EOS 5D MK II workaround: Converting 30p into 24p/25p

30p to 24p is one of those scenarios where it just doesn't make much sense at all. It's a case of twisting something to do a job it was never designed to do. If someone wants to devote weeks and weeks to the task because they couldn't afford the right tool for the job, and they think that time expenditure is worth it to them, then nobody will tell them "no". But for anyone who values their time, just shooting on the right tool will be infinitely less painful.


To complete what is implied . . . .

But for anyone who values their time more than their art.


I am not sure that film making should ever be about saving time, I would - without a second thought - dedicate weeks and weeks to getting something just how I want it if my budget limited my choice of tools. Obviously for those who cant afford the right tool for the job, the choices are limited, to dedicate my time is a small price to pay.

Of course I agree that if you can afford the right tools (something that can match the 5D2 in image quality and shoots 24/25) then that would be a better choice.

But right now, with a little patience and dedication you can generate great 24/25p from the 5D2 and of course the process does not take weeks and weeks.

. . . . . .


On a different note:

I wonder if this is the first device - in the modern history of camcorders - from any manufacturer to be released into PAL territories without the local frame rate ?
 
Last edited:
I don't understand your use of the word 'but' ?

I live in a PAL region (UK) - we broadcast at 50hz, so I wanted to see how 30p converts to 25p - the results of the first test looks pretty good to me.




Yes, 24p is 1fps slower. :2vrolijk_08:
I'd just like to know if is it possible to get the same or acceptable outcome converting to 24p? Thanks.
 
Last edited:
To complete what is implied . . . .

But for anyone who values their time more than their art.
No, what was implied was "anyone who values their time more than just spending more to buy the products that will work out of the box rather than trying to come up with some post solution to make a product do what it wouldn't otherwise normally do."

I very well understand the appeal of trying to extract maximum performance out of an affordable product. I mean, when I was 15 I dismantled the muffler off a mo-ped to reduce back pressure and took apart the whole engine to polish the piston to try to get an extra mph or two out of it. But nowadays it just seems a huge waste of time. You want something faster, buy something faster and be done with it, right? When I was 16 I had a Honda 400, and I used to pester the parts shop about getting in a big-bore kit for it. Nowadays I look back and realize that was just crazy; I should have just spent a little more and gotten the 650. But at the time I was all agog with the idea of changing/twisting/adapting whatever I had, into whatever it was that I actually wanted. Nitrous? Turbo? Whatever.

But what I learned sooner rather than later was that all that ends up happening is you end up screwing yourself by voiding your warranty, hacking apart a perfectly-working, properly-engineered product to try to turn it into something it was never intended to be.

So - Gordon and Emmanuel asked what I thought of this. My quick response, as listed above, is: why bother? It seems like a tremendous amount of hoop-jumping to go through, and soon enough there'll be an alternative. Until then, just get a 150 with a Lex and shoot your opus.

(and this being said from a guy who's going to buy an EOS 5D Mark II on Monday morning. But not to shoot video with.)
 
I ask the same.
I've seen Twixtor do some amazing things. I've also downloaded the demo and tried it on my own footage and seen it repeatedly screw stuff up. I don't trust it. I would hate to commit tens of thousands of dollars of budget, and everyone's time and effort, and then end up with something that I couldn't make work.

Others may feel differently, go right ahead and feel differently, nobody's stopping you. But I wouldn't do it. If they announce an EOS 6D with 24P mode, then maybe. But by the time that comes out the Scarlet will probably be out, and the whole conversation will completely disappear.
 
Thanks for your help. But...

1) What's your impression about Lee's 25p conversion tests?

2) If they are acceptable in your opinion, do you think we can get the same result going with a conversion to 24p?

3) And comparing them with Twixtor... once you said about Twixtor: "I've never heard anyone report good results in anything other than limited test shots". Do you think the same about Lee's method?
 
It seems like a tremendous amount of hoop-jumping to go through, and soon enough there'll be an alternative.

It really is straightforward.

The steps for converting 30 to 24/25p takes moments to set up - import your footage - enter the export settings - hit 'go' ! 5 minutes work (and 5 hours rendering ! :Drogar-BigGrin(DBG) )

There is little more to it than that, no hoops to be jumped.

Like I say I agree with your basic point, if you can get you hands on a machine that shoots 24/25 in the first place (and has the look you need) - then all the better.
 
Thanks Barry for your PM. He has a favorable opinion about your tests, Lee. What a few of us are wondering... will your conversion method show an acceptable outcome in ANY situation (for both 25p and 24p cases) ?
 
