A month shooting with the BM4K

I received one of the early units here in the UK and have used the BM4K on 3 professional shoots over the last month, twice as A-cam and once as B-cam.

My findings are in-line with most of the information out there, but I just wanted to add the problems I've had with my camera, which I sincerely hope BMD can fix with firmware soon:

- 800 ISO is basically unusable.
- 400 ISO appears to be the native sweet spot, but feels distinctly 'dark'. For a usable comparable exposure I've been treating 400 as actually 320.
- Dynamic range is very disappointing. Feels like 5D or even less. I haven't done chart tests, but the 'feel' is that of 8 stops, possibly even less.
- Highlight falloff is particularly harsh and digital feeling.
- Sun spots are completely unacceptable in a "production camera".
- A couple of times the camera has failed to boot up. Distinct memories of the first RED cameras.
- I believe genuine 2K, 2.5K and 4K ProRes would be far more useful than RAW for most users.
- High speed 1080p also seems like an obvious omission.
- Full control of colour balance would be a simple and welcome addition.
- In extreme low-light the camera falls apart as a usable capturing device.
- Internal battery laughable, external batteries negate portability as an advantage of using BM in the first place.

In favourable conditions, a nice image is attainable from the camera. All of the problems above should be fixable by firmware and would go a long way to making this a proper production camera. In it's current state it feels like a smaller, quicker and obviously far cheaper RED ONE from 2010, which won't do BMD's reputation any favours on set.

This feels like a public beta test of a camera, rather than a finished product.


Eben DP
ebenbolter.com
Represented by Wizzo Features

8 stops dynamic range? Unless your particular camera has a serious defect, you're way wrong on that number. You're making sure to shoot in "film" mode right? 8 stops isn't anywhere close to being correct. It's easily 10, and feels more like 11-12 to me. Some of your other points are fair though.

8 stops dynamic range? Unless your particular camera has a serious defect, you're way wrong on that number. You're making sure to shoot in "film" mode right? 8 stops isn't anywhere close to being correct. It's easily 10, and feels more like 11-12 to me. Some of your other points are fair though.

I don't know one way or another (don't have a BMPC4K), but having spent some time comparing BMCC and RED DRAGON, I have learned that there is definitely a wide range of interpretations when it comes to dynamic range (DR). My explanation is this: there is "measurable DR" (the point at which one can or cannot distinguish chips on a calibrated chart) and "useable DR" (the point at which one gives up dealing with the amount of noise present at or beyond a given chip in the chart).

Davinci Resolve and other post-processing tools can do some fairly amazing things to address noise. If one allows for that, the DR numbers are going to look better than somebody who says "I just want to count what the camera itself gives me". If the camera gives you bad noise in your bottom 3 chips, you're going to downgrade an 11 stop DR rating down to 8.

Unless and until there's a standard for exactly what levels of noise we are willing to tolerate, people are going to continue to disagree about the range of a given camera's actual DR.

This feels like a public beta test of a camera, rather than a finished product.

I feel exactly the same way.

There's Dynamic Range-- which is what the camera can record off of the sensor; and then there is, as mentioned, Latitude, which is what you can actually use. The BMs have a very large dynamic range, it is true, and that can really help you with the highlights, sometimes, but the latitude of the camera is never as high as the Dynamic Range. I would say I have a good solid 10 stops Latitude on the pocket/cinema. While I haven't yet used the 4K, I'd say 9 stops latitude is probably right around where most would peg it.

Dynamic Range is not subjective - it is equal to the best possible signal-to-noise ratio (DR=SNR) and SNR is defined as the difference in decibels between the noise floor and the maximum value the sensor can represent. This system is used in both audio and RF applications.

As an example, the SNR on a PMW-EX3 is rated at 54db on the Y channel, which is 9 stops of dynamic range (for luminance).

The BMCC is claimed to have 13 stops of DR, which translates to 78db SNR.

As a side note, binary is a base-2 number system, which means additional bits result in doublings in the possible representations of data. You could therefore say that if you have a camera with 12 stops of DR, you'd need 12 bits per channel to accurately represent that data; any more would be a waste and any less would result in loss of precision.

It looks more like 11 stops but definately more than 8 stops and more than 5DMkII.
Yes the BMPC has issues that need to be address. I cant believe it need that much power consumption.

Thanks for the replies guys.

Yeah I'm probably being slightly hyperbolic when saying 8 stops, but why shouldn't I be when BMD are rounding up so generously on the other end of the scale.

From a pure usability perspective, I have found that the BMPK4k has an incredibly narrow range of exposure between clipping whites and losing blacks (of course in film mode), and most situations require 'help' more than I've recently been used to. Since my last BM shoot I've shot Alexa twice and I'm currently on a C300 shoot and even with the C300 it's so liberating to be able to react more organically to environments and situations without constantly concentrating on clipping and recovering highlights.

On shoots where I have a full crew and a truck full of lighting there's no way I'd choose the BM4K over an Alexa or film. I bought the camera for the lower budget short films etc that I do between bigger productions, where I have to react more to natural situations and do less lighting.

As I said before I think it's capable of producing gorgeous imagery for the price in certain situations, but in my opinion those situations aren't nearly common enough.

Very interesting post ebenholter, it appears that the 4K cam has significantly narrower DR compared to the BMCC 2.5k, have you tried the BMCC ?

Hi Guys,

I have a same question with Ebenbolter that my BMPC 4K only provides narrow dynamic range in highlights area, I've tested with a grey chart and the results is -5 to +2 which gave only 8 stops of dynamic range. The correct exposed grey area went clipped after 3 stops above the normal expose F stop.

Results as below:
https://drive.google.com/file/d/0B8nSt6 ... sp=sharing

Setting:
ASA 400
ProRes 4K
Dynamic: Film

I start wondering is that a problem with my camera, anyone else has the same siutation?