July 27, 2012

First Blackmagic Cinema Camera Footage Since NAB, Camera Won't Have Optical Low-Pass Filter

Earlier in the week John Brawley, a director of photography helping Blackmagic with the development of their Cinema Camera, posted some full resolution screenshots of footage he'd been shooting with that camera -- the first such material since NAB. Now we've got the next best thing -- he's posted a minute of footage online. Video samples from this camera have been rather scarce, according to Brawley, since Blackmagic has been working on finalizing the sensor calibration and wanted to wait until they felt good that the calibration and firmware were fully mature. There has been quite an interesting revelation in the past day, however, and it has been revealed that the final camera will not have an optical low-pass filter (or anti-aliasing/blur filter as it is sometimes called). We all know that moire can be a problem on many of the cameras we shoot with (mostly DSLRs), so the big question many are wondering is, how will moire be on the camera now that it won't feature the OLPF?

The footage has been made available only on the Vimeo page for John Brawley, so you can get there by clicking the video image below:

Here are some of the details about the shoot from Brawley's blog:

Lenses were the lovely Zeiss Compact Primes...18mm for the wide, 35mm for the panning shot of Casey at the table and most of the closer shots were the 50mm or the 85mm. Exposures were mostly T2.8 or T8 for when I was looking back at the windows. Shooting at ISO800 to ProRes 422 (HQ) in film mode.

The clip was edited in FCPX and then it was exported using XML into Resolve 9, and graded using only single nodes and primaries. The lighting was very minimal, and with such little done to the footage in post, it's remarkable how nice this camera is rendering colors. The skin tones in the clip have a nice balance without too much red, yet the shirt, her lips, and the red pool balls show excellent saturation and a full range of tones. This type of shot is difficult to replicate with DSLRs without introducing noise somewhere. Often, you can shoot for nicer looking skin tones, but you'll lose the fullness in the reds, and if you try to bring those back to what they were, there's a good chance you will be introducing some noise. This is not the case with this footage, thanks to the high bitrate and exceptional color space (ProRes HQ at a maximum of 220mbps and 4:2:2 color). Even the compressed Vimeo footage shows clear separation and excellent tonality.

We have now learned a major detail about the camera, that it won't feature an optical low-pass filter. I don't believe this will change, as this is what John Brawley has stated (I'll be sure to update if this is wrong). It may not suffer from moire and aliasing nearly as badly as Canon's cameras have in the past. They achieve their image in a very different way, and some of the problems come from the fact that those cameras have to figure out a way to get from 5K down to basically 2K. What this does mean, however, is that very fine detail patterns have a much greater potential of posing an issue. If you're still reading and have no idea what's going on, you can read more about the Nyquist and sampling theorem here (have fun with that one if you're not an engineer, it's a very in-depth read).

With the way this camera derives its image, I believe the positives will far outweigh the negatives. The huge positive of not having the optical low-pass filter should be improved sharpness. That's because OLPFs are a trade-off. Make one that's too strong, and you'll lose fine detail that wouldn't have been affected by moire, and make one too weak and you'll let a lot of moire through. By not having one at all, similar to digital medium format cameras, you should be able to get slightly more overall image quality. This is not too dissimilar to the way the Nikon D800E works, but in that case, Nikon actually used two filters that cancelled each other out. All of these cameras have to have an infrared filter, to block out the infrared wavelength which pollutes much of the color in daylight images.

We'll see what this will mean for the image soon enough (and how bad moire will actually be), but it may just be another trade-off for maximum image quality and lowest cost.

[via John Brawley]

Your Comment

78 Comments

I don't understand the lack of downloadable files. The camera is supposed to be released in a couple of weeks and it still seems like it's in early beta testing. It seemed so promising at first but in the months leading up to release it sure seems like Black Magic isn't really ready to release this camera yet. I hope I'm wrong but you have to admit they are going about this in a very unorthodox manner. Let's hope that means the camera is unorthodox (awesome).

