Page 1 of 1

Why camera makers do not do proress? Some one please explain

PostPosted: Thu May 02, 2013 1:15 pm
by jonathanyonkers
Can some one please explain to me why camera manufactures do not do proress on their cameras?
This is a real question. I really just dont understand why.

RED is against it because they believe their codec is better. I agree and that I understand and respect. RED offers the option for you to do it separately. BUT, What about Canon, Sony, Panasonic, etc, etc.

Is it because cards (SDXC, CF) are not fast enough? Then how Black magic finally did do it? Or why not just use SS drive as media?
Is it because manufactures would have to pay too much money to someone in royalties? They how can Black Magic do it in camera that only sells for $1000?

Having a camera that shoots 1080p in proress at up to 60fps into any media has made sense for so long. Yet it never did haven. My question is: WHY?

Some one please do explain it to me.

Re: Why camera makers do not do proress? Some one please exp

PostPosted: Thu May 02, 2013 1:22 pm
by AdrianSierkowski
It's owned by Apple and to implement it you need to work out deal with, and pay, apple inc. Many companies don't see the need nor reason to nor have the want to enter into those negotiations. It's not personal, it's just business.

Re: Why camera makers do not do proress? Some one please exp

PostPosted: Thu May 02, 2013 1:29 pm
by Doug Guerra
It likely has to do with the hardware onboard the camera, and yes, the royalties (Though DNxHD is supposedly free).

ProRes and DNxHD are both extremely similar wavelet codecs, so creating an ASIC that can encode in both formats is not probably not too hard (ie, if you can do one, the other is very easily implemented).

Most cameras that can record internally, likely only have the hardware for the codec of their choice, and how it's implemented. For example, DSLRs use H.264, but with a very poor implementation that is incredibly inefficient (48Mbits looks like that?). Sony's implementation is about half that, while looking far superior. Fact of the matter is, the codec is built into the hardware and cannot really be changed. The throughput might be an issue, but fast enough media should do the job.

It's ultimately up to what he manufacturer wants to do, and where the camera is aimed in the market. For the C300, the camera encodes in 4:2:2 MPEG-2 50Mbit, which is broadcast spec, while providing ample recording time compared to ProRes. Different needs really. I guarantee you that a well-crafted H.264 codec at the same bitrate would wipe the floor with it, however.

Re: Why camera makers do not do proress? Some one please exp

PostPosted: Thu May 02, 2013 3:05 pm
by Frank Glencairn
Doug Guerra wrote:ProRes and DNxHD are both extremely similar wavelet codecs, so creating an ASIC that can encode in both formats is not probably not too hard (ie, if you can do one, the other is very easily implemented).



For the record: They are not wavelet codecs they are both DCT based, very similar to motionJPEG.
RedRaw and Cineform are wavelet codecs.

Re: Why camera makers do not do proress? Some one please exp

PostPosted: Thu May 02, 2013 3:13 pm
by Doug Guerra
Frank Glencairn wrote:
Doug Guerra wrote:ProRes and DNxHD are both extremely similar wavelet codecs, so creating an ASIC that can encode in both formats is not probably not too hard (ie, if you can do one, the other is very easily implemented).



For the record: They are not wavelet codecs they are both DCT based, very similar to motionJPEG.
RedRaw and Cineform are wavelet codecs.


Ahh! Confusion at its worst, I should know better. Was thinking of REDCODE's wavelet iteration and how it relates to JPEG2000!

Re: Why camera makers do not do proress? Some one please exp

PostPosted: Thu May 02, 2013 3:28 pm
by Peter J. DeCrescenzo
ARRI Alexa has been able to record ProRes for years, and external ProRes recorders have been used with a wide variety of cams for years also.

John Brawley recently posted these astonishing, but not unusual, statistics about his recent ProRes usage:

==============================
"... I just wrapped a drama series that was all ProRes. 75 production days. 78TB. Nearly 19,000 mins of rushes for 13 x 1 hours. Camera letter codes went to "i" and yes that included a pocket cinema camera. Was mostly EPIC but probably about 10% of our footage was BMCC. ..."
==============================

Sony, Panasonic & Canon have always used their own, or co-developed/"cabal" codecs because they believe it helps lock users into their product lines, and also allows them to mete out small incremental "improvements" under their strict control.

As the other poster said, it's not personal, it's business. For better and worse.

Re: Why camera makers do not do proress? Some one please exp

PostPosted: Thu May 02, 2013 3:52 pm
by Chris Hocking
RED also now offers a proxy module, which offers ProRes & DNxHD:

http://www.reduser.net/forum/showthread ... oxy-Module

Re: Why camera makers do not do proress? Some one please exp

PostPosted: Thu May 02, 2013 10:47 pm
by John Brawley
Data rate is one of the first hurdles.

Dealing with Apple is the second. Look at how many iOS developers complain about the way the app store works...

jb

Re: Why camera makers do not do proress? Some one please exp

PostPosted: Fri May 03, 2013 10:16 pm
by Aaron Scheiner
Not everyone needs ProRes... I admit I don't do hugely fancy projects, but the only time my clients ask me for ProRes (or better) is for VFX or very short sequences (like adverts)... almost everything else is still shot as MPEG2 36mbps - and the fact is it works and takes up very little space (as JB said, bit rate). I'm an example of a market segment that generally doesn't need better than MPEG2/36mbps or even H.264 24mbps... as sad as that is. The fact that most of Sony's customers don't need better than those codecs/bitrates allows them to avoid the licensing fees associated with Apple, the hardware requirements of high-bitrate codecs and the hardware costs associated with chips that can handle multiple codecs (they generally use ASICs, BMD is probably using an FPGA/s). It also allows them to segment their equipment for different markets so they can make more profit (as someone else hinted at).

Doug wrote :
For the C300, the camera encodes in 4:2:2 MPEG-2 50Mbit, which is broadcast spec... a well-crafted H.264 codec at the same bitrate would wipe the floor with it

H.264 is an amazing codec, it's not the codec that's the issue, it's a combination of the quality of the encoder, the bitrate and the codec itself (all relative to the content being encoded). People are very quick to judge H.264 as BAD because it's easy to lump everything into the same box, especially when someone is ignorant. This is very similar to the Full-Frame 35 argument... it's not full frame, it must be BAD.