BMCC6K Facts and Findings

Have there been any good in-depth reviews of this camera already? The usual suspects (cined, cvp) seem to be working on it still.

Looks like Atlas got a camera and is testing their anamorphic lenses on it:

Nothing very technical. Just a new impressions video with the BMCC6k.

Dan, interesting observation that claims the colour is improved over the Pocket cameras. Unfortunately no side-by-side comparisons.

Hello. Are there plans to add 6K 16:9 50fps to the camera?

I shoot a lot at 50fps on the BMPCC6K camera (((( I was upset that the BMCC6K only has 48fps, why couldn’t they add another 2 fps, is it really that difficult.

If the firmware comes out with the addition of 50fps, I will definitely buy this camera.

Does 4K DCI have a crop factor?

It all comes down to the sensor clock speed. So yes it is that difficult. The top speed is usually the top speed. It’s not usually arbitrarily chosen. It’s an agonising compromise.

This is why the higher frame rates tend to be found on more expensive cameras.

JB

Dan, interesting observation that claims the colour is improved over the Pocket cameras. Unfortunately no side-by-side comparisons.

Rick, that was exactly my take away from his impressions.

A mini documentary shoot on the BMCC6K Full Frame by French/Australian filmmaker and photographer Florent Piovesan (Of Two Lands)

Settings: BRAW 8:1, 23.98 & 36 fps.
Lenses used: Sirui Jupiter Macro Primes 35, 50, 75 & 100mm, T2 (with Sigma Adapter)
NISI IRND filters
Graded with FilmConvert

The framerates of the Cinema 6K leave much to be desired, I’ll admit.

The sensor should’ve been capable of 4K 2.39:1 up to 75 fps, and 2.6K 16:9 up to 120 fps at least. I believe BMD’s own framerates, especially the excellent ones found on the Pocket 4K, should be the standardized starting point for choosing sensor clock speed.

Full marks for the OLPF. Its effect on amazing color rendition, especially the vegetation, is self-evident. I hope this change in BMD’s design philosophy is encoded permanently. The pixel pitch of 5.9 µm is also an exceptional attribute – the low-light footage looks promising.

The framerates of the Cinema 6K leave much to be desired, I’ll admit.

The sensor should’ve been capable of 4K 2.39:1 up to 75 fps, and 2.6K 16:9 up to 120 fps at least. I believe BMD’s own framerates, especially the excellent ones found on the Pocket 4K, should be the standardized starting point for choosing sensor clock speed.

Full marks for the OLPF. Its effect on amazing color rendition, especially the vegetation, is self-evident. I hope this change in BMD’s design philosophy is encoded permanently. The pixel pitch of 5.9 µm is also an exceptional attribute – the low-light footage looks promising.

This is the downside of using off-the-shelf sensors. You don't have control over every aspect of the sensor and you basically make it work with what you have. The upside is that it is a much more cost effective solution compared to designing your own sensor and let a sensor manufacturer bake it for your (a la Arri, Red etc).

As much as I love the images coming from this camera, I think I'm going to wait for the next BMD body. I'm just so content with my P6kPro as my travel/low-profile camera and the 12k as the 'big boy' camera. The 12k (OLPF/rawlite) image is just so nice and the P6kPro is so compact!

What is the readout on this camera in terms of rolling shutter? pocket 6k was worse than 4k. What about this one?

What is the readout on this camera in terms of rolling shutter? pocket 6k was worse than 4k. What about this one?

This topic is pinned two subjects above.

viewtopic.php?f=2&t=156200

JB

What is the readout on this camera in terms of rolling shutter? pocket 6k was worse than 4k. What about this one?

This topic is pinned two subjects above.

viewtopic.php?f=2&t=156200

JB

Thank you. So that's 25.08ms at the highest res. Would love to see footage of tests related to this. Will look up if anyone's done that. It's even higher than 6k pro.

What is the readout on this camera in terms of rolling shutter? pocket 6k was worse than 4k. What about this one?

This topic is pinned two subjects above.

viewtopic.php?f=2&t=156200

JB

Thank you. So that's 25.08ms at the highest res. Would love to see footage of tests related to this. Will look up if anyone's done that. It's even higher than 6k pro.

