• ✅ Technical and creative solutions for your film.
    ✅ Screenplay formatting help, plot and story guidance.
    ✅ A respectful community of professionals and newbies.
    ✅ Network with composers, editors, cast, crew, and more!
    🎬 IndieTalk - Filmmaking and Screenwriting help site and community.
    By filmmakers, for filmmakers since 2003

24p vs. cine-look through editing

Hi everyone.

After researching camcorders I feel inclined to purchase one that can shoot 24p. However, I just read a camcorder review in which the author claimed that editing programs can create the cine-look just as well as 24p, therefore leaving me to infer that the extra investment for a camcorder with 24p mode might not be worth it.
What do you guys think? Can you shoot at various frames per second with any camcorder and still achieve a 24p-quality cine-look through editing alone?

By the way, do shooting modes that are different than, but very close to 24p look that much different?

Thanks

-horseradish
 
I'm going to say "no", you cannot achieve the same look of 24p @ 1/48 second shutter shot on the camera from something like 60i with 1/60 second shutter. It may be possible, with some sophisticated software, that you can enhance motion blur and reconstruct 24 frames from 60 fields, but it simply will not be the same. Furthermore, how much will it cost for the software to do this? How much processing time will it take?

25p is very close to 24p, and if you're shooting PAL, you don't need to fret about telecine. However, conforming the audio to 24p is hardly worth the trouble when the telecine of the 24p can be removed at capture time so you don't need to screw around with any conversions.

Since I shoot mostly for commercials, I was shooting 30f until recently I started experimenting with 24f to see what sort of workflow issues I might have and to verify that the audio will remain in sync, etc. I was pleasantly surprised at the quality of the 24f/p footage, and getting there was as simple as setting the camera to 24f and capturing with the correct settings.

Ok, having said all of that, if all you can afford is a camcorder that shoots 60i or 50i, don't sweat it. Just do yourself a favor and get a good deal on a used camera. Get out there and start shooting. Learn about lighting and sound. You'll find you can get your story just as well on 60i as you can with 24p. It looks great on DVDs, and if you learn to shoot well, it will take you far.

Like we keep saying, over and over, camera skill, good lighting, and high quality audio will do more for your movie than 24p will ever do.

Doug
 
You're right, Doug. Talent, creativity, craftsmanship, and technical knowledge all take precedence over the quality of the gear you may have. I was never so naive as to think that high quality gear alone is all that is needed to spawn a great film. As a longtime visual artist and writer, I know first hand the value of artistic skill over the quality of the medium being used, and I'm confident that I'll be able to produce interesting films regardless of the type of film gear I have to settle for. After all, the same eye that is required to make stunning drawings or paintings (or even sculptures) is also required to make stunning film shots. However, needless to say, if one can come into possession of high quality equipment without having to sacrifice their leg, one might as well do so. I may finally have come upon such an opportunity.

Sony's DCR VX2100, although outside my price range if paid in full, is available through Dell for a monthly rate of $77.00. This I can afford, thanks to my new job. Moreover, the reviews for this camcorder are VERY positive all across the board. I think I'll purchase it.

What do you, Doug, and anyone else who'd like to contribute, think of the VX2100? Although it doesn't shoot 24p, can it nonetheless produce a look that is close to, or perhaps even indistinguishable from film? Also, I read that this camcorder has the tendency to produce a cold, blue look. Is this true, and what thoughts do you guys have about this?

"It is only by way of the serpentine path through immorality and darkness that epiphanies can emerge."-
John Candy

-horseradish
 
I have no personal experience with the VX2100.

No pro-sumer video camera can shoot video that is indistinguishable from film; even the ones that sell for $10,000. However, the only thing I see missing from that camera for serious shooting is XLR inputs. You will probably want an XLR adaptor for it soon enough.

B&H Photo has 58 user reviews and the average rating is 5 out of 5 stars. That says a lot about the camera's usability and quality.

Good luck with your new job!

Doug
 
I'll put it to you this way, I've tried shooting both ways, and I still can't say for sure which I prefer.

Here are my results:

"Sometimes They Come Loose" -- 24p -- the only post effect we did was ad a glow to make the orange pop.

"Overtime Boogaloo" -- 60i with various shutter speeds (depending on what I wanted to achieve with the shot) -- I also understood more about depth of field and how to manipulate a camera's settings, as I shot this 2 years later than STCL.

Some days I think STCL looks better, others I think the better one is Boogaloo.

Then there is the mess of "The Stream, The Cave, Jim, and Dave" which I shot the same ways as Boogaloo but could not find a way to make the "film look" work correctly in post.
 
