Help a newby understand Cineform (and Blackmagic)

[Jason Goldberg]

Hi, I'm an enthusiast/hobbiest who is getting back into digital video after a 4 year absence. I've picked up a Sony HC-7 HDV camera and a license of Vegas Pro 8.
I'm running a basic PC with a 2.66GHz quad core 2 duo, 4 gigs RAM, a GeForce 8800GTX, 7200 RPM drives, and unfortunately Vista 32-bit. I'm reasonably technical, but pretty out of date.

I've read a lot of references to Cineform, and so visited site to figure out what it was. It sounds like Cineform is a less lossy codec for lower CPU utilization and higher quality.

So basically, you use more HD space (and presumably HD speed) in exchange for less multi-generational loss and less CPU utilization. Is that about right? As I understand it, HDV is 1440x1080 (non-square pixels) at 8-bit with 4:2:0 color which is using MPEG2 compression (DCT with heavy interframe) at like 25mbit/sec.

The Neo HDV variant of Cineframe is 1440x1080 at 8-bit with 4:2:2 color, using DWT/Wavelet intraframe with light or no interframe compression in a .MOV or .AVI wrapper at around 75 mbit/sec?

I assume Neo HDV installs as a codec and a utility to capture, and that I need to either capture using the Cineform tool to get NeoHDV .avi's, or I can capture .m2t in Vegas and then use the Cineform tool to transcode to NeoHDV .avi's?

Assuming I have all this right my first set of questions is around capture:

Can the Cineform tool capture .m2t's and .avi's in one process (in case I want to archive the .m2t's)?
Does Cineform capture from my camera in real-time?
If I capture .m2t's can it transcode to NeoHDV in better than real time?
Does the Coneform tool have any advantages in scene detection or device control over vegas?
Does the either capture tool, add any meta data to clips (tape name, timecode in and out, user entered clip descriptions)?
Does either tool have any flexibility for clip naming convention (when capturing a whole tape with w/ scene detection)?

My other questions are about application:

If I were doing a project with just light cuts, I assume I'd be better off with native .m2t's because large pieces of my video might not need to be re-encoded/re-rendered?
As I apply more color-correction, filters, etc... I'd get more more benefit from the less lossy NewHDV codec?
Besides CPU use and quality, any other benefits to NeoHDV over HDV?
If using the NeoHDV codec, can you still print to tape?
Can Vegas still preview out the firewire to my camera to use my cameras HDMI out for a preview?
If I moved up to NeoHD, I'd get 10-bit and 1920x1080. Since I don't have any other video sources besides my HC-8, I assume the big benifit to NeoHD would be in adding 1920x1080 stills/grafics to the native timeline?

My last questions are about Blackmagics Intensity card (maybe the wrong forum):

It seems like this is mainly a capture card for HDMI streams. It claims to offer quality advantages over native HDV via firewire, with a camera like the Sony HC3 (or my HC7). Can that really be true? I know that HDMI has full 1920x1080, better color, etc... but coming out of a Sony HDV camera it's all interpolated up from the HDV data on the cameras sensor, right? On my camera, the data coming live from the HDMI port is no better than the data that would come from the HDMI if I played back an HDV tape.

It seems like the Blackmagic is mainly useful for giving a HDMI preview, and/or for live capture from fancier cameras that have full 1920x1080 sensors. Do I have all that right?

I know this is a ton of questions, and I'll keep digging for more info on my own but if anyone's inclined to answer even some of my questions, I'd be be very greatful. Thanks for your patience!

Jason

[Carl Middleton]

Okay... Let's see what I can do with this.

Originally Posted by Jason Goldberg

Hi, I'm an enthusiast/hobbiest who is getting back into digital video after a 4 year absence. I've picked up a Sony HC-7 HDV camera and a license of Vegas Pro 8.
I'm running a basic PC with a 2.66GHz quad core 2 duo, 4 gigs RAM, a GeForce 8800GTX, 7200 RPM drives, and unfortunately Vista 32-bit. I'm reasonably technical, but pretty out of date.

I've read a lot of references to Cineform, and so visited site to figure out what it was. It sounds like Cineform is a less lossy codec for lower CPU utilization and higher quality.

So basically, you use more HD space (and presumably HD speed) in exchange for less multi-generational loss and less CPU utilization. Is that about right? As I understand it, HDV is 1440x1080 (non-square pixels) at 8-bit with 4:2:0 color which is using MPEG2 compression (DCT with heavy interframe) at like 25mbit/sec.

