EVO 4k 30fps-vs-60fps test

[YuKay]

Following some discussion here (and in many other places) about the quality difference between video shot in 30fps and 60fps, I am running some subjective tests which I will upload to Youtube.

My comparisons are not scientific as the EVO, although hovering, was drifting a little in the wind so the images are not identical. They were however shot within 10 seconds of each other and on the same flight (with me changing only the frame rate at halfway).

For starters, here's a full-width slice of a still image grabbed from each unedited NTSC H.264 video and saved losslessly and without any manipulation (other than downscaling by approx 15% to get under the file size upload limit) as a PNG file. Are there any clues as to which half of the image is from the 30fps video and which from the 60fps footage?

(If you click on the image below, a larger image will open and if you alt-click on that one and select Open in New Window, the full-size image will appear.)

 

[Jagerbomb52]

There seems to be a tad more detail on the bottom one to me but I really had to look for it.

 

[Ansia]

Well I can say I don't have an eye to these things. So I'm going to shoot. The 60fps is the bottom one. @YuKay, if you do this again, you can set the Evo on top of a table or chair and simply take the photo from a stationary place.

 

[rkyleh]

bottom

 

[YuKay]

Well I can say I don't have an eye to these things. So I'm going to shoot. The 60fps is the bottom one. @YuKay, if you do this again, you can set the Evo on top of a table or chair and simply take the photo from a stationary place.

Take your point but I want to compare video rather than still life images. Trouble is, I don't know how to replicate a short flight so that my 30fps video follows precisely the same route as my 60fps video. Can I save waypoints/height/speed/camera angles etc for repeat use?

 

[YuKay]

Well, I agree that the bottom image is slightly superior - but it's the top half which was shot at 60fps.

This is likely due to the greater bitrate allocated to each frame of a 30fps video, compared to 60fps which crams twice as many frames into the same processing 'bandwidth'.

If anyone needs to see the original video clips, I will upload them. Otherwise, I might start the project over by shooting as near as possible identical motion videos for comparison…if I can figure out how to do it.

I'm thinking the best way might be not to fly (as per Ansia's recommendation) but to replicate some live action in front of the camera. Or buy another EVO to film the same action once in two cameras at two different frame rates.

 

[Ansia]

Take your point but I want to compare video rather than still life images. Trouble is, I don't know how to replicate a short flight so that my 30fps video follows precisely the same route as my 60fps video. Can I save waypoints/height/speed/camera angles etc for repeat use?

Yes.

 

[HiloHawaiian]

Yes indeed, there is more detail in the bottom image, even knowing it’s the 30fps image before I looked at it — it’s sharper. There’s a few variables that could be in play to explain it, but it is subtle for sure...

 

[Quad808]

60fps decreases motion blur by a lot. That alone is worth it, and why I always use 60. With 30, you need to yaw slowly. I like the crispness of 60fps. It stands to reason that 30 may be a tad bit more saturated, but are we talking stills from video, or actual moving video?

 

[YuKay]

60fps decreases motion blur by a lot. That alone is worth it, and why I always use 60. With 30, you need to yaw slowly. I like the crispness of 60fps. It stands to reason that 30 may be a tad bit more saturated, but are we talking stills from video, or actual moving video?

But motion blur is reality - it's what we see with our own eyes. To be lifelike and convincing, moving video should include motion blur; otherwise it's an animation/cartoon/XBox game.

You are obviously entitled to your preference but publishing 60fps video carries overheads (in processing time, upload time and disk space) and risks (a significant percentage of your target audience won't have the hardware to play back 4k/60fps video - and possibly even 1080p/60fps video…and many more will experience dropped frames, stuttering, etc).

And for smooth video, I need to yaw slowly at 60fps too.

Anyway, this comparison was just the hors d'oeuvres. The stills were taken from moving video, albeit the movement was unwanted hover drift. I've uploaded both clips to Youtube but I intend to find a way to compare otherwise identical real-world video at 30/60fps.

Youtube shrub clip at 30fps
Youtube shrub clip at 60fps

 

[Jagerbomb52]

Maybe shoot a email off to Autel and ask them why they think 4K 60fps is a good setting to use. This was posted on their facebook page this morning.

