View Full Version : Panasonic Cagey About HVX200 1080p?


Steven Schuldt
April 21st, 2005, 12:10 PM
Did anyone else get a sense from NAB that Panasonic was hedging their earlier stated 1080p support? At the Apple event on Sunday 1080i and 720p were explicitly mentioned by name but 1080p only appeared on a slide.

At the Panasonic event, the slide that showed supported formats was intentionally (sleazily?) shown for just a split second so unless you had a photographic memory you couldn't verify what was on it.

Don't get me wrong, I love this camera, but I walked away from NAB concerned that Panny was setting us up for a bait-and-switch on 1080p later this year.

Aaron Shaw
April 21st, 2005, 01:11 PM
That's because the camera doesn't record 1080p it records 1080p to an interlaced stream just as the DVX does with SD. Remove the pulldown and you end up with your original 24 frames.

Barry Green
April 21st, 2005, 06:17 PM
Panasonic is absolutely, completely, and thoroughly committed to deliver 1080/24p and 1080/30p. There is no caginess, no hesitation, no "sleaziness" going on. It's the real deal and it's what they're going to produce.

You can see 1080/24p and 1080/30p listed clearly at http://catalog2.panasonic.com/webapp/wcs/stores/servlet/ModelDetail?displayTab=O&storeId=11201&catalogId=13051&itemId=93120&catGroupId=14569&modelNo=AG-HVX200&surfModel=AG-HVX200

Daniel Hollister
April 22nd, 2005, 02:42 AM
I've never understood this...

So Panasonic calls it 1080i because technically it's interlaced with a pulldown. Like the DXX. OK. So then why is the 720p different? It does not work with a pulldown? It records directly in 24fps progressive? Is there a real difference in quality between the pulldown and direct 24p since you still end up with the same frames? The image degradation can't be that dramatic with 2:3:3:2...

It just seems like they're downplaying the capabilities of the 1080i/p so they can make the 720p sound better. I think it's a bad move. It keeps confusing people (like me) who want to know if I'm getting 24p from this. Pulldown or not... i just want 24p.

Dylan Pank
April 22nd, 2005, 03:45 AM
720p is NEVER interlaced at any stage. There is a pulldown of sorts but it is padded with complete progressive frames at the full 920*720 (for DVCproHD) codec. Therefore 720p is 720p is 720p. (at 24, 25, 30, 50 or 60 fps)

with 1080, the CCD maybe scanning at 24p or 30p, but it's always recording 60i and either printing the progressive frame to both fields or introducing a pulldown (for 24p).

This is why I think Panasonic is being a little cagey about claiming 24p. In the SD world, there is no standard for 24p. so when Panasonic introduced 24p it was as much 24p as you could ever get in a SD format, it could never be bested.

However with HD there IS the opportunity for a true 1080 24p format (no interlace no pulldown) so it seems that while the 24p format in the HVX seems a perfectly good approach, they are a being a little bit careful when there are cameras such as the Genesis, the Viper, the D20 and the F900 series in existence which offer absolutely true 24p.

Aaron Koolen
April 22nd, 2005, 05:07 AM
It is scanned from the CCD's as TRUE 24p

It is stored interlaced with pulldown (2:3:3:2 and I assume 2:3??)

There is NO degradation from what you'd get from another system that scanned AND recorded 1080p.

One downside, is that because it's stored with pulldown, you get fields that you won't use on the P2/HDD and so storage space is not optimal.

Aaron

Barry Green
April 22nd, 2005, 02:10 PM
You will get 100% absolute pure unadulterated legitimate 24p in both the 1080 mode and the 720 mode.

720 is a progressive-only format. 1080 is recorded as interlaced, so the way they do it is they store pure raw progressive frames in two fields, and then they add one "pad" frame to bring the data rate back into line with what the 1080i recording mode is expecting. But knowing how it works is really irrelevant, all you have to know is that when you bring it into your nonlinear editor you will be seeing 24 distinct, pure, progressive frames every second, with no interlacing of any type. Those frames will have been imaged 24 times per second, at 1/24 second between frames, exactly like you would expect.

Shannon Rawls
April 22nd, 2005, 11:54 PM
1080 is recorded as interlaced, so the way they do it is they store pure raw progressive frames in two fields...
Hmmmm, that's interesting.......So does that mean it's really 2160p?? If you say it's two full progressive frames in 2 seperate fields interlaced together wouldn't that make that 2160 with each field at full resolution, right? Heaven forbid each of those frames are only HALF RESOLUTION (540) and interpolating the two fields to make it 1080i....hmmmmmm. Something to think about? Sounds familiar.

...knowing how it works is really "irrelevant".
It is?? I think it's very relevant. All of a sudden the microscope is put away for the Pany camera and now we gotta hush up and and listen?? *smile* No question asking?? Everybody was hell bent on showing the "relevancy" of the sony camera and how it works...why stop now?? *smile*

- Shannon W. Rawls

OK

Barry Green
April 23rd, 2005, 12:28 AM
Shannon, I gotta wonder -- are you interested in actual answers, or just in stirring up trouble? I mean, I know you're goading me, so I don't want to respond, but if I don't, then will you say I'm "ducking the question?" Please, work with me here on this, okay?