It has been done:

I would honestly consider buying a BMCC6K but the one thing that would hold me back is the rolling shutter. Motion is such an under appreciated aspect of a camera's look and I just don't like the images that come from sensors with ≥16ms shutters. It'll be fine for lockoffs though.

New video from Matteo:

It's even higher than 6k pro.

it’s really the same at the same ASPECT ratio.

When you change the resolution the RS times change.

JB

It has been done:

…this open gate test was shot sperical. If you actually shoot anamorphic that jello is going to be 2x worse.

Good Luck

So the consensus of opinion is that it is unlikely that the 36fps limit of 6K open gate recording, will ever be increased in a future firware update?

I don´t even use super high frame rates, but I do love to flick between 24fps and 48fps for commercial work.

Is 48fps at 6K open gate on a future firmware update just never going to happen?? Or is 36fps the upper limit of the sensor/camera?

It's not gonna increase. If you need/want that speed look elsewhere. Even if it were possible in firmware, I think relying on a theoretically possible but not announced major feature update like that would be a mistake; even for features that are announced you're really rolling the dice if it's not currently shipping. That goes for any mfg really.

… I don´t even use super high frame rates, but I do love to flick between 24fps and 48fps for commercial work…

Ian, although it’s not your ideal solution (shooting open gate 3:2), you do have the option of shooting 48 fps if you record in 6048x3200 (17:9) which has the benefit of a faster shutter. If you know which sections you want at 24 fps, you could select open gate for those shots and then switching to 6048x3200 for takes using 48 fps capture.

Is that a feasible option for you?

If it was important to keep everything open gate, you could run a test on your current camera where you mix some 24 fps with 36 fps footage and observe how that plays back in 1.5x slow motion. Psychologically it may achieve the desired effect on the viewer as your usual 2x slow motion.

Depending upon your use case, there’s also the option of shooting in 6K 2.4:1 6048x2520 at up to 60 fps with even less rolling shutter if that’s a concern. Whichever option you take may depend upon the required resolution and aspect ratio of your deliverables.

Your thoughts?

So the consensus of opinion is that it is unlikely that the 36fps limit of 6K open gate recording, will ever be increased in a future firware update?

Won’t happen.

If you actually shoot anamorphic that jello is going to be 2x worse.

In the horizontal axis only, yes. Just like every other camera.

JB

Looking at the CVP review I don't see any improvement compared to the BMPCC 6K pro. We'll see, there was only one side by side test. (It's actually worse at high ISO)

Looking at the CVP review I don't see any improvement compared to the BMPCC 6K pro. We'll see, there was only one side by side test. (It's actually worse at high ISO)

Depends what you’re looking for doesn’t it? And I mean YOU are looking for.

L mount
135 format full frame
CFe media.

These are things that the existing BMD cameras don’t have. They might matter to others. Everyone has a different context for the fit of this camera.

JB

CVP Hands on Review.

Looking at the CVP review I don't see any improvement compared to the BMPCC 6K pro. We'll see, there was only one side by side test. (It's actually worse at high ISO)

Improvement is kind of an arbitrary word. It is more - different cameras that focus on different things.
There's no one camera that is perfect for every task or job, especially at this price point.

It depends on what you want or value. EF mount vs L mount. Cfast vs CFexpress cards. Internal ND filters (which are not as good as actual exterior filters for color accuracy) vs. OLPF (Which I think is a big deal and my preference). 135 or LF sensor vs. 35mm Sensor - which is very different and a matter of need or preference. And we will have to see, but I imagine the overall look of the Image may be a bit different. Not necessarily better or worse but again different. So that is a few changes honestly. This camera will offer users some new options or features at that same low price point. It may or may not be better for a person's workflow than the previous models - it just depends on what they need. But now you have the choice. And choice is good.

It doesn't mean the 6K Pro stops being great for those that prefer its features. But now people that want or need something different also have an option that may better fill their needs.

I am still interested in this as a Super 16 camera: the rolling shutter in Super 16 is better in this camera than on any BMD camera other than the Ursa 12K in Super 16 crop, and L-mount would allow me to use lots of interesting lenses apart from Super 16 ones, including my tiny but amazing rangefinder lenses in M and LTM mount.

