New toy for us camera geeks

Stacky

Member
I suspect the ability to use live view and other canon/nikon tethered shooting software would also mean a 2 man operation is more necessary for stills photography. Even now shooting AP stills with an intervalometer taking the shots for you, its still a nervy operation doing it on your own.

Ok, now that sounds cool. So live view over wifi with no need for a video transmitter?

Only problem is... Wifi is 2.4GHz, same as my control system receiver. Can they co-exist on a single airframe? Or will the wifi link swamp the Rx?
 

DucktileMedia

Drone Enthusiast
Yeah, there's really only 2 ways to do AP, shoot and pray/trial and error or have a 2nd person. On production shoots the DP usually wants to be part of it anyways. So it could be really neat to actually let them have full control of the camera. Its not hard at all, they only have to understand that the camera needs to be left alone if the pilot has an orientation panic. The way we do it, is to leave it alone if it is way off coming in for landing or snap it back to its correct orientation quickly if it is less than 90deg off. It can be so damn hard to focus on the heli when that gimbal is spinning in circles. I think it would be really cool to actually have those retracts just for this reason. Less to confuse the pilot. but this isnt what the thread is about.
 

jes1111

Active Member
In a decade the DSLR wont be the perceived "serious" camera. With the improvements in technology and screen display speed the need for a single reflex mirror system are now starting to be of less importance which is good for all of us, not just in terms of weight. Having the rear element of a lens closer to the image sensor on higher end cameras with all the advantages of interchangeable lenses will be nice.
You're probably right - but that's a long way off, I reckon. Nikon/Canon glass bought today will need to be useable (at its "35mm equiv" focal length) for many years to come if they want to avoid a revolution. Ithink its obvious that both Nikon and Canon are "playing safe" with the V1 and EOS-M launches - creeping in at the bottom end. Interesting that Canon called it the "EOS-M" - what will the next model up the ladder be called? - this naming pattern offers no obvious path.

Ok, now that sounds cool. So live view over wifi with no need for a video transmitter?

Only problem is... Wifi is 2.4GHz, same as my control system receiver. Can they co-exist on a single airframe? Or will the wifi link swamp the Rx?
Yep - no video transmitter. The 802.11n wi-fi standard is 5GHz, so no conflict with 2.4GHz. Although I gather (I haven't tested) that RC and 2.4GHz wi-fi can coexist if you have sufficient space between aerials. The 5GHz band is also much less over-crowded than 2.4GHz and has better rejection of interference.

Interestingly, "Live view" (both Canon and Nikon) is not a video signal at all: it's actually a stream of lo-res jpegs - easier on the bandwidth.

I suspect the ability to use live view and other canon/nikon tethered shooting software would also mean a 2 man operation is more necessary for stills photography. Even now shooting AP stills with an intervalometer taking the shots for you, its still a nervy operation doing it on your own.
If you're flying with regular RC sticks, yes - maybe. But the next step forwards is "tablet control" (with thumb-operated joysticks either side of a touch-screen). Combined with reliable position hold and "moveable waypoint" flying, single operator would be a cinch. Significantly, both Nikon and Canon offer full SDKs for their camera control capabilities - so the camera control can be built right in to the GCS.
Yeah, there's really only 2 ways to do AP, shoot and pray/trial and error or have a 2nd person. On production shoots the DP usually wants to be part of it anyways. So it could be really neat to actually let them have full control of the camera. Its not hard at all, they only have to understand that the camera needs to be left alone if the pilot has an orientation panic. The way we do it, is to leave it alone if it is way off coming in for landing or snap it back to its correct orientation quickly if it is less than 90deg off. It can be so damn hard to focus on the heli when that gimbal is spinning in circles. I think it would be really cool to actually have those retracts just for this reason. Less to confuse the pilot. but this isnt what the thread is about.
Given a tablet-based system, two-operator control would be as simple as having a second tablet on the same ad-hoc local network. What controls/capabilities are assigned to each station would be a software configuration issue only. And you could have a "lock the other guy out" button.
 


R_Lefebvre

Arducopter Developer
Yeah, there's really only 2 ways to do AP, shoot and pray/trial and error or have a 2nd person. On production shoots the DP usually wants to be part of it anyways. So it could be really neat to actually let them have full control of the camera. Its not hard at all, they only have to understand that the camera needs to be left alone if the pilot has an orientation panic. The way we do it, is to leave it alone if it is way off coming in for landing or snap it back to its correct orientation quickly if it is less than 90deg off. It can be so damn hard to focus on the heli when that gimbal is spinning in circles. I think it would be really cool to actually have those retracts just for this reason. Less to confuse the pilot. but this isnt what the thread is about.

Don't you just fly in simple mode? Then orientation doesn't matter at all.

Yep - no video transmitter. The 802.11n wi-fi standard is 5GHz, so no conflict with 2.4GHz. Although I gather (I haven't tested) that RC and 2.4GHz wi-fi can coexist if you have sufficient space between aerials. The 5GHz band is also much less over-crowded than 2.4GHz and has better rejection of interference.

Interestingly, "Live view" (both Canon and Nikon) is not a video signal at all: it's actually a stream of lo-res jpegs - easier on the bandwidth.

