I currently have a hex which is my first build. I'm looking to build a new copter for aerial video and I would like to know which configuration (eg. hex, octa, y6) would be best. If there is no definite answer please let me know what your opinion is.
As far as I see it right now Y6 is best for reliability and field of view but the coaxial setup causes more vibrations.
Octa has the most stability but has a higher chance of failure as there are more points of failure. Also i'm concerned about the arms getting in the frame with an octa; can the camera's be mounted low enough to not have this issue?
I plan to use a Sony nex 6 once i get it running reliable enough for my liking.
I do not know yet from my own experience but from everything I have seen, it's hex and up. I am currently using a F450 with 2 axis GoPro Gimbal and it is quite a task to try to get usable footage outdoors. Inside I have gotten it pretty steady with rubber dampeners, balancing, etc. But outside I am still getting vibration. I am currently using APC props and am planning to try Graupners along with a few more tweeks. But from all my testing so far, I believe that a larger rig all together will help.
I also want to use the NEX line. I have a Canon 50D, but that is pretty darn hefty. Take a look at DJI's new S800 for a design that clears the props out of the way. That thing is too pricey for me but it's a cool design. I am personally looking to go to hex or octo as well but haven't quite decided.
Regarding the failure points, also keep in mind that the octo can recover from failures more-so that a hex. An motor failure on a quad and you are down. A hex and you lose some control but can possibly recover. An octo can continue to fly. All this pending the payload.
Just remember to dampen the camera from vibrations. Would love to get away from rolling shutter on camera's all together. Good luck.
I'm starting to lean towards an octo with the S800 design of angled arms but i plan on building it myself.
My current hex takes decent video with the GoPro but since it was my first build I see a lot of room for improvement on the design.
Some quick thoughts.
The GoPro is not a good camera for aerial video in my opinion. It's angle of view is too wide, no stabilization, rolling shutter (jello) issues. I prefer an angle more like 80 to 90 degrees like the Sony NEX-5 with the 16mm pancake lens.
The problem with getting the gear or props in the picture is because the camera is hung under the middle of the craft. I extend the camera out front on a boom and counter balance with the battery behind. I have used three cameras; GoPro, Sony CX-160 and Sony NEX-5N. The two Sonys provide much better video.
The more arms on the copter, the narrower the space between them so it's harder to get a clear shot with camera.
Reliability is important but does it really override having a simpler craft? Six or eight motors are more points of failure than four.
Dampening vibration is far trickier than most folks think. I like using an actual video camera with built in optical stabilization so whatever vibration gets through to the camera is reduced further.
And beyond vibration, the issue of copter rolling and tilting looks bad in video. So it's important to compensate for roll and tilt. Roll is more distracting than tilt so address that first.
My 3DR quad weighs in at a bit more than 2 kg with a Sony Handycam and flies for about six minutes. That suits my purposes well. I plan to increase my battery from 2200 mah to 4000 for longer flights approaching ten minutes.
My videos are not on YouTube, only on my own ftp. Msg me if you'd like to download them.
Until I get the new build running perfectly I plan on using the GoPro then I will switch to the NEX-6 and get some real quality video.
As for the arms i'm think of using a 15 degree upward angle to have them out of the way. I haven't seen any discussion regarding angled arms does anyone know a reason not to do it that way?
Angle the arms upward? That seems like a lot of work. Would you then angle the motors 15 degrees to the arms so they are vertical? As long as the motors are all vertical, I don't see why not. But why angle six or eight arms when you could just put the camera out front? Keep it simple.
Yes the motors would end up positioned as normal in the end.
It seems to me when putting the camera out front it would cause an imbalance. Also how far out front does it need to be to avoid getting the arms in the shot? Do you have a picture of your setup as an example?
The camera mounting hole is about 14cm from the center of my quad. With GoPro the props still show in the top corners when the camera is angled downward maybe 15 deg. Nothing shows with Sony cams with narrower field of view.
My rig is a work in progress. I'm now 'hacking' this together to make sure it works before I 'do it right' with better hardware.
I've included two photos. One shows the boom mount and roll articulation. This is before I added servos. The cam boom is hinged onto a plate that in turn is screwed onto four rubber iso mounts which are attached to underside of the 3DR quad main body (or whatever they're called). For the roll hinges I've adapted RC plane nose gear steering blocks.
The second photo is the underside as it is now with roll and tilt servos hacked on. They function ok but I have not flown it yet.
To balance the camera, I attach a second 2200 mAh pack onto the rear of the boom. In my experience, the balance is not super critical. I just hold the quad by the top plate of the stack at the mid point fore and aft and see if it hangs level.
Once this design gets more firmed up I'll decide on final battery (batteries?) and do a proper job with the linkages. I have some ball joint links on order. I also plan to cut a new 'boom mounting plate' from air cell PVC instead of ABS.
Did I cover your points? I hope I've included the photos properly.
How smooth is the moment of your camera with that mount?
I was planning on using a gimbal so I wouldn't be able to really mount it forward as you have although your method seems like it might be a good alternative.
Regarding reliability versus simpler craft, I guess that can be argued both ways but I lean towards the reliable craft. I am still not completely clear as to whether 6-8 props also translated into more stabilized shots, but if that is the case it is a major factor for me. Right now, with my F450 Quad shots looking pretty stable indoors but jittery outdoors in light wind, I am either thinking that the quad is not a good platform or there are some software stability issues in the APM2 that I need to learn to tune out.
@Colin, re: smoothness of movement, I haven't flown the quad with roll servo yet. I've only flown with the boom tied off, fixed. I get pretty good video, nil jello or vibes. But the roll and tilt aiming of the camera keep it from being "pro video smooth" like a SteadiCam rig or pro boom.
I'm interested in your results between indoor quiet air and outdoors in light wind. When you say "jittery" are you talking about vibration or shake?
I think of vibration as small, fast oscillations and shake as larger, slower movements. I don't connect wind with vibration.
It does make sense to me that more motors mean more flight stability. And more motors could make for smoother vibrations (does that make sense?). For example, eight motors vibrating at different frequencies make a smoother "sound" (white noise) than four with resonances and harmonics that "screech" or "squeak".
The sound of an octo compared to a quad is probably "smoother". I've not seen, held or flown an octo live in person so I'm guessing about this.
I have seen excellent video examples from quads, hexes and octos, so there's got to be more to image stabilization than just number of arms/motors. Look at what this prolific guy has done. He's tried several and a quad is his platform of choice in the end.
It would be helpful to agree on terminology used to describe video image problems. I think separating copter terms from camera terms would be good. Terms like these are tossed (heh, heh) around: jittery, shaky, jello, rolling shutter, stabilize, smooth etc. There's a distinction to be made between copter stability and image stability; each should be considered separately. Does anyone know of a discussion on terminology?
What does all of this mean? Ultimate video image quality depends on two separate issues: copter movement and camera movement. Each needs to be addressed with different techniques. Well, of course, there is the issue of getting your copter in the shot, but that's straight forward to deal with.
Sorry if this is too rambling...
Well, I am going to convert my DJI F450 into an F550 Y6. Depending on the results, I may go a step farther and test a F550 X8 in a 120 degree pattern. But I am going to start Y6 first and see how it compares.