BMCC6K Facts and Findings

I wanted to start a fresh page dedicated to the new BMCC 6k. For those of us who want to compare notes and discuss this camera in detail.

A few questions to get the ball rolling.

1. For those who have had the opportunity to shoot with the camera. How does the color, IQ, etc. of this new camera compare with the 12k and other Pocket cameras that came before it.

2. How will the new OLPF will affect the use of NDs. Do we no longer need an IRND? Will they cause problems now. Will they still be useful. Or will a plain ND be better suited to the BMCC6k?

2. How will the new OLPF will affect the use of NDs. Do we no longer need an IRND? Will they cause problems now. Will they still be useful. Or will a plain ND be better suited to the BMCC6k?

Just to address this: As someone who works with Venice2s, Alexas, Raptors, etc., you should just be using IRNDs all the time if you can. It’s a reality of digital cinema that you need proper IR filtration. Obviously those bodies all have IR rejection in their internal NDs so you can kinda cheat a little on lower strength external filters, but I still feel like you should have IRNDs from 0.9 upward.

In an effort to consolidate some of the info all ready out there, I will drop a few links here in case someone has not seen them yet.

First Impressions from Florent.

Test Footage from Florent.

First Impressions from Matteo.

Test Footage from Matteo

I also found the breakdowns that JB, Florent and Matteo gave of their test footage on the gallery page interesting.
https://www.blackmagicdesign.com/produc ... ra/gallery

2. How will the new OLPF will affect the use of NDs. Do we no longer need an IRND? Will they cause problems now. Will they still be useful. Or will a plain ND be better suited to the BMCC6k?

Just to address this: As someone who works with Venice2s, Alexas, Raptors, etc., you should just be using IRNDs all the time if you can. It’s a reality of digital cinema that you need proper IR filtration. Obviously those bodies all have IR rejection in their internal NDs so you can kinda cheat a little on lower strength external filters, but I still feel like you should have IRNDs from 0.9 upward.

Alex thanks for sharing your thoughts. I appreciate it.

2. How will the new OLPF will affect the use of NDs. Do we no longer need an IRND? Will they cause problems now. Will they still be useful. Or will a plain ND be better suited to the BMCC6k?

Just to address this: As someone who works with Venice2s, Alexas, Raptors, etc., you should just be using IRNDs all the time if you can. It’s a reality of digital cinema that you need proper IR filtration. Obviously those bodies all have IR rejection in their internal NDs so you can kinda cheat a little on lower strength external filters, but I still feel like you should have IRNDs from 0.9 upward.

Agreed. For any digital cinema camera, any ND from 1.5 up (even 1.2) can lead to IR chaos in the image. It’s always going to be better to cut IR light and visible light as equally as possible so you don’t get the nasty red pollution that you can’t fix later.

Unless BMD suggests otherwise, keep cutting IR with quality filters.

At IBC2023, a BMD rep gave some of the rationale for design decisions BMD has made over the years. Often they were attempting to make their new cameras attractive by reducing the buyer’s overall costs. An example was the EF mount on the BMPCC6K since it was assumed most people had several EF lenses they could use without additional cost. So this conversation was around the body style selected for the BMCC6K. They made it very similar (but not identical) to the BMPCC6K Pro so that most accessories (battery clip, viewfinder) including most camera cages wouldn’t need to be purchased. Even though the BMPCC4K probably has sold more copies and those cages likely won’t fit.

Personally I think that logic is myopic for a camera with several outstanding features and that releasing an URSA Micro instead of a clone of the BMPCC6K would have not been seen as a negative aspect. Don’t call it URSA (the big bear) but give the new camera a new family name that designates a new line of business that shares those new features: L-Mount, CFExpress, 36x24 sensor, OLPF, 2x anamorphic lenses, etc. As Matthew Allard stated, BMD must stop this years long cycle of camera refinement (and go boldly into the future).

Back in the film days, the ASC listed recommended panning speeds per focal length to avoid a kind of staccato, stuttering image during camera movement. That was different than what we're dealing with here—rolling shutter—but at 25ms for a full height readout I am curious whether or not those speeds would still induce this kind of wobbly jello effect.