Ah, ok, that should be fine then. I do have a 5.8GHz FPV system, but then this would become much less necessary. I had thought it would be handy for the pilot to still have a small, simple, FPV cam, and the camera operator uses the main cam. Just for basic things like "am I above that tree or not".

I hadn't considered the EOS live-view when considering the Nex5, so thanks for pointing it out. That is a real game-changer. Sony doesn't have anything similar?

So what additional equipment is required to do this? Just a wireless router?

If you're flying with regular RC sticks, yes - maybe. But the next step forwards is "tablet control" (with thumb-operated joysticks either side of a touch-screen). Combined with reliable position hold and "moveable waypoint" flying, single operator would be a cinch.

We have this working now. It's pretty cool. You basically have a laptop in your hand showing Google Earth imagery, double-tap, and it flies there. Simple as that. Can also be done on a laptop obvoiusly, which would also allow simple multi-tasking between the GCS and the Live View software. You could even potentially set up an extended screen with the instance of GCS and the Camera software running on each. Now that would be perfect. IMO, that's actually the way to go because at this point we still need the laptop running in addition to the tablet anyway. The tablet is just sort of cool-factor at this point, I think. The reason is the laptop is basically a relay. Tablet-->AdHoc Wifi-->Laptop-->900MHz Radio Modem-->Copter
 


jes1111

Active Member
Do you need to add use this $850 device to do this?

http://www.usa.canon.com/cusa/consumer/products/cameras/consumer_cameras_wft/wireless_file_transmitter_wft_e7a#Overview

That's... ouch.

How does this end up comparing to using A/V out to a transmitter, and then an IR control for the camera? I'm not sure what settings you can access with a wireless transmitter. Obviously varies by camera.
No, you don't need that. Haven't tried that new model but all the previous efforts, from both Nikon and Canon, were completely pants ;)

The trick is to use the USB connection on the camera and a "USB Device Server" (basically a variation on serial over IP using VCOM). At the moment I'm using a Silex SX-3000GB USB device server into a Ubiquiti 5GHz "Bullet M" (with modified connectors). The USB device server has two ports, so I have a little FTDI converter to feed serial telemetry from OP CC to the GCS over the same link. On the ground is another Bullet M, so they are set up as a "point-to-point" AP.

Both the Nikon (paid) and Canon (free) software allows complete control of the camera - more or less as if you had the camera in your hands - change mode, shutter speed, aperture, ISO, focus point, exposure compensation, bracketing, interval shooting, etc. - everything :) That's a very different proposition to setting "auto everything" and hoping that all your shots come out the way you want them.

Only Nikon and Canon have this capability - unless you want to fly a Hasselblad ;) - I've not seen any other makes implementing this kind of remote control.
 

R_Lefebvre

Arducopter Developer
So how much is that setup? And what is range like?

I wonder if the same thing could be done with two Xbees. I'm not sure what the baud requirements are for the Canon. I'm running Xbees at 56k baud, not sure how high they can go though.
 

jes1111

Active Member
So how much is that setup? And what is range like?

I wonder if the same thing could be done with two Xbees. I'm not sure what the baud requirements are for the Canon. I'm running Xbees at 56k baud, not sure how high they can go though.
About $200. Range - I've tested it to 500m line-of-sight across open ground. Data rate drops, but that's the beauty of live view - it's a stream of individual jpegs so as the bandwidth drops it just reduces the effective frame rate. In practice why would ever need to be more than 100-200m from the aircraft's position?

Problem with XBees will be packaging the USB feed to go across that link. Why bother? WiFi has all this worked out for us already.
 

R_Lefebvre

Arducopter Developer
For range, what I'm interested in doing is shooting in remote areas. Not intending to fly hundreds of miles, ;) but I may want to be able to shoot pictures from further than 1-200m away. Landscape photography in remote locations, that kind of thing.
 

jes1111

Active Member
This is the pivot point, I believe. At anything more than 100 meters the craft is no more than a dot. At 500m it would be all but invisible. So if one is looking to capture images at that sort of range then you have to be flying full IFR anyway. So this is a cheap, short range, off-the-shelf solution. Longer wavelength radios can do the distance but you face the problem of packaging/unpackaging the USB over that link - the Canon/Nikon software application (or SDK) wants to see and identify the camera as a USB device (and we can't change that). Probably possible but would require a lot of work, I'm sure. You could also set up a relay system with "assist" craft hovering at interval points between you and the camera - but, again, that's a whole new level of complexity. Note, also, that the Liveview has an inherent lag, so it's definitely not safe to use it as FPV.
 

R_Lefebvre

Arducopter Developer
Ok, great info. I had seen somebody else us an IP webcam for FPV, but the there is lag on that as well which can be problematic.

Using simple A/V out of the camera, it's easy to put together a system to achieve at least 1km of range if not more, using a directional antenna on a tracker. And I know the cams I have played with will output image details on link if you want. But I think you'd have to be full auto for shooting.

I'd be curious if any of the cams can be made to change settings using the IR link which can be hacked and replaced with a microcontroller. I don't have much info in this area.
 

Top