MultiWii / NAZE / SP Racing F3 Anyone using APM 2.5 for aerial filming?

[RTRyder]

I'm finally starting to get caught up on the backlog of projects I've had sitting on the shelf, this week has been designated APM week.

First thing I did was install the "real" 3DR APM 2.5 I've had sitting around into the FlipFPV frame in place of the RCTimer version mainly to see if there was any discernable difference, answer is not really. The 3DR version has the hard shell case so it's mounted a bit differently than the RCT open board setup but otherwise I notice absolutely no difference in flight beyond the fact I spent a bit more time refining the parameter settings so it is smoother than before overall. Amazing how little change it takes to make a difference vs. what I usually do when tuning a Multiwii board, the difference between flying well and locked in is often a .001 or .0015 change on the APM.

Second part of the project just saw flight for the first time, the RCTimer board was installed on my Droidworx AD6 H/L in place of the Hoverfly Pro board that's been on there recently. The RCT board got the eprom cleared and the most recent version of APM firmware uploaded for an I hex prior to install, first flight was with default settings and I have to say the difference between the APM and the Hoverfly Pro is quite noticeable. Also, the AD6 flew near perfectly on the default settings, basic flight in stable mode is rock solid as is altitude hold which actually works as well or better than any of the other flight controllers I own, including the MK 2.2 upgraded F/C and both DJI WKM and Naza. Only thing I can see from that first flight that needs attention is it oscillates a bit when engaging loiter mode though that should be relatively easy to tune out. I am quite surprised and very happy with the way it flys "out of the box", probably the best first flight experience I've ever had in terms of how well it flew and how little it actually needs to be dialed in. I will admit that I've had this frame for a couple years now and it is thoroughly debugged and properly setup so it 'should' fly well with any F/C at this point, but it's amazing at how stable this thing is in the air vs. the flight controller it replaced that I've owned as long or longer than the Droidworx frame.

The real difference in this case is the firmware, the last two releases of the APM code being extremely stable and all functions working well. It's only been with the most recent Hoverfly firmware that the Pro board I have has actually been reliable and stable enough to actually use it without having to retune everything from one flying session to the next, still it should have been at this point well over a year ago, it's just now that the autolevel and altitude hold work well enough to be able to say they do work as they should. In a feature to feature comparison the H/F is woefully behind just about every other board on the market though and given the cost to feature ratio of boards like the APM, H/F better step on the gas and catch up or they may be completely out of the game before too long.

Next step is to have a tuning session to get the hex thoroughly dialed in, after which the brushless gimbal hanging from the AD6 will get powered up and a Hero 3 installed, then the real flight testing will begin!

Ken

 

[SamaraMedia]

I'm finally starting to get caught up on the backlog of projects I've had sitting on the shelf, this week has been designated APM week.

First thing I did was install the "real" 3DR APM 2.5 I've had sitting around into the FlipFPV frame in place of the RCTimer version mainly to see if there was any discernable difference, answer is not really. The 3DR version has the hard shell case so it's mounted a bit differently than the RCT open board setup but otherwise I notice absolutely no difference in flight beyond the fact I spent a bit more time refining the parameter settings so it is smoother than before overall. Amazing how little change it takes to make a difference vs. what I usually do when tuning a Multiwii board, the difference between flying well and locked in is often a .001 or .0015 change on the APM.

Second part of the project just saw flight for the first time, the RCTimer board was installed on my Droidworx AD6 H/L in place of the Hoverfly Pro board that's been on there recently. The RCT board got the eprom cleared and the most recent version of APM firmware uploaded for an I hex prior to install, first flight was with default settings and I have to say the difference between the APM and the Hoverfly Pro is quite noticeable. Also, the AD6 flew near perfectly on the default settings, basic flight in stable mode is rock solid as is altitude hold which actually works as well or better than any of the other flight controllers I own, including the MK 2.2 upgraded F/C and both DJI WKM and Naza. Only thing I can see from that first flight that needs attention is it oscillates a bit when engaging loiter mode though that should be relatively easy to tune out. I am quite surprised and very happy with the way it flys "out of the box", probably the best first flight experience I've ever had in terms of how well it flew and how little it actually needs to be dialed in. I will admit that I've had this frame for a couple years now and it is thoroughly debugged and properly setup so it 'should' fly well with any F/C at this point, but it's amazing at how stable this thing is in the air vs. the flight controller it replaced that I've owned as long or longer than the Droidworx frame.