I get that this is a specific kind shot and not everyone is going to be doing "That 70's Show" style circular whip pans but goddamn; this would bug the hell out of me.

Alex, that video was what I didn’t like about the BMCC6K but there have been recent posts here that using the gyro stabilization with a very small strength like 0.008 apparently greatly removed the rolling shutter. But you might want to try it with the actual BMCC6K camera.

Alex, that video was what I didn’t like about the BMCC6K but there have been recent posts here that using the gyro stabilization with a very small strength like 0.008 apparently greatly removed the rolling shutter. But you might want to try it with the actual BMCC6K camera.

Controversial opinion: I should not have to add post stabilization to a tripod mounted camera for a medium speed pan.

On dynamic range:
"What we're seeing in the detail and noise levels with this camera is a very clear and very concise 13 stops". (3:55min)

On OLPF (moire & IR):
"...has a specifically designed OLPF for the sensor. What you see there is a reduction in moire and a quite aggressive IR cut as well." (4:25min)

BMCC6k looks like something BMD built before the Pocket 6kPro. Now it looks like one step forward and two steps back.

I currently have several Panasonic L-Mount lenses, along with a Sigma 150-600mm sports.

My question is: Will the new BMCC 6K support the built in lens profiles for the removal of CA, distortion etc?

I'm presently using the S5 & S1 with the Video Assist 12g 5 inch and it appears that the lens profiles are being applied before RAW output as I've never experienced any issues.

I was just in the process of creating a pre-order for the new camera when a shiver went down the spine. The thought of dealing with CA etc in post production is a definite obstruction.

Any help much appreciated.

https://youtu.be/B0ExwwY4pg0

On dynamic range:
"What we're seeing in the detail and noise levels with this camera is a very clear and very concise 13 stops". (3:55min)

On OLPF (moire & IR):
"...has a specifically designed OLPF for the sensor. What you see there is a reduction in moire and a quite aggressive IR cut as well." (4:25min)

Thanks for sharing this one. I had not seen it yet. It will be interesting to see how the aggressive IR cut affects shooting with NDs going forward.

IQ?

It’s self evident I think. I’m sure more tests will come but I think it looks great and it does things no other BMD camera can do.

The OLPF does have a more aggressive IR cut.

Ever since the fist BMD camera the recipe for the sensor cover glass was exactly the same.

Only with the 12k when they introduced the new OLPF did they change it.

It now has the same IR cut as the 12k OLPF and I presume it will be the same for all cameras.

It doesn’t mean you don’t need IR ND filtration. And I suspect we will see more issues as people use less expensive NDs.

JB

Paul, I don’t think the camera supports specific lens profiles but there are manual corrections you can apply within DaVinci Resolve Studio / Colour including Resolve FX Revival to address chromatic aberration and distortion. The Edit page also has a Lens Correction to analyze and correct distortion. Not as convenient as Panasonic’s in-camera approach but if you proceed with the purchase (or rental) of the BMCC6K, there are options that may work. An automated lens correction profile according to the lens identified in metadata would be a possible future enhancement to Resolve.

I’d suggest you ask a version of this question in the Resolve forum as you may already find someone that have used these options in Resolve Studio 18.6.

IQ?

It’s self evident I think. I’m sure more tests will come but I think it looks great and it does things no other BMD camera can do.

The OLPF does have a more aggressive IR cut.

Ever since the fist BMD camera the recipe for the sensor cover glass was exactly the same.

Only with the 12k when they introduced the new OLPF did they change it.

It now has the same IR cut as the 12k OLPF and I presume it will be the same for all cameras.

It doesn’t mean you don’t need IR ND filtration. And I suspect we will see more issues as people use less expensive NDs.

JB

JB - Thanks for your input, I appreciate it. I really liked the test footage as well. And that is helpful to know on the ND filters.

I had a question for you.

At the Palmdale Diner some shots were in anamorphic and listed as Open Gate Anamorphic. At the risk of sounding like a dolt. Were those shot in the 6/5 anamorphic setting or with an Anamorphic lens on the 3/2 Open Gate setting. I am not sure if that is even possible. And if so what would be the advantages or differences for either setting when shooting Anamorphic. As always thanks for your time and input.

I had a question for you.

