Jump to content


Photo

External Recorders, Bit Rates, and Sampling


  • Please log in to reply
5 replies to this topic

#1 Marco S King

Marco S King

    New

  • Basic Members
  • Pip
  • 1 posts
  • 1st Assistant Camera
  • Johannesburg

Posted 10 October 2011 - 09:17 AM

Hi There

I have searched this forum about the following question but have had no luck finding answers:

Facts: Using an Ag-af100 with a nanoflash external recorder the bit rate goes from the standard 25mbps up to 250mbps. The colour sampling rate goes from 4:2:0 to 4:2:2.

Questions: having not observed any noticeable difference between the two images, standard and nano-infused, do external recorders coming through an HD SDI port somehow bypass the cameras compression settings and create a souped up image? Or does it take the cameras compressed image and digitally ramp up the bit rate?

How does the bit-rate actually affect the overall image?

It is all just a bit confusing and I thank you in advance for your answers and patience for my ignorance.
  • 0

#2 Freya Black

Freya Black
  • Basic Members
  • PipPipPipPip
  • 2933 posts
  • Other
  • Europe

Posted 10 October 2011 - 09:38 AM

The video coming from the HDMI or SDI port is coming from the camera before it hits the compression stage of the camera, so the video coming from these ports could be considered uncompressed in some sense (not entirely true but lets just pretend it is for now) As a result the external recorder can do its own compression that isn't as extreme as the one that takes place inside the camera. So no it's not taking the compressed image and ramping up the bit rate as that would degrade the quality of the image further.

Advantages of a less compressed image are that a 4:2:2 colour space is easier to key for chroma key stuff. The milder compression means the footage is easier for a colourist to grade without the image going bad. Also less compression means less video compression artifacts.

If you are trying to compare footage you have seen on the internet, keep in mind it has all been compressed to hell before uploading to youtube or Vimeo or whatever anyway. Also sometimes the differences can be subtle, and lastly, some people are by this point, so used to seeing compressed video, they don't notice any of the artifacts.

love

Freya

Edited by Freya Black, 10 October 2011 - 09:39 AM.

  • 0

#3 Phil Rhodes

Phil Rhodes
  • Sustaining Members
  • 9498 posts
  • Other

Posted 10 October 2011 - 11:34 AM

There are some subtleties to this.

Some broadcasters specify an absolute minimum bitrate for certain types of acquisition, often 50Mbps.

This is idiocy, because you will get better pictures with (for instance) h.264 at a given bitrate than you would with MPEG-2, and you'll get better pictures with a good encoder that costs a lot of money and consumes a lot of power, as opposed to one that's cheap and runs on double-A batteries. It also puts things like Panasonic's P2 format in an interesting position because although the bitrate for that format is commonly taken to be 100Mbps, at 24fps it's actually only recording 40.

Given that various codecs can also create problems with compatibility in postproduction, personally I prefer to get everything into uncompressed as soon as possible. The hardware required to handle it is no longer exotic. Recent developments such as Blackmagic's Hyperdeck Shuttle recorder make this so cheap and easy it almost seems churlish to turn it down. At some point, compression is just an excuse.

P
  • 0

#4 John Sprung

John Sprung
  • Sustaining Members
  • 4635 posts
  • Other

Posted 10 October 2011 - 01:03 PM

The problem is mostly moving and storage in post. Our shows typically shoot about 500 GB per day on Apple ProRes on the Alexas. If we tried do do that uncompressed, it would be north of 3 TB per day. And we'd be stuck with those archaic DPX files -- each frame is a separate file, and there's no sound.



-- J.S.
  • 0

#5 Phil Rhodes

Phil Rhodes
  • Sustaining Members
  • 9498 posts
  • Other

Posted 10 October 2011 - 04:39 PM

What exactly is the workflow that's used there?

I think once you get up into the hundreds of gigabytes it invariably boils down to an LTO solution if you aren't just willing to splurge on heaps of flash. I tend to engineer data handling to deal with the worst feasible case, working away from civilisation somewhere with at best a 12V power supply available. Of course if you're willing to assume more support than that, it gets easier.

P
  • 0

#6 Simon Jon Knight

Simon Jon Knight
  • Basic Members
  • PipPip
  • 28 posts
  • Cinematographer
  • Bristol

Posted 28 October 2012 - 04:21 AM

In answer to Marcos original post and to agree with Freya..

You really should look at native footage shot on a camera. I have a trusty Sony Z5 and went from HDV tape to using it with an Atomos Ninja. The difference is incredible. Exactly as Freya says, the coloring in Final Cut is solo much better. You can shoot at nearly a stop slower and still pull tons out of the lowlights. Made the Z5 with 0.5 lux into a midnight wonder cam... Well almost.!

Best £600 you can spend on a small-bitrate camera... And you are HD channel ready should something you shoot get used!

Enjoy..

Simon Knight
MediaHound Films
  • 0




Paralinx LLC

Aerial Filmworks

K5600 Lighting

Lemo Connectors

Robert Starling

Visual Products

Zylight

Cinelicious

CineLab

Abel Cine

System Associates

rebotnix Technologies

Rig Wheels Passport

The Slider

Ritter Battery

Glidecam

Pro 8mm

Cadrage Directors Viewfinder

NIBL

CineTape

The Slider

Abel Cine

rebotnix Technologies

Zylight

K5600 Lighting

CineLab

NIBL

Visual Products

Rig Wheels Passport

System Associates

Pro 8mm

Lemo Connectors

Cadrage Directors Viewfinder

CineTape

Glidecam

Paralinx LLC

Ritter Battery

Aerial Filmworks

Cinelicious

Robert Starling