Last edited:
I don't think anyone should attempt this on a small or non existent budget without ALOT of skill. In the first place 24P is an attempt to look like film which is practically impossible, and then to try to take 30 and pull out frames to make it look like 24P to make it look like film.....it's just too much. Do like me. I'm making the switch to film as we speak. With some study, and some intelligent E-bay usage you can have a pretty nice, albeit olschool, 16mm setup for pretty darn cheap.
 
I don't blame you, Vince. I've done some experimenting with this, and the results are not that bad. I'm now thinking of buying a 35mm adapter. I'm afraid, though, that if I try it on my old camera, I might wish I'd just saved the money on the new camera. So far, all of the commercial work I've done (which is where the business justification comes from) has been in SD. I haven't even needed the 16x9 format for anything commercial, and I have an anamorphic adapter for the old camera if I did.

Sometimes, the new technology just isn't justified, and of course, there is always something better just around the corner.

Doug
 
Vince...are you releasing this on film, web or DVD? If DVD is the answer and you're looking to screen it where DVD is the delivery method. It will end up at 60i. Everything done to it will be a mathematical approximation of the 24p no matter what.

I like to shoot my XL1s at 30p (f in canon's little world). to get a progressive feel to the footage, but even then, it's stuffed onto a 60i tape (so is 24p). The extra work to re-conform the 60i back to 24p for editing just doesn't fit my workflow or delivery method. And I don't see that enough difference between shooting 24 and 30 progressive to account for the extra time.

Taking 60i and stuffing it down to 24p doesn't look good to me, it's all a rough approximation of motion that never holds up under scrutiny. Basically, the processor has to take apart the fields of the 60i image (at 50% resolution) and figure out what moves from field to field, then calculate (read: fake) motion blur to account for the motion between one field and another.

The film look lies in; lighting, makeup, costume, location choice, framing, set dressing, acting, casting, editing, camera movement... none of these things is frame rate. I've seen film that looks like digital through the misuse of these elements... I've seen 60i video that I couldn't tell (or didn't care) wasn't film based on the use of these things.

My advice:
1) always shoot 30f (unless you intend to use slow mo with the footage
2) backup and zoom in
2a) use ND filters to get the iris open almost to full (not all the way as there's some color fringing at the extremes) to blur the background a bit and compress space
2b) put more space between the subject and the background to help with this
3) put the camera on a dolly and/or a crane to show off the 3-d ness of the space in front of the lens
3a) does the environment need some smoke to make the room look deeper?
4) study what really makes movies look like movies... hint: it has jack to do with frame rate.
5) get the lines in the script into camera, then do 2-5 more takes of each setup just to let the actors explore the parts more.
6) you did get actors right? not just friends (no offense to your friends)... unless your friends are seriously studying acting as a craft.
7) take the time to look at the frame before shooting (add 1/2 hour to each setup so you can do all the little things) Is the background or foreground too distracting, too empty, colored incorrectly? These little things matter. Are your actors moving too jerky or suddenly for the camera to pick up (Bruce Lee actually had to slow down so the camera could pick up his punches)? Are the movements fluid and STAR looking?
8) Is the footage you're getting in camera good to edit with? Are you giving yourself options in the edit bay? Different pacing, different amplitude, different attitudes?
9) When editing, screw the script, does the next shot move the story along, is it necessary? When watching the edited piece do you "FEEL" the next cut and it isn't there? Have you had a group of critical friends watch the piece with a timecode at the bottom of the screen and notebooks in hand so you can beat the hell out of the first final edit?
 
Knightly, great summary and advice.

For now, everything I'm outputting is 30p (progressive scan 480p NTSC DVD). So I still have to deinterlace it in post no matter what. I have seen some really crappy consumer electronics devices (and software video players) that can screw up interlaced content pretty well. The only real difference between the XL1s and the XL2 (aside from the native 16:9 CCDs) is that it can do the 24p conversion in-camera. Well, it would cost me a heck of a lot more than $200 to change out my camera body to an XL2. Plus, I'm ready when I want to start playing with film transfer. I'm not making the investment yet. My comments were in reference to Nick's comments that you should not attempt this yourself. And my response was that I won't. It will cost me $200 to get the tool to do it for me.

More and more televisions and projectors are becoming available that can handle 120Hz and, thus, 24p content without "judder". HD content can handle 24p natively. So, a combination of 60i converted to 24p with an upconversion to 1080p with good pixel interpolation and color enhancement would look fairly decent and for considerably less initial investment than buying an equivelent HD camera.

And I really don't want to shoot 30f because I don't want to give up half my resolution in-camera. When you display this on a 73" DLP rear projecion TV screen, it's very noticeable. Imagine if it were projected in a theater. I plan to use the 16:9 guides and crop in post, so I need all that resolution. For YouTube, I'd definitely consider it, but not for anything I'd like to send to a festival when that day comes. Of course, by then I may really want to spring for a better camera (or rent one) anyway, so who knows.

