Hi group! While I have used the philopod technique from time to time, I have never managed to get good enough results from it. I always had difficulties with the pendulum motion of the weight at the end of the string and would take a long time to get it steady. Another issue is that when you tilt the camera up and down, the spot where the string is attached to the lens moves away from the vertical line down from the NPP, and up or down, respectively, from where it rests when the camera is horizontal, resulting in errors. And I found fixing the string was also bothersome.
Nevertheless, the technique is very attractive for being simple, cheap, lightweight and potentially accurate. And it has a nice 'manual' touch to it, requiring skill :) So I wondered how it's drawbacks could be lessened. And I've found a simple way to improve the technique, which noticeablly diminishes the problems with pendulum motion and the tilting issue, and, on top of that, opens up a few new options which are missing from the original technique. In a nutshell, the modifications are as follows: - instead of a line, a loop of thread is run around the lens and through the weight - the thread can slip freely through the weight - optionally, a simple brace keeps the loop in place on it's way over the lens Using a loop instead of a single thread fixes the weight with two lengths of thread instead of just one, reducing the degrees of freedom for the pendulum motion. Since the weight can slip, it will always slip so that it's as low as possible, and ideally just under the NPP. The optional brace, used to guide the thread over the top of the lens, has points through which the thread passes out and down. These fixed points are positioned so that the NPP is precisely half way between them. They are positioned to be at the same height when the camera is in pano shooting position (portrait for me), and keep the thread far enough from the lens to allow sufficient tilting to capture the zenith and nadir. Guiding the thread with the brace as described will keep the weight under the NPP even if the camera is tilted up or down. The loop can be attached to the brace or slip through it, the effect is the same. The brace is only needed if the thread will not remain in place where it runs over the lens, or if the shooting pattern involves tilting of the lens. If it's just a matter of a single-row panorama with a compact camera, loop and weight are enough, though the brace never does harm. As a thread I use thin nylon line - it should be just thick enough to not show elasticity when weighed down with the weight and not to curl up. It should not be made from twisted strands. Like thin fishing line. You can test a rough version of the technique by simply taking any weight with a hole in it (like, a smooth ring), and some three, four metres of fishing line - for a quick trial even sewing thread will suffice, though it's bothersome to handle and less precise. The hole in the weight where the thread passes through should let the thread slip through with as little friction as possible. Thread the line through the weight, close the loop by knotting the thread's two ends together, and hang the loop over your lens roughly where the NPP is. On my Samyang 8mm stereographic fisheye, there is a handy groove just about there. Take a coin as a simple target and put it on the floor. Now you lift the camera so that the weight hovers just over the target. You will notice that the hovering is much easier to accomplish and keep steady than with just a single line: There is practically no lateral pendulum motion, and the residual friction in the weight is so small that the weight can always slip along the thread to keep both halves of the loop taut. Now shoot a horizontal pattern. Before every exposure, make the weight hover just ove the target, lower it until it touches the target and just then take your picture. Obviosly, with the thread not fixed to the lens, when tilting the camera it will not be optimally positioned any more on the lens: it will slip or be deflected by the lens. This is where you see how a brace on the lens can help by preventing the slippage and touching of the lens and allow other than horizontal shooting patterns. A simple brace is quickly made: cut a ring of thin card, which has as it's inner diameter the diameter of your lens just where the NPP is, and an outer diameter a good two centimetres or so larger than the inner. Fold it along a radial line. Fix this brace to the top (like a cap) of your lens so it's in-plane with the NPP and it's surface is at a right angle to the optical axis. Thread the loop of thread in between the two halves of the folded card ring. The thread now exits just to the left and right of the NPP and gravity keeps the whole length of thread, and therefore the weight, it in plane with the NPP. This sounds more complicated than it is, if you just try it out you'll instantly see the point, and the improvements over the simple philopod technique, if you have ever tried it. With the brace in place, you can now tilt the camera far enough to easily capture nadir and zenith. While a coin or logo are fair enough targets, You can make more of the target. And you can make more of the weight: Fix a light horizontal bar on the weight. Since the weight cannot rotate (except for an admittedly annoying rotating penulum motion around the vertical) the bar will always be at the same angle to the direction the camera points to. So if you use a target with, for exaple, a star pattern on it, you can see how making the bar parallel to the axes of the star will help you implement a shooting pattern more precisely then by just doing it by gut feeling. A very handy thing to put into the center of the target is a compass: That way you get at least a rough idea of absolute yaw. You can orient the shooting-pattern-star with the compass. The handiest thing to have as a target is probably a smartphone or tablet, since it can be programmed to display the target (star, compass and all) next to other helpful bits and bobs like the GPS position, time of day, or QR-codes for later interaction with the image-processing system... and the touchscreen could register the weight touching down on the target and trigger the camera's exposure. It might even take a picture at the same time, theoretically allowing an approximation of the camera's position and orientation at the moment of the exposure, which could be used to recitfy the images before stitching... So far I see only one disadvantage of the technique: since the thread is doubled and in-plane with the NPP, it gets in the way more, requiring a bit more masking then a single string. Kay -- A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ --- You received this message because you are subscribed to the Google Groups "hugin and other free panoramic software" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/hugin-ptx/8fb9efa8-83cb-499d-ad7c-b85f609bdabc%40googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
