Many thanks, Sonic Dennis. It seems to be impossible to 'stump' SF experts even with cutting-edge products that aren't even on the market yet (I'm told October 2002.) These are exciting times for DV filmmakers, almost like anti-gravity floating cars had been invented. As to VV3, I am absolutely, totally hooked. Keep up the good work.

Sonic,

How do you transfer your 24p-telecined-to-60i edit back to the original 24p format tape of the Panasonic (or other 24p) system? Will the conversion process from 24p to 60i (for use in VV) change the quality of the original footage?

I ask, because I noticed that when I render a standard DV-captured VV project to 24, then re-render it back to NTSC, the final look is very similar to 16mm film (transfered to videotape), even without using any of the "film" filters. The overall look is a bit soft, and not nearly as detailed as 35mm film. Is that typical? Is it because the resolution of the native DV is too low to approximate 35mm film? Or, are there tweaks I am unaware of to help the material look more like 35mm? Most of these questions regard issues of matching and intercutting DV originated video into a film project, be it 16mm or 35mm, hence the desire to know how to "push" DV into a 35mm realm - it already seems decent for 16mm.

Finally, if the original 24p source material retains it's quality during a 60i edit, would a 24p camera, such as Panasonic's (or Sony's HD CineAlta package, which I believe is a higher-end system) take care of the "look" I refer to, assuming care is taken to light and shoot the footage well?

EW:

Boy, that's a lot of questions, I'll see if I can answer them all.

I'm posting in parts because I'm having problem with the forum software.

Part I

> How do you transfer your 24p-telecined-to-60i edit back to the original 24p format tape of the Panasonic (or other 24p) system?

There is no 24p DV format. This camera captures 24p and does 2-3 pulldown in-camera, and records 60i on tape. To transfer your edited footage back, just print-to-tape to this camera or any other DV camera. When played on a TV, your video will have the "film transfered to video" (telecined) look; the "cadence" of film.

The advantage of doing reverse 2-3 pulldown and working in 24p is that your FX and titles are in 24p.

> Will the conversion process from 24p to 60i (for use in VV) change the quality of the original footage?

There are no changes to the footage, therefore no changes to the quality.

///d@

EW:

Part II

> I noticed that when I render a standard DV-captured VV project to 24, then re-render it back to NTSC, the final look is very similar to 16mm film (transfered to videotape), even without using any of the "film" filters. The overall look is a bit soft, and not nearly as detailed as 35mm film. Is that typical?

Without seeing the footage, it's hard to say, but I know there is some "softening" in the 60i to 24p conversion as specified in our "film look" tutorial due to the field deinterlacing ("interpolate" mode) step.

You can't emulate everything about film in DV, such as how film captures light, it's resolution and dynamic range. The "film look" tutorial does show, however, how to get the 24p "telecined" look that film transfered to video has, without having to start with film. The tutorial also shows got to match film grain look, which some people feel adds to the "film look" experience.

///d@

EW:

Part III

> Finally, if the original 24p source material retains it's quality during a 60i edit, would a 24p camera, such as Panasonic's (or Sony's HD CineAlta package, which I believe is a higher-end system) take care of the "look" I refer to, assuming care is taken to light and shoot the footage well?

The Panasonic camera will give you the film-look cadence, and they have some special gamma curves that simulate the latitude of film so, yes, you'll get a lot of the "film look" with that camera, or so I'm told (I've never touched one). I know even less about the Sony HD CineAlta, but I'm guessing it too has similar features. I'm also imagine that it's very expensive <g>.

///d@

Thank you very much for the replys! The Sony CineAlta IS very expensive. In the area of about $120,000! It rents for about $1500-$2000 per day!

One more thing...
>The advantage of doing reverse 2-3 pulldown and working in 24p is that your FX and titles are in 24p.>

As opposed to creating them in 60i, then converting them to 24p? What ends up happening when you DON'T do your FX and titles natively in 24p? Does working interlaced make them look any worse?

Thanks again!

EW:

> What ends up happening when you DON'T do your FX and titles natively in 24p? Does working interlaced make them look any worse?

Then they are rendered at 60i. It will work, but they will have a different cadence than the media, which you've already converted to 24p. Presently, if you want the 24p look for everything, you have to render to 24p, then render that to 60i for display on TV.

If you're following the 'film look' tutorial, you could save the 24p to 60i render as the last step, and do all your work in 24p, doing deinterlacing of the 60i on the fly while you're editing. Note that the 'film look' tutorial is not doing 3-2 pulldown removal, it's just grabbing a single field for each 24p frame. This is because it's working on arbitrary 60i material. If you have pulldown material, you should find a way to do pulldown removal to 24p instead, for a sharper picture.