At the Palmdale Diner some shots were in anamorphic and listed as Open Gate Anamorphic. At the risk of sounding like a dolt. Were those shot in the 6/5 anamorphic setting or with an Anamorphic lens on the 3/2 Open Gate setting. I am not sure if that is even possible. And if so what would be the advantages or differences for either setting when shooting Anamorphic. As always thanks for your time and input.

In the publicly available footage we had two anamorphic lenses.

Orions FF / 135 / Vista covering Mercury lenses have a 1.5x squeeze. I selected open h gate which is another way of saying the whole sensor.

We also shot a few shots with the Orion Atlas anamophic which are a different anamophic lens from the same company. This one is designed for S35 / academy open gate. In other words the original 35mm cinema standard which is a 2x squeeze.

In both cases the advantage is having a full height or tall sensor.

For many cameras they have a S35 sensor but that doesn’t always mean it is a full height sensor. You need the height if you’re wanting to shoot anamophic otherwise you’re effectively cropping the lens and with anamophic all the interesting optical things tend to happen on the sides.

I see a lot of people posting RS times compared to other FF / 135 / Vista cameras too but a lot of those times are also for a 17:9 cut of the sensor, therefore we call the whole sensor open gate. Not a 17:9 DCI cut of an open gate sensor for example.

In that scenario the open gate BMD will render a different field of view on an anamorphic lens.

JB

I currently have several Panasonic L-Mount lenses, along with a Sigma 150-600mm sports.

My question is: Will the new BMCC 6K support the built in lens profiles for the removal of CA, distortion etc?

I'm presently using the S5 & S1 with the Video Assist 12g 5 inch and it appears that the lens profiles are being applied before RAW output as I've never experienced any issues.

I was just in the process of creating a pre-order for the new camera when a shiver went down the spine. The thought of dealing with CA etc in post production is a definite obstruction.

Any help much appreciated.

They don’t.

In fact there is no way to do this with built in raw shooting on any camera platform that I know of. Anytime these corrections are done it’s “baked” into the compressed or encoded image and therefore can’t be undone either.

And I doubt they are applying it on the HDMI output if you’re receding to BRAW either.

So if you’re happy with what you get now, it’s most likely uncorrected. I could be wrong but that’s my best guess.

JB

Back in the film days, the ASC listed recommended panning speeds per focal length to avoid a kind of staccato, stuttering image during camera movement.

That hasn’t gone away.

I had an Alexa mini Steadicam shot that got bounced for stuttering on a fast pan. HDR mastering has made it worse. When you have bright and dark alternating tones on a fast pan it becomes way more obvious.

Impossible to fix in post too.

JB

The new CFexpress 4.0 standard means this media is pursuing some very fast recording bitrates:

https://www.newsshooter.com/2023/09/17/ ... -standard/

I had a question for you.

At the Palmdale Diner some shots were in anamorphic and listed as Open Gate Anamorphic. At the risk of sounding like a dolt. Were those shot in the 6/5 anamorphic setting or with an Anamorphic lens on the 3/2 Open Gate setting. I am not sure if that is even possible. And if so what would be the advantages or differences for either setting when shooting Anamorphic. As always thanks for your time and input.

In the publicly available footage we had two anamorphic lenses.

Orions FF / 135 / Vista covering Mercury lenses have a 1.5x squeeze. I selected open h gate which is another way of saying the whole sensor.

We also shot a few shots with the Orion Atlas anamophic which are a different anamophic lens from the same company. This one is designed for S35 / academy open gate. In other words the original 35mm cinema standard which is a 2x squeeze.

In both cases the advantage is having a full height or tall sensor.

For many cameras they have a S35 sensor but that doesn’t always mean it is a full height sensor. You need the height if you’re wanting to shoot anamophic otherwise you’re effectively cropping the lens and with anamophic all the interesting optical things tend to happen on the sides.

I see a lot of people posting RS times compared to other FF / 135 / Vista cameras too but a lot of those times are also for a 17:9 cut of the sensor, therefore we call the whole sensor open gate. Not a 17:9 DCI cut of an open gate sensor for example.

In that scenario the open gate BMD will render a different field of view on an anamorphic lens.

JB

This is great! Thanks so much for the explanation. I guess I am even more excited now that I understand it a bit more.