For me improvement means, faster, easier, better dynamic range, better battery performance, better noise performance, better colour, lighter, better screen, better viewfinder, menu, better sensor readout. Not sure if I'm missing something there but so far I don't see any improvements in any of those things.
I'm a big Blackmagic fan, if it does turn out to be better I'll be first in line but so far, I'm not compelled.

I am still interested in this as a Super 16 camera: the rolling shutter in Super 16 is better in this camera than on any BMD camera other than the Ursa 12K in Super 16 crop, and L-mount would allow me to use lots of interesting lenses apart from Super 16 ones, including my tiny but amazing rangefinder lenses in M and LTM mount.

If you go to 5.7K in the BMPCC6K pro you get 16.26ms readout which is basically the same as when you go to 4096 x 2160 on the new camera.

If you go to 5.7K in the BMPCC6K pro you get 16.26ms readout which is basically the same as when you go to 4096 x 2160 on the new camera.

On the new BMCC6K, Blackmagic Design lists Super 16 as 2112 x 1184, and the readout speed is 8.29 ms. Actually the Pocket 4K is a little faster in Super 16 crop.

See https://www.blackmagicdesign.com/produc ... s/W-CIN-20

and

viewtopic.php?f=2&t=156200

If you go to 5.7K in the BMPCC6K pro you get 16.26ms readout which is basically the same as when you go to 4096 x 2160 on the new camera.

On the new BMCC6K, Blackmagic Design lists Super 16 as 2112 x 1184, and the readout speed is 8.29 ms. Actually the Pocket 4K is a little faster in Super 16 crop.

See https://www.blackmagicdesign.com/produc ... s/W-CIN-20

and

viewtopic.php?f=2&t=156200

Weird that the readout speeds aren't getting faster. Though if you just put some gyro stabilisation on it it almost gets rid of rolling shutter so it's not so much of a big deal anymore.

… I don´t even use super high frame rates, but I do love to flick between 24fps and 48fps for commercial work…

Ian, although it’s not your ideal solution (shooting open gate 3:2), you do have the option of shooting 48 fps if you record in 6048x3200 (17:9) which has the benefit of a faster shutter. If you know which sections you want at 24 fps, you could select open gate for those shots and then switching to 6048x3200 for takes using 48 fps capture.

Is that a feasible option for you?

If it was important to keep everything open gate, you could run a test on your current camera where you mix some 24 fps with 36 fps footage and observe how that plays back in 1.5x slow motion. Psychologically it may achieve the desired effect on the viewer as your usual 2x slow motion.

Depending upon your use case, there’s also the option of shooting in 6K 2.4:1 6048x2520 at up to 60 fps with even less rolling shutter if that’s a concern. Whichever option you take may depend upon the required resolution and aspect ratio of your deliverables.

Your thoughts?

Hi Rick sorry for the late reply and thank you for the suggestions! I hadn´t thought of just using the 36fps instead of the 48fps. Good idea that viewers might not really notice the difference between the two slower speeds!

I did a 36fps test this evening (im based in Spain) like you said, but unfortunately it´s not quite slow enough when slowed to 24fps for what I need on commercials, etc.

Rolling shutter isn´t a problem for me, as I don´t use fast pans, but using the other aspect ratios is an option I guess...I´m gonna have to think about this one!

It's not gonna increase. If you need/want that speed look elsewhere. Even if it were possible in firmware, I think relying on a theoretically possible but not announced major feature update like that would be a mistake; even for features that are announced you're really rolling the dice if it's not currently shipping. That goes for any mfg really.

Good advice!! The camera hasn´t even been released yet, so e thinking about updates already is probably a sign the camera perhaps isn´t for me.

CVP did mention IQ improvement especially the colour being closer to the UM12k though did not do a proper side by side comparison with the Pocket 6k in that regard. What suprise me was they think the 1080p crop of the Pocket 6k LF is cleaner and much higher quality, if that is so I be tempted to uprezzed that to 4k in post.

Interesting to know using gyro metadata can reduce rolling shutter but still at the expense of cropping. Rather BMD focus on creating faster readout without this compromise.