The real difference in this case is the firmware, the last two releases of the APM code being extremely stable and all functions working well. It's only been with the most recent Hoverfly firmware that the Pro board I have has actually been reliable and stable enough to actually use it without having to retune everything from one flying session to the next, still it should have been at this point well over a year ago, it's just now that the autolevel and altitude hold work well enough to be able to say they do work as they should. In a feature to feature comparison the H/F is woefully behind just about every other board on the market though and given the cost to feature ratio of boards like the APM, H/F better step on the gas and catch up or they may be completely out of the game before too long.

Next step is to have a tuning session to get the hex thoroughly dialed in, after which the brushless gimbal hanging from the AD6 will get powered up and a Hero 3 installed, then the real flight testing will begin!

Ken

Can't wait for the results. I recently purchased a used H/F Pro board for a UAP-1 hexa I built over the winter and planned to use it on that but I may have to reconsider and sell the H/F and purchase the 3DR AMP 2.5. Which SBG are you mounting underneath?

John

 

[RTRyder]

Can't wait for the results. I recently purchased a used H/F Pro board for a UAP-1 hexa I built over the winter and planned to use it on that but I may have to reconsider and sell the H/F and purchase the 3DR AMP 2.5. Which SBG are you mounting underneath?

John

Right now I have an RCTimer GoPro brushless on the AD6 and I recently bought another one that I have installed on my MK hexa. Depending on how the flight tests go I may swap the small RCT gimbal for the larger Nex 5n size gimbal currently hanging under the Cinestar 6 to see how well that combo works.

Batteries are charging for the AD6 right now, time to do some parameter tuning...

Ken

 

[Bartman]

Can't wait for the results. I recently purchased a used H/F Pro board for a UAP-1 hexa I built over the winter and planned to use it on that but I may have to reconsider and sell the H/F and purchase the 3DR AMP 2.5. Which SBG are you mounting underneath?

John

John,

If you read between the lines, Ken said the HF PRO flies great but it's short on features. Ken's angst with the HF product goes back a long way but the board has flown near perfect for more than just the last firmware revision. It's the flight control system of choice for may pro operators because of its flying characteristics where advanced features are just dead weight. It sets up in minutes and then works awesome, flight after flight. APM 2.5+, the subject of this thread, may be comparable (with some tuning), I have one on my bench so I'm not against the idea, but it's important to keep Ken's comments in context as his HF angst goes back a long way.

Before you make any decisions you should fly the Hoverfly board and see for yourself.

Bart

 

[RTRyder]

John,

If you read between the lines, Ken said the HF PRO flies great but it's short on features. Ken's angst with the HF product goes back a long way but the board has flown near perfect for more than just the last firmware revision. It's the flight control system of choice for may pro operators because of its flying characteristics where advanced features are just dead weight. It sets up in minutes and then works awesome, flight after flight. APM 2.5+, the subject of this thread, may be comparable (with some tuning), I have one on my bench so I'm not against the idea, but it's important to keep Ken's comments in context as his HF angst goes back a long way.

Before you make any decisions you should fly the Hoverfly board and see for yourself.

Bart

Bart,

I've been flying the H/F board on the same platform that now has the APM for the past couple weeks, I have to say even with the greater reliability and stable firmware I'm underwhelmed. It does fly well enough but I have at least 3 or 4 other flight controllers here from various manufacturers that will equal or exceed the reliability of the H/F for half the cost and also include a bunch of advanced features that a user may or may not want to use, but at least there're included in the cost and actually work the way you expect them to. I'll also note that my first APM has been 100% reliable and hasn't needed any further tuning since I got it setup on the FlipFPV frame, plus it's capable of doing everything that either my DJI WKM or MK 2.2 F/Cs can do and more.

Yeah, I get it about the ease of use, and for a number of APV professionals that aren't pilots or engineering types I see the attraction but when you come at it from the other side of the equation the attraction becomes a huge limitation, you pay a high price for a controller that can only do well what any $50 Multiwii board can do. Taking it a step further, tell me how much aggravation it was when you couldn't access the H/F server to load firmware over Memorial Day holiday, what if you had a job to get done that required being able to connect and reload that board? There are HUGE advantages to open source platforms even overlooking the low cost of the hardware and all the features you get as part of the package. No, not everyone cares about being able to have a copy of the source code as well as the latest version of firmware resident on their laptop, but it sure makes it easy to reload the board or make major configuration changes without ever having to worry about an internet connection or a remote server being available and if you have the ability to code you can even add your own custom features and functions if it floats your boat.

