DV100 vs DNxHD vs CFHD

[Graeme Nattress]

So how is converting DVCproHD to an intermediate compressed codec going to improve matters? This is just the same as people who edit DV uncompressed thinking they’re getting a better picture out at the end – they don’t! Even if you’re mastering to D5 or HDCAM SR, you’re still going to edit your DVCproHD footage as DVCproHD, flip your codec settings to uncompressed when you’re finished, do a final render then out to tape. Where does an intermediate compressed codec help you here?? If you’re plugging in a deck worthy of rendering uncompressed, you’ve certainly got the money for the RAID and HD card to handle it.

Graeme

– http://www.nattress.com – Film Effects for FCP

[Luis Caffesse]

[Graeme Nattress] “o how is converting DVCproHD to an intermediate compressed codec going to improve matters?”

That’s what I was looking for Graeme.
Especially when it comes to P2, transcoding to some other ‘intermediate’ codec seems like a lot of work for little if no return.

Luis Caffesse
Studio 3 Productions, Inc.
Austin, Texas

[Graeme Nattress]

At that point, surely, you edit uncompressed and have the advantage of faster renders due to much less processor overhead.

Graeme

– http://www.nattress.com – Film Effects for FCP

[Luis Caffesse]

[David Cherniack] ”
I believe the DVCProHD codec and the CFHD codec use about the same amount of bandwidth. Last time I looked they did. “

Right, but if we were to transcode all of our footage from DVCProHD we would need twice the hard drive space. Remember that on P2 everything is being kept on solid state, and then transfered directly to your hard drive for editing (with no digitizing).

If you were to transcode 1GB of footage, you would need 1GB to store the DVCProHD footage, and 1GB onto which you would save the CFHD footage.
Hence, twice the hard drive space.

That’s all I meant.

Luis Caffesse
Studio 3 Productions, Inc.
Austin, Texas

[Graeme Nattress]

My thoughts exactly. Intermediate codecs are for when you’re shooting with a distribution codec. DVCproHD is more than suitable for both acquisition and editing. Then you convert to a delivery codec, or render out uncompressed for D5 or HDCAM SR. No intermediate codec needed at any point here.

Graeme

– http://www.nattress.com – Film Effects for FCP

[Graeme Nattress]

All the HD deliver formats are not full raster. Broadcast HD is not full raster. If you’re doing film work on HD, you’re doing it totally uncompressed, so no need for an intermediate codec there.

Cineform’s quality white paper is heavily biassed, and because of their workflow, they repeatedly squash and stretch the image from 1280×1080 to 1920×1080 then back again as well as apply the DVCproHD compression – of course it’s going to look soft if you do that to it. But if you’re editing DVCproHD, then this does not happen. If you dub DVCproHD from one deck to another you do it over firewire and loose no generational loss at all.

Again, why are we talking about an intermediate codec designed for HDV use (because HDV as MPEG2 is painful to edit) when you can quite easily edit DVCproHD without converting to another codec? I really don’t see the point here. As for DVCproHD not being online quality – it’s no worse a codec than the HDCAM one, and any HDCAM workflow, even editing uncompressed isn’t going to produce better results really, and both HDCAM and DVCproHD are more than good enough to wow the viewers at home and keep the broadcasters happy. As I say, if you’re going for a more exacting environment, like for digital cinema, you’ll be going totally uncompressed anyway.

Graeme

– http://www.nattress.com – Film Effects for FCP

[Karl Holt]

pardon my ignorance if Ive got it wrong, but once you’ve captured in DVCPROHD then the damage (if you want to call it that) has been done. you’re then at 920×720 instead of 1280×720. Converting to CFHD isnt really going to help from that point (unless you’re talking about recompressing lots).

I keep getting the impression that a hack or firmware upgrade may be able to make the cam record in a different format. CFHD is probably a too labour intensive codec to capture real-time too, but it wouldnt prevent the camera from accepting an upgrade to the current DVCPROHD codec to capture the full HD signal in the future. The DVCPROHD codec maybe limited in the way it is simply because of a tape transport mechinism – it wouldnt suprise me if it changes to accomodate a larger res now P2 is here.

If you want to edit in CFHD for better recompression – then Im sure you can convert it all later in your NLE; but I’d like to see the source material use the full pixel resolution from the start.

Karl

[Vsv]

Jason J Rodriguez
DNxHD is DCT-based, not wavelet.

Thank you, Jason, for clarification.
Which NLE is better for editing CFHD 10bit in YUV2 under WinXP?
Sony Vegas can not to render to 10 bit YUV2 ? How about Avid Express Pro HD?
Under winXP no choice, only PPro1.5?

[Graeme Nattress]

The DVCproHD codec is done using 4 DVCpro codec chips, and DVCpro50 is done using 2 DVCpro codec chips. DVCpro just uses one of them :-).

There is a specific data rate that each chip can handle. Therefore resolution of image is limited. 720p is double the pixels of DV, and 1080i is double the pixels of 720p in DVCproHD format.

These 4 chips are not user-reprogamable, so don’t expect “upgrades” to them.

As for capturing the “full” resolution – I think you’re expecting too much here. Every camera has a resolution filter on the front of it – the lens. Lenses that let more detail through are more expensive. I thoroughly expect that even if you could rig up the camera to capture the full raster, although you’d get more resolution, I severely doubt you’d get more actual real detail. It’s detail that counts, not resolution. You can take VHS and blow it up to HD size, and now it has 1920×1080 pixels, but it has no more DETAIL than which it started with.

Graeme

– http://www.nattress.com – Film Effects for FCP

[Karl Holt]

doesn’t this come down to how good the lens on this camera will be?