So the consensus of opinion is that it is unlikely that the 36fps limit of 6K open gate recording, will ever be increased in a future firware update?

I don´t even use super high frame rates, but I do love to flick between 24fps and 48fps for commercial work.

Is 48fps at 6K open gate on a future firmware update just never going to happen?? Or is 36fps the upper limit of the sensor/camera?

Just out of interest, as I´m not a super technical person when it comes to how the cameras actually work, does anybody know what the limiting factor is of not getting higher than 36fps??

Is it the electronics of the camera can´t process the images quickly enough?
The ´CPU´of the camera?
The new FF sensor readout speed??
Data transfer to the recording media?
Or something else!

What´s the likely bottleneck here on the 6K full frame in open gate, that doesn´t allow an extra 12fps to take it up to 48fps?

I don´t get it! Surely an extra 12fps isn´t that much more work.

Or would there be a trade off of worse rolling shutter, or less dynamic range or something??

Sony's approach is down res from higher resolution with a little bit of cropping as oppose to just cropping to get higher fps something I prefer over BMD's method.

I don´t get it! Surely an extra 12fps isn´t that much more work.

I see people saying this all over the web and I'm like... Do you think that there's just some button the BMD engineers forgot to press labelled "Make the camera faster"?

Sensors have limits in how quickly they can read out, and while I haven't read the data sheet on the one in the CC6K I get the impression that it is operating at its limit. It feels like this is functionally the same sensor design as the one in the Panasonic S-Series mirrorless bodies and those hit a wall at roughly the same point. It is what it is. Sometimes sensors can read out faster if they do so at a lower precision (e.g., 12-bit vs 10-bit), but if that sensor has a lower precision mode I doubt BMD wants to engage it considering the dip in image quality.

I am still interested in this as a Super 16 camera: the rolling shutter in Super 16 is better in this camera than on any BMD camera other than the Ursa 12K in Super 16 crop, and L-mount would allow me to use lots of interesting lenses apart from Super 16 ones, including my tiny but amazing rangefinder lenses in M and LTM mount.

Like other folks have mentioned, the original BMPCC4K has a faster readout in its Super16 crop than the BMCC6K. If you're mostly interested in the ideal Super16 camera in the BMD lineup, I'd still recommend going that direction. You get a higher resolution window and also a higher ceiling for slow motion for less money, although you obviously lose out on a few other very nice features that the BMCC6K has.

BMD has extolled the virtues of shooting in open gate with more vertical resolution to frame your deliverables and that’s great but it comes at the cost of a higher sensor readout and therefore potentially rolling shutter issues. Gyro provides some compensation for the rolling shutter, but in very many use cases, the rolling shutter at open gate is not going to be an issue.

Ian has pointed out that the concomitant issue of a slower maximum frame rate of 36 fps is a serious issue for some use cases where 48 fps is needed. That leads me to wonder if a purchase decision is all about the tradeoff of the aspect ratio and frame rate maximum:

6K 6048x4032 3:2 aspect ratio, 36 fps
6K 6048x3200 17:9 aspect ratio, 48 fps
4K 4096x3072 4:3 aspect ratio, 50 fps
6K 6048x2520 2.4:1 aspect ratio, 60 fps
4K 4096x2160 17:9 aspect ratio 60 fps
Super 16 100 fps
HD 120 fps

I assume most of us are delivering 16:9 or 17:9 or 2.39:1 aspect ratios with 9:16 gaining in popularity for social media. For those who are looking for more vertical real estate in frame, both 3:2 and 4:3 may be preferred capture ratios. The 4:3 ratio shouldn’t be overlooked as it offers capture up to 50 fps.

I’m interested in how Ian ultimately responds to the trade off as I think frame rates will prove to be more important than rolling shutter in selecting a camera with a minimum 4K resolution desired.

BMD has extolled the virtues of shooting in open gate with more vertical resolution to frame your deliverables and that’s great but it comes at the cost of a higher sensor readout and therefore potentially rolling shutter issues. Gyro provides some compensation for the rolling shutter, but in very many use cases, the rolling shutter at open gate is not going to be an issue.