The new CFexpress 4.0 standard means this media is pursuing some very fast recording bitrates:

https://www.newsshooter.com/2023/09/17/ ... -standard/

I saw that Rick. Seems like good news on that front.

Those shots are fantastic, really impressive.

The new CFexpress 4.0 standard means this media is pursuing some very fast recording bitrates:

https://www.newsshooter.com/2023/09/17/ ... -standard/

While CFE Type B v4.0 is rad, the nice thing for BMCC6K users is that this camera does not seem to push media especially hard. The settings that stress bandwidth the most would be:

6K DCI
48 FPS
Q0

Which comes out to 704MB/s if the encoder is pushed to its limit. As long as BMD actually approves it, Angelbird's lowest cost 512GB mag should be able to cover that, no problem. $130USD for a mag this size and this speed is rad.

That’s way more bitrate capacity than I would require. Still I think going to CFexpress is a good move that gives the shooter more flexibility in the future. For narrative work destined for television, shooting open gate at 30 fps Q0 for complex scenes with a bitrate perhaps over 450 MB/s (really not advisable on a CFast2 card), a 2TB card recording for 80 minutes would be less than a day’s shoot. At Q3, recording would be 2.5 hours per 2TB card. I think the move to CFexpress is the right thing to do for a 6K 3:2 camera as it will not restrict the quality of your captured media. But such high quality means lots of time pressure if shooting in a situation where dailies are reviewed.

My typical shoots would be shooting 6K DCI Q1 30 fps recording for well over 2 hours on a 2TB card with a bitrate under 250 MB/s to deliver 4K 2.39:1.

For long events I’d shoot 6K 2.4:1 Q5 recording for up to 8 hours on a 2TB card with a 1TB card in reserve if it’s needed.

I currently have several Panasonic L-Mount lenses, along with a Sigma 150-600mm sports.

My question is: Will the new BMCC 6K support the built in lens profiles for the removal of CA, distortion etc?

I'm presently using the S5 & S1 with the Video Assist 12g 5 inch and it appears that the lens profiles are being applied before RAW output as I've never experienced any issues.

I was just in the process of creating a pre-order for the new camera when a shiver went down the spine. The thought of dealing with CA etc in post production is a definite obstruction.

Any help much appreciated.

They don’t.

In fact there is no way to do this with built in raw shooting on any camera platform that I know of. Anytime these corrections are done it’s “baked” into the compressed or encoded image and therefore can’t be undone either.

And I doubt they are applying it on the HDMI output if you’re receding to BRAW either.

So if you’re happy with what you get now, it’s most likely uncorrected. I could be wrong but that’s my best guess.

JB

That's fairly much what I was thinking. My only reservation is I understand that with the S series cameras the camera and the VA share some of the processing in the creation of the raw file. I wonder if part of that processing includes the lens profile corrections.

This video from Blackmagic is what got me thinking, you may have already seen it.

Thanks for the input, appreciated.

I wonder if part of that processing includes the lens profile corrections.

I can't speak for those cameras specifically, but micro 4/3 cameras rely rather heavily on lens corrections in post to get good performance particularly from the less expensive consumer-level lenses, and those corrections are not usually applied to the RAW files. Most commonly the software that processes the RAW files is expected to apply the corrections at the time the RAW data is processed, so the correction data is generally included in the RAW file to make it available to the software.

The cameras usually apply the corrections when generating JPEGs, video clips, or other formats that are not RAW.

I'm more interested to see how all the focus-by-wire lenses (which I think almost all electronic L mount lenses are) function on this new camera. I know Panasonic cameras have a function to focus electronic lenses linearly, which seems like a must-have function for the cinema cameras as well.

That's fairly much what I was thinking. My only reservation is I understand that with the S series cameras the camera and the VA share some of the processing in the creation of the raw file. I wonder if part of that processing includes the lens profile corrections.

This video from Blackmagic is what got me thinking, you may have already seen it.

Thanks for the input, appreciated.

I use Panasonic S cameras and if you shoot video in any of the RAW formats available via Ninja V or Blackmagic Video Assist there is no corrections done.

Some low light footage from Matteo.

Some 6/5 anamorphic footage from Flo.