///d@

>>> If you have pulldown material, you should find a way to do pulldown removal to 24p instead, for a sharper picture.

I think there are some free (or at least very inexpensive) programs that will do this. can anyone post specific names? can VirtualDub handle that?

it strikes me that we won't be able render to the SF DV codec from other programs, but that doesn't matter because it will be 24p and DV can't do that so it will have to be uncompressed or perhaps Huffy... not sure there.

so just to be clear... the steps to get 24p footage into Vegas for FX processing are:

1 - Capture DV video normally over Firewire
2 - Render captured AVI to 24p format using special utility
3 - Edit / add FX in Vegas

then to output back to DV:
1 - Render to 24p format
2 - Render resulting file to 60i DV
3 - Print resulting DV AVI to tape

wow it looks like a lot of steps would be eliminated and a lot of disk space would be saved if VV could do the 3-2 pulldown processing internally! but then I guess there won't be many of us using 24p DV as source material... not yet anyway.

If the final destination is Dvd, can I just skip the conversion to 60i? I mean, the player should take care of that, right?

>>
so just to be clear... the steps to get 24p footage into Vegas for FX processing are:

1 - Capture DV video normally over Firewire
2 - Render captured AVI to 24p format using special utility
3 - Edit / add FX in Vegas
>>

Rather than using VV to do the conversion as described in their tutorial, this "special utility" to convert to 24p is something that would produce better results than VV can?

Also, as I understand it, Panasonic's 24p miniDV camera telecines while recording, putting a special "mark" on the duplicate fields, and with inverse telecine software, like VirtualDub (or something made by Panasonic) those duplicates can be accurately removed to leave you with the exact progressive frames needed for the 24p version. But applying inverse telecine to footage originally recorded and captured at 60i, has no such guarantee, and when it is inaccurate you get less detail in the image due to occassional times when a single field is deinterlaced into a single frame, or by combining two fields that doesn't actually match.

>>Panasonic and Apple, however, will support a FireWire transfer mode that brings 24p (actually, 23.976fps) video into Final Cut Pro. After editing at 23.976fps, your production can be exported to videotape either at 23.976fps (to a 24p VTR) or – after pulldown is applied – at 29.97fps (to a 30i VTR).>>

Any chance that VV will add something like this?

If I understand correctly, the footage when it is shot and laid to tape is 60i. Correct? The camera itself doesn't know if you are using Premiere, Vegas or FCP. It just sends what has been recorded through 1394 to the NLE. Right? If this is the case, then FCP is not getting the 23.97 fps and editing it natively but rather getting 60i and performing the inverse telecine automatically. Am I wrong?
It appears (to me anyway)that the only thing different you need to do with Vegas is rerender captured footage to 24p whereas FCP does this automatically. Regardless of how it is done, if the data on the tape is has a pulldown added when it was recorded, then it doesn't look like FCP is getting 24 frames per second via firewire while everyone else gets 29.97 unless the camera has a way of recognizing FCP's ability to edit 24p and removes the pulldown itself (something I find doubtful).

Anyone know different?

James Green

Don't want to sound commercial here, but I'm intrigued by a company in Austin, TX which puts out DV-Film transfer software and sells an interesting book that gives the pitfalls to be aware of when trying for the 'film look.' They offer a downloadable demo and sample DV footage to see for yourself. Here's the link: http://www.dvfilm.com/

EW:

As far as I know, that's correct.

There's more info here:

http://www.panasonic.com/PBDS/subcat/dvproline/DVworld_24qs.html

Direct from Panasonic.

I'm frankly amazed over the buzz on this camera. Canon has been doing 30p for years, now Panasonic does 24p and everyone goes ape. I can't wait for film to go 72 fps (or whatever) and here we're slowing down video from 60i to 24p. Oh well.

///d@

"I can't wait for film to go 72 fps (or whatever) and here we're slowing down video from 60i to 24p."

If I recall correcly, Douglas Trumbull developed something called "Showscan", which ran at 60fps. But it never took off, it didn't "look like film".

>>then it doesn't look like FCP is getting 24 frames per second via firewire while everyone else gets 29.97 unless the camera has a way of recognizing FCP's ability to edit 24p and removes the pulldown itself (something I find doubtful).>>

It sounded to me like Panasonic was working with FCP to create a firewire solution for capturing 24p

Read this:
http://www.apple.com/pr/library/2002/apr/07panasonic.html

Here is a section from that April 2002 article

"Panasonic’s new AG-DVX100 24P mini-DV camcorder was announced today at NAB, and Apple has committed to supporting it in a future version of Final Cut Pro. This integration will allow video professionals to acquire, digitally transfer via high speed FireWire, and edit content in Final Cut Pro at the camcorder’s native 24 frames-per-second, 480-line progressive-scan resolution."

Since FCP can already capture 60i via firewire, then this "support" must extend to something different, namely the 24p native format...

EW:

No, it won't be for 24p over Firewire, it will be inverse telecine during or after capture. Repeat after me: There is no 24p DV format.

Bottom line is: All DV NLEs support this camera TODAY. They can capture, edit, and print "telecined" 24p as 60i. The only downside I can think of is that FX and titles will be at 60i. With additional work, NLEs can add inverse telecine support for this camera, work in 24p, and render telecined 60i so that FX and titles work at 24p. Much of this can be done, again TODAY, using separate steps (sometimes with additional tools) and intermediate renders. Future NLEs will likely do much of this on-the-fly or as an included process in the capture/render stages.

///d@

SonicDennis:

>>No, it won't be for 24p over Firewire, it will be inverse telecine during or after capture. Repeat after me: There is no 24p DV format.>>

Actually, that was my assumption, but I felt it neccessary to post the question regarding FCP, firewire and 24p to make sure I got a "once-and-for-all" response from someone who knows more about this than I do.

Thanks.

>but I'm intrigued by a company in Austin, TX which puts out DV-Film transfer >software and sells an interesting book that gives the pitfalls to be aware of when >trying for the 'film look.'

they also do tape to film transfers. i've seen their sample reel in LA. they worked with a lab on sunset (near 101) . when you're in the area email www.dvfilm.com and they will set up a screening (35mm) for you ... transfers are good ... HD looked the BEST , followed by sony dsr 500ws .. they have PAL & NTSC of some camera's -IMO there is NOT a night and day difference between PAL & NTSC (OK i couldn't see the difference)- after seeing the reel i decided to shoot all projects NTSC and not even discuss PAL ...

EW,

I checked out the quote you posted. It is written to sound as if FCP will aquire native 24p. First of all, consider the source. This is coming from Apple so of course they are going to make it sound as if they are going where no NLE can go.

"...digitally transfer via high speed FireWire, and edit content in Final Cut Pro at the camcorder’s native 24 frames-per-second, 480-line progressive-scan resolution." You've gotta remember that Apple pays the guys that write this stuff top dollar. Notice the comma placement after the word "firewire". These guys were very careful to delineate between the aquisition and the statement about the "editing natively" portion of the sentance. What they are really saying is that the video is aquired through firewire. Then they are saying that FCP will be able to edit at 24 fps. Putting the two statements together though skews the perspective of the reader as it is intended to do.

If FCP was able to capture, edit and output native 24p, they would succeed in doing with a $3500 camera what Panasonic is not even doing with their HD cams. Even their HD cameras video is laid down as 60i.

One other thing to remember is that if Pana and Apple had developed a way to aquire 24p over firewire, you would be forced to capture directly from the camera as no decks (at least none that I've heard of) will be able to output the 24p. Most serious videographers have decks and don't want to subject their cameras to the unnecessary wear and tear of capturing.

A simple way of looing at the issue is that a 1394 transfer is basically a file transfer between the source and the NLE. If your source is 60i, then the computer is getting 60i, leaving only the NLE to perform the inverse telecine.

What I wonder is this:
1. Is FCP doing a reverse pulldown on the fly as the footage is captured? It seems that doing this is more processor intensive than just a simple capture over 1394 resulting in captures taking longer. (Perhaps SonicDennis can comment on this.) If this is the case, then there is no advantage to aquiring normally, then manually rerendering to 24p.
OR
2. Is FCP actually creating some sort of "overlay" that allows the editor to view the footage as 24p (without the reverse pulldown) until the final product is ready to be rendered.

Guess we'll have to wait and see...Wow, that was a longer post than I intended

James Green

oh, for number one, I meant that there is no advantage (speed-wise) OVER aquiring normally then rerendering...


James

I have no idea whether FCP is doing inv-tele during capture or as a "media interpretation" layer. The former would make a non-DV file, so probably not. The latter can be pretty efficient, especially with the "special" telecine that this camera is rumoured to do. I know how I'd do it.

But, again, why the hype over 24p when 30p has been around for years? You can't seriously say that all these people plan on actually producing film as final output, can you? That's the only case where 24p has an advantage, it's going to convert to film better than 30p.

///d@