-
Remove pulldown?
Hi Guys, I am confused about the pull down issue. I hear about it being removed sometimes but am not sure why. Are there times when you want pull down and other times that you want it removed? Please explain. Thanks a lot for all your knowledge.
-
2 Replies
-
If you want it you don’t remove it, if you don’t want it you remove it. How you remove it is based on what method of pull down you used. 24pA gets removed when captured, 24p needs to be removed with Compressor or CinemaTools.
You remove it if you want to edit in a 23.98 sequence, you leave it in when you want to edit in a 29.97 sequence.
-
3:2 pulldown was invented to allow 24fps film footage to be recorded and displayed on NTSC 29.97 interlaced video equipment. Any theatrical release film that you’ve seen on an NTSC SD broadcast has had 3:2 pulldown added. When viewing footage with 3:2 pulldown on an NTSC CRT monitor, it’s not really apparent to the naked eye when watching the signal played back in realtime that there is any pulldown added. However, if you were to step through the frame-by-frame, you would see the cadence of three progressive frames followed by two split interlaced frames. To be technically accurate, all of the frames are actually interlaced, it’s just that in three of every five frames, both fields are displaying the same source frame, so the interlacing is not visually apparent, while the remaining two frames will “twitter” as they are showing two different source frames displayed on alternating scan lines.
Video camera manufacturers introduced 24p recording modes on cameras to mimic the look of film’s frame rate. Tape-based cameras need to record 24p using 3:2 pulldown because tape formats are limited to recording broadcast spec formats (NTSC SD or 1080i60). When cameras that could record to solid state media became available, they gained the ability to record 24p in its native format, without pulldown. Some solid state cameras can record 24p either with 3:2 pulldown or natively (24pN).
Two flavors of 3:2 pulldown exist – Standard and Advanced. An excellent summary of the differences between the two types of pulldown can be found here:
https://www.adamwilt.com/24p/index.html#24pRecording
Shooting in the 24pA mode requires that pulldown be removed prior to editing, as not doing so will result in horribly stuttering, unwatchable footage. If a project will be edited in a 29.97i sequence, Advanced pulldown should not be used.
The decision whether or not to remove pulldown is mainly determined by delivery format. 3:2 pulldown becomes unnecessary (and indeed, undesirable) when footage will be viewed on the web or on DVD. The only cases in which removing 3:2 pulldown is not desirable are (A) if you plan to edit in a 29.97 sequence and output for broadcast, or (B) if you have to mix 24p (with 3:2 pulldown) and 29.97i footage in the same project. In most other cases, removing pulldown prior to editing and then editing/finishing in native 24p is desirable for many reasons. Also keep in mind that if needed, it’s trivially easy to add 3:2 pulldown to 24p source footage either via software (Compressor, After Effects, etc.) or via hardware (such as an AJA Kona or Blackmagic Decklink video capture card), so even if a project is bound for broadcast, a 24p workflow can be used up until the final output phase. Using a native 24p workflow by default is often the most flexible way to proceed, as it allows the greatest range of options for final delivery.
In the case of web delivery, leaving 3:2 pulldown intact is the worst possible choice. Footage with 3:2 pulldown is a hybrid, neither progressive nor interlaced. Deinterlacing footage with 3:2 pulldown will damage the resolution and clarity of the progressive frames, while not deinterlacing it will leave obvious interlacing artifacts, since progressive displays show both fields of an interlaced frame simultaneously instead of alternately. Any project shot in 24p that’s bound for web delivery should use a 24p native workflow for best results.
In the case of DVD delivery, there are several reasons that a native 24p workflow is superior. MPEG-2 encoders produce better results with progressive source footage, and at the same bit rate a 24p video will receive greater encoding bandwidth per frame than a 29.97 interlaced video. Alternately, you may be able to encode 24p footage at a slightly lower bit rate with successful results compared to 29.97 footage, allowing you to fit more content on a DVD. All set top DVD players have circuitry that adds 3:2 pulldown in realtime to the video output signal when the player is connected to an NTSC monitor. When playing back a native 24p disc, DVD players with progressive component outputs bypass the 3:2 pulldown circuitry and instead deliver a native 24p signal to the monitor. All DVDs of Hollywood theatrical release films are native 24p discs. Native 24p DVDs deliver the best image quality on all devices.
On computers, DVD playback software varies greatly in quality. Some software DVD playback applications use realtime deinterlacing algorithms that produce acceptable results on progressive computer monitors when viewing interlaced or mixed footage. Other DVD playback software isn’t so great. It’s much easier to ensure great looking playback if DVDs are authored as a native 24p discs.
If your destination format suggests that a native 24p workflow is beneficial to you, it is critically important to remove 3:2 pulldown from footage prior to editing. I’ll repeat that for emphasis:
If your destination format suggests that a native 24p workflow is beneficial to you, IT IS CRITICALLY IMPORTANT TO REMOVE 3:2 PULLDOWN PRIOR TO EDITING.
The resulting native 24p footage must then be edited in a native 24p sequence, and all subsequent workflow (output, compression, etc.) will treat the footage as native 24p.
While most editing applications can automatically remove Advanced pulldown on capture, they can’t remove non-Advanced pulldown automatically. Software like Cinema Tools or After Effects can remove non-Advanced 3:2 pulldown from footage by having the user manually guide the process. The user must specify the pulldown cadence of the source footage in order for the pulldown to be removed. For example, if a progressive frame is indicated as “P” and a split interlaced frame is indicated as “S,” a clip may have a cadence of PPPSSPPPSS. Another clip may be SSPPPSSPPP, etc. The pulldown cadence must remain constant and unbroken within a clip for the pulldown removal to be successful. This is why it’s important to remove pulldown prior to editing. If footage with 3:2 pulldown is edited in a 29.97i sequence, the cadence of the 3:2 pulldown will most likely change on every cut point in the edit. This changing cadence will make it impossible to remove the pulldown from the edited video without the use of a very expensive hardware processor like a Teranex.
In summary:
When shooting 24p video with 3:2 pulldown, adopt the default workflow of first removing the pulldown and then editing in a native 24p sequence unless you have a specific need to do otherwise, such as editing in 29.97i for broadcast output or mixing 24p footage (with 3:2 pulldown) with 29.97i footage in the same sequence.
Reply to this Discussion! Login or Sign Up