My biggest annoyance with DV is the aliasing you get with poor cadence detection and deinterlacing. So I'll do everything I can to ensure the image stays decent quality once it leaves my computer.

EDIT: A side note, both productions I've done so far had still photos with Ken Burns effect, and having those natively progressive made a huge difference in the final product. You have a copy of one of those projects. Pause it on one of the still images and notice how decent it looks (no combing artifacts). The first outputs were interlaced, and it tore those still images to shreds when they moved.
 
Last edited:
And I really don't want to shoot 30f because I don't want to give up half my resolution in-camera. When you display this on a 73" DLP rear projecion TV screen, it's very noticeable. Imagine if it were projected in a theater. I plan to use the 16:9 guides and crop in post, so I need all that resolution.

I've had my footage projected in a theater. The footage I've shot in 30f (~30% resolution loss) looks much better than the de-interlaced footage (50% resolution loss since the bits between lines are interpolated - often poorly), and the motion looks much more real with the progressive as well. At full movie screen height projected from DVD, the progressive holds up better both in resolution and motion artifacting. The problems with the de-interlacing algorithms become much more evident when you blow the pixels up to the size of your hand.

If your destination will be progressive anyway, the motion is more cinematic with 30f and you keep more resolution (not initially in camera, but after the de-interlacing process ends).
 
I've had my footage projected in a theater. The footage I've shot in 30f (~30% resolution loss) looks much better than the de-interlaced footage (50% resolution loss since the bits between lines are interpolated - often poorly), and the motion looks much more real with the progressive as well. At full movie screen height projected from DVD, the progressive holds up better both in resolution and motion artifacting. The problems with the de-interlacing algorithms become much more evident when you blow the pixels up to the size of your hand.

If your destination will be progressive anyway, the motion is more cinematic with 30f and you keep more resolution (not initially in camera, but after the de-interlacing process ends).

I don't understand. Are you saying that Magic Bullet works by stripping out one field and then just interpolating? I don't believe that is accurate, but I haven't really studied their method yet. That would be a waste if they're just discarding half the data and rebuilding it.
 
Last edited:
No, at the risk of speaking for Knightly, I believe he is referring to the fact that Canon's 30f mode on the XL series is not true progressive, and it results in reduced resolution. Even on the Canon XH-A1, there is a subtle loss in image sharpness using the frame modes.
 
Yes, Magic Bullet strips out and interpolates %50 of the parts of the image that are in motion. They do funky math that figures they can combine the fields in the still parts to make a progressive image while keeping the resolution there, but to avoid the combed edges with in frame motion, they discard %50 of the region of the frame that is in motion.
 
I did some tests with Magic Bullet and came to the same conclusion. I also didn't see any discernible advantage to using their SD to HD upscaling. It may work better with different subject matter.

I may make this my next challenge, to create good 24p from 60i, using as much information is available and as many clever tricks as I can come up with. I suppose there would be some market for it. If not with new material which can easily by shot @ 24p these days, but with older material that someone wants to mix with their new 24p stuff, or whatever.

It's and interesting challenge, anyway.

Doug
 
I'd be curious to see the results...especially with actual 24p of the same scene/shot captured in camera for comparison. I don't think you'll be able to get the motion cadence/blur to reproduce in a realistic/convincing manner... but I'm excited to see what you can do. I know shake and compressor (cinema tools?) do a pretty good job of retiming and interpolating motion blur.
 
I tried Cinema tools to convert 60i to 24p. The result was Ok, but there was still some comb (interlacing artifacts) after the conversion. I read the docs for Cinema tools and it would appear that they expect you to be removing telecine, not converting standard 60i to 24p. Of course, there is a subtle difference between 24p that has been telecined to 60i and footage that was shot at 60i.

I must say, I was disappointed, because I thought it would only be a small adjustment from the inverse telecine to just retime the 60 fields into 24 frames. I'd even accept that some frames would be made from a single field (1/2 resolution). Of course, if I write it, I will start by selecting fields closest to the 24p timing, then start trying to work in additional fields to enhance motion and resolution. The original 60 to 24 conversion, selecting the closest mapped field for each target frame seems relatively simple. Audiosync will be an entirely new experience for me, so I'm hoping the NLE will retime the audio if I retime the video frames in a manner that is supported by the API.

I've got a library of 60i footage that I'd love to mix with some of my newer stuff. I'm sure some other people do also.

Doug
 
Thanks, guys, for saving me $200 and costing me a couple thousand to upgrade the camera at some point. :P

Unless Doug can create a package that will do a much better job...
 
Last edited:
Back
Top