Videomaker – Learn video production and editing, camera reviews › Forums › General › Video and Film Discussion › 24p native: Worth upgrading to it? › Rob: the video refresh rat
Rob: the video refresh rate for a PC is relatively moot. Most online video streaming services, like YouTube, transcode your bideo, and not just in format, but also frame rate, to optimize their storage. It’s common for online material to be rendered to 24p for delivery. The monitor refresh rate doesn’t matter, because your playback isn’t beam-synched to the monitor anyway. And PC displays are universally progressive, but never below 60p. LCDs are generally 60p, CRTs are usually 75p-85p.
Everyone: 24p stored as 60i is still very much real 24p. This is an old timey trick from the tape era. When DV was defined, it was with a fixed frame rate, bitrate, and tape speed. Period. And no real consideration for anything but 60i/50i. So when camcorder manufacturers wanted to add 24p, they were forced to map that 24p into a 60i tape format. It’s still real 24p, just broken up into fields, some of them doubled. The only loss of quality is that the compression CODEC is run on the field, not the frame… and you’re storing redundant data, versus what could have been stored if you did a natural 24p compression. Both are fairly minor issues.
If your NLE is sophisticated enough, it’ll automatically detect the pulldown and ignore it, so you only ever see 24p. Vegas these days is hit or miss… it does this well on DV with most cameras, not so much with AVCHD. It’s fairly stupid for camera companies to still do 24p in a 60i wrapper for tapeless formats, but some persist.
You do actually need more bits per frame to do 24p well, so there can be an advantage to native 24p. The reason is the way MPEG family CODECs (including MPEG-1/2, AVC, and VC-1) work. In these, you store one independent frame (I-Frame), not much different than a JPEG photo. Then you store 14 more frames (usually), which are composed of differences from that first frame. A motion estimation algorithm runs on the difference between frame N and N+1, and encodes essentially how to move things from frame N to N+1 to restore the N+1 image. The reason you want a higher bit budget here for 24p is simple: you’re going to get more motion between frames at 24p than 30p or 60p. The motion estimation stuff is the same, but the bulk of what’s stored is the difference… the encoder applies the motion vectors to frame N, compares it to frame N+1, and that difference also has to be stored. If there’s too much motion, that difference will be overcompressed, and the result fairly ugly… that’s when camcorder video starts to “pixelize” (not really, but that’s what people like to call it.. you start to see the DCT blocks).
I don’t think the difference between 24p and “24p in a 60i wrapper” is enough of a change to justify a camera upgrade, alone. But it’s a valid issue when factoring in the features of one camera vs. another.