Comments

dreamlx wrote on 12/2/2004, 11:03 AM
I think this is depending on how the acquisition software captures the footage. If you capture the whole tape, and then do everything in Vegas, this problem should be gone.
farss wrote on 12/2/2004, 12:23 PM
I get the same thing happening (but not that many frames) even with DVCAM. You'd probably find the camera continues to write data after you button off, Probably the video stops being written before the audio. None of this matters so long as the thing is in sync although you need to trim the audio before butting another clip up to the end of the first.
Bob.
mbelli wrote on 12/2/2004, 6:16 PM
busterkeaton wrote on 12/2/2004, 9:18 PM
Can we use the Cineform codec to work with this? Or does that have to happen at acquisition.
dreamlx wrote on 12/2/2004, 9:25 PM
Yes, you can use cineform. Cinform also allows conversion of existing m2t files.
Spot|DSE wrote on 12/2/2004, 9:46 PM
You can use Cineform, or convert to Huffyuv (or anything else you'd like to convert to that supports the frame)
My conclusions, and supported by the co-author of the HDV book, is that Cineform's codec is3 times faster than native M2T files, and up to 10 times faster than m2t with multiple streams.
busterkeaton wrote on 12/2/2004, 10:28 PM
with the files you posted, how does one use Cineform? Do you render to it? Do you set up a new project based on it?
Spot|DSE wrote on 12/2/2004, 10:36 PM
Here's a step by step:
1. Download footage
2. Download Cineform Connect HD demo and install it.
3. Open Connect HD, convert the raw m2t to avi. I'd recommend using the default of Medium.
4. In Vegas be sure to set your project properties to 1920 x 1080, 50i (PAL)
5. Import avi to Vegas, either to timeline or media pool. Set aspect ratio to 1.333 in either the event or in the media pool.
6. Set Preview window to Preview/Auto or Preview/Full

Have fun.

[edt]Keep in mind, you can't see the full quality of this unless you've got a 1900 x 1200 monitor. But you can still get a dang good idea.
for giggles, set your preview to Best/Full, and drag it to fill your screen.
then Render to RAM. (CTRL+B)[edt]
busterkeaton wrote on 12/2/2004, 10:38 PM
OK, I thought it happened within Vegas. I didn't know it was standalone.
Spot|DSE wrote on 12/2/2004, 10:42 PM
Not at the moment....
Also, experiment with the different Cineform wavelet qualities. you'll be surprised, I think.
rfm58 wrote on 12/2/2004, 11:10 PM
"Can we use the Cineform codec to work with this? Or does that have to happen at acquisition."

I've been playing around with the Mallard one. I just changed the .m2t extension to .mpg and it opened fine in Vegas 4.

It looks gorgeous, even on my tired 17" crt monitor. I've got my screen res set at 1152x864 so I obviously can't view it in it's full size. I used a 1920x1080 project size and rendered it out as 1280x720 .avi with huffyuv (which went outside my screen size) and an 854x480 .avi using XviD (not true HD, but it still looks sweet). And, it's fun to have something to play with in Vegas.

Thanks for these files!
Grazie wrote on 12/2/2004, 11:24 PM
Oh wow Spot . .. oh dear . .. .

I'm only looking at the WMV sample of the geese and thew mallard . . the DOF is excellent .. but it is the whole .. how can I say? Immeadiacy of it all that is so stunning . .. The landscapes ans the skyscapes . . another shade has been removed from my vision .. . oh deary dear . . and all in switchable 4:3 switchable 16:9 . . . oh dear me . .. too good . . Even viewing on my 15" Sony LCD I can see the difference . .. There is a section where there is a rock sticking out of the water ... the "lines" of clarity around the goose/waterbirds and the water and the ripples .. too good ...

I need some BIGger paying projects for this next leap - UG!

Respect!

Grazie
JJKizak wrote on 12/3/2004, 5:12 AM
You can also change the file ending from m2t to tp and view them on the MY-HD video card to an HDTV. But in my case some of them were a bit glitchy.

JJK
Spot|DSE wrote on 12/3/2004, 6:18 AM
Thanks Dreamix, it's his work, but he's graciously allowed us to host the files so that people can start to check out HDV. Stupid me, I haven't shot anything with the Z1 that isn't copyrighted, and I don't have access to the cam very often. We'll have more soon.
FrigidNDEditing wrote on 12/3/2004, 1:41 PM
Is there going to be any difference in the quality of video between the FX-1and the Z1? They are going to obviously be using HDV so that should remain the same, but are the chips going to be the same?, or are the only differences going to be in the extra goodies that you get with the Z1?

PS, what does the Z1/FX-1 capture at? 4:1:1:, 4:2:0, or are we lucky enough to get 4:2:2?
Thanks guys.
Skevos_Mavros wrote on 12/4/2004, 6:26 AM
Thanks for posting these Spot, I downloaded them and them forgot all about them until tonight. After watching them I have a question or two (I take my time asking them though, just so you know where I'm coming from). I should perhaps post this question to a seperate thread, but I guess it belongs here, I hope you see it.

Wow. These images are a lot better than they have any right to be from such an affordable camera. I will confess that I had very very low expectations from the HDV format - after all, how good could 1080 be if it's squeezed into the same bandwidth as DV? My expectations were so low that I was one of the minority that thought that the footage captured from the JVC HDV camera looked okay, for the price. I thought the widespread disapointment with that camera was unrealistic given it's target market and price.

Yet these raw captures you've posted from the Sony camera, even more than the Japanese ones linked to here a month or two ago, are really making me think about HDV in a whole new way - as a realistic alternative to 16mm or even 35mm acquisition for low budget filmmakers looking for a film finish (not the target market I realise).

Sure, the Sony camera is not perfect for everyone in my market. The lack of interchangeable lenses is a shame, and no support for true progressive scan or 24fps is a pity for folks like me, looking for a cheap way to finish on film. Then again, it's hard to fault the built in lens from what I've seen so far, and the places that print interlaced DV to film do a great job de-interlacing it first, so that's not a really a big deal given all the extra detail they will be starting with when converting HDV to film.

Anyway, I do have a question for you Spot. The clips you've posted are the first I've seen that were shot in 25fps on the Sony HDVs, and although Vegas Explorer reports that the footage is interlaced, it doesn't appear to be. That is, although the footage may be technically interlaced as far as the way the data is laid to tape, I can see no motion between the fields on any of the moving objects in the footage you've posted. Which is potentially very cool.

A month or so ago, when I saw the results from using the 24, 25, and 30 "frame" modes from the NTSC model of the camera, I was underwhelmed - I'm sure they would look "film like" on an HD monitor, but they are no good for people like me looking for a way to acquire 24p or 25p for printing to film since they still displayed interlacing effects, though more randomly than the normal interlaced modes. But the footage you've posted appears to be truly progressive.

Why do I care? Because I've shot stuff at 25p and had it printed frame to frame to film for projection at 24fps, and although it requires stretching the audio to fit, no one notices the 4% slow down in image and sound. Yes, it's not the purest way of doing it, but if you've shot 25p it's the only way really, and it looks remarkably good because not deinterlacing means you get to use every pixel in the film transfer.

My question - can I dare to hope that this camera captures actual honest-to-goodness 25p at 1440x1080? The specs for the camera on Sony's website do not state it can do 25p, only that 25 "frame" mode. But what I see with my eyes is progressive - no motion between fields at all.

I've taken a close look at your clips in Media Player and Vegas. I love what I see, especially if it's straight from the camera with no software deinterlacing going on. If my eyes aren't deceiving me, I could hypothetically shoot with this camera, edit in Vegas using no more hard drive space than if I had shot in DV, render to a lossless Quicktime file or two (or ten) and take it to a film scanning place for a simple frame to frame print, yes?

Lovely. Please tell me I'm not seeing things...

All the best,



Skevos Mavros
http://www.mavart.com
dreamlx wrote on 12/4/2004, 7:04 AM
Well these scenes have been shot in Cineframe25 using the PAL model HDR-FX1E. Cineframe25 effectively writes progressive footage in 1440x1080 to tape. I say writes as I think only Sony knows what they actually read from CCDs and how they arrive to the 25p. The fact that Vegas recognizes them interlaced may have multiple reasons. It may be that the camera itself marks the stream as interlaced, maybe also that the used capture software writes a wrong header when saving files, or even maybe that they are recognized the wrong way by Vegas. After all, there is a simple workaround for now to this by simple manually changing the media file properties to progressive in Vegas. After all, I can only say, fantastic work Sony. Once for the camera, and once for Vegas :-)
Skevos_Mavros wrote on 12/4/2004, 7:12 AM
Interesting, thanks for the reply. The Cineframe25 I saw from the Japanese NTSC camera looked awful, with interlacing effects aplenty. But this looks great, and is actually usable. Far better than deinterlacing in software as far as I can see.

Hmm, so does than mean that "close-enough-to-true" 24p can extracted from the Cineframe24 setting? Or am I being greedy now? :-)


Skevos Mavros
http://www.mavart.com
mhbstevens wrote on 12/4/2004, 7:32 AM
AND does this mean that reagrding the FX1 (nott he Z1) it is better to get the PAL camera (even in USA)? No matter what version of the FX1 you have, NTSC or PAL, vegas can capture and output any format but the PAL camera has 25p (or so you seem to be saying) when the NTSC FX1 has only the fake 24p. Is this what is being said? is it true?

The Z1 is the same world-wide being switchable NTSC/PAL, right?

dreamlx wrote on 12/4/2004, 7:33 AM
The PAL model HDR-FX1E can only do Cineframe25. I don't know exactly how cinframe24 on the NTSC model works. Maybe someone having the NTSC version can comment here. For the cineframe25 mode, converting to 25fps is rather easy. Simple mark the file as 24fps and slow down audio a bit.
mhbstevens wrote on 12/4/2004, 7:35 AM
AND does this mean that reagrding the FX1 (nott he Z1) it is better to get the PAL camera (even in USA)? No matter what version of the FX1 you have, NTSC or PAL, vegas can capture and output any format but the PAL camera has 25p (or so you seem to be saying) when the NTSC FX1 has only the fake 24p. Is this what is being said? is it true?

The Z1 is the same world-wide being switchable NTSC/PAL, right?

dreamlx wrote on 12/4/2004, 7:35 AM
Well the edit function seems not work. "converting to 25fps" should of course be "converting to 24fps"
dreamlx wrote on 12/4/2004, 7:41 AM
mhbstevens: I can only say for the HDR-FX1 25p is written to tape in Cineframe25 mode. However I don't know how Sony obtains the 25p. The m2t files on Spot's site are exactly how they are coming from the camera in Cineframe25 and have not been altered in any way. As I had no NTSC model for testing, I cannot tell anything about the NTSC model.
Spot|DSE wrote on 12/4/2004, 7:51 AM
Cineframe 25p is arrived at in the Z1 in the same way it's done with the FX1.
The Z1 is switchable to PAL/NTSC, and does CF 24, 24, and 30.