High resolution, fixed focal length camera for aerial photography

[Amp34]

A friend has asked me for some recommendations for a camera to put in a plane for aerial photography. It would be used for wildlife censuses so needs to be high resolution and able to cope with taking and saving a photo every second or two (several hundred+ so buffers will run out).

He's currently looking at the sigma DP range but are there any other options? From his calculations he wants something in the 40mp range and with a fairly wide angle lens (ff eq of around 20-30mm). I'm guessing battery life will be important as well, along with weight and cost (not overly expensive, plane crashes... Say up to £1k).

With a decent SD/CF card would 40mb a second real world write speeds be possible?

 

[D.P.]

The sigma cameras don't actually have the resolution they claim. Each pixel capture R+G+B instead of one or the other, so there is a higher chrominance resolution but equal luminescence resolution compared to a bayer design. Sigma simply multiply the pixel count by 3 as some kind of equivalence to bayer sensors. However, as we know form video compression, color resolution is actually much less important than luminescence (in apart why regular color filters have twice as many green pixels). The Foveon sensors are very nice but i wouldn't pick a camera over other due to the claim megapixels, too many other variables.


What your friend actually need to do is work out what size the animal is, what height the plane will be at and how many pixels will be needed to identify each animal. With a bit of math you can then work out required field of view and sensor resolution.


One of the important things with any plane mounted camera is rolling shutter issues, a camera with a global shutter really makes things much better.
Something like this can be the results of a rolling shutter: https://www.youtube.com/watch?v=TId15XAumqI

 

[King4aDay]

A friend of mine uses m43 gear on drones for seal censuses, although I'm hazy on the details. You probably need to know precisely your friend's weight, buffer, battery, focal length and resolution requirements to pick the best compromise.

Probably better off contacting authors of similar studies than asking on a random computer forum to be honest!

 

[Amp34]

Thanks for the comment on the actual resolution D.P.. I will relay that.

I think he knows a fair bit about what he wants, he knows what resolution he needs and the approximate range of focal length. That said I will see about contacting some authors to give him a hand.

m43 cameras is what I suggested to him, but at the moment I haven't looked into which would be best, and this is one of the reasons for this thread. Buffer is essentially irrelevant because the number of photos being taken will quickly overload it, hence the question about whether a fast memory card would help. I guess I could do the calculations myself with that one, looking at real world write times for fast cards, but was hoping people would have real world experience.

Basically if I can get a good list of cameras that could fit the needs in the OP then we could go through them and have a look and see if any of them are suitable.

 

[Rilot]

What is he mounting this camera on? Is he using a motorised or brushless gimbal? If so, then the gimbal will largely dictate what cameras and lenses he can use.
Are we talking a real plane or a RC plane?

 

[Genoma]

Was wondering the same. If it's a real plane how will he be shooting?

 

[Amp34]

Sorry, yes, an RC plane! The camera will be static inside the plane, pointing downwards, as far as I'm aware there is no gimble needed.

Basically i'm just looking for a list of cameras that would fit the brief, fairly small, with a high resolution and either a fixed wideish angle lens or the option to fit one. Weight is a factor but not the biggest issue as we aren't talking about DSLR's.

Something like the Sigma DP series (although that appears to have sensor issues), or the Sony 7R II (but a little on the expensive side). Just wondering if there are similar cameras I don't know of?

 

[nolimit]

I have the original Sigma DP1 (now dead) and also Sigma DP3m. In regards to resolution the DPm range does produce super sharp results and images can be exported double size to around 30MP, and it will still look better then most 24MP dSLR's.

The biggest issue with DP is the battery life and the buffer, as it takes roughly 5-8 seconds to write to the card because each file is roughly 50mb due to the way the sensor captures the RGB separately at 14MP each.

Therefore after around 70-90 shots the camera will run out of battery in mid air.

Also worth noting that the workflow can be tedious if you plan to shoot loads of images as Lightroom and most popular authoring tools does not support the X3F file format. The only application that will give the best result is the Sigma Photo Pro and it is bit slow.

 

[Rilot]

Panasonic RX8?
It's the highest resolution micro 4/3 camera and is pretty light. With a nice 12mm f/2 Olympus lens will give very sharp images.

Weight is going to be the key here.

 

[Tomsk]

Mobile phone?
Nokia 1020 has 41Mp

 

[Genoma]

I might be wrong but I can't see how you are going to take sharp photos from a static camera in an RC plane given the amount of vibration and movement of the plane? How big is this model to carry such camera? Have you considered a drone instead? Might be worth you coming have a chat with us in the multirotor forum. There are folk in there doing this for a living

 

[Amp34]

This afternoon I've been looking into some published papers on the subject and looking at the study results I don't see a particular specific need for any more than 20 something MP, at least for basic identification purposes. If that is actually the case it opens things up a lot. One of the cameras that could be of interest there is the new EOS M, that said the M43 cameras are probably going to be up there. I'm also wondering if the fixed focal length "compact" prosumer cameras may be a good option to.

Im pretty sure a drone won't work, I did have a fairly lengthy debate about that to him on the first conversation (fighting the case for drones), but the need for multiple hours of flight time pretty much negate anything but a fixed wing with current technology. The benefits of quads, such as maneuverability, is redundant when the plane will be going in straight lines for kilometers on an automated flight path*. The studies I've read so far are also all using fixed wings. Aircraft he's testing at the moment have 1.5-2 m wing spans but he's looking at getting one with upwards of 3m. The vibration isn't really an issue from testing because they are static images and shutter speeds are usually in the four figure range.

It's an interesting project but at the moment I don't know too much about the finer details, but was looking into cameras that may fit the bill he thinks he needs.

*That said if you do know of any quads/hex/octs that can carry several kg of payload (several devices) and stay in the air for a couple of hours, at fair speeds, then I'll definitely give him the info to chew over.

Edit: this is the kind of thing he's aiming for http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0054700 . That's with elephants but he's looking at much smaller animals like deer, which is why the 10MP camera won't cut if for this particular system.

Edit2: actually, one of the other factors to consider is ruggedness. The camera will have a hard life so decent build/weather sealing is very beneficial.

 

[Rilot]

What about a GoPro, shooting 4k video and then taking the stills from the video? Small, light, rugged.

 

[Amp34]

Resolution not high enough and the lens distortion is too high apparently. That's something I suggested as well.