The Neo HDV variant of Cineframe is 1440x1080 at 8-bit with 4:2:2 color, using DWT/Wavelet intraframe with light or no interframe compression in a .MOV or .AVI wrapper at around 75 mbit/sec?

I assume Neo HDV installs as a codec and a utility to capture, and that I need to either capture using the Cineform tool to get NeoHDV .avi's, or I can capture .m2t in Vegas and then use the Cineform tool to transcode to NeoHDV .avi's?

You are prettymuch correct through here. The bitrate will vary depending on quality settings you select, though. And I think Cineform prefers the term 'Visually Lossless' ;)

Assuming I have all this right my first set of questions is around capture:

Can the Cineform tool capture .m2t's and .avi's in one process (in case I want to archive the .m2t's)?

Yessir.

Does Cineform capture from my camera in real-time?

Yep!

If I capture .m2t's can it transcode to NeoHDV in better than real time?

Correct again, my friend.

Does the Coneform tool have any advantages in scene detection or device control over vegas?

I'm not familiar with Vegas, but the scene detection does work quite well. My personal workflow is to capture entire tapes.

Does the either capture tool, add any meta data to clips (tape name, timecode in and out, user entered clip descriptions)?

This is another one I'm not sure about. I don't use metadata often.

Does either tool have any flexibility for clip naming convention (when capturing a whole tape with w/ scene detection)?

No, but HDVSplit has the ability to use the timecode from the camera when naming tapes, HDLink uses a numbering system.

My other questions are about application:

If I were doing a project with just light cuts, I assume I'd be better off with native .m2t's because large pieces of my video might not need to be re-encoded/re-rendered?

Incorrect. With Long-GOP compression, as is the case with m2t's, every time you create a cut, etc, it requires a rerender of some sort or another. As well, at least in Premiere, it will force you to render the whole thing to output, even as m2t.

As I apply more color-correction, filters, etc... I'd get more more benefit from the less lossy NewHDV codec?

The less-compressed Cineform files are very easy to work with, though again I use Premiere, and have been spoiled by the realtime effects.

Besides CPU use and quality, any other benefits to NeoHDV over HDV?

The external tool to export m2t to tape is rather nice. No render, just a straight throw to tape.

If using the NeoHDV codec, can you still print to tape?

You'd output an m2t file, and bring it back into HDlink.

Can Vegas still preview out the firewire to my camera to use my cameras HDMI out for a preview?

Not sure there, man... sorry...

If I moved up to NeoHD, I'd get 10-bit and 1920x1080. Since I don't have any other video sources besides my HC-8, I assume the big benifit to NeoHD would be in adding 1920x1080 stills/grafics to the native timeline?

10 bit color correction is a beautiful thing. You'll get better quality for your CC touchups. As well, The ability to work in true full size HD is great if you are looking to possibly output to bluray or any other full res format.

My last questions are about Blackmagics Intensity card (maybe the wrong forum):

It seems like this is mainly a capture card for HDMI streams. It claims to offer quality advantages over native HDV via firewire, with a camera like the Sony HC3 (or my HC7). Can that really be true? I know that HDMI has full 1920x1080, better color, etc... but coming out of a Sony HDV camera it's all interpolated up from the HDV data on the cameras sensor, right? On my camera, the data coming live from the HDMI port is no better than the data that would come from the HDMI if I played back an HDV tape.

I do not *believe* that is correct. You can use that to capture live if recording directly to your computer, bypassing the m2t compression, which is the *most* painful thing in post with HDV.

It seems like the Blackmagic is mainly useful for giving a HDMI preview, and/or for live capture from fancier cameras that have full 1920x1080 sensors. Do I have all that right?

See above. :)

I know this is a ton of questions, and I'll keep digging for more info on my own but if anyone's inclined to answer even some of my questions, I'd be be very greatful. Thanks for your patience!

Jason

[Andrew Swihart]

This guy (link below) feels very strongly that a lot of detail is lost during HDV compression. I can only imagine how he must feel about the AVCDV compression in newer cameras!

http://dvinfo.net/conf/showpost.php?...&postcount=195

[Graham Morton]

he's right... not only in initial compression but then all the time saved in editing waiting for rendering but also space on an hd and all the other cool stuff, i have been shooting hdv since for over three years now and cineform's tools have been a godsend.