That's fairly much what I was thinking. My only reservation is I understand that with the S series cameras the camera and the VA share some of the processing in the creation of the raw file. I wonder if part of that processing includes the lens profile corrections.

As far as I know there’s no way to accomodate a lens profile in BRAW.

Just like in every other raw cinema format. They don’t support lens profiles because the problem is that some of these things are dynamic.

Think about CA for example. It’s on contrast right? What happens in a shot if the shot starts with no CA, pans into a scene that has high contrast and produces CA and then pans off again?

It’s very very difficult to dynamically do CA correction as anyone who has tried doing it in the cDNG days will tell you.

Other lens geometry corrections can mayyyybe can be applied but again, you’re going to likely see weird things happening on vertical lines as you pan through something because the correction CHANGES based on what’s in the shot AND which part of the lens is being corrected.

There’s also no real mechanism for intra frame metadata. For example, tracking focus distance accurately OVER TIME within a clip. There are only external animation techniques that are guaranteed to work.

These kinds of areas are what we hope the BRAW developers are working on because it’s only one you track all this info action by frame rather than clip that you can maybe start to do these other processing techniques.

JB

What is the spec on the HDMI out?

What is the spec on the HDMI out?

The Tech specs list: 1 x HDMI up to 1080p60

Original BMCC 2.5k has SDI out, not this and only 1080p HDMI out. If it looks like a Pocket, behaves like a Pocket, it must be a Pocket

It should be called Pocket 6k LF

Original BMCC 2.5k has SDI out, not this and only 1080p HDMI out. If it looks like a Pocket, behaves like a Pocket, it must be a Pocket

It should be called Pocket 6k LF

Original pocket output to sdi only 1080p,
Bmpc4k output uhd not 4k dci from its sdi…
I had both of it…
Ursa mini pro output uhd from sdi or dual sdi raw only for BMD raw recorder.

From what I know no one cinema camera from Blackmagic Design output raw but Ursa line only and only with dual sdi.
Different production line, different power (cpu and energy) to manage all data’s.

This is a 2500$ cinema camera that record internally in braw.
If I want to record braw from my Fuji x-h2 I should add Va12g and ssd, up other 1000$

Inviato dal mio iPhone utilizzando Tapatalk

… Original pocket output to sdi only 1080p,
Bmpc4k output uhd not 4k dci from its sdi…

I think you meant to say HDMI, not SDI.

… Original pocket output to sdi only 1080p,
Bmpc4k output uhd not 4k dci from its sdi…

I think you meant to say HDMI, not SDI.

I talk about original black magic production camera 4k


Inviato dal mio iPhone utilizzando Tapatalk

My mistake, apologies.

The original raw (CDNG) recorder used a single SDI for raw recording as the camera only had a single 12g output.

The Micro 4k G1 also output raw over SDI.

Good Luck

Original BMCC 2.5k has SDI out, not this and only 1080p HDMI out. If it looks like a Pocket, behaves like a Pocket, it must be a Pocket

It should be called Pocket 6k LF

Original pocket output to sdi only 1080p,
Bmpc4k output uhd not 4k dci from its sdi…
I had both of it…
Ursa mini pro output uhd from sdi or dual sdi raw only for BMD raw recorder.

From what I know no one cinema camera from Blackmagic Design output raw but Ursa line only and only with dual sdi.
Different production line, different power (cpu and energy) to manage all data’s.

This is a 2500$ cinema camera that record internally in braw.
If I want to record braw from my Fuji x-h2 I should add Va12g and ssd, up other 1000$

Inviato dal mio iPhone utilizzando Tapatalk

BMCC 2.5k is not UHD or 4k so it would be an ask getting the SDI to output 4k. With respect the new 6k camera we were told by BMD rep only externally does it looks like the Pocket range but it is a different beast. If so and being the next gen BMCC in 10 years it should be capable at least UHD SDI out. Right now its just another one of their Pocket range regardless of what is inside it does the same or even less with just a bigger sensor. Pocket LF it should be.

I don't want to be a luddite but why do people want 4K/UHD output from the CC6K again? For external recording?

