Here is a link to stand alone and Vegas scripts for removing pulldown in an HV20 24P stream.

__________________
Guy

Camcorder User Network

Sorry for what might be an amature question but what is pulldown?

For conversion between 24fps and the 30fps (29.97 actually) NTSC rate (which is further split into 60 interlaced /fields/ per second), one has to convert four 24fps frames into five 30fps frames. This is done by duplicating one frame of each four (or one field out of each four, though this results in a lot of field reversals when an "odd" field is duplicated as an "even", and the next frame "even" now has to become "odd", etc.).

The HV20 records only 30fps (60 interlaced) to tape, so the camera has already added the duplicate frame when recording in 24P mode. Pull-down removal reverses this process, removing the duplicated frame and retiming the others to fill out the space. This allows one to recover the true 24p frames. (The XH A1, in 24p HDV, actually records 24p on tape and transfers it as is -- pull-down is only applied to playback out of the A/V port; in standard definition it records to tape in 2:3 or 2:3:3:2 pull-down -- the latter making it easier to remove the pull-down as it puts two duplicated fields adjacent, avoiding odd/even reversals, and letting the field pair be removed cleanly)

Is the difference THAT noticeable to the untrained eye?

What difference?

i have a question
when you shoot with the hv20 at a shutterspeed higher than 60, say something like 250, will there be any weird ghosting, double images because of the change from 60i back to 24p? Or since the camera originally shot in 24p will it just look fine.

24P is a frame rate... 24 frames are /exposed/ in one second.

1/250s is a shutter speed. It means that, for each frame, the shutter is only open for that duration.

If using a really fast shutter speed, you might notice moving objects seeming to flicker. This is because a fast shutter speed takes a sharp snapshot of the object. Then, 1/24s later, another snapshot is taken.

On playback then, you see the object "here", and 1/24s later the object is "there", with no illusion of having moved between the two spots.

With a slow shutter speed, you get motion blur in each "snapshot", but when played at normal speed, the motion blur in the first frame covers the distance the object moved, and the end of the blur track is right where the next frame blur track begins.

Cinematic film cameras use a spinning disk as the shutter. The disk typically has a opening for 180deg rotation. The film is advanced while the disk blocks the light, and held in place while the opening exposes the film. A 180deg shutter means the exposure is 1/48s for the 1/24s frame... Motion blur only covers half the distance between frames.

{"film look" is not just running in 24p mode -- it really requires applying lots of other things: shutter speed of 1/48s to emulate the 180deg shutter disk [I suppose a film camera could have the disk swapped out with, say a 90deg opening, to get an exposure of 1/96s, but that is a physical modification to the camera]; changes to the gamma curve used by the video processing chips to emulate the response curve of film [as with slide film, highlights are quite easy to wash out, so one might slightly underexpose a scene -- making for darker shadows to save the highlight details]; some changes in color balance...}

thank you for the response.

What I'm talking about specifically is a look that happens when converting to/from 60i to 24p. When I turn footage from my 60i pvgs500 into 24p with vegas, the footage will look terrible if a shutter speed above 60 is used. Im not talking about the stuttering effect caused from high shutter speeds. I'm talking about how because of the pulldown process, how it blends frames and fields. You might end up with one frame with multiple "ghost" images that are a result of the blending.

My question is just that since the hv20 can shoot native 24p, will the footage look like this or not?

thank you for your help

I have no idea why shutter speed would have any effect on that -- except for the gaps in motion blur caused by faster shutters... At 1/60s, each field has motion blur from start to end of the time, so while the scan lines shift for odd/even, the moving object is "connected"

It might be more effective to first convert to 30fps progressive, with some sort of blending option as needed... THEN convert down to 24fps progressive. Going from 30 to 24 fps requires losing one out of every five frames. Going direct from 30fps interlaced ("60i" is a misnomer as it implies 60 full frames per second consisting of two interlaced fields) may have the conversion dropping one out of every five /fields/ and /then/ converting to progressive (though how it handles making a progressive frame from two odd or two even fields I have no idea).

Creating 24P from 60i shot with a GS500 is different than reproducing 24p from a previously recorded 3:2 pulldown in 60i shot with a Canon HV20 (or any other camcorder that does this, like the XL2 or DVX100). With the GS500, you are asking the software to create progressive frames where none previously existed. In order to do this, it has to deinterlace and combine fields. You are going to get blurring and other artifacts with this method. On the other hand, where you have progressive frames already recorded in pulldown, it is simply a matter of recombining the fields and discarding the redundant fields with the right software and you have pristine 24P.

Upping the shutter speed too much will create a staccato-like video even in 60i. Much like shooting in the dark using a strobe light. With 24P that effect will be exacerbated.

__________________
Guy

Camcorder User Network

I would like to add that GV/Canopus Edius Pro 4.5x does Inverse Telecine in RT (Real-Time) when you use an OHCI HDV 23.98p project profile. No need to convert.

As a test I first opened an OHCI HDV 59.94i project and frame stepping I would see the 2 interlaced frames on the HDV 24p recording. Opening a new project with OHCI HDV 23.94p all interlaced frames are gone during the editing process.

Anyone can download the demo from GV/Canopus Edius site.

To correctly edit 24p, Inverse Telecine must be applied. Then to correctly view 24p adding Telecine to convert back to 60i frame rate so be played back correctly on a NTSC monitor.

There are people that edit the 24p as 60i which is fine, if it is unnoticeable to the user.

...Angelo

1) Would the HV30 ( has 24p and 30p ) eliminate the need to do this conversion?
2) Regarding converting for HV20, is my understanding correct:
a) when you playback using the camera, the camera does the conversion for you and gets 24p as output.
b) For transferring to computer and obtaining a 24p avi you need to do the convert.

Tx.

30p should work without conversion, just like the 25p. Sadly, this information is missing from the CCInfo test.

When you play 24p from the camera, you get 60i from all outputs, including Firewire, HDMI and component. There is no real 24p in the camera after it has been written to tape.

I have a MiniDV camera, but looking for upgrade to HD. I am looking for 30p mode as the best compomize between film-like motion, absense of interlace artefacts, and ease of editing. This should be as simple as 25p for PAL users.