Ian has pointed out that the concomitant issue of a slower maximum frame rate of 36 fps is a serious issue for some use cases where 48 fps is needed. That leads me to wonder if a purchase decision is all about the tradeoff of the aspect ratio and frame rate maximum:

6K 6048x4032 3:2 aspect ratio, 36 fps
6K 6048x3200 17:9 aspect ratio, 48 fps
4K 4096x3072 4:3 aspect ratio, 50 fps
6K 6048x2520 2.4:1 aspect ratio, 60 fps
4K 4096x2160 17:9 aspect ratio 60 fps
Super 16 100 fps
HD 120 fps

I assume most of us are delivering 16:9 or 17:9 or 2.39:1 aspect ratios with 9:16 gaining in popularity for social media. For those who are looking for more vertical real estate in frame, both 3:2 and 4:3 may be preferred capture ratios. The 4:3 ratio shouldn’t be overlooked as it offers capture up to 50 fps.

I’m interested in how Ian ultimately responds to the trade off as I think frame rates will prove to be more important than rolling shutter in selecting a camera with a minimum 4K resolution desired.

For me I'd shoot 3:2 for a lot more cinematic content now since it's much closer to 1.44:1 IMAX ratio. And, I also really dig the 3:2 aspect ratio. But the trade off in frame rates is only for when you do need slow motion. Most narrative doesn't necessitate high frame rates as 24, 25, and 30 are the standard rates.

But otherwise use a crop of 1.89:1 or 2.4:1 for cinematic aspect ratios as well and get 48-60 FPS. The advantage of using 3:2 open gate then becomes to help with extra room for cropping later to one of those two standard cinema aspect ratios. Maybe then with crop guides a sound person can see where there boom is before it enters frame. Again, look back to how Hitchcock used the 3:2 frame.

I'm all for higher frame rates in an URSA version of the VistaVision sensor. That camera should have better cooling and processing speeds and power. But that also means the sensor needs to be faster so it is likely a different sensor. At that level may as well get the other body improvements of an URSA and the production friendly design for set life.

I read someone writing that they used "a cheap full frame sensor on it (the full frame sensor they used is found in $1100 mirrorless cameras from Nikon)"

How true is this? And what implication does that have if at all? I don't fully know the truth of this.

Just out of interest, as I´m not a super technical person when it comes to how the cameras actually work, does anybody know what the limiting factor is of not getting higher than 36fps??

One of the first choices you make designing a sensor is the clock speed.

Like the Mhz rating of a processor in a computer. It defines the speed at which the sensor can operate.

The faster the MHz, the faster the RS and the higher the maximum frame rate. Other things come into play too, like memory, speed of memory, number of lanes (or pipes) for the data to leave the sensor and processing and memory in the camera itself.

Basically, lower cost sensor = lower performance spec in terms of RS and maximum frame rates.

JB

Question: Would the BMCC6k's (LF or 135 sensor) offer the same advantages over the 6k Pro sensor as far as Highlight Rolloff etc. that Roger Deakins talks about as an advantage of the Arri Alexa LF vs. the standard Arri Alexa in this Arri Interview? Is that truly a characteristic difference between the two different sensor sizes or is this more specific to Arri cameras?

It's not inherent to the sensor, but depends on the way the camera deals with its information.
Generally, a larger sensor should have more dynamic range, given everything else is the same. But if you use that for a nice highlight rolloff or for less noise in the shadows is up to the manufacturer. It's closely connected to the ISO they'd indicate as 'native'.

There’s this old myth that the larger the sensor the better the DR noise etc.

It seems to be tied into equivalence theory that a lot of FF photographic evangelists love to espouse.

http://www.josephjamesphotography.com/equivalence/

It’s just not true in the real world because there’s a gazillion other things that come into play that affect DR more than simple size.

If the logic was true then my Leica S should be better than my IPhones DR but it isn’t. One is a medium format larger than 135 format FF sensor and the iPhone is tiny and yet that’s just not how it works.

JB

Fully agree. That’s why I wrote something like ‚everything else being the same‘. From the side of physics, a larger array of photocells will collect more photons. But apart from technical progress, there's a lot happening in the camera.

