24 fps or 30 fps?

[PageXone]

Hi, so I have P3A and Gopro 4 Silver on F550.
I see a lot of people shooting in 24, others in 30.
Which on is better? I prefer a cinematic look.
I use ND filters all the time and shot at 2.7k.
Since Gopro does not let me control the shutter speed it is difficult to achieve 24 fps?

 

[Dheorl]

Hi, so I have P3A and Gopro 4 Silver on F550.
I see a lot of people shooting in 24, others in 30.
Which on is better? I prefer a cinematic look.
I use ND filters all the time and shot at 2.7k.
Since Gopro does not let me control the shutter speed it is difficult to achieve 24 fps?

The shutter speed is not directly related to the fps. For a "cinematic" look, you normally want to get a shutter-speed about double the frame rate, irregardless of what that frame rate is. This is where you likely need to start using ND filters. Obviously this is more of a guideline than a rule, but normally a good thing to stick to until you know why you want to change it.

As for 24p vs 30p; I personally don't think 24p is needed to get a cinematic look (although I'm sure many will disagree). Go for whatever is best for your output. One thing I do however really quite like is 30p slowed down to 24p. It gives everything a very sedate feel, without giving it too much of an actual drawn out slow-mo feel.

 

[MPPilot]

24 fps is supposed to give the film look, but I shoot it for one simple reason, storage. 24 fps uses 17% less storage for the same time period shot over 30 fps. I know, I know, storage is cheap. I agree, but when filming at 4k I'm thinking about the resources required when post processing. I built a really nice editing station (i7 4GHz, 32GB RAM, GTX 980 with 6GB RAM) and it barely handles 4k at a reasonable rate. I could shoot at lower resolutions and save more space but I always like to work with the greatest amount of data I can. I can alway lower resolution on my outputs, it is not as easy to start low and output high. I think I got off on a tangent here, but again I shoot at 24 fps to save a little space. Does it really matter in the long run? I don't know. It's my choice, take it for what its worth. Keep in mind that I also bet on Amiga, OS/2 Warp, and HD DVD and we saw how well that went.

 

[Dheorl]

24 fps is supposed to give the film look, but I shoot it for one simple reason, storage. 24 fps uses 17% less storage for the same time period shot over 30 fps. I know, I know, storage is cheap. I agree, but when filming at 4k I'm thinking about the resources required when post processing. I built a really nice editing station (i7 4GHz, 32GB RAM, GTX 980 with 6GB RAM) and it barely handles 4k at a reasonable rate. I could shoot at lower resolutions and save more space but I always like to work with the greatest amount of data I can. I can alway lower resolution on my outputs, it is not as easy to start low and output high. I think I got off on a tangent here, but again I shoot at 24 fps to save a little space. Does it really matter in the long run? I don't know. It's my choice, take it for what its worth. Keep in mind that I also bet on Amiga, OS/2 Warp, and HD DVD and we saw how well that went.

I'm curious, have you actually tested this with identical clips, or have a source that has? I can definitely see you're logic behind it, but as the phantoms record compressed video I would have thought, given a busy scene, the 24p would simply show less compression artefacts for the same bitrate, and therefore HD space.

 

[PageXone]

This is where you likely need to start using ND filters.

About that, see this video bellow, I shot with P3A at 1080p 60fps and ND8 filter.
The shot got blurry, maybe because of the flare? Some people say bad ND filter, other say I shouldnt use ND filters in this shot. Mine are TACO-RC

 

[MPPilot]

I'm curious, have you actually tested this with identical clips, or have a source that has? I can definitely see you're logic behind it, but as the phantoms record compressed video I would have thought, given a busy scene, the 24p would simply show less compression artefacts for the same bitrate, and therefore HD space.

Good point. No I have not. Now you have me curious. I think I shall try this later today and I will compair the results. I'll let you know what I find.

 

[Flipsonic]

I read somewhere that 3840x2160 30 fps is still 4K quality but should give you less judder during panning shots compared to 4096x2160 24fps. I've tried this and did notice the difference. My suggestion is to shoot identical scenes with panning at both settings and compare. Let us know what you think.

 

[matti]

Me living in a PAL land the question is whether to stubbornly continue using base25 fps or switch to base30 (30/1.002 in ancient analog world) frame rate.

iOS devices' current max frame rate is 240fps which is nicely dividable by 30 but sadly not by 25. So mixing 30-60-120-240 fps footage is a piece of cake while adding 25base footage to the mix is not.

24 vs 25 fps in my perspective is futile because they are so close.

 

[MPPilot]

I'm curious, have you actually tested this with identical clips, or have a source that has? I can definitely see you're logic behind it, but as the phantoms record compressed video I would have thought, given a busy scene, the 24p would simply show less compression artefacts for the same bitrate, and therefore HD space.

OK, here is what I found out, filming the same thing at different rates for 60 seconds (give or take a second):

1920x1080 24fps: 181.4 MB
1920x1080 30fps: 290.9 MB
1920x1080 60fps: 438.4 MB
2704x1520 24fps: 328.3 MB
2704x1520 30fps: 337.3 MB
3840x2160 24fps: 438.7 MB
3840x2160 30fps: 436.7 MB
4096x2160 24fps: 440.8 MB (only fps available at full 4k)

Take if for what it is worth. Interesting thing I noticed was that 3820 24fps was a bit bigger than 30fps, but to be fair, it did run one second longer. Significant size savings are made at the lower frame rates in the lower resolutions. Once you hit UHD and full 4k realms it does not seem to make much of a difference.

 

[Dheorl]

OK, here is what I found out, filming the same thing at different rates for 60 seconds (give or take a second):

1920x1080 24fps: 181.4 MB
1920x1080 30fps: 290.9 MB
1920x1080 60fps: 438.4 MB
2704x1520 24fps: 328.3 MB
2704x1520 30fps: 337.3 MB
3840x2160 24fps: 438.7 MB
3840x2160 30fps: 436.7 MB
4096x2160 24fps: 440.8 MB (only fps available at full 4k)

Take if for what it is worth. Interesting thing I noticed was that 3820 24fps was a bit bigger than 30fps, but to be fair, it did run one second longer. Significant size savings are made at the lower frame rates in the lower resolutions. Once you hit UHD and full 4k realms it does not seem to make much of a difference.

Interesting. Can I ask what the scene was?

 

[MPPilot]

Interesting. Can I ask what the scene was?

The ugliest thing I could find to film, me sitting on my couch. Weather is too miserable outside to take it for a flight.

 

[Dheorl]

The ugliest thing I could find to film, me sitting on my couch. Weather is too miserable outside to take it for a flight.

Fair enough. I was just wondering because something stationary where not much changes between the i frames can be very efficiently compressed. I wonder how large the differences will be when the scene challenges the bitrate limit, for instance close to trees in the wind.

(dw, I'm not expecting you to go out and do this, you're results are still interesting, but if you do be sure to post it here )

 

[RocketBrew]

Keep in mind that I also bet on Amiga, OS/2 Warp, and HD DVD and we saw how well that went.

We still have lots of systems in the field that run on OS/2 where I work... (just upgraded a bunch to Windows 7 earlier this year!)


Sent from my iPad using PhantomPilots mobile app

 

[PageXone]

I tried to shot today in 2.7k 24fps and when I do some fast yaw the image gets ugly.
I noticed if I choose 24fps I need to do some really slow and smooth movements, am I correct?

 

[Dheorl]

About that, see this video bellow, I shot with P3A at 1080p 60fps and ND8 filter.
The shot got blurry, maybe because of the flare? Some people say bad ND filter, other say I shouldnt use ND filters in this shot. Mine are TACO-RC

Hi, sorry I didn't reply to this earlier. As to whether or not you need an ND filter the sunny 16 rule can be handy (old exposure guideline from film photography days). You end up getting into a bit of exposure maths here, but basically on a sunny day, at f16, your ISO should match the shutter speed. I.e. 100ISO = 1/100s shutter.

Now the phantom has an f2.8 lens, so at 100ISO you need 1/3200 shutter. This is a lot faster than you really want, and although that scene doesn't feature sky and doesn't look like the brightest of days, I'd definitely recommend one.

Tbh I'm not sure what you mean by "the shot got blurry". On my screen I don't notice the quality drop off.

 

[Dheorl]

I tried to shot today in 2.7k 24fps and when I do some fast yaw the image gets ugly.
I noticed if I choose 24fps I need to do some really slow and smooth movements, am I correct?

Again, it can sometimes come down to shutter speed. If you're shooting 24fps with a fast shutter speed and start doing fast movements, then your footage will end up with an almost strobe like effect, with stuff jerking from frame to frame. This could potentially lead to compression funniness, but even if it doesn't it isn't normally a nice effect.

 

[matti]

A general rule of thumb to prevent any motion blur in still images while hand-holding cameras is to match the lens 35mm equivalent focal length to an inverse shutter speed.

The P3P lens with a 35mm equivalent focal length of 20mm would therefore require a shutter speed of at least 1/20th of a second to prevent any blur when hand-held (if the lens or camera has image stabilisation, you might decrease the shutter speed by up to four stops for hand-held shots).

 

[PageXone]

Again, it can sometimes come down to shutter speed. If you're shooting 24fps with a fast shutter speed and start doing fast movements, then your footage will end up with an almost strobe like effect, with stuff jerking from frame to frame. This could potentially lead to compression funniness, but even if it doesn't it isn't normally a nice effect.

I had shutter speed only to 50. I really start to dislike the 24fps.

 

[WetDog]

About that, see this video bellow, I shot with P3A at 1080p 60fps and ND8 filter.
The shot got blurry, maybe because of the flare? Some people say bad ND filter, other say I shouldnt use ND filters in this shot. Mine are TACO-RC

YouTube is terrible for trying to figure out video problems because of the compression and variable bit rates which depend on your line capacity, where the server is and various other factors.

That said it appears to be displaying at less than 1080 resolution although something happens towards the end and the shot cleans up. There is no reason to use 60 FPS for this sort of shot, 30 is fine and you can cut down the speed to 24 if you need the look. You can always ADD blur in post, taking it away is quite a bit harder.

But the best way is to shoot at the speed you want to display at. Which means you have to know ahead of time.

But you have to know why you are doing this. This video is a tad long, but interesting.

 

[SimonH78]

Also makes a big difference on quality is the bit rate the video was exported to in post production...