They come and go by day...
...and they come and go by night.
The owls will win control of the nest box, but not before the starlings have made complete and utter pests of themselves, as long time viewers can attest. Also, the starlings continuously dump new bedding material into the box. I cleaned out the box this weekend to ensure that it wouldn't pile-up so high that it would block the view of eggs (when those start to arrive in a month or so), or bury any of the infrared illuminators (LEDs) that make nighttime pictures possible. Unfortunately, it appears that little short of daily cleanings would ensure that, and there just isn't time enough for that. On the plus side, once the starlings are persuaded to give-up, owl activity should trample and compress most of that material into a less problematic mass.
Meanwhile, I experiment with a means of performing video capture that is as hardware, software, and platform independent as possible. The ultimate in such solutions seems to be stand-alone boxes that accept analog audio/video input, digitize it, encode it using standard codecs, and transmit it over Ethernet using standard streaming protocols. With modifications, my custom owl cam software should be able read those frames, clean, adjust and timestamp them just as it has for years, and produce the familiar still images.
That's great in principle, but making it happen may be a bit tricky with the video server box (an S2071 from 3Svision) I'm currently experimenting with. Among other problems, customers can't get firmware updates [they have since provided one in response to an email request], the manufacturer's specifications are wrong in several areas (as is often the case), and those just happen to be the areas I was depending on (as is often the case). The problems can still be solved, but the software won't be as simple, or (possibly) as platform independent, and that means more potential points of failure (if not now, then in years to come), which is disappointing.
So, two things: (1) If anyone has experience with better video servers, at comparable price-points, please tell me about them. (2) The owls could begin nesting any time, and my system is not yet adapted to the new gadget, so images from the box are likely to begin a bit late this year. Sorry about that; I don't like it either.
On the other hand, the S2071 has some very nice features (one of which was left undocumented in the specifications, but may make the S2701 almost as good, in practice, as having a DVR recording everything), a price-point per video channel digitized that's lower than any other devices I've used, and it produces higher quality (very good noise reduction and de-interlacing), and slightly higher resolution, images than any video digitizers I've ever used.
Finally, there's this tantalizing possibility: Provided I upgrade my home broadband service to something with higher and more dependable upload speeds than Time Warner is currently providing me (my cable bill is already stunning, so giving TW even more money is not a good option), and I can find some well-connected, high capacity server to replicate and re-serve the video streams that the S2701 produces (ideally without reducing their quality or size, or presenting them surrounded them with advertisements), I might finally be able to offer to viewers live audio & video of the same quality I see, which, believe me, will transform what you will learn about the nesting process as you watch the cam'. That said, please, do not get your hopes up. There are a number of big "ifs" associated with this potentiality, and no guarantee of solving the various technical, financial, quality and other problems. All I can say is that using a video server like S2071 should solve the first of the problems, which is an obvious prerequisite for solving the next problem, and the next, and so on. So, for the first time, I'm in with a chance.