I've found this:


«You can edit in 30P, then when finished, run the footage through a conversion to 24P. Apple Shake and Motion have Optical Flow technology that works very well for this. It is not frame blending, the frames are completely redrawn using an advanced technology called Motion Vectors. The result is cinema quality. If your footage going in is 180 degree shutter, that will be the angle generated. It works best with Progressive to Progressive conversions since the Interlacing can lead to some flickering between frames.

I do this quite often for TV commercials and DVD's that often look better at 24P in Standard def. That said, I prefer the look of 24P in standard def since DVD's and Plasma TV's will remove the pulldown automatically and thereby increase resolution. In native HDTV environments I prefer the look of 30P for most footage since it just locks in with most Display refresh rates better with less flicker.

Shooting in 30P is the better choice all around since the Optical Flow conversion works best when "down converting" frames rather than raising the frame rate.
»


http://cinema5d.com/viewtopic.php?f=13&t=25&p=284&hilit=conversion#p284


«As I've said about 30 times around here lately, (it*) can be easily and perfectly retimed to 24P if you have Apple Final Cut Studio's Motion app, or Apple Shake. The technology is called Optical Flow, and alternately may be called Motion Vectors.

30P is a serious limitation of the camera only if you don't have the right software. Most if not all indie filmmakers have this software. It's like the Photoshop of film making.

In the vast majority of footage, 30P will look better on an HDTV. 24P is only superior in SD when delivered over DVD compression.
»

* typo (in the original text it is mentioned 24p) as far as I could understand


http://forums.dpreview.com/forums/read.asp?forum=1032&message=29496902


~ Matthew Thomas

http://www.thomasarts.com (company website)
http://gallery.me.com/thomasarts (company photo working gallery)
http://gallery.me.com/mthomas
 
it takes *a lot* of rendering time and the results aren't always perfect. that's what the debate is about. it's a fact that it *can* be done and quite easily so.

/matt
 
For a true film look, 24p and 1/48th ARE necessary

For a true film look, 24p and 1/48th ARE necessary

The answer comes here:


«You make it sound so easy! If only it was.. :)

I work in Visual effects, and I've been working in the field for years. I've personally worked with most major compositing packages over this time. Shake and Nuke both have optical flow methods and both are very good (we use them mainly to retime shots, make them longer etc.) and for certain shots, you would be hard pressed to tell it wasn't originally shot like that.

BUT, in order for optical flow to work successfully at all, you have to have footage that lends itself to be retimed nicely in the first place. A good example of a shot that is horrible for retiming, is a crowd shot from waist level stacked on top of eachother. The algorithm to calculate the motion vectors that optical flow uses, will have a hard time tracking what belongs to what, creating artifacts in the rendered frames. This is but one example, many shots with countering motion between parallaxing objects are headaches to deal with.

(...)

Taking my notes above, my point is that software will only work when the footage is agreeable. And even then, it isn't quite the same as film with a 180 degree shutter.. But certain types of shots work much better than others.

I also want to note, that not all 30p footage is the same. The motion blur from standard film is about 2 times faster than 24p, so about 1/48th second per frame. This creates the familiar motion blur we associate with "film". Many 30p sources are shot at faster shutters, creating more crisp "video" looking footage, like Vincent Laforet's latest short film "Reverie". That's the main reason it still looked like video, as beautiful as it was.

30p shot at a shutter speed of 1/48th would be much closer to a film look than using the shutter that they had used at but with 24p output. 24p AND 1/48th are necessary for the look.

Back to the topic, my point here is that to achieve a "film" look, and not just convert video footage to different frame rates, you need to capture the footage the real way. Software as you pointed out, can help, but they are limited and need a lot going for them to be truly successful.
»


~ Nordin [ LINK for profile] http://rahhali.com


EDIT -- Here's the link:

http://forums.dpreview.com/forums/read.asp?forum=1032&message=29497059
 
Last edited:
How would I download the 24p version ?
That clip is not available for download, you can check at Vimeo. You could and can extract the 24p version from your temporary internet files. What exactly did you do it? Where did you get the original 30p file? If I can ask...
 
Cesar, my question was rhetorical (it was not a genuine question).


Here are the clues that may lead you to the answer you seek for your question:
"Did you convert from the original 30p file"

Clue 1: You cannot (easily or at full quality) download Vimeo's 24p conversion.

Clue 2: You can (and very easily) download the original 30p source footage.

Clue 3: My example runs at 25p - I entered the correct percentages (in After Effects) to change 30p to 25p.

Clue 4: I know what I am doing, as I used to do this kind of thing for a living ! :)




Where did you get the original 30p file? If I can ask...

When you are logged in to Vimeo, go to the movie page you are interested in, in the bottom/right of the page is a grey bar that says: 'Downloads' and under that a link that you can click on and download the original uploaded file.
 
Back
Top