and record my fpv quadcopter flights.
    No worries about the hijack. I'm glad to start a useful conversation.

    My converted 808#16D is still for sale if you know anyone interested. $70 shipped.
  • I have never tried NDVI!
    PLOTS cover it in some detail: http://publiclab.org/wiki/ndvi-plots-ir-kit but I have no feel for how useful it would be for our work.
  • I could see it being useful for identifying buried archaeological features, similar to how you're using IR to gauge soil depth by the health of the above-ground vegetation. I'd love to post an example, but we don't have a lot of buried archaeological sites here. The only example I have was made on the ground using a tripod as an example of using NDVI to gauge the health of the vegetation for agricultural purposes. Not really the ideal example.

    Thanks for posting the PLOTS page, John. I was puzzling over how to do NDVI or something similar using a single camera. But the idea of using an IR converted camera with a superblue filter on it made a lot of sense, and helped answer some of the questions I had from Ned's post above. Time to read up on filters! (Hey, we've got some Schott BG3 at work. I gotta play!!)

  • Sold my camera to one of the KAP users. Thanks.
  • I forgot about this thread until I followed Tom's link from his filter post. NDVI has been used by archaeologists for the reasons Tom mentions. NDVI is much better at detecting changes in vegetation vigor than using NIR on its own. Vegetation is bright in NIR but so are a lot of other features so NIR images tend not to have much contrast. NDVI accentuates plant vigor because a healthy plant reflects a lot of NIR and very little red and as it is stressed it still reflects a good bit of NIR but the reflected red signal increases.

    For what it's worth, a blue filter doesn't work quite as well as a red filter since the reflected blue light remains fairly low when a plant is stressed so the NDVI of healthy and stressed vegetation is not as different as it is when using a red filter. I tried to explain this in a PLOTS note here: http://publiclab.org/notes/nedhorning/11-01-2013/why-a-red-filter-should-work-well-for-ndvi
  • Thanks for posting that link, Ned. I've got a Wratten #25 in my bag I can play with.

    I was excited to see the URLs for the spectral libraries. We collect these for a really limited subset of materials at work, mostly stuff that's either very black or very white in reflection, and things that are very opaque in transmission. A good bit of my job over the years has been to sit in front of a spectrophotometer scanning anything from lenses and optical filters to paint samples, construction paper, tape, or anything else we might use to make something black. The ones that work go in our database of usable materials. The ones that don't get a note: "Don't use me!"

    In all that time I never thought to stick a leaf in the machine! GAAAH! But that would've been useful information to have before wandering down this garden path. I'm glad to see you took a more methodical approach. I'm looking forward to trying the Wratten #25 filter.


Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion