Is there such a thing as progressive DV50?

[Aaron Neitz]

technically DV50 is an interlaced codec: designed to be read upper field first and then lower (or the other way around, I forget).

BUT if you were to render DV50/24fps from after effects, that Quicktime for *all intents and purposes* is progressive. you can edit with it in a 24fps FCP timeline (set your field dominance to None) and export it self-contained and it’s stays progressive. Your web compressed Flash movie will be progressive too – no interlace issues that need to be removed.

I don’t think there’s a camera that can do 24fps DV50 without some tricks and fakery – post processing to remove interlaced frames and all that.

[Bret Williams]

NTSC is interlaced always. You can’t change that. Your ntsc television set plays back interlaced.

BUT, what the two fields are comprised of is what makes it progressive. Field rendering or not field rendering is simply progressive or interlaced. If it’s progressive then each field represents two parts of the whole from the SAME moment in time. If it’s rendered interlaced, then they represent subsequent moments in time.

If you render it progressive, it doesn’t matter for squat if FCP thinks it’s field rendered. For all purposes, it is. It’s going to be played back as fields by your TV anyhow. But if you didn’t field render it, then you only have 60 images to deal with. Doesn’t matter how FCP interprets it. If you DID field render something, then FCP or AE definitely need to know how to interpret it, because playing the field order back incorrectly means playing frame 2, then 1, then 4, then 3, etc. If you pretend each field is a frame.

Making your sequence progressive means that everything created in the sequence won’t be field rendered. You may not want that. Or you may. Up to you. If you made the sequence field rendered, then text you create in FCP moving across the screen will be field rendered, and the video that is behind it will remain progressive. You’re not going to magically create fields when there aren’t any to begin with.

HTH. I’m sure the technical isn’t spot on, but the gist of it seems to be lost on many. And of course HD is a whole new ballpark.

[Michael Sacci]

[Kevin Murray] “in other words, cameras use hardware algorithms to “fake” progressive DV50 but that interlacing is hardwired into the codec”
Panasonic cameras, DVX100, HVX200 and many others record TRUE progressive frames but they have to put them on interlaced tape with pull down (pA mode), when you capture this footage correctly and remove the pull down you have true progressive frames.

DV50 (or DV25) 23.98 is a progressive as long as it is on the computer there is never a need to introduce interlacing. You can also have true 30p if the camera supports it (if going to tape it is recorded with pulldown that must be removed but once captured correctly it is progressive as a file. Anytime you go to tape or video out to TV interlace is added by the video card or computer. [Kevin Murray] “in other words, cameras use hardware algorithms to “fake” progressive DV50 but that interlacing is hardwired into the codec”

[Kevin Murray]

Thanks to everyone for the responses. I’m not sure if my question has been completely answered and maybe my situation wasn’t clear.

My goal is to move 24p animation to the web (not TV) with as little alteration as possible, although the file delivered has to be video, not SWF. I unfortunately don’t have control over the end process but I can offer input. Right now, the person in charge wants to convert from my progressive ProRes files to DVCPro50 before encoding for the web (a different process is used for the broadcast version, which doesn’t concern me).

What bothers me is, as Aaron said, “DV50 is an interlaced codec”. Even when I tell After Effects to render my non-interlaced footage to a non-interlaced file, both AE and FCP see that file as interlaced. I assume that the compression software might see it that way, too, and want to de-interlace. This would be a bad thing, given that it doesn’t need to be interlaced in the first place. I’ve seen lines on previous tests and I think it’s because of using a DV master.

I think that either Animation or a best-quality h.264 would be the best option since I don’t have to worry about whether they introduce fields or not.

Kevin
Kevin

[Gary Adcock]

[Michael Sacci] “Panasonic cameras, DVX100, HVX200 and many others record TRUE progressive frames but they have to put them on interlaced “

That would be by definition, a Progressive Segmented Frame,

“True P” content is never split into fields for playback, All 720 is a True P format, but NO SD and only a miniscule amount of HD ( from RED or Phantoms) is actually what can be considered “True Progressive”

gary adcock
Studio37
HD & Film Consultation
Post and Production Workflows

Inside look at the IoHD
https://library.creativecow.net/articles/adcock_gary/AJAIOHD.php

[Jeremy Garchow]

[Kevin Murray] “Even when I tell After Effects to render my non-interlaced footage to a non-interlaced file, both AE and FCP see that file as interlaced.”

Then you are either interpreting it incorrectly or, your footage really is interlaced.

From AE, you render with no fields and make sure your composition is set for 23.976 fps.

You bring that in to FCP. If your movie really is 23.976, FCP should mark the clips with a ‘Field Dominance’ of none.

You then put that in a 23.98 timeline and export.

There should be no fields.

Now, why transcode from ProRes, to dv50, to flash? Why not skip the middle man? A ProRes decoder is available form Apple for free for both Windows and Mac.

I think what Aaron meant by dv50 being an interlaced codec was that it is typically used an interlaced environment. The codec itself will do whatever you want it to by telling the NLE/graphics program how to render the file, fields or not.

Jeremy