Why 24P is Important from RED Centre Podcast

[Tony Tibbetts]

Originally Posted by Peer Landa

Actually, if it was possible to get, say, 96000 frames per second (as we do in sound) this would indeed be better -- the higher the sampling rate, the more interpolation possibilities. But as of yet, the sampling rates of audio vs video aren't even in the same ballpark.

...and yet nobody shooting video seems to be doing this and nor do they seem to want to.

Originally Posted by Peer Landa

fine, I'll try help you out by rehashing this one more time:
You stated that "if it looks good to the naked eye" there shouldn't be any "problem" and that "there are plenty of examples floating around the web" to prove this point.

To this I replied that this measurement -- "if it looks good to the naked eye" -- didn't sound very scientific to me, and that it reminded me of some people who said exactly the same thing to us ("if it sounds good to the naked ear") when we were developing 24bit/96kHz audio tools early in the 90's.

So, confusing 60i with 30p sounds scientific to you?

Originally Posted by Peer Landa

see, some of us are so anal that we use oscilloscopes, spectrograms, and even raw bit-stream printouts to determine "if it sounds good enough". Hence, if my field of work had been in video, I doubt I would rely on your "if it looks good to the naked eye" or use those "plenty of examples floating around the web" as my measuring tool.

-- peer

24p is still 24p no matter what format it originated on. If it looks good on a frame by frame examination. Then yes it is "good enough" for me and for a post house. I'm not saying it is without some minor flaws, but the benefits of shooting with this camera far eclipse the rare minor flaw in the conversion process.

Here's the inescapable fact Peer. Plenty of projects with less than stellar format acquisition and delivery have made it past these supposed post facility standards, but have also ended up on TV and in theaters. Stuff with FAR less quality than what is shot on a 5D.

I've shot on a 5D. I've converted and delivered the end result in 24p, to more than happy clients, and for my own personal projects. To reiterate, it sounds as if the people on the podcast never took the time to do the same thing. Me and many other people have done this exact workflow with great results. Have you? Have the guys on the podcast done so? The answer seems to be a resounding NO. Yet, here you are carping on about scientific process from a position of ignorance. There's that pesky part of the scientific method, yet again rearing it's ugly head. TESTING!

If I took the time to worry about the minute level of anal nitpicking you ascribe as important. I would never get any projects off the ground, and neither would a lot of other people on these boards.

Bruce Springsteen recorded "Nebraska" on a Tascam 4-Track cassette recorder. It is a masterpiece. That album sounds "good enough" to me.

[Nick Hiltgen]

Hey Tony, I think you may actually need to listen to the podcast again. I finally did after all of the discussion here. When they were talking about fields they refferring to the "old days" of getting a 24 frame look from 60i. Instead of it being a simple 3:2 pull down, they were saying that what actually happened was you would use part of the upper field of say frame 1 and the lower field of frame 2 in order to get a matched frame that was sharper. I don't think they meant that this is how the 5dm2 works.

The issue they seem to have is that in order to get that look of 24 it's actually really difficult, because if you throw away every 5th frame it looks more jittery then it should, to me this makes since.

They may in fact be shills for RED I don't know I don't listen to them (other then this podcast) I will say that it seems silly for anyone to propose that a theoretical camera (scarlet) is better then a camera you can shoot with today. Don't get me wrong the day that I can go down to my local rental house and pick up a scarlet and shoot a movie in 24p 3k or whatever yes, it will have the 5dm2 beat. But I think it's silly to have a discussion about a camera that doesn't exist (yet?) and one that does.

Also I don't understand what the controversy is, Is it really that earth shattering to think that a lof of people on a film making forum would want 24p? No of course not. And yes 30p does look more "TV" then 24p but it also looks more 24p then 60i.

[Tony Tibbetts]

Originally Posted by Nick Hiltgen

Hey Tony, I think you may actually need to listen to the podcast again. I finally did after all of the discussion here. When they were talking about fields they refferring to the "old days" of getting a 24 frame look from 60i. Instead of it being a simple 3:2 pull down, they were saying that what actually happened was you would use part of the upper field of say frame 1 and the lower field of frame 2 in order to get a matched frame that was sharper. I don't think they meant that this is how the 5dm2 works.

