HDV Capture and Intermediate files

edmellnik wrote on 8/4/2005, 4:05 PM

IT is turning out that Vegas may be a real dissapointment as far as HDV goes.
I was hoping to be able to capture in DV as a lower res and then print to tape in HDV but it looks like that is not possible unless I am missing something.

I also tried the intermediate feature....capture the HDV stream then render intermediate files for editing.... I set the intermediate as a DV file. The DV file was four times the size of the HDV file. So using intermediate files sure does not save any space.
Since the Z1U will downconvert in DV, it would seem a no brainer to capture everything donconverted to dV and when you are finished editing just batch capture the EDL or project in HDV then print to tape.....but if that is possible I have not found it yet.
Ed

Comments

Spot|DSE wrote on 8/4/2005, 4:28 PM

No, you can't do this. Did someone somewhere suggest you can?
Some where you've got some information wires crossed.
1. if you want HDV, you MUST capture as HDV. (m2t)
2. If you want to use DV proxies, they must be rendered from the m2t files, otherwise they're not relevant.

There is no batch capture for any application for HDV, therefore the concept of editing with an EDL and then recapturing the HDV, is not possible in any application on any platform.

A DV file cannot be 4 times larger than an HDV file. They are the same filesize (approximately) as 1080i is the same bitrate as DV is. (25Mbps)
HDV is approx 13 gig an hour, so is DV. DV is intraframe whereas HDV is interframe, so it's not quite possible to be as exact as DV is, but they're approximately the same size, unless perchance you're using a JVC HDV cam where it's 720p vs 1080i. (720 is 19Mbps)

I'm not sure how you're thinking you can capture a low res file and turn it into an HD file, and how you set the intermediate as a DV file if you captured a DV file?
Maybe I'm confused about your workflow, but am pleased to help if I can.

Wolfgang S. wrote on 8/5/2005, 12:21 AM

I think there is no softwar, that will scale down HDV to SD-resolution during capturing. What you can do, is to scale down to SD-resolution with the camcorder during capturing - the FX1 has the ability to deliver DV-avi if you switch the camcorder to that mode.

But that is a capability of the camcorder, not of Vegas.

philfort wrote on 8/5/2005, 11:29 AM

Is it really true that you can't print HDV back to tape? I just assumed you could. What is the technical limitation there?

As for generating intermediate files - are there any good instructions for this somewhere? The Vegas help file isn't very helpful. I tried rendering to the cineform codec (which fails with "an error", if you don't set the exact size of the video correctly), but I must have messed something up, because the resulting file had messed up aspect ratios, and extra black space around the video. I was hoping for the workflow to be more automatic - I mean, it should be a pretty brain dead operation: "just make a cineform intermediate file from the .m2t". I should haven't to drag it onto the timeline and render, and figure out render settings, etc...

Spot|DSE wrote on 8/5/2005, 12:24 PM

Of course you can print HDV back to tape.

Unless you use GearShift or the CineForm capture tool, you must drag the m2t file to the timeline, and render. What you can do is make a region of each clip, use the batch render tool, and then use the intermediate template that is already provided for you. Nothing to figure out.

This is what GearShift does for you, it loads files from the Media Pool, Timeline, or Explorer, and creates CineForm, Proxies, 4:2:2 YUV, or any combination thereof for you. it can work from files or from regions you've defined.

The batch render tool can do this as well, albeit without the same proxies, conversions, colorspace setup, etc.

edmellnik wrote on 8/8/2005, 3:16 PM

I thought the whole Idea of an intermediate file was that you were then working with a more computer friendly file that would play without choking. This it does But...The intermediate file IS four times the size of the HDV file so I must have done something wrong.

One should be able download the footage in DV using the Camera capability to downconvert to DV....create a project then batch capture the HDV footage in HDV the render and print to tape.
This would save all the time of rendering a new set of intermediate files to work from.
IT is also unfortuneate that EDLs cant be saved in the HDV mode.
For a SOny camera and a SOny NLE there is very little help in the HDV mode.
Ed

Liam_Vegas wrote on 8/8/2005, 4:00 PM

I thought the whole Idea of an intermediate file was that you were then working with a more computer friendly file that would play without choking.

That's exactly the idea.

This it does But...The intermediate file IS four times the size of the HDV file so I must have done something wrong.

Nope... that is the price you pay. The intermediate file is more "CPU" friendly that working on the "long GOP" structure of the m2t file.

This is why Gearshift can be a better way... as it allows you to work on DV format intermediaries with smaller file sizes than the Cineform intermediate.

Spot|DSE wrote on 8/8/2005, 11:19 PM

Ed,
It's got nothing to do with the Sony camera/Sony NLE combination. It has everything to do with the format. Frame accuracy in HDV currently isn't really doable over firewire, so Batch capture isn't available. This is why it's important to start with the m2t and convert from there.
It's expected that the intermediary file is more than 4 times the size of your DV stream; so is the resolution. NTSC-DV has just under 346K pixels. 1080i/60 HD has just over 1.5M pixels. Huge difference. This is why proxy editing is great for folks that have older, or slower computers. CineForm's tool is outstanding, but if you've got limited diskspace, then proxy-based editing is the only way to fly, because you only need double the space of the original m2t to do it.
1 hour of m2t=13 GB, approx, and one hour of NTSC DV=13GB, approx. Therefore, one hour of both file formats=26GB, which is still nearly half the 40GB you're needing for 4:2:2 short GOP files.

Wolfgang S. wrote on 8/9/2005, 2:50 AM

Ed, take care to distinguish in a clear way between Proxys and Intermediates. Intermediates keep the full HDV quality - since it is assumed that you render to your final video using those intermediates. And editing is easier, since this Intermediates are based on a frame internal compression only, compared with m2t footage and the GOP structure.

Proxys are smaller files, e.g. widescreen DV-avi proxys, or I tend to use mjepg-avi proxys for some pixel aspect and sizing reasons. Proxys are much easier to edit then intermediates. But it is assumed that you switch back to the original m2t footage before you render your final video. Gearshift is a good way to do that.

So, what is better for you depends on your PC capabilities, and on your requirements. But also on your desired workflow.