In closing, just for full disclosure it should also be noted that you have built the APV platform you're selling around the H/F board, so while I may have some angst about the lost time, effort, and $ I've endured in my two plus year love/hate relationship with H/F, I'd say you're equally biased in the other direction by having somewhat of a vested interest in the product.

Oh yes, and by all means I encourage anyone that wants to know for themselves to absolutely do a hands-on comparison of a well tuned H/F and a well tuned APM or any other flight controller for that matter and then decide for themselves, I already have and I prefer the opensource alternative.

Ken

 

[DucktileMedia]

I am completely unbiased and my XY8 under HFP control is hands down as solid as I can imagine one being able to get. The GPS does suck from my experience but they are working on it. I tend to fall more into the "lazy" region of the builder enthusiasm meter. I see further than the fluid basic performance of the HFP, I see an American made product that can compete with the rest of the world. I also like the fact there arent a lot of functions to tune and go wrong from a reliability stand point. Perhaps its not the same as something mechanical but mentally I see simpler being better when you have $10-20k+ in the air. I also think Hoverfly has proven itself to work really well with all size helis where some controllers are very finicky about the sizing/weight they control. The guys at Hoverfly are also paving the bumpy roads to a mediation with the government. They are also the only company that will pick up the phone when you call, albeit not always but enough to satisfy most. As this thread is about the APM and not comparing it to HFP I dont want to contribute to the hijack but my experiences with HF have been the best of any I have tried.

There isnt one controller on the market that is bullet proof and does it all. We pick our poisons. Ken, I think it's admirable that you spend so much time and money being one of the first to test out new electronics and reporting back to us. I tend to wait until things are ironed out before jumping in. But you also have a much more in depth understanding of how these things work than the average guy on here. So maybe whats easy and obtainable for you is realistically out of reach for non-engineers.

I can say with confidence that, although Bart may have interest in supporting Hoverfly, I know he would not promote an unworthy product.

 

[RTRyder]

I am completely unbiased and my XY8 under HFP control is hands down as solid as I can imagine one being able to get. The GPS does suck from my experience but they are working on it. I tend to fall more into the "lazy" region of the builder enthusiasm meter. I see further than the fluid basic performance of the HFP, I see an American made product that can compete with the rest of the world. I also like the fact there arent a lot of functions to tune and go wrong from a reliability stand point. Perhaps its not the same as something mechanical but mentally I see simpler being better when you have $10-20k+ in the air. I also think Hoverfly has proven itself to work really well with all size helis where some controllers are very finicky about the sizing/weight they control. The guys at Hoverfly are also paving the bumpy roads to a mediation with the government. They are also the only company that will pick up the phone when you call, albeit not always but enough to satisfy most. As this thread is about the APM and not comparing it to HFP I dont want to contribute to the hijack but my experiences with HF have been the best of any I have tried.

There isnt one controller on the market that is bullet proof and does it all. We pick our poisons. Ken, I think it's admirable that you spend so much time and money being one of the first to test out new electronics and reporting back to us. I tend to wait until things are ironed out before jumping in. But you also have a much more in depth understanding of how these things work than the average guy on here. So maybe whats easy and obtainable for you is realistically out of reach for non-engineers.

I can say with confidence that, although Bart may have interest in supporting Hoverfly, I know he would not promote an unworthy product.

FYI, 3D Robotics, the company that designs and develops the APM is based in San Diego, California, also an American company... http://3drobotics.com/about-us/

Ken

 

[Bartman]

Ken,

My relationship with Hoverfly began when I called them to ask for a Gimbal controller ASAP (this was pre-delivery of the first batch). I had the idea for a hand held gimbal and needed a controller that could work without any external inputs once it was turned on. In the conversation that followed with Al he convinced me to try a PRO, so I did (I'm still kicking myself for never seeing the hand held gimbal through, i hear they're a good idea after all!).

In choosing the Hoverfly PRO as the basis for my own business I've had to balance all of the factors so that I could comfortably make a recommendation to a potential client. Would a new client want to know that their firmware (and hardware in some cases) was being freelanced by volunteers? Would a new client want to have to learn PID's to tune things or hang their copter from strings to tune it? Would a new heli pilot understand why their new power distribution board went up in smoke for no apparent reason or why their horizon doesn't stay level in roll after a few seconds? How would I explain that their kopter had an underlying and compelling desire to fly home to its country of origin? At the time that I was making a decision there were a lot of negatives to go along with the positives of each system so my decision was to support Hoverfly while also keeping in mind that each user requires an independent assessment before making a flight control recommendation. I still keep the option open to build with MK, KK, APM 2.5+, whatever so long as the end user understands the goods and bads of the decision.

