Honey, I shrunk the camera!

[George Tasick]

I'm sure I'll get a lot of flack for asking such questions, but assuming the new Blackmagic Micro cameras use the same sensors as their larger counterparts. Why would anyone need something more than these palm sized cameras (kitted out, of course)?

To make things even more inflammatory; why spend $30K on a Red or Alexa, when the current Blackmagic cameras can give you 95% of the quality for 10% of the cost?

Sure, you can't fairly compare $1000 camera with a $30K camera today ... but what about in 5 years?

With the current pace of technology, how fast do you think that gap will close? Bear in mind, 15 years ago we where still recording 4:3 SD video to magnetic media.

[Brian Drysdale]

Dynamic range is one answer, those other cameras are offering 14 stops. If you're shooting high end productions, you want robust, reliable camera systems and BlackMagic doesn't have that reputation with their cameras.

The Micro uses a Super 16 sized sensor, so you don't have the shallower depth of field that is currently desired. The smaller size commonly results in more noise.

That's not to say these productions won't use these BlackMagic cameras, they'll use them as required for specialised shots.

In 5 years the cutting edge will have moved on and that edge is always expensive, especially for those last few percentage points. .

[Warren Kawamoto]

The difference between a cheap camera an an expensive one of equal resolution is video noise and grain. That is my understanding.

[George Tasick]

Originally Posted by Brian Drysdale

In 5 years the cutting edge will have moved on and that edge is always expensive, especially for those last few percentage points.

True. However, the human eye can only see 20 stops of dynamic range; the Red Dragon sensor is advertising 16.5 stops for $30,000 and the Blackmagic sensors offer 13 stops for $1,000.

What happens when the high end hits 20 stops?

How long do you think it will take for the low end to catch up?

When technology meets our physical limitation, what separates low end from high end?

[George Tasick]

I agree that dynamic range is the biggest factor separating image sensors, and not only that, but how the sensors handle the transition to the extremes of their range. The Alexa handles hit's highlights very well (as well as film) while the Blackmagic sensors aren't as elegant in this area.

But I have to ask myself; what can I do with 14 stops of range that I can't do with 13? Am I constantly running my cameras up against the edge of their rage? Or will I use lighting to keep the things more toward the middle?

Ultimately, I suppose what this boils down to is, can a good DP get $30,000 worth of image out of a $1,000 camera? And, is the audience's experience lessened by the technical differences between the two price points?

[Mike Watson]

Originally Posted by George Tasick

When technology meets our physical limitation, what separates low end from high end?

Let's say that you're right (and you're not, but let's say you are) and that in 10 more years we have some magic camera that can see more dynamic range and more colors at greater clarity than the human eye, and that it costs $1k.

Where will camera stabilization be? Will it need stabilization? This would differentiate between cameras.

How's the audio? Once we are able to have this camera that sees like the eye, will it record audio like the ear? Will peaking audio still be a problem? Over driving? Options here will differentiate between cameras.

What's the form factor? Does it sit on your shoulder? Do you hold it in your hands? Can you clip it to your helmet? This will differentiate.

Long lens? Short lens? Fixed lens? Shallow DOF? Narrow DOF? Or will this camera do all of this automatically?

[George Tasick]

I'm not arguing a point, I just want to have a discussion to see what other people think.

Almost every major brand is getting modular with their cameras. Offering a "brain" that houses the sensor, a lens mount and various connections to build out the functionality of the camera.

As far as things like audio, stabilization, lenses and form factor; that's customize-able for whatever your situation demands. Even a $30,000 Red brain needs another $15,000 worth of accessories to have practical functionality. Besides the huge difference in cost, Blackmagic cameras are no different.

Once manufactures can produce a sensor that matches the limits of human sensory input, I can't see a reason to develop beyond that.

Personally, I'm really looking forward to that day. We can all focus on being artists instead of technicians.

[Brian Drysdale]

You'll also need a display system that offers 20 stops, which they currently don't.

Currently the marketing is on resolution rather than dynamic range, so that's what the newer high end cameras are going for. That usually means oversampling, so that you have a true resolution of 4k or 8k.For feature films this being the cinema 4k, rather than UHD used on most cameras. Plus a larger colour space.