You profess to own a DVX100 -- surely you're familiar with 24p as implemented on it, right? And if you've edited on a 24p timeline, you know you're getting 24 distinct frames per second, at full resolution. The HVX does the exact same thing. It's been explained over and over. Heck, the discussion of "24p vs. 24pa", or "2:3 vs. 2:3:3:2", is probably one of the most over-discussed, over-microscoped topics that's ever been brought up about the DVX, and then the XL2, and I'm sure it'll continue with the HVX.

One frame is written to two fields. Raw resolution, not flicker-reduced interlaced video. When editing, it is reconstituted back to progressive frames. This method is employed on every 24P-capable camera. This is just like the $100,000 Sony CineAlta does. Just like the $25,000 SDX does. Just like the XL2 does. Just like the DVX does. The only difference between how the CineAlta does it, and how the DVX/XL2/SDX/HVX do it, is the CineAlta only writes 48 fields per second to tape, whereas the other cameras write a full 60 fields to tape, and drop those unnecessary fields in the edit suite. And the only reason they do that is to maintain compatibility with television and editing systems.

There is *no* drawback. It is full progressive. It is 24 hz capture. It is exactly what people want. There is no mystery, there is no hidden microscope, there's no conspiracy, there's no aliens involved.

Everybody was hell bent on showing the "relevancy" of the sony camera and how it works...why stop now?? *smile*
Because the Sony camera's CF24 doesn't work. *smile*

At least for 24p acquisition, it doesn't work... what it does is *not* 24p. And it's been adequately documented, over and over, demonstrated, proven, dissected, even dissertated upon by Adam Wilt himself, in extensive detail.

Not to prove "superiority", but to get to the bottom of the question. What I was saying is, we've *gotten* to the bottom of the question. We've been to the bottom of the question, inspected it, dissected it, written reports, filed it away, sent it on to our superiors, gotten subpoena'd, had the records opened up, examined, verified, brought in "expert witnesses", gotten a jury verdict, been tried in the media, and finally, finally, finally, just when we think we can finally put it to rest -- someone has to throw a firebomb in and try to stir it all up again.

Plus, we've been down this path for over three years. The path is a deep, deep, deep trench, it's been walked down so many times. The question has been answered.

Aaron Koolen
April 23rd, 2005, 12:36 AM
Shannon, Barry didn't say two (2) frames, he said 2 fields.

Aaron

Shannon Rawls
April 23rd, 2005, 08:35 AM
Oh hush up Barry & Aaron, y'all know my questions are the best! *smile*

They encourage people to *think* on their own and make correct decisions. My question above caused you to break it down for everybody, rather then saying "shut up and don't ask no questions, how it works is irrelevant"....now...since you just explained it to everybody.....they all know. Yes, I knew the answer, yes I have a DVX...but others didn't and don't. Only 15% of this websites visitors actually engage in discussion. The other 85% of people that come here just read and learn. *smile* SO if we gonna disect one camera, let's disect em all. Put it out there. Say it in plain english. SO everybody knows. Especially on a new camera. It don't matter that we went to court on an old exsisting camera. You and I both know each new camera released has to go to trial! *smile*

- ShannonRawls.com

Here's a photo of me after I wrapped everybody one late night shooting that someone took with their camera phone!! *smile* It's me and my beloved DVX100a:

http://www.shannonrawls.com/images/photogallery/albums/userpics/shannon@4am.jpg

Luis Caffesse
April 23rd, 2005, 11:06 AM
"You and I both know each new camera released has to go to trial! *smile*"

Maybe we should wait to start the trial until the camera is released.
Right now we're dissecting specs and marketing information.
Rest assurred Shannon, the trial will begin in full swing fourth quarter of this year when the final design is released and put through the paces.

y'all know my questions are the best! *smile*

You want to talk about something that should be put on trial, how about that statement? :)

Just playing around Shannon.

Steven White
April 23rd, 2005, 11:40 AM
There is *no* drawback.

Actually Barry - if you're here I might as well ask.

If you split a progressive frame into 2 fields, how is it then compressed? Do you compress each field seperately? (I assume the answer is "yes" due to the sequential nature of field storage) If you do compress each field separately, is there a hit in performance or compressor efficiency because of this on a progressive stream?

Barry Green
April 23rd, 2005, 02:35 PM
If you split a progressive frame into 2 fields, how is it then compressed? Do you compress each field seperately? (I assume the answer is "yes" due to the sequential nature of field storage) If you do compress each field separately, is there a hit in performance or compressor efficiency because of this on a progressive stream?
Well, that's the thing -- DV actually employs two compression techniques. It can compress an entire frame, or it can compress on fields. It analyzes the difference between fields to determine which method would be more suitable. If there's little motion between fields, it selects the more efficient frame-based compression, and treats both fields as a full frame. If it finds a lot of motion between the fields, it selects field-based compression, getting a more efficient compression on that type of material.

So when shooting 2:3:3:2, it will *always* select the more efficient frame-based compression. And because DVCPRO-HD is a derivative of DV (being, basically, four DV codecs teamed up together), the same techniques should apply.

So yes, even though it's recording the progressive frames as fields, you will still gain the benefit of full-frame progressive compression.

Of course, in 720p it's never an issue, as 720p is always stored and compressed as full progressive frames; there's never anything field-based about 720p.