24P and 24F (whats the difference??)

[Steve Witt]

I am a new HV20 owner and trying to learn all I can about HDV in gerneral. I know that 24P is 24 frames per second progressive scan and I sort of understand what is going on with it. But then I see "24F" mentioned in some of the material I have been reading. Could someone explain the Difference? (OR is it the same thing?)

[Austin Meyers]

24F typically, in terms of canon, refers to the 24 Frame mode that the prosumer cams shoot. the hv20 can play back this format, but cannot shoot it, as the hv20's 24p implementation is different (24-in-60)

[Peter J Alessandria]

They are not the same.

24F is Canon's implementation of interlaced 24 fps. It outputs from the A/G-1 as 24 frames per second over Firewire. It is not progressive, though with a slight loss of resolution [~ 15-20% depending on who you ask] it can be converted to progressive by most NLE's.

24P as created by your HV20 wraps true progressive frames within a 60i wrapper. It outputs as 60i via Firewire. To get to 24p you'll need to remove pulldown. 24F files have no pulldown. They natively edit as 24 fps.

[Thomas Smet]

Both have tradeoffs of course.

24F = Yes it is true that there is a certain process used to create progressive frames. While they may not be as detailed as a true 24p image they sre still progressive looking in nature and in the way they are compressed. 24F is encoded as true progressive frames in the HDV stream so they look and act exactly like 24p. Other then a slight softness 24F is basically 24p. Not only is true progressive chroma much better for progressive images but 24F also encodes much easier then 60i HDV. This is first of all due to less frames to encode in the 25mbits/s. Second it encodes better because mpeg2 has an easier time predicting motion on progressive encodes.

24F if you have the right NLE is pretty easy to work with. It just loads in as 24p and you are all set to go.

24P (from the HV20) = These are true 24p frames sampled in the camera. However once they are encoded to HDV they are no longer a progressive stream. You have to use special software to remove the pulldown pattern. Once you do this you will again have progressive frames. With HDV there is a slight loss when removing the pulldown. First of all the stream is encoded as interlaced. 4:2:0 chroma is in a slightly different pattern between interlaced and true progressive. This means the chroma has to be converted back to the progressive form of chroma. Not all decoders will do this so you may still end up with jagged color edges. So even though the HV20 processes 24p there is a slight quality reduction by the time you actually get to work with the 24p frames. The best 24p solution with the HV20 is of course to capture live uncompressed so you do not loose any quality when you remove the pulldown. Once you get the true 24p frames you can convert to an easier to use format such as HDV or any other form of HD compressed codec.

HV20 24p is much harder to work with and it will load into most NLE's as 60i.


24F really has two parts to it.
1. The way it processes the 24p frames.
2. The way it encodes the 24p frames.

For example if the HV20 would of had a 24F record mode it could have used the true 24p frames and encoded them as true 24p mpeg2. This would have made for some very sexy 24p video that would have been very easy to work with. Sadly however this was not the case and it seems as though Canon is reserving the true 24p mpeg2 encoding for the higher end cameras.

[Chris Hurd]

Also, since it relies partly upon the Pixel Shift process, Canon's 24F is possible only in their three-chip XL and XH series camcorders (it could not exist in a one-chip camcorder due to the way it works).