There are manufacturing aspects which adds to the costs, like quality checks, burn in time, more robust buttons and controls, specialised firmware.

It can be similar to buying power tools, there are the cheaper ones which are OK for lighter domestic use and the industrial ones which can withstand years of heavy duty work.

In the current BlackMagic range, the URSA Mini with their new viewfinder seems to be the one.

[David Heath]

Originally Posted by George Tasick

True. However, the human eye can only see 20 stops of dynamic range; the Red Dragon sensor is advertising 16.5 stops for $30,000 and the Blackmagic sensors offer 13 stops for $1,000.

What happens when the high end hits 20 stops?

It's one thing to measure dynamic range of a camera - quite another for the human eye.

It's reckoned that in the static sense the human eye can only manage about 7 stops - which is why for straightforward display 8 bit is more than good enough. But the reality is far more complicated than that.

Looking at a high contrast scene the eye "targets" on only a small portion of the scene at a time - and for each portion the dynamic range is about 7 stops. But if you look back and forth between such as a brightly lit area and the shadows, the eye adapts via a variety of means to give an IMPRESSION of much higher dynamic range. Crucial to this is what is known as saccadic suppression or masking - the brain "suppressing" the image between targets - and also suppressing the adaptation.

Nett result is that the eye/brain combination SEEMS to have a much higher than 7 stop DR at any one time.

I think that the ABSOLUTE dynamic range of the eye is something like 45 stops. But that's the difference between the absolute minimum that is perceivable from complete blackness, and the absolute maximum it can possibly handle. That's not saying it could handle any given scene with such a range (it would take a long time to adapt for one thing - think of going into a very dark room from sunlight).

But a camera is different. It "sees" everything at the same time, and if any camera system really did have a dynamic range of 14 stops from input to output - such that an object 2^14 brighter in the scene was 2^14 brighter in the image - it would look far too contrasty to the eye. The eye would be taking the whole image in at one time.

Consequently, if a camera system has an INPUT dynamic range of (say) 14 stops, it has to present such to the viewer in a far more limited range - one the eye can handle in one go. So typically about 7 stops. To do this, it uses the principle that if the scene brightness doubles, then whilst the presented brightness does increase - it's not by as much as a doubling. And (typically) the brighter/darker the parts of scene, the smaller the change will be relative to the middle. This is the whole subject of gamma curves, "knee" etc etc.

In a camera giving a viewable output, this dynamic range compression (totally different from data compression!!) handles in a fixed way according to camera lineup. What RAW does is capture more or less the full sensor range and let it be decided later how the 14 to 7-8 stop transformation is to be done.

[Bruce Watson]

Originally Posted by George Tasick

But I have to ask myself; what can I do with 14 stops of range that I can't do with 13? Am I constantly running my cameras up against the edge of their rage? Or will I use lighting to keep the things more toward the middle?

What you can do with the extra stop of dynamic range is cut your lighting in half. Say, if you are lighting an interior to match the sunlight streaming through a window. If you need 8k of HMIs with your 13 stops of dynamic range camera, you can use 4k of HMI with your 14 stops of dynamic range camera.

Saving on lighting is huge. HMIs are really expensive, and you can cut your inventory in half. That includes half your generator capacity, half the cables, half the storage space, half the truck space, etc. Failing that you can cut your rental bills in half-ish. Not that the lighting rental places will appreciate the drop in business.

And yes, you could just gel the windows. If you have time, and if you have the capability. Some scenes take place in front of window walls, and sometimes you don't get a previous day to setup. And to make gel look good (that is, to be completely unnoticed) takes an interesting amount of skill that many people just can't develop. I'm one of those. Sigh...

Since the scene I describe is a fairly common one, I can see why producers and cinematographers are willing to pay way more for a camera that delivers more of what they need. And why they continue to beat on the camera makers (just like they used to beat on the film makers) for more speed, more dynamic range, more sharpness, better tonality, etc., etc., etc.

[Brian Drysdale]

I suspect lighting rental companies won't be going away for some time. On large productions you want control of the light over the day and unfortunately the big light in the sky moves position.

The extra dynamic range helps to give roll off in the highlights and reduce/prevent clipping.