WIth APM 2.5+ we're just seeing the kind of performance and stability of the firmware development process to make it deserve more attention. Some people will slam it for being based on Arduino as if it presents a software related performance limitation. That's above my paygrade but it's another question I'd have to have an answer for at some point. With Hoverfly, it works, it's damn reliable, and I can offer it to someone knowing they won't have to have an advanced degree or a 15 yr. old on hand to keep it working properly.

So, regarding APM 2.5+, I hope it turns out great and I hope 3DR continues to manage the firmware development so that users and dealers can make plans around the system and see continued improvements.

FWIW, here's a flight control assessment done earlier this year. I asked for input from the site's moderators in order to complete it but it didn't quite get the attention I had hoped it would.


View attachment 15562

 

[RTRyder]

Ken,

My relationship with Hoverfly began when I called them to ask for a Gimbal controller ASAP (this was pre-delivery of the first batch). I had the idea for a hand held gimbal and needed a controller that could work without any external inputs once it was turned on. In the conversation that followed with Al he convinced me to try a PRO, so I did (I'm still kicking myself for never seeing the hand held gimbal through, i hear they're a good idea after all!).

In choosing the Hoverfly PRO as the basis for my own business I've had to balance all of the factors so that I could comfortably make a recommendation to a potential client. Would a new client want to know that their firmware (and hardware in some cases) was being freelanced by volunteers? Would a new client want to have to learn PID's to tune things or hang their copter from strings to tune it? Would a new heli pilot understand why their new power distribution board went up in smoke for no apparent reason or why their horizon doesn't stay level in roll after a few seconds? How would I explain that their kopter had an underlying and compelling desire to fly home to its country of origin? At the time that I was making a decision there were a lot of negatives to go along with the positives of each system so my decision was to support Hoverfly while also keeping in mind that each user requires an independent assessment before making a flight control recommendation. I still keep the option open to build with MK, KK, APM 2.5+, whatever so long as the end user understands the goods and bads of the decision.

WIth APM 2.5+ we're just seeing the kind of performance and stability of the firmware development process to make it deserve more attention. Some people will slam it for being based on Arduino as if it presents a software related performance limitation. That's above my paygrade but it's another question I'd have to have an answer for at some point. With Hoverfly, it works, it's damn reliable, and I can offer it to someone knowing they won't have to have an advanced degree or a 15 yr. old on hand to keep it working properly.

So, regarding APM 2.5+, I hope it turns out great and I hope 3DR continues to manage the firmware development so that users and dealers can make plans around the system and see continued improvements.

FWIW, here's a flight control assessment done earlier this year. I asked for input from the site's moderators in order to complete it but it didn't quite get the attention I had hoped it would.

Not to worry about the Arduino processor, it's reached the limits of what it can do. Going forward there will be more and more boards based on ARM processors, 3DR is already there... http://store.3drobotics.com/products/px4fmu-1 my next build will be based on this hardware.

I'd also say the open source development effort by a group of volunteers allows for some of the best and brightest to collaborate and move the project ahead further and quicker than closed source, look at how far some companies have gotten with their platform in two years vs how far the open source platforms have come in terms of the features and functionality that work as expected. Make no mistake the firmware is far more thoroughly tested and debugged before release than some other companies that come to mind, I have far confidence in a Multiwii or APM software update than I ever did or still do from MK, DJI, etc.

If it weren't for open source and collaboration there wouldn't be any hobby grade brushless gimbals just as an example, only $3500 Zenmuse and $15,000 Movi gimbals.

Ken

 

[SamaraMedia]

John,

If you read between the lines, Ken said the HF PRO flies great but it's short on features. Ken's angst with the HF product goes back a long way but the board has flown near perfect for more than just the last firmware revision. It's the flight control system of choice for may pro operators because of its flying characteristics where advanced features are just dead weight. It sets up in minutes and then works awesome, flight after flight. APM 2.5+, the subject of this thread, may be comparable (with some tuning), I have one on my bench so I'm not against the idea, but it's important to keep Ken's comments in context as his HF angst goes back a long way.

Before you make any decisions you should fly the Hoverfly board and see for yourself.

