BMCC Dynamic Range Test -- INSANE!!!

Well there are two issues to that..

-you must have access to a bunch of lighting if you can control DR fully in the field
-you must have some serious on location monitoring

With out those things IMO every one needs some wiggle room

I consider myself a 'modern' shooter and 'shooting for post' is not the same as 'saving something in post' :)

S
 
-you must have access to a bunch of lighting if you can control DR fully in the field
-you must have some serious on location monitoring
Click to expand...


Point 1. That has always been the case .
Nothing changes there except the easy access to great filmaking gear these days, has introduced some users who have not grasped the concept and look elsewhere for solutions to lighting for DR!

Point 2. We never had that in the film days and coped....but its a mighty pleasure to have this tool now in the digital age.


I consider myself a 'modern' shooter and 'shooting for post' is not the same as 'saving something in post' :)
Click to expand...


I concur. :thumbsup:
 
In our realm of the production business "Raw" is a synonym for unexposed motion picture film or un-used tape. . . "Raw Stock". . . and doesn't refer to exposure: rather that is referred to as "blown out" or "over-exposed".
 
Last edited:
No one I know refers to exposure as "raw"

It is a term that referss to RAW data.

eg.

A camera raw image file contains minimally processed data from the image sensor of either a digital camera, image scanner, or motion picture film scanner. Raw files are named so because they are not yet processed and therefore are not ready to be printed or edited with a bitmap graphics editor or NLE edit system.

...and doesn't refer to exposure: rather that is referred to as "blown out" or "over-exposed".".
Click to expand...

Thats an unusual way to refer to "exposure". I generally call that a problem!
 
rjp85 said:
It boggles my mind why it took so long for video cameras to catch up to the way still cameras give you a raw workflow.
Click to expand...

It boggles my mind how many people over look the significant difference between 'still' read out and 'moving picture readout'...

For your examples below... 1 and 2 Mpx frames we have the following 'simple' calculations... I'll just do the arithmetic for the HD case...

2Mpx = 48,000,000 bits for a 8-bit per R,G,B image.

with 30 such frames per second one has 1.44 Gbits/s bit rate... that is about 150% of a 1 Gbit stream. Most 'consumers' do not have HD-SDI interfaces for either camera or other transfer devices...

We can go on to such things as 12 bits per channel, having about 12 stops worth of 'Linear' values, to yield a 2Gbit wire is required...

How many SD cards have 2Gbit data rate?

Sure with lossless compression one can get compression rates of 50% perhaps less with some fancy techniques. Sure one can diverge from 'lossless' and say that well 1 bit of 'change' at an intensity value of 200 is not going to be noticeable, and so encode the data as 'log' data, or encoded it with 'floating point' numbers... but recall to do so means that for say, a 16 bit number, 4 bits are used for the exponent, and that leaves 13 bits (hidden bit trick to get 13 bits out of 12 bits of mantissa...), for encoding the data. But that is an even faster bit rate if the values are so encoded.

Enter in 'lossy'... how much is dependent on the technology and how much someone want's to pay...

Back in the 'olden' days the Wife and I use to shoot almost exclusively Jpeg because even though 'raw' was an option, the Compact Flash cards that were in any reasonable price, would only allow for perhaps 10 raw images, and 100 Jpegs... simple math... and the 'action' of a Wedding required such a compromise...

Did we shoot brides in blasting white dresses, and the groom in a black tux... sure... enter in Fill flash... an old technique that works... (For the longest time we did use the Hasselblad for the formals so that 1) we could enlarge to 'huge' sizes, and 2) retain values in black tux and blasted white dress, if that was the situation...

So, it is no mystery at all to me 'why' it took technology to advance to the state where within reasonable budget one can get a fairly powerful camera... which requires an understanding of 'reading' the light... something that was required for Film without the benefit of a realtime histogram or the like... since most of the time polaroid test shots were not really an effective option, unless one was in a studio... or an Ansel Adams scene (I don't think he did a test shot of "Moonrise over Hernandez"... I think he saw the light and shot accordingly...


As for the 'shoot to the right (or left depending) of center'... that to me is due to a couple of things... Even with Film film the 'standard' for a proper exposure was different from the 'manufacturer's recommendation', and ASA/ISO rating. The standard tests that determined the ASA/ISO value were typically different from the way most photographers processed their film.

With negative film the maxim was 'shoot for the shadows and develop for the highlights'. In the case of digital, the maxim tends to be more like the maxim for positive film, that is 'shoot for the highlights and let the shadows fall where they will'.

With better digitizers, with higher bit rate transfer and memory devices, one can get these days a 12-14 stop 'still image' that retains as much sensor data as possible. low end pro and consumer devices aren't there yet for 'moving pictures', mostly for the aforementioned bit rates and the high cost of high speed interfaces and acquisition devices.

The Black Magic is getting there, in terms of lower price and higher performance for motion picture capture, but then it is not really directed to compete with the DSLR market, only the 'video' portion of the DSLR market, and the portion of that market that would choose say a high end Canon solely for 'video' purposes. (Given their first offering is with an EF mount...).
 
https://vimeo.com/74387127

Stumbled upon this thread loping for more info on the BMCC, looks like a lot of people really don't get the concept of ETTR. Watch the video and see what the guys that are getting it all "wrong" are talking about.
 
The DR on this camera look so nice. I hate everything else about it. But image is king. And the way it handles light intensity is very nice, just wonderful. Up there with the best. I may pick one up as a b-cam some day, even if it isn't the whole package, just wonderful.
 
I'm in no hurry to get a new camera, but yes, this one has a lovely image... and it's really a pain to use.

Anyway, my *usable* DR tests, showing BMPCC+RAW to be the king in its price range by a nice margin: http://www.similaar.com/foto/pocket-dr/bmpcc.html

dr.jpg
 
Some weird semantic / terminology arguments. "Correct" is what the camera needs. "Recovery" is a mischaracterization, as there is nothing to "recover." It's simply stored data that you can plainly see in post. You may not be able to see it on the lower DR field monitor. Why would "correct" have to be gauged by the DR properties of field monitors?

ETTR has some value as a principle in avoiding noise. It is dangerous, however. Clipping of channels may not be as obvious as clipping when all three channels are over. So a 100% zebra setting can give false sense of security. Blackmagic should address that, as it is their most crucial exposure tool.
 
You must log in or register to reply here.
Back
Top