How to get the Hollywood DVD quality?

[jasonjoseph]

Before you say, "Aiyah Hollywood shoot on 35mm/Viper/Grass Valley/Red, surely better quality when burn to DVD lah."
and also before you say
"Google it"

my simple question is how do we get the best looking quality DVD from whatever camera or system we shoot on (no matter how crappy the dp was)? they must be doing something differently in hollywood right?

this is my Apple/FCP workflow
Footage shot on Sony NXcam -> Edit in FCP (1920) -> Send 1920 Seq to comepressor -> export best quality 90mins 2 pass vbr, max mbps as far as it can go-> import mpegs into DVDSP -> Burn.

it produces an ok looking dvd, not fantastic.

Just wondering whats other CSers workflows and your comments?

 

[jaegersing]

Of course they are doing it differently in Hollywood. For example, they take many hours setting up the lighting for every shot.

How does the footage look before you convert to MPEG2 and make it into a DVD? If it is already so-so due to indifference towards lighting the set (or other reasons) then the DVD creation process is not going to transform it into something better.

There are tools such as Magic Bullet and Colorista that allow you to change the appearance of the video, quite dramatically if that's what you want. But they are also applied before the DVD authoring stage.

http://www.redgiantsoftware.com/products/all/magic-bullet-colorista-II/

 

[jasonjoseph]

aiyoh...

maybe i should have simplified my request.

my question is how do you guys author the best quality DVD? or what is your dvd authoring workflow?
regardless of the footage that you acquired. or the number of effects you put on it.

reason i ask this, is because on this forum, we have all kinds of shooters, wedding videographers, industry freelancers, hobbyists. so we will have all kinds of footage, lit, unlit, HD, SD, 3/4 Chip, 1/2 Chip, chocolate chip . But when we export to dvd we're all basically dealing with the same export medium and most of us will be using simillar platforms.

and yes i know stu maschwitz's products very well, thank you

 

[jaegersing]

OK, I understand now what you are asking. What they do in Hollywood is probably not so relevant then. Below is what I normally do, on PC platform obviously.

For SD video source I use Vegas or Edius for editing, then from timeline to Procoder to create the MPEG2 files for DVD. Usually I compress at 7 or 8 Mbps CBR which gives great results.

For HD source I mainly use Edius because it is much better at handling HD playback in realtime. If quality is not so critical I go from timeline to Procoder as before. But for best results, especially where there is a lot of fine details with movement in the HD source video, I will go via Virtualdub to do high quality downsizing (HD to SD) before using Procoder to create the MPEG files. Procoder has a very good compression engine, but the resizing algo is not so great hence the use of this extra step.

I prefer to use CBR if possible because the quality does not degrade as it does with VBR. "If possible" means where the DVD duration is not so long, so I can use a high bitrate. Of course for longer DVDs at lower bitrates, VBR is better.

Up to one hour I use CBR video and uncompressed audio. Up to 90 minutes I use CBR video and Dolby stereo for audio. Above 90 minutes I use VBR and Dolby stereo.

If the video source is noisy enough to cause quality drop in the MPEG compression, I will maybe use noise reduction or temporal noise reduction to clean it up before rendering.

 

[DXNMedia]

Of course the source matters when you're performing the encoding process...heard of garbage in = garbage out?

For usual mid-end DVD mastering meant for press/duplication, we capture & master uncompressed video off Digital Betacam tape source....ie. 10bit Uncompressed 4:2:2 SD
From there, we pretty much followed similar procedures in Compressor using the highest quality 2-pass encoding.

For higher end DVD mastering, we pass the Digital Betacam SDI signal via a Snell and Wilcox box for image re-processing, ie. smoothing out noise & compressional artefacts, and also to stabilise the video signal...then we pass thru a Miranda MPEG2 dedicated encoder box for file generation, then we take the encoded file for mastering.
:thumbsup:

 

[jasonjoseph]

woah names i dun recognise

whats a snell & wilcox box and whats a miranda mpeg2 encoder ? what do they do ?

 

[flyingbee]

google it? hehe

 

[jaegersing]

No need to google it. The answers are in Dixon's post, you just need to read that.