Bart

Bart, Ken,

Sorry to have caused such a commotion on the merits of either the H/F or AMP and for highjacking the thread, it wasn't my intention. I've learned so much from all of you pro's and respect all your opinions and advice. I've been involved in flying and building for just about a year and a half starting with a Gaui X330 then upgraded to a HoverThings VC450, first a MultiWii and then a NAZA. After that I built a DJI F550 and added an Aerial Media Pros extender kit with an AV130 with RSGS. Results were OK, but I account for 90% of my troubles on user error since I'm new to the RC field. The Gaui control was simple, the MW & Naza was my first foray into adjustable gain settings, Arduino, VAR and other such foreign files, it made my head spin. But whenever I had a problem I could quickly turn to MRF's to find answers. This past winter I built a HT FPV with a CC3D and a UAP-1 with original OP CC. When I finally got the HT FPV flying good and taking my first steps into FPV flight I was gassed. I couldn't have gotten to that point without the help and info found here.

I don't have a huge electronics background but I do have technical training as a tool and die/machine builder. But that was 20 years ago, today I'm a photojournalist with the desire to take my visual talents to new heights (sorry couldn't help myself there) for when the newspaper industry totally collapses. I recently modified the HT FPV to accept a DIY SBG with Alexmos control. Now the PID's need to be re-adjusted again to reflect the added weight in the front which in turn forced me into using a larger and heavier battery in back to get the COG where it should be. The Alexmos in itself is a nightmare for those of us that have limited knowledge of PID settings, let alone wind my own motors. It took me six hours to figure out that my ACER PC didn't have JAVA installed, being a MAC guy, I thought every computer had JAVA installed. I don't like spending so much time tuning or learning the latest firmware updates. I'm not real confident in DJI since the F550 never seemed to get all 6 motors to start up and spin around the same RPM until it reached 50%, 3 of them always seemed to be running at half speed until then. People assured me that was normal and once you got it cranked all was well.

I've always wanted to try the H/F because listening to Ben, Bart, Iris and Kloner talk about the virtues of the stability and quality support from Al I finally bought a used one with the intent to mount it on the UAP-1 hexa I built this winter, not that the OP CC isn't getting close to tuned, but do plan on mounting it soon with the AV130 or a SBG for GoPro.

What does amuse me is that how operator capabilities can make anything work great or not at all. When Jussi introduced the first flight with his SBG, people were awe struck and a new era of smooth aerial video was born. Soon after the product began to hit the open market, everybody was abandoning their AV130's and 200's for this revolutionary product, and low and behold average Joe's (like myself) were having success or complete failures trying to get their gimbals to work. Balance is the key to get things working properly and many just throw something together and think its going the work like all the YouTube and Vimeo videos show. I can't even begin to count the number of people complaining about horizon "drift", the main complaint for the PhotoHigher servo gimbals.

I know part of my biggest problem trying to figure out one controller before moving on to another "better" control. I do like the fact that both the AMP and H/F are built in the US of A and I plan on supporting both in the future. Hearing that there is an 87 page manual for the H/F is daunting and listening to some talk about the tuning on the AMP is also a bit much but I realize in order to get better I need to experiment.

Sorry for the lengthy rant but I do want you all to know that if it weren't for dedicated pilots, builders, engineers, testers and pros like yourselves I wouldn't have the confidence to jump into the water with both feet.

Thank you,

John

 

[DennyR]

I'm with you Ken, Stuff has moved on and the high prices no longer reflect what you get. I don't give a flying duck where it is made and I have no political alliance with any manufacturer. I want what does the job correctly and a system that is lacking in autonomous position hold capabilities does not cut the mustard for professional aerial filming. That said, one does not need to fly in this mode continuously but for some jobs it is essential.

My old mate Bobby Violet made hand built scale jet kits in Florida, they were great quality and well ahead of their time, but in the end every aspect of what he did was eclipsed by the Chinese due to manufacturing costs and relentless R&D to constantly improve.

APM has been a long time in the making and has finally reached the point where it is on a par with the best out there. Due to the work of dozens of great developers around the world who collectively made it happen.

 

[Bartman]

............But whenever I had a problem I could quickly turn to MRF's to find answers.

Kumbaya baby! Kumbaya!

Thanks for that and thanks to our moderators Ken and Dave (Dave and Ken)!!

What better than a place to find out that you really do have choices and where you can actually freely discuss them???