Agreed, I even said as much in an earlier post. It seemed to me that the portion where they were talking about the 5D specifically, was the point at which they brought up fields.

Originally Posted by Nick Hiltgen

Also I don't understand what the controversy is, Is it really that earth shattering to think that a lof of people on a film making forum would want 24p? No of course not. And yes 30p does look more "TV" then 24p but it also looks more 24p then 60i.

For sure. I wasn't even in disagreement on this point. I would jump for joy if they released a 24p update to the 5D. My point has always been that you can deliver good 24p from stuff shot on a 5D.

24p acquisition is better, even preferred, but nowhere near as important as some people make it. A few years ago it might have been a bigger issue, but with the advances in hardware and software it isn't.

[Peer Landa]

Originally Posted by Tony Tibbetts

So, confusing 60i with 30p sounds scientific to you?

No one is "confusing" 30p with 60i. I suggest that you listen to the podcast again.

Yet, here you are carping on about scientific process from a position of ignorance. [...] Bruce Springsteen recorded "Nebraska" on a Tascam 4-Track cassette recorder. It is a masterpiece. That album sounds "good enough" to me.

Well, this is pretty much what I expected -- you are confusing artistic abilities with engineering/scientific aptitudes. Yes, Springsteen managed to make his masterpiece using a subpar analog Tascam 4-channel cassette recorder, and Beethoven managed to write his 9th symphony while being completely deaf. However, none of this has anything to do with the issue at hand -- namely, "why 24p is important".

By the way, I know for a fact that the brilliant engineers at Tascam use both oscilloscopes and spectrograms (and even tools that I've helped developing). The "naked ear" isn't good enough for those guys either.

-- peer

[Jon Fairhurst]

Originally Posted by Tony Tibbetts

24p acquisition is better, even preferred, but nowhere near as important as some people make it. A few years ago it might have been a bigger issue, but with the advances in hardware and software it isn't.

Having produced an international BD from a variety of frame rates, I think 24/25p is still a huge issue. The quality generally suffers when making the conversion, and the better processes are slow. When working on deadline, the last thing I want to face is a framerate conversion. (I certainly don't plan to convert to 24p for this year's 48 hour film project.)

So, yeah, conversion is possible, but nothing about it is desirable.

Sorry if this comes off as argumentative. Clearly from your first sentence, you also prefer 24p acquisition. By implementing 24/25p, Canon can save us significant time and increase the end quality of our productions.

[Brian Luce]

Originally Posted by Peer Landa

When it comes to this kind of work, "the naked eye" is trivial, as in; everyone has one. To satisfy peoples "naked eyes", science and engineering rely on more sophisticated tools (in conjunction with the eyes).

-- peer

You can crunch numbers and measure everything you want, but film and video are ultimately sensual experiences. All the 0's and 1's in the universe will never get you over the finish line and tell you what makes a given piece of content compelling. Dismiss the eye, the ear and the heart at your peril my friend.

[Jon Fairhurst]

The best of all worlds is to consider the human experience as well as the numbers. If you can correlate the two, you can make good decisions.

For instance, I know not to buy an f/8 lens for low light work. I don't need to waste my time buying it to see how it feels. On the other hand, I might buy an f/1.4 lens that makes ugly pictures, in which case I'd want to sell it to get something better.

Numbers don't tell the whole story, but they can help keep us from going down a fruitless path.

[Brian Luce]

Originally Posted by Jon Fairhurst

The best of all worlds is to consider the human experience as well as the numbers. If you can correlate the two, you can make good decisions.

That's right Jon. Under most circumstances, you need both, but Peer is dismissive of the eye, saying "When it comes to this kind of work, the naked eye is trivial."

[Robert Martens]

I'd like to take the chance to try and clarify myself before the thread is finally locked, if that's all right:

Originally Posted by Peer Landa

No one is "confusing" 30p with 60i. I suggest that you listen to the podcast again.