Autel Robotics
5 hrs ·
Shop here --> http://bit.ly/2Qfw597 EVO's powerful camera is the choice pick for award-winning cinematographers and photographers. Boasting 60 frames per second for buttery smooth 4k resolution footage, it records faster, smoother, and cleaner than any other compact, foldable drone. ///// 4.5 Stars on Amazon /////

 

[YuKay]

The EVO is also "buttery smooth" at 30fps -- but the trouble for a lot of potential viewers is that online - and local - playback of 4k60 video isn't buttery smooth…and Youtube don't automatically step it down to 4k30. That's not Autel's problem though.

The real value of 60fps footage is to give a lot more flexibility in post - before final output at ideally 30fps.

 

[Ansia]

I'm changing the subject a little bit, but I can't play the Evo's 4K60. Yet I can play my Phantom's 4K+60 without problems. This is puzzling me, since it's hard to work in post wheb you can't watch it correctly.

 

[YuKay]

I'm changing the subject a little bit, but I can't play the Evo's 4K60. Yet I can play my Phantom's 4K+60 without problems. This is puzzling me, since it's hard to work in post wheb you can't watch it correctly.

Does the EVO 4k60 open on your machine and then play badly or can't you even open the file? Do you get any alert message? What machine are you using? What codec did you select in the camera? Did you try opening in an alternative player such as VLC?

 

[Apollo11capcom]

I'm changing the subject a little bit, but I can't play the Evo's 4K60. Yet I can play my Phantom's 4K+60 without problems. This is puzzling me, since it's hard to work in post wheb you can't watch it correctly.

I wasn't able to play them on my laptop either, and it had an i5 processor, plenty of ram, and a Nvidia video card, it would stutter and skip like crazy. I upgraded to a gamer laptop and now I can play them fine. It just takes alot of hardware muscle to play them correctly.

 

[Lyk2fly]

I'm no expert but after reading and watching a lot of literature and videos it seems the consensus leans towards 30 fps for general video...because of better image quality. But if you plan on slowing things down then 60 fps is the "chosen one".
It all depends on what you are trying to accomplish with your video. Either way you should try to double your shutter speed over your frame rate for a more natural looking motion blur...so @ 60 fps=1/120 shutter and 30 fps=1/60 shutter.
60 fps and 1/120 shutter is easier to use on a bright day but even it will require nd filters to not over expose.

 

[YuKay]

Does the EVO 4k60 open on your machine and then play badly or can't you even open the file? Do you get any alert message? What machine are you using? What codec did you select in the camera? Did you try opening in an alternative player such as VLC?

Just to be clear, are you describing the situation with all 4k60 videos or just mine? Because I've just noticed that my two clips which I thought were H.264 are actually H.265 (HEVC) which makes playback even more difficult on some machines.

This is a known AE bug (possibly exclusive to IOS) because the camera setting looks to be on H.264 but it doesn't stick unless you open up that panel and select it again before every flight.

 

[mixchief]

Well, I agree that the bottom image is slightly superior - but it's the top half which was shot at 60fps.

This is likely due to the greater bitrate allocated to each frame of a 30fps video, compared to 60fps which crams twice as many frames into the same processing 'bandwidth'.

If anyone needs to see the original video clips, I will upload them. Otherwise, I might start the project over by shooting as near as possible identical motion videos for comparison…if I can figure out how to do it.

I'm thinking the best way might be not to fly (as per Ansia's recommendation) but to replicate some live action in front of the camera. Or buy another EVO to film the same action once in two cameras at two different frame rates.

use 2 evos, one flyng the other duck taped to the belly of the first one

 

[YuKay]

use 2 evos, one flyng the other duck taped to the belly of the first one

It's actually a daft comparison I'm trying to make because I don't need to compare native 30fps footage with native 60fps. The valid comparison is between 60fps and 30fps generated from the same 60fps video clip - and that's easy to do. In fact I have done it in another thread.

 

[jstacy]

Correct me if I am wrong here (and I very well could be) but isn't the 4K part that gives you detail and the frame rate between 30 and 60 gives you smoother transitions not more detail. Basically more frames per second means your getting more still images in your video so rather than seeing a choppy transition when the drone is in motion it smoothes out the higher the frame rate. You video is just a bunch of still images moving at a defined rate the more images you cram into the gaps the less choppy your video.

On the note of uploading to youtube remember they degrade the quality of your video to improve streaming abilities

i could be wrong on some of it or all of it... its just how I understand it.