 

[Bartman]

APM has been a long time in the making and has finally reached the point where it is on a par with the best out there.

From what I understand, the progress lately is due largely to the fact that 3DR is more actively managing the firmware development and release. As a user I had a lot of questions about how I was supposed to count on an open community to provide me with firmware that would work but also wouldn't be screwed up by unconstrained updates by multiple parties. It's more organized than that now and 3DR is managing the process.

Bart

 

[R_Lefebvre]

From what I understand, the progress lately is due largely to the fact that 3DR is more actively managing the firmware development and release. As a user I had a lot of questions about how I was supposed to count on an open community to provide me with firmware that would work but also wouldn't be screwed up by unconstrained updates by multiple parties. It's more organized than that now and 3DR is managing the process.

Yes, that is correct. Randy has been working on the code development full-time (and probably more) for about a year now. And not coincidentally, that's when things really turned around. It is no longer a free-for-all, and the stability of the code now reflects that. However, strength builds upon strength, and we've recently had an influx of really talented developers who are doing really great things. I guess you could say the program has reached critical mass. I mean, who wants to develop an advanced autopilot program, when the thing barely flies in the first place. 2.8 was a watershed release for us, as that is when Leonard really straightened out the Stabilization code, allowing it to fly manually with a precision and crispness we had never had before. With 2.9, Leonard made the Alt Hold mode work properly, and now 3.0 will be bringing the same excellent inertial hold to the other two axes. This code, still in Beta but should be out soon, flies shockingly good in Auto. It can fly in auto better than most novice or even intermediate pilots.

From here, things will just get better. Based on the things we talked about at the convention we had last weekend, we'll be working to make the auto waypoint flying even better, with more ability to script turn radiuses, etc.

Now, I've seen talk about the PX4, but the developers are absolutely dedicated to continuing to support the APM2.5 as long as possible. Part of the reason is that the effort needed to make what we have right now work on the old Atmel chip, has resulted in an extremely well thought out, tight code. Continuing to keep it efficient, will allow us more headroom on the PX4 for really advanced stuff, like collision avoidance, etc.

 

[aaidan602]

Aerial photography typically uses aircraft and satellites to fly over and take pictures. Aerial shots show a larger area and the surroundings.

____________
​

aerial banner
​

 

[Kilby]

is anyone using APM for larger machines right now? I'm building a quad with 15inch props that I'm looking for a flight controller for.

 

[R_Lefebvre]

Aerial photography typically uses aircraft and satellites to fly over and take pictures. Aerialshots show a larger area and the surroundings.

​

What kind of point are you trying to make? Aerial Photography means taking photos from the air. Period.

is anyone using APM for larger machines right now? I'm building a quad with 15inch props that I'm looking for a flight controller for.

I'm not sure on the exact prop size, but Olivier Adler is flying a fairly large Octo with the system.

View attachment 12793

There's no reason the system couldn't fly a 747 if you wanted to. We have not unnecessarily restricted tuning parameters to limit airframe size, the way some manufacturers do.

 

[Kilby]

I might have to give this a go. I've been going back and forth between using a Wookong or my CC3D. It's a large (800mm motor to motor) quad that is meant to be flown via FPV. I always loved the OpenPilot stuff for FPV because it let you set the maximum pitch angle while in stabilize mode, but it would be great to also have GPS. APM does both, so it seems like a win win.

 

[R_Lefebvre]

I might have to give this a go. I've been going back and forth between using a Wookong or my CC3D. It's a large (800mm motor to motor) quad that is meant to be flown via FPV. I always loved the OpenPilot stuff for FPV because it let you set the maximum pitch angle while in stabilize mode, but it would be great to also have GPS. APM does both, so it seems like a win win.

I'm completely biased, but to me it's a complete no-brainer. The stability of the APM matches everything else out there, and then it offers so many more features, for such a low price. The GPS modes in 3.0 are simply amazing. Last night I was doing the first waypoint flying with tradheli, and it's just awesome. Accelerating at just shy of a half-G up to 76 km/h on a 450 heli, and it does a better job than I can do manually because it's butter smooth.

 

[Kilby]

I've been following the development for sometime now. I'm part of a meetup group in Washington, DC that is filled with lots of APM users. When they first got started it seemed that some of the guys would get random glitches while hovering in loiter mode. Nothing major that caused any issues, but a sudden unexplained jerk here and there didn't leave a warm and fuzzy feeling in my gut. It looks like that is fixed now though, so I'm about ready to give it a go myself.