DV Info Net

DV Info Net (https://www.dvinfo.net/forum/)
-   Convergent Design Odyssey (https://www.dvinfo.net/forum/convergent-design-odyssey/)
-   -   nanoFlash Codec compared to ProRes (https://www.dvinfo.net/forum/convergent-design-odyssey/492746-nanoflash-codec-compared-prores.html)

Dan Keaton March 7th, 2011 08:59 AM

nanoFlash Codec compared to ProRes
 
Dear Friends,

Based on questions that I have received, I feel that there is a common misconception about ProRes.

The common perception is that ProRes HQ is 220 Megabits per second (Mbps).

Based on the Apple ProRes White Paper, the target bit-rates vary depending on the resolution and the frame rate.

The Target Bit-Rates for Apple ProRes are:

1080p24......................176 Mbps
1080p25,50i................184 Mbps
1080p30,60i................220 Mbps

720p24.........................88 Mbps
720p25.........................92 Mbps
720p30.......................110 Mbps
720p50.......................184 Mbps
720p60.......................220 Mbps

Note: The actual bit rate may be lower (if the video is easy to encode)
but will never exceed the target rate by 10%.


For the nanoFlash, the bit-rate that you select will be the actual bit rate.

Thus for 720p at 220 Mbps you allways get 220 Mbps regardless of the frame rate.

The same applies for 1080p at 220 Mbps.


Depending on the resolution and frame rate, the difference in actual bit-rates can be very significant.

This helps explain why the nanoFlash can produce such excellent images.



Notes:
Any bandwidth for the audio, in the nanoFlash, is in addition to the selected bit-rate, thus you always get the bit-rate you requested for the video.

There are a few technical exceptions, such as when we are converting 720p60 to 720p30, or cranking, since the duplicate frames are eliminated after the frame has been encoded.


For more details on the Apple ProRes codec, please refer to page 20 of the following:

http://images.apple.com/finalcutstud..._July_2009.pdf

Brian Walstad March 10th, 2011 01:59 PM

Re: nanoFlash Codec compared to ProRes
 
This is great information that really sets the Nanoflash apart in my opinion. Dan I think you should do more to get this information out to the masses. I think maybe there should be a new way of calculating bitrates, instead of Mbps why not Mbpf (Megabits per frame)?

ProRes 422 (HQ) via KI Pro/Mini or other device = up to 7.333 Mbpf for 1080p or up to 3.667 Mbpf for 720p
Unthrottled XDCAM 422 via Nanoflash = up to 11.667 Mbpf for either 1080p or 720p

With the Nanoflash you can get 1.59 times information per 1080p frame than you can with any device recording ProRes 422 (HQ) and 3.181 times more per 720p frame. In fact according to my calculations the Nanoflash can even exceed ProRes 4444 in 720 24p in information per frame by 2.121 times and in 1080 24p by 1.06.

Bottom line, if you are shooting 24p, you're better off buying the Nanoflash especially if you are shooting 720.

Thanks CD for an awesome product.

Dan Keaton March 10th, 2011 02:58 PM

Re: nanoFlash Codec compared to ProRes
 
Dear Brian,

Thank you!

Nice Post!

Yes, the nanoFlash is rather unique and the higher "Megabits per frame" really adds to creating excellent images.

While the numbers are impressive, the real proof is in examining the images.

What we get most often from people seeing our images for the first time is "Wow, that is really clean!".

Brian Walstad March 10th, 2011 03:07 PM

Re: nanoFlash Codec compared to ProRes
 
Yeah that is my experience when using it a few times with my HPX170 and AF100. Still haven't bought yet tho only rented. Probably my next purchase.

Andrew Stone March 11th, 2011 03:19 PM

Re: nanoFlash Codec compared to ProRes
 
Fascinating Dan. Thank you for that info about the auto scaling bitrate of ProRes. Lord help you if you are shooting in 720p24 and using ProRes and expect the results associated with 220 mbit.

Dan do those bitrates scale down to those figures regardless of whether the signal hitting it is 8 bit or 10 bit?

Dan Keaton March 11th, 2011 10:24 PM

Re: nanoFlash Codec compared to ProRes
 
Dear Andrew,

All of the detail is in Apple's ProRes White Paper.

I do not know if it answers your question in the White Paper.

Rafael Amador March 12th, 2011 06:10 AM

Re: nanoFlash Codec compared to ProRes
 
Quote:

Originally Posted by Andrew Stone (Post 1627038)
.... Lord help you if you are shooting in 720p24 and using ProRes and expect the results associated with 220 mbit. ?

You get exactly the same quality.
In the end what matters is the data/picture.
In 720p24 (88Mbps) you get the same data/picture than in 720p60 (220) so the same quality.
In 720p60 you have 2,5 x more images to transcode. You need 2.5 x the data rate to keep the same quality.

Quote:

Originally Posted by Andrew Stone (Post 1627038)
....
Dan do those bitrates scale down to those figures regardless of whether the signal hitting it is 8 bit or 10 bit?

Prores is always 10b.
The NANO is always 8b.
rafael

Andrew Stone March 12th, 2011 03:41 PM

Re: nanoFlash Codec compared to ProRes
 
Hi Rafael,

Yes I understand that. My comments were based on thinking through scenarios of how the Ki Pro Mini would deal with data streams that were either 8 bit or 10 bit and how that would effect what bitrate the Ki Pro Mini would invoke under a given scenario.

But since we are now on the topic... an 88 mbit encoding on 720p24 10 bit data stream is going to have a lot more trouble reproducing a faithful image than an 8 bit stream capturing 720p24 at say 180 mbit. I would hazard a guess the 8 bit stream captured at 180 mbit data rate would be more pleasing to the eye.


All times are GMT -6. The time now is 06:04 PM.

DV Info Net -- Real Names, Real People, Real Info!
1998-2024 The Digital Video Information Network