What exactly does 24f do?

DavidBeier

Veteran
I've heard it mentioned a lot but what exactly is the process (as non technical as possible) and what are the drawbacks?
 
There is no documentation on the process, Canon refuses to divulge their technique.

Graeme Nattress has said he believes it's a realtime de-interlace, where there's one valid field and one interpolated field. Adam Wilt has said that the results look consistent with a simple one-field system (field duplication). I think it's probably some variation on Frame Mode.

The drawbacks are:
1) it's not as high-resolution as the 60i mode. 60i looks like it delivers upwards of 750 lines, 24F delivers around 550-600.

2) 24F mode is incompatible with any other equipment, it can only be played back on a Canon XLH1.

The benefits are that it appears to deliver a temporal signature equivalent to 24p.
 
Barry_Green said:
There is no documentation on the process, Canon refuses to divulge their technique.

Graeme Nattress has said he believes it's a realtime de-interlace, where there's one valid field and one interpolated field. Adam Wilt has said that the results look consistent with a simple one-field system (field duplication). I think it's probably some variation on Frame Mode.

The drawbacks are:
1) it's not as high-resolution as the 60i mode. 60i looks like it delivers upwards of 750 lines, 24F delivers around 550-600.

2) 24F mode is incompatible with any other equipment, it can only be played back on a Canon XLH1.

The benefits are that it appears to deliver a temporal signature equivalent to 24p.


We really need a "Stump The Barry Green Thread".

-Jeremy
 
btw David, did you watch those h.264's? Did the motion look correct and did it still look good?

I don't know exactly how it's achieved but IMHO, it's highly accurate and highly defined.
 
David Jimerson said:
You would think to quell speculation, Canon would release its specs.

I happen to believe they might have something patent pending. Even if not, why would they want to divulge a competitive advantage over, say, the Sony? Their R'n D spending is fairly evident to me in this camera.
 
Elton, I'm pretty sure you misread me a while back. I said that 24F motion should be identical to 24p and that there should be no motion artifacts.
 
Elton, thanks a lot for your footage. The Magic Hour stuff looked fantastic (really). That said, since I'm using a PC, I have trouble having such large quicktime files play at a steady rate (unless I bring them into Avid). I'm gonna try it on my friend's G4 to get a better look.

1) it's not as high-resolution as the 60i mode. 60i looks like it delivers upwards of 750 lines, 24F delivers around 550-600.

I remember reading that the HVX200 had roughly 540 lines and the pansonic about 650. If this is the case, why does the Canon footage still look sharper?
 
In vertical resolution they're about the same. The XLH1 holds an edge in horizontal resolution, and that's why the images appear sharper.
 
David Jimerson said:
Elton, I'm pretty sure you misread me a while back. I said that 24F motion should be identical to 24p and that there should be no motion artifacts.

That's cool man, I think I understood you there, it's just that (maybe I misread) implying that it was only a field from 48i didn't jibe with me.

It's all good in the 'hood!
 
Are we talking uncompressed recording, or the 1440x1080 HDV format? The H1 pixel shifts horizontally to achieve a full 1920x1080 image via SDI. I believe THAT is the image that is being subsampled to the 1440 HDV. To my eyes there's more than 15% there.
 
DavidBeier said:
I remember reading that the HVX200 had roughly 540 lines and the pansonic about 650. If this is the case, why does the Canon footage still look sharper?

I have a purely layman's theory about why it's so sharp.

1) It pixels shifts horizontally to get the full 1920 pixels needed for the SDI spec.

2) Cramming 1.5 megapixels into a 1/3" chip is such an incredible density physically that it's just inherently SHARP, period.
 
Back
Top