Call me crazy but if it keeps the cost down or it contributes to the reliability of the camera, 1080p is perfect. I'm not saying we won't make the switch eventually but we're still monitoring in 1080p on Tier-A union shows and I'm not sure when that'll change. Hell, I feel like the only person on set even close enough to their monitor to benefit from it would be the 1st ACs, and 1080p works well enough that they're usually more concerned about making sure their signal is stable and clean.

I don't want to be a luddite but why do people want 4K/UHD output from the CC6K again? For external recording?

1080p output mean downscaling operation, processing power requested, it's better if i can't have it.

many people, me with them, would like to have more resolution for many reasons :
- external backup recording, when you record something like wedding, documentary, sport you cannot repeat the action.
- better focus action, you see framing on camera, and focus on external monitor where you can zoom if you good external output. focus peaking is good, bu sometimes you have camera in a unconfortable position (gimbal or crane) and a good output allow to you, allow to focus puller to work better.

additional... if from my fuji x-h2 i can have braw from hdmi, may be possibile from hdmi of pockets or 6kLF to have a backup?

1080p output mean downscaling operation, processing power requested, it's better if i can't have it.

The camera is already demosaicing and scaling the image regardless. Outputting 1080p isn't making the camera draw more power. Besides, the camera would still be scaling the image with a 4K/UHD output too...

Again, a 4K/UHD output would be nice and it's not like the camera is better without it but if the cost of putting an HDMI 2.0/2.1 port in there would've been substantial then I get why it isn't in there. 1080p is still the standard on sets.

Good observations Alex, but I think there are some benefits for a revision of the BMCC6K to include 4K in some form. When the BMVA12G was launched it supported HD resolutions from BMD cameras using SDI and/or HDMI such as the BMPCC4K and the UM4.6K. However the monitor can handle 4K from many third-party camera manufacturers. It seems overdue for BMD to support 4K output from their new cameras.

I feel it’s time to promote 4K delivery where feasible.

I don’t know the penetration of 4KTV versus HDTV in consumer’ homes or businesses, but I would imagine it is growing. So good to support it without getting sidetracked by arguments regarding if viewers will see a difference. I still only have HDTV personally and like it.

I’ve recently started upscaling 2K captures to deliver 4K via Resolve’s excellent 2x Enhanced Super Scale functionality, but I suspect it’s better working in post with 6K captures and delivering 4K.

What is the spec on the HDMI out?

The Tech specs list: 1 x HDMI up to 1080p60

That's actually good news for me, as all my monitors are 1080p; I was worried I'd have to upgrade.

… Original pocket output to sdi only 1080p,
Bmpc4k output uhd not 4k dci from its sdi…

I think you meant to say HDMI, not SDI.

I talk about original black magic production camera 4k


Inviato dal mio iPhone utilizzando Tapatalk

What was the output for that camera? I bought a SDI cable but it didn't work correctly for me.

I don't want to be a luddite but why do people want 4K/UHD output from the CC6K again? For external recording?

Call me crazy but if it keeps the cost down or it contributes to the reliability of the camera, 1080p is perfect. I'm not saying we won't make the switch eventually but we're still monitoring in 1080p on Tier-A union shows and I'm not sure when that'll change. Hell, I feel like the only person on set even close enough to their monitor to benefit from it would be the 1st ACs, and 1080p works well enough that they're usually more concerned about making sure their signal is stable and clean.

It's good for using with the Video Assist.

Also when streaming it's nice to have a resolute signal out for cropping etc.

Back in the film days, the ASC listed recommended panning speeds per focal length to avoid a kind of staccato, stuttering image during camera movement.

That hasn’t gone away.

I had an Alexa mini Steadicam shot that got bounced for stuttering on a fast pan. HDR mastering has made it worse. When you have bright and dark alternating tones on a fast pan it becomes way more obvious.

Impossible to fix in post too.

JB

And it got worse with digital. In analog cameras with a mechanical shutter, the edges of that shutter are not in the image plane, so the beginning and the end of the motion blur is softened.

In digital sensors the edges of the motion blur are sharp, so they get more visible. Arri tried to cure this in the Arri Alexa Studio with a mechanical shutter, but it was no big success in the market.

Right now, the only solution is the Motion Mount by Red. Actually, made by Tessive, but they bought the rights. It works, but it has quite a few other issues, the worst one being the polariser, just like in variable NDs.