Have you used software stabilisation for the Vision?

[digicrew]

Hi everyone, I am new here and am considering getting a Phantom for my video production business.

I was excited to learn the Vision was in Australia in the next few days but of course disappointed to learn it didn't include a Gimbal.

I watched a couple of videos on YouTube on the Vision and was disappointed except for this one:

http://www.youtube.com/watch?v=rGJQJet5ImY

And I would love to learn how they got the video looking so darned smooth/good.

Then I thought of software stabilisation as a compromise. Has anyone used it yet on the video from a Vision?

I played around with some vision from BuzzFlyer and got these results:

http://youtu.be/0o4tTvGKAGI

Apologies the video quality isn't great but you get the idea that software stabilisation does indeed work quite well as a compromise.

You will note that the amount of correction can be seen when the BuzzFlyer logo appears.

I understand that there is some cropping and the software method is nowhere near as good as a Gimbal, but, to me, it's a pretty good compromise. I would love to see it used on full quality vision.

Thoughts?

 

[ethemtolga]

Footage definitely looks more stabilized after editing. Very promising. A better video quality would give us more but it looks good anyway. What software did you use to play with it?

 

[digicrew]

Sorry, I should have included that!

It was cut in FCPX using the built-in stabilisation tool, default settings.

I guess you would do a couple of fly-bys for each shot, pick the best one and then stabilise it.

I am excited by the Vision's ease of use compared to the original Phantom with a GoPro. Sure, there are benefits to using the GoPro, but for my work - especially if you can get stable pics, the Vision is enough.

Yes - would LOVE to try stabilisation on full quality HD vision!

 

[Gizmo3000]

video from the Vision will no doubt need stabilization to look decent

 

[digicrew]

I guess it won't worry me too much, the vision I get from my GoPro mounted on the cars that I film regularly needs stabilsation as well.

 

[martcerv]

That video shows pretty heavy software stabilization and only about 30 seconds of footage is actually from the vision in the aerial media pros video. If your wanting good stabilized video getting a regular phantom or v2 phantom when that comes with a gimbal and FPV will do much better then any video you can get from a vision. Software stabilization is ok but never great in anything but dead calm conditions and even using this you need very smooth flying for best results.

A gimbal will let you get the shot in much less takes or just 1 attempt where shooting without especially in wind and if your not a very experienced quad pilot will be pretty tough to get any decent footage.

Here is some 100% phantom video I shot on a gopro 3 black with no gimbal just smooth flying in calm conditions and some post stabilization.

[youtube]http://www.youtube.com/watch?v=HkjNGQ04Q8c[/youtube]
[youtube]http://www.youtube.com/watch?v=ZtCkU-_qY1A[/youtube]
[youtube]http://www.youtube.com/watch?v=IZsQRMHwLq0[/youtube]

I still ended up getting a gimbal but havent flown it yet will do soon though, no gimbal makes slower smooth flying for video in any wind pretty much impossible. Sure if you are fine with only flying in dead calm conditions and you are pretty good at flying

 

[digicrew]

Thanks martcerv.

We all know a gimbal is the answer. But what I wanted to know was is it possible to get acceptable vision from the Vision using software stabilisation. I did it without even trying with someone else's highly-compressed YouTube vision. Imagine what I could do with full HD. Has anyone else done it and been satisfied? Looks like appleboy has.

Any videos I shoot would not be 100% Phantom anyway, they would all be a mixture of cameras with some quick Phantom thrown in for another pov. So I believe I would be happy with the Vision, using some software stabilisation, after choosing the best shots.

Another thing is, I use our only GoPro for other things during a shoot, so it would be inconvenient for me to keep changing the GP from a car to the Phantom and back, etc etc. So, unless I decide to buy another GoPro, having a dedicated camera on a copter would be great.

 

[martcerv]

In the right conditions flying smooth with no wind you should be able to get ok video much like my examples all shot without a gimbal though image quality wont match a h3 black with protune and graded. Get any sort of wind and if you purely rely on flying in GPS mode you will have pretty major stability issues. Software stabilization will only do so much and the smoother your source footage the better your results will be. If your only planning to use just a few shots from your footage then out of a 20 minute flight you should be able to get a bit of useable footage. The hardest part of shooting without a gimbal is you basically just need to shoot lots of footage and hope some of it works then use the best shots. With a gimbal you can plan shots and be much more likely of getting the shot you planned without needing lots of footage hoping some of it works.

Here is another example of software stabilization using adobe warp stabilzer, this test was in fairly windy conditions and has worked ok but here I used 120p so the stabilizer had more frames to work with and so got a better end result. You can see even though it is quite smooth after stabilization flying in the wind you just wont get a stable horizon no matter what you do and this is the biggest limitation for such a video platform.

[youtube]http://www.youtube.com/watch?v=bf_AeQ07fPE[/youtube]

Not sure what the max frame rates in the vision will be yet but also video wont be up to hero3 black protune footage but from what I have seen the vision can get close to regular hero2 footage.

