Bitrate in crank mode

[Piotr Wozniacki]

Dear Dan,

Now that my serious problems with using even the basic nanoFlash functions are gone thanks to the Second Beta modifications to the firmware, I'm testing more advanced functionality of this sweet device. One of them is the Crank function. Usually, I set my EX1 to 720/50p, and the nano to 25 (base fps) and 50 (overcranked fps). All is fine, but I've got this question:

- As we know, the nano will use I-frame mode for cranking (even if it's set to Long-GoP). I don't quite understand the "rules" on which bitrates it uses then. For instance, if it's set to 180 Mbps Long-GoP for normal recording (no cranking), "180M" is displayed in the nano's LCD status area. When I engage overcranking, it changes to "90M" - why?

When using I-frame, I'd expect nanoFlash to use a higher bitrate than that for Long-GoP!

[Dan Keaton]

Dear Piotr,

Good Question.

I will attempt to get an answer for you today.

[Piotr Wozniacki]

Bump

:)

[Dan Keaton]

Dear Piotr,

I spoke with Mike Schell about this yesterday.

Our engineers are looking into this. Of course, 90 Mbps does not make sense, we do not evern have a 90 Mbps mode.

[Fletcher Yeung]

Hi,

I too am starting to have issues using a nanoflash in cranking. I set my EX-1 to 720p60 then on the Nanoflash, I set it to 720p 25, with a frame rate of 60. Changed bit rate to 100Mbps, hit record, and the Mbps drops to 50mbps.
I'm using Sandisk 32gb Extreme III cards, and in normal recordings, they are fine, it stays constant at 100Mpbs. But just like Piotr, it only happens now in cranking mode.
I tried to check the footage in FCP, and I get green and black flashes throughout the video.
Any ideas?!

Thanks!

[Ray Paunovich]

I had the same thing happen to me ( EX3 ) but I was in 720P60 mode and 24 and 60 crank mode so what I did was bump up the bitrate to 180. here is Toms reply..

Hi Ray,

it's related to the overcrank / undercrank, the manual is not so obvious about this but there's a statement in the crank discussion:

> The bit rate of cranked recording is based on the incoming video frame rate. For example, a 720p60 incoming signal @ 280 Mbit will allocate the 280 Mbit across every 60 frames.


So this means that, for example, a 720p60 set to a record rate of 720p24 @ 100 Mbits / sec will spread the 100 Mbits across 60 frames,
meaning for every 24 frames the bit rate is 40 Mbits /sec (which is rounded up to a minimum 50 Mbit).

We need to base the bit rate on incoming rate, not the recorded rate, because of the way our hardware and codec handle cranking.

So yes the bit rate will be reduced based on the ratio of recorded frame rate to incoming frame rate.

Tommy


Ray Paunovich

[Rafael Amador]

When we do overcranking with the EX-1, we end up with a file (p30/25/24)) that shows a data rate of 35 Mbps. This is when the file is played.
But when we are recording that in the SxS (p60 or p50), in fact we are writing at least 2 x 35 Mbps.
With the NANO happens the same. If you record a signal at 100 Mbps and you play it at 50% speed, the final date rate you are getting is 50%.
If your want that your final p30 clip to be 50Mbps, you need to record the p60 at 100 Mbps.
rafael

[Fletcher Yeung]

Cool! I was wondering how I was able to do get 100mbps before, didn't think that the actual bit rate was answer! Thanks for info!!!

[Piotr Wozniacki]

Originally Posted by Rafael Amador

When we do overcranking with the EX-1, we end up with a file (p30/25/24)) that shows a data rate of 35 Mbps. This is when the file is played.
But when we are recording that in the SxS (p60 or p50), in fact we are writing at least 2 x 35 Mbps.
With the NANO happens the same. If you record a signal at 100 Mbps and you play it at 50% speed, the final date rate you are getting is 50%.
If your want that your final p30 clip to be 50Mbps, you need to record the p60 at 100 Mbps.
rafael

Thanks Rafael - this analogy explains it in a most convincing fashion.

I only wonder (pure curiosity), whether at the nominal 180 Mbps (see my OP), it's indeed recorded at the (displayed) 90Mbps (which is not one of the available speeds), or at the nearest one (which would be either 100 or 80 Mbps)...

[Fletcher Yeung]

I did a few tests and at 100mbps (down to 50mbps) it's unusable, as it had black and green drop-outs in the vision. However, at 180mbps (down to 90mbps, it played fine in FCP (Using Long-GOP of course). Good-luck!

[Robin Probyn]

Hi there

I shot a test on HDX900 720p 50p... Nano base rate 25.. crank rate 50p at 50Mbps QT LongGOP. Didint check but presumably 25Mbps then.
Played back fine on a Macbook with Calibrated software on quicktime player.. no green spots etc..

Why would that be?

[Rafael Amador]

Originally Posted by Piotr Wozniacki

Thanks Rafael - this analogy explains it in a most convincing fashion.

I only wonder (pure curiosity), whether at the nominal 180 Mbps (see my OP), it's indeed recorded at the (displayed) 90Mbps (which is not one of the available speeds), or at the nearest one (which would be either 100 or 80 Mbps)...

Piotr,
Forget that we are talking about video. Think that we are just writing a digital file in an HD.
If you set 180 in the NANO, that's the data rate that the NANO is writing in the CFs, BUT:
Your 180 Mbps on recording (720p60) may end up in 4 different data rate files:
- 180 Mbps if you play it at p60
- 90 Mbps if you play it at p30
- 75 Mbps if you play it at p25
- 72 Mbps at p24.
These are the same results that you would get from any p60 at 180 Mbps if you Conform (CinemaTools) to the different time-bases (30, 25, 24).
The NANO just Conform on the Fly the p60/180 Mbps to the selected Overcranking time base (30, 25,24)

Robin,
Your starting data rate would be OK for GOPs, but is too low for any Intraframe encoding.
To get something decent, the minimum you should set would be some 140Mbps. Your final p25 would be 70 Mbps.
That should provide a quality around (possibly lower) than standard HDV.
rafael

[Robin Probyn]

Ok thanks...