shooting 30f: 30, or 29.97?

[Jim Newberry]

Is 30f mode actually 30 frames a second, or 29.97, or something else? If you're shooting with the A-1 at 30F and a recording sound seperately, what frame rate would you set the audio recorder to--29.97 or 30?

[Robert Sanders]

Think of 30F as 29.97P. Set your field recorder(s) to 29.97 and you should stay in sync.

Personally, I would shoot at 29.97 and then de-interlace the footage to 30P in post. While the results look identical to the camera's 30F (in terms of temporal motion), I think the de-interlaced footage looks decidedly sharper. Just my opinion.

[Jim Newberry]

Originally Posted by Robert Sanders

Think of 30F as 29.97P. Set your field recorder(s) to 29.97 and you should stay in sync.

Personally, I would shoot at 29.97 and then de-interlace the footage to 30P in post. While the results look identical to the camera's 30F (in terms of temporal motion), I think the de-interlaced footage looks decidedly sharper. Just my opinion.

How do you shoot with the A1 at 29.97? I only see 24, 30, and 60i as frame rate options.

[Nick Hiltgen]

Jim when he says "shoot in 29.97" he means 60i (which is really 59.94).

[Jim Newberry]

Originally Posted by Nick Hiltgen

Jim when he says "shoot in 29.97" he means 60i (which is really 59.94).

Ah ha, thanks for the clarification. Who knew there'd be so much math?

[Marty Hudzik]

So you guys think 60i deinterlaced looks much sharper than 30P directly? How about compression? typically 30 discrete frame would much easier on the codec than 60 half frames.....24f is even better....but resolution wise I do see the benefits of 60i.

[Robert Sanders]

Originally Posted by Marty Hudzik

So you guys think 60i deinterlaced looks much sharper than 30P directly? How about compression? typically 30 discrete frame would much easier on the codec than 60 half frames.....24f is even better....but resolution wise I do see the benefits of 60i.

You bring up an interesting point Marty. Theoretically long GOP compression works better with progressive frames. But I've learned that Canon's HDV implementation is so rock solid that no one's really broken it. However, that doesn't change the fundamentals of compression and interlaced imagery. So while you might get a sharpness increase shooting 60i you may also take a compression hit.

On one hand...

And on the other...

[A. J. deLange]

HDV in NTSC has one framerate: 29.97 (approximately - it's actually 30*1000/1001). Irrespective of the selected frame rate (60i, 30P, 24P) 60*1000/1001 fields will be written to tape every second. A separate audio recorder will record 48ksps (or 96 ksps or 192 ksps) irrespective of the selected frame rate. The camera and the recorder will each produce time code at the rate of 30 frames per second which slip relative to the video frames by the factor 1000/1001. Thus the only reason the recorder needs to know the camera frame rate is so that it (or an NLE using files from it) can reconcile time code with frame number for display and especially for file timestamping. Equally as important as the frame rate is whether it is drop frame or non drop frame. Not all NLEs can process all frame rates equally well. It is thus generally considered best to set timecode to 29.97ND when recording audio seprately with HDV. If experiment with recorder and NLE indicate ability to use DF and you prefer that then by all means have at it.

[Jim Newberry]

Originally Posted by A. J. deLange

HDV in NTSC has one framerate: 29.97 (approximately - it's actually 30*1000/1001). Irrespective of the selected frame rate (60i, 30P, 24P) 60*1000/1001 fields will be written to tape every second. A separate audio recorder will record 48ksps (or 96 ksps or 192 ksps) irrespective of the selected frame rate. The camera and the recorder will each produce time code at the rate of 30 frames per second which slip relative to the video frames by the factor 1000/1001. Thus the only reason the recorder needs to know the camera frame rate is so that it (or an NLE using files from it) can reconcile time code with frame number for display and especially for file timestamping. Equally as important as the frame rate is whether it is drop frame or non drop frame. Not all NLEs can process all frame rates equally well. It is thus generally considered best to set timecode to 29.97ND when recording audio seprately with HDV. If experiment with recorder and NLE indicate ability to use DF and you prefer that then by all means have at it.

Thanks for the very informative reply!