 

[sundara nēvalā]

I watched a couple of videos on YouTube on the Vision and was disappointed except for this one:

http://www.youtube.com/watch?v=rGJQJet5ImY

And I would love to learn how they got the video looking so darned smooth/good.

Not smooth good IMO. Obviously stabilized in post as the distorted margins clearly illustrate.

It's quite telling that they have already abandoned the stock camera (with a spec that does not come remotely close to the GPH3B spec) in favor of retrofitting a real gimbal solution. Despite being a really spammy situation on their part.

 

[Gizmo3000]

But what I wanted to know was is it possible to get acceptable vision from the Vision using software stabilisation. I did it without even trying with someone else's highly-compressed YouTube vision. Imagine what I could do with full HD. Has anyone else done it and been satisfied?

The key to getting "acceptable" footage from non-gimbal stabilized footage is to carefully fly in windless conditions, in ATTI mode, with gentle stick movement. - which then will produce the most ideal footage to apply stabilization software to.

 

[vbluguitar]

But what I wanted to know was is it possible to get acceptable vision from the Vision using software stabilisation. I did it without even trying with someone else's highly-compressed YouTube vision. Imagine what I could do with full HD. Has anyone else done it and been satisfied?

The key to getting "acceptable" footage from non-gimbal stabilized footage is to carefully fly in windless conditions, in ATTI mode, with gentle stick movement. - which then will produce the most ideal footage to apply stabilization software to.

I think the second key is to keep in mind that we, as video and photo guys, pick up on every little error in photos and video. The end user "viewer" is much less likely to pick up on camera shake or stabilization glitches. Every single one of us says "ewww, that looks gross!" and every single end viewer sees the same clip and says "wow, cool video". LOL. Very few people notice that kind of stuff until you point it out to them.

For example, I did a walk through video for a client recently and had two errors. My arm and sleeve showed when I opened the front door (just for a second or two) and I left a vacuum in the closet which is a no no. The clients response was "thank you so much, this video is AWESOME". lol. Because I am so attuned to the video and any errors, I kept thinking "they are going to make me reshoot this", but they were happy with the product.

 

[digicrew]

These last few posts have me wondering how much we should expect of ourselves.

We have to keep in mind that these devices are sub-$2000 and give us incredible results that we couldn't get a couple of years ago. As a professional camera operator with 25 years of shooting under my belt, I feel the results - with software stabilisation on reasonable vision form a Phantom Vision - is indeed acceptable for, say, low-budget property videos and other low budget corporate videos. In my opinion, the stabilised vision in the BMX video was very acceptable for what they were using it for.

Just the fact that there is another, new "flying around" angle available, would quite possibly make clients "pee" themselves regardless of the quality of stabilisation.

That said, I do like the true gimablised look because, I believe, that is the holy-grail look we are after. Even better when it's sub-$2k!!

I believe a Phantom will revitalise a stale market in my area. I know as a producer I am now looking for something special to give my videos a kick...and I think aerial footage could be the answer. I also believe my clients are looking something different...something to give their products and brands an edge. I also therefore believe aerial vision will be enough of a hook to get me some new clients that were taking their sweet time to jump on board.

I might be wrong...as I am a producer first and marketing genus a very distant second!! ;-)

So, I guess what I was saying initially was this...is the Vision good enough for me to get started with with the help of sw stabilisation for the foray into aerial shots to get people's mouths watering...or should I go the extra and go gimbal?

I would go the Phantom with a gimballed GoPro, but the low battery life (6-12 mins max with FPV) concerns me and - to be honest - having to change GoPro from Phantom/to Phantom all the time during a shoot would annoy me enough to warrant a dedicated GoPro...so add another $400 to the package. We're now looking at $2500 instead of $1400.

I just have to decide if the extra $1000 (or more with extra batteries and other accessories) is warranted at this stage for me and my business.

Finally, I would indeed love to hear about the talk of "gimbalising" a new Phantom Vision. Any more info on this?

Cheers and thanks for the replies.

 

[vbluguitar]

The way this market works, its pretty safe to say that there will be a 3d gimble available for the vision camera, and it will probably be available fairly quickly, but we shall see.

Once again, I think most people will pee themselves rather than asking if you used a gimble and after effects. They just won't see the flaws like we do. Most content gets played on youtube at 340 or 480 nowadays anyway. I am not saying its not good to get the best available, or to do the best job you can, but sometimes we work on stuff that nobody can see, and won't know the difference either way.

 

[Gizmo3000]

I would go the Phantom with a gimballed GoPro, but the low battery life (6-12 mins max with FPV) concerns me

I'd take 6 minutes of stabilized footage over 20 minutes of unstable footage any day!

 

[digicrew]

Good point, Gizmo 3000.

By the way I forgot to ask:

Any news on the Phantom 2? Release date? Features/improvements? I Heard this last week it would possibly be released in about 2 weeks...or did that person mean "announced in two weeks"?

I am so excited by the concept of aerial video for my business, I don't think I can wait until the next version...