I never meant to imply that they were flat out confusing the concepts of 'progressive' and 'interlaced' (nor did I mean to attack their credibility; this thread's taken a nasty turn that I certainly didn't anticipate), only that the impression I was left with after listening to the podcast--three times, just to see if I misunderstood them--was that they were associating twenty-four frames per second with a cinematic, filmic look, and thirty frames per second with the look of video. My point was simply that I think this association is unfair, as what most of us in NTSC countries have grown accustomed to as the silky-smooth, high sample rate motion of the "video look" is not thirty frames per second in the way most people think of the word "frame". Since it's interlaced, what we're all used to seeing is actually comprised of sixty individual images per second, not thirty; as such, an image capture rate that's a mere six samples per second more than twenty-four will not make your footage leap to the realm of TV commercials, the ten o'clock news, and amateur recordings of childrens' birthday parties. To do that, you'd need to instead record thirty-six more images per second than the twenty-four we're using as a base for comparison.

Put more simply, if I may take what Nick said a few posts back just a bit further, although 30p provides smoother motion than 24p, it's nowhere near enough to scream "video" to viewers.

That may be blindingly obvious to everyone, but no matter how many times I review the podcast I get the impression that the hosts are misstating the difference between the two framerates. I'm sorry if it seemed more inflammatory than that.

[Jon Fairhurst]

Originally Posted by Robert Martens

...although 30p provides smoother motion than 24p, it's nowhere near enough to scream "video" to viewers...

Along those lines, I have a theory (untested) that shooting and playing back 30p at 1/80 will have a similar effect as 24p at 1/48.

30p is a bit smoother than 24p. By increasing the shutter speed to 1/80 @30p, the gap (shutter closed time) is actually 1/48. This will give a bit more stutter to a 30p video to help compensate for 30p's inherent additional smoothness. We will also get about half the motion blur of 24p 1/48, which could be a good thing.

On the other hand, if you will slow 30p down to 24p, shoot at 1/60 for a perfect 180 degree look. And if you will drop frames to get down to 24p, shooting at 1/48 (1/50) is probably the right approach.

If you will use advanced processing to go from 30p to 24p - test it! 1/48 might be the most natural, but sometimes the conversion barfs on soft edges. In that case 1/60 or 1/80 might be best. It really depends on the conversion algorithm and your content.

The day that we can record audio and 24p (fingers crossed) will be blissful...

[Olof Ekbergh]

I agree with Jon.

If this cam could do real audio, preferably 48k (mono would even be OK), with meters, volume adjustment and headphone monitoring. And 24p.

It would be incredible @ this pricepoint.

It would rival the Reds.

This is only the beginning, just think of what the future will bring.

[Peer Landa]

Originally Posted by Brian Luce

You can crunch numbers and measure everything you want, but film and video are ultimately sensual experiences. All the 0's and 1's in the universe will never get you over the finish line and tell you what makes a given piece of content compelling. Dismiss the eye, the ear and the heart at your peril my friend.

I sincerely, thoroughly, absolutely and wholeheartedly agree.

And the best way to transfer 30p to 24p would be...?

-- peer

[Brian Luce]

Originally Posted by Peer Landa

I sincerely, thoroughly, absolutely and wholeheartedly agree.

And the best way to transfer 30p to 24p would be...?

-- peer

I don't know what the "Best" way is, but I do know there's no *good* way. You can try using After Effects.

[Erik Andersen]

To take a cue from Peer and try to address the title of this thread...

Isn't there simply an aesthetic reason why 24p is so different, and often "better," than 30p or 60i? I'm thinking of 24p as a low framerate, that looks removed from reality, while 30p seems closer to what the eye sees. Definitely a subtle difference. Of course 60i is very different from both.

Think of super8 footage, which is 18fps. Aside from the graininess and pastelly colours, isn't there something dreamlike in the stuttery motion?

Or think of the opening credits of "Juno," which looks like 12fps or maybe less. Totally cartoonish and childish looking, due to the graphic style also, but largely the framerate.

IMHO 24p is perfect to create a visual experience removed from everyday reality, that would not be mistaken for all that garbage taking place outside the movie theater!

[Christian Ionescu]

I have a question and I am not going to start another thread as it is closely related to this one.

How is Canon going to allow 24p in camera? The 30fps is hardware implemented in camera. If the chip is doing only 30p does it mean they are going to make a fake 24p, they are going to get 24p out of 30p instead of getting direct 24p from sensor? It looks like they are trying to implement a "hack" that does kind of conversion to 24p as we do in post. Correct me if I am wrong.

Honestly,
Christian I. Ionescu