I downloaded all the footage and have been playing it with a variety of players. It looks tremendous to my eyes. There is quite a difference between players though. I have an old version of Media player (Media Player Classic) that seems to work better than version 10.

Another interesting thing: if you play it back at half size, the progressive frame is still bigger than an XL2's or DVX100's progressive image!

On the con side, Vegas performance with these clips on my P4 3.06 is not good enough to be able to edit with, at least in the mts format.

Using a software player on a computer is not the best way to judge the quality of these clips. Probably the best way to judge them is to use a hardware decoder and a monitor capable of native 1080i. I've been looking at all of these clips on the MyHD card and my 34" HD monitor and, I must say, they look very good. By the way, there are some new FX1 clips posted at Kaku's site (see the link in the first message of this thread.)

John

I notice that, when brought into v5, it's being detected as "upper field first". Is this a PAL camera? Can't be because it's 29.97 fps. So, intentionally selecting lower field first seems to minimize, altho' doesn't eliminate, the motion jaggies.It's also detecting the PAR as 1.333

Procoder2 is reporting that it is upper field first as well, but upper field first in an MPEG2 file is not unusual.

Here is a .gif image of Procoder's report on the HDV source file parameters:

Procoder HDV report

I have seen no "motion jaggies" when viewing any of these files on my HD monitor.

John

No offense to anyone, but a simple point, everyone needs to remember that Mpeg2 is an encoded format, which really means that the encoder can actually produce excellent, good or poor ouput based on control parameters, as long as the decoder can handle the information. New encoding software can produce an encoded stream that approaches lossless compression if you change the parameters at the encoder stage. Sony has thought this out, they are not new to this issue, the Mpeg2 is an old format, but math has improve the encoding technology.

Fyi: Native interlaced Sony HDV footage will always be upper field first, par 1.3333

A Pixel Aspect Ratio of 1.3333 makes sense... 1440 / 1080 = 1.3333

Is it really upper field first? I re-rendered it as "lower field first" and the motion artifacts seemed a little better. Am I deluding myself?

Well I just did a test and the FX1 footage renders into outstanding SD widescreen. A little bit is chopped off the top and bottom but the difference in aspect ratios would be negligable, and the interlacing pattern is like it should be. Also, the footage doesn't seem to take that 25% line averaging performance hit that plagues interlaced cameras.

My only problem with the footage is how much it taxes my CPU in Vegas. Does anyone know if Sony is going to go for the hardware accelerator approach or the faster codec approach as is done by the Cineform codec?

Bill,

Yes, it seems that it really is upper field first. You really shouldn't be seeing any motion artifacts, what modifications are you making to the footage, how are you rendering it and how are you viewing the result?

John

John...

I'm taking the raw m2t files and playing them directly with Zoomplayer. If I use the Elecard mpeg2 codec, it shows severe motion jaggies. If I use ffdshow, the jaggies look morelike motion blur. Importing directly into V5 and accepting the default clip properties shows bad motion jaggies.....just like interlaced video is supposed to look on a progressive monitor. Beause it looked so bad, I then tried lower filed first and rendered to DV (720x480)i....no other changes.

Stay tuned for some HDV/Vegas announcements but as of right now we point to the Cineform Connect HD tool as our go-to solution for anything HDV. Most of you HDV-savvy types know we've been working with Cineform for well over a year, together we've worked out many kinks in the HDV worlkflow, and if you've ever tried CHD you know there's a huge performance boost over native TS with no measurable quality degradation at all.

Sorry I cannot find the link again, anyway it seems someone has written a rather neat conversion tool for downsampling HDV. Instead of simply reducing the res this thing makes use of the additional data to produce a lower res but better sampled AVI, for example down convertering 1080i at 4:2:0 to 720 gives you 4:2:2 and I think you can get SD at 4:4:4.
Seems to take a LONG time to do it's trillions of calcs but results were reported to be pretty damn good.
If anyone knows where the link is please post, this sounds very interesting. I'm suspecting someone will be offering the same thing in hardware (Miranda) at a matching price but for free this sounds good.

It seems to me like you should just be able to drop lines in an interlace pattern aware way and accomplish the same thing with no heavy math. I just set the project properties to widescreen dv, right clicked on the framing of the clip and set it to "match output format" and got great quality widescreen SD right off the bat.

Apparently there is a tool called CapDVHS (can be found via google) that can also be used to capture hdv data from the HDR-FX1 into m2t files usable by Vegas. As it will still take some weeks until I receive my HDR-FX1 I have no possibility to test this. Maybe someone from Sony is interested in testing as this would be a free of charge workaround until Vegas is updated.

Yes you can do that however you are throwing away chroma data. Basically all you end up with is DV25 at 4:1:1, sure it's very good looking DV25 however you could go to 4:2:2 and if you are going out to DVD there should be a noticable improvement if you can stand the time it takes for the conversion. Also things like CC may work better.

Bob.