An iPhone is only foreshadowing what computational photography will be…

If anybody would like to read more about what developing a top-notch movie camera really means, here you go:
https://www.fdtimes.com/issues/june-202 ... 35-report/
Those interviews are quite revealing.

P.S. Right at the beginning, BM is listed among "TITANS OF THE INDUSTRY".

Question: Would the BMCC6k's (LF or 135 sensor) offer the same advantages over the 6k Pro sensor as far as Highlight Rolloff etc. that Roger Deakins talks about as an advantage of the Arri Alexa LF vs. the standard Arri Alexa in this Arri Interview? Is that truly a characteristic difference between the two different sensor sizes or is this more specific to Arri cameras?

The Pocket 6k LF sensor I would expect to have more photosites to that gathers lights and more details to work with including widers tone and produuing more accurate colours compared to the none LF Pocket 6k S35 sensor cameras.

Question: Would the BMCC6k's (LF or 135 sensor) offer the same advantages over the 6k Pro sensor as far as Highlight Rolloff etc. that Roger Deakins talks about as an advantage of the Arri Alexa LF vs. the standard Arri Alexa in this Arri Interview? Is that truly a characteristic difference between the two different sensor sizes or is this more specific to Arri cameras?

Kind of... a little bit.

In terms of lenses and rendering geometry—and be aware that this is a gross oversimplification—it's simpler for lens manufacturers to achieve straight lines when designing for larger sensors. There are also tradeoffs when you're covering a much larger image circle, and it's not like it's impossible to have good geometry on smaller image circles, but these things are fairly reliable rules of thumb.

In terms of "highlight rolloff" and such, things are more murky. The nice thing about moving from the Alexa to the LF was that it was basically the same sensor tech and the same photosite design, but just more of it. Think of it like turning two Alexa sensors 90 degrees and then smashing them together and you're roughly there. Considering how good the ALEV III sensor already was, a more modern version of it with twice the number of pixels is generally going to give you a better image.

For the BMCC6K it's not so straight forward. It's a completely different sensor design so it's less of an apples to apples comparison with the BMPCC6K/4K. Is it going to have better highlight rolloff? Probably, but we don't really have a great handle right now on whether or not it'll be transformative enough to notice outside of a side-by-side comparison. All we know is that BMD rates it as having slightly better dynamic range than either the Pocket 6K/4K.

@Simon Chen
While most folks use 'pixels' instead of 'photocells' (or 'sensels'), the numbers are the same in this respect. This does not translate into resolution, though.
So, two cameras with 6K have the same number of photocells if both dimensions are the same.
Of course, the new 6K, being 3:2, has more photocells, but they are used for different parts of the image.

This does not apply to some smartphones, drones or action cameras, where a large number of photocells is averaged together, usually in groups of four, and stored as a smaller number of pixels. This can be changed, of course. That's how Apple changed the number of pixels stored between 14 and 15, even if some models got the same sensor.

I’m not aware of any forensic confirmation of the sensor and electronic details used in the BMCC6K. It may be identical to other off-the-shelf sensors or it may have some design characteristics that are specific to BMD that applied to an otherwise off-the-shelf sensor. BMD has not claimed they designed the sensor like they did for the UMP12K, but it still could be tweaked by BMD. The sensor is only part of the story though as the dynamic range is going to also be affected by design of the image processing that occurs based on the light collected by the sensor photocells.

The surface area of the sensor photocell is fairly large at 5.94 microns which implies a larger electron well that could mean less noise and increased dynamic range, but this is speculation only. The key is the image processing and BMD has not shed any light on that and has only modest claims with a tiny increase of dynamic range. How that is applied to the latitude and roll off will gradually become known as the camera images are compared to other cameras.

AFAIK CineD has not yet posted their image analysis which is a start in determining real world differences; although rather limited in terms of lighting, it does compare images in the same way across many cameras. Real world use is still the best guide.

You can quite confidently assume that they use the Sony IMX410 sensor, just as other camera manufacturers do (Panasonic, Nikon for example). Maybe with some customization, but it isn't going to be anything major taking the price of the camera into account. Sony Imaging (or any other sensor manufacturer) isn't going to bake a completely new sensor for the price this camera is selling. We have to be realistic about that.

You’re probably right, Ruben. Thanks for the comment.