July 27, 2012

-1
Reply

I honestly can't tell you, I mean no company ever releases downloadable footage for their cameras. There is a plausible deniability to web footage. I actually don't think that's the issue, I think bandwidth could be more of an issue, but if you look at it from their point of view, they are damned if they do and damned if they don't. It should be common practice to release full quality footage, but this just sadly doesn't happen.

July 27, 2012

-1
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

I guess I'm more worried about the overall lack of information. No cameras being released early for test/review. Firmware and little changes like no OLPF coming days before supposed ship dates.

July 27, 2012

0
Reply

I don't know if it was a change, it may just have been something they weren't sure if they were going to implement. July 31st is not a hard ship date for the camera, I think it will be a few weeks roll-out before it gets into any serious volume. I think this just proves how hard camera design is, it's not like the thing isn't working, I would imagine they are being extra cautious since this is their first camera and they can't make a second impression.

July 27, 2012

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

I want it to be a great camera. I really do. I'm rooting for it because IF it does end up being an amazing cam then it forces the hand of bigger manufacturers and spawns smaller companies to get in the game. Good for everyone.

July 27, 2012

0
Reply

The reason they said they didn't release the footage yet (from what i remember) was because they keep improving the camera and tweaking it. I still would like some tbh. but owell.

July 27, 2012

0
Reply
Roy

I've actually mentioned from very early on that the use of an AA filter was very much up for grabs.

jb

July 28, 2012

0
Reply
John Brawley

@roy oh my, perpetual beta. Where have I heard that before?

July 28, 2012

-1
Reply

I know right? Lol. Just saying what I think I remember them saying.

July 28, 2012

1
Reply
Roy

Hey BMD aren't those guys. I'm the only one that get's late night email rants from the CEO of BMD ! ;-)

In all seriousness, they plan to get the camera out the door and get the camera making nice pictures in a basic way (and it doesn't get more basic than RAW).

Depending on feedback, they will implement feature request where they can based on demand (how loud people yell and jump up and down)

jb

July 29, 2012

0
Reply
John Brawley

The doubters of this camera and people who just love to be negative and point out flaws in general are about to rip this camera to shreads over this smh. Either way for the pricepoint the image i'm seeing is revolutionary. I'm glad they don't have slow motion on this camera as it gives me an excuse not to buy one (I only shoot music videos where slow motion is a necessity) and I can wait for them to fix the bugs of the first generation. If they through at least a 4/3rds mount (aps-c size even better) add slow motion and get rid of aliasing I would glady pay even more than they are charging for this now.

July 27, 2012

0
Reply
Jacob

There you go! Maybe for a few dozen people this is new news :-)

July 27, 2012

0
Reply
Chet

Well the no OLPF is news I haven't heard anywhere else, I felt that was significant.

July 27, 2012

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

bmcuser.com, where you hear it first :-) Very informative over there.

July 27, 2012

0
Reply
Chet

So why did you take the time to read it here? Do you get a kick out of wasting your own time? Ha ha!

July 27, 2012

0
Reply

I'm not here to read.... here to make sure Joe stays on point, son.

July 27, 2012

0
Reply
Chet

Ahhhhh. I see. So you're wasting your own time AND Joes. Very constructive. Carry on.

July 27, 2012

0
Reply

Hop off my dick kid

July 27, 2012

0
Reply
Chet

Interesting. "Son". "Kid". "Dick". You should see if any Psychiatry schools in your area need test subjects.

July 27, 2012

0
Reply

Damn, you must be bored as hell.

July 27, 2012

0
Reply
Chet

OK, I think we can just move on from this one. Plenty personal enough :)

Also, not everyone has time to read through forums and other sites, which is why we sort through all that stuff if you don't want to.

July 27, 2012

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

Why do you even care haha?

August 2, 2012

0
Reply
Robert

That actress has a great ass. John Jeffreys seal of approval. I'd shoot that all day. All day.

July 27, 2012

1
Reply
john jeffreys

Now in regards to the actual topic, this recent series of stills and footage from Brawley is really, really, tempting me to sell my 5d kit and put it towards a BMC. My biggest concern was a very small-sensor-ish depth of field but it seems perfectly fine in that clip, AND he was only shooting at 2.8. The colors and dynamic range is unbelievable, so filmic. Crop factor is my only concern right now.

July 27, 2012

0
Reply
john jeffreys

Jacob:
"I would glady pay even more than they are charging for this now"
You people will always complain about the cameras even if there is everything you mentioned above..

July 27, 2012

0
Reply
Black dynamite

Did anyone else notice the color or aperture shift/pop around 0:40 - 0:42? Is that the camera or the Vimeo player?

July 27, 2012

0
Reply

What are you talking about? I did not notice anything...

July 27, 2012

1
Reply
Black dynamite

It's likely a compression artifact, because it's a slight gamma shift.

July 27, 2012

0
Reply
kholi

it´s not a compression artifact, when you wach again carefully that shot is made from two shots , and their gamma/gain doesn´t mach perfectly

July 27, 2012

0
Reply
matus amrich

Didn't I just say its a Gamma shift? Mpeg. Compression or conversion does this to footage all the time.

July 27, 2012

0
Reply

It's not in the original I'll re-upload this tonight when I get the chance.

jb

July 28, 2012

0
Reply
John Brawley

I'm not particularly excited about this camera, but the dynamic range is certainly impressive.

July 27, 2012

0
Reply
Kilgore

For completeness of this article, what is the native rez of the BMCC sensor again? Without an OLPF the aliasing/moire will be introduced at that rez, and any downsampling after that can amplify or hide artifacts introduced at the higher rez. There will be a downsampling in post as well from 2.5K to 1080p, though usually better downsampling algo's are used in post (because their authors don't have a captive audience like the camera manu's do), and moire/aliasing shouldn't be introduced much then. But again that downsampling can amplify artifacts upstream.

I am sure the moment someone even vaguely scientific gets their hands on a BMCC the tests will be done. If the news is good, Blackmagic should release its own proper tests preemptively. If the news is bad, they should try to filter out geeks from the list of people who receive their bodies first, and try to sell on subjectives. I'm someone who balances both factors but I want to see the objective data before spending $3,000.

July 27, 2012

1
Reply
Peter

Full RAW resolution is 2432 x 1366. Moire/aliasing are also dependent on pixel fill without the OLPF. The more pixels you have theoretically the less moire/aliasing will occur because it's resolving that fine detail.

July 27, 2012

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

Pretty impressive. Didn't realize how much the BMC looked like Alexa footage, are they seriously selling the body for only 3k...? At one point I was tempted by the Red Scarlet... but damn, this looks far more impressive than anything I've seen shot on Epic or Scarlet in regards to regular 24p.

This is pretty much the most cinematic image I've seen a digital camera produce aside from an Alexa. Where the hell was I during the pre-order...?

July 27, 2012

0
Reply
Todd

Just for your information this was actually filmed in 25p.

July 27, 2012

0
Reply
Raphael

How do you know it was shot in 25p? Also i wonder why they would not of shot in 24p.

July 27, 2012

0
Reply
Matt

Because 25p is PAL standard in Australia, where John is based.

July 28, 2012

0
Reply
Tim

Will the first person capable of telling the difference between 24 and 25 fps please stand up?

July 28, 2012

-1
Reply

After doing more research i see John shoots for TV in AU. Thanks for the info.

July 28, 2012

0
Reply
Matt

John Brawley said so and yes, it's also the PAL standart in Australia.
And Richard it depends on the person, I probably couldn't but i know an old DP who could guess after say, one minute of footage just to be sure. I guess it's a simple matter of experience and knowing what to expect beforehand.

July 28, 2012

0
Reply
Raphael

The first lot of footage WAS shot 24 because that's all the first prototype could do ! But yeah, even for cinema, we often still shoot 25 FPS in this part of the world.

jb

July 29, 2012

0
Reply
John Brawley

NO OLPF= BAD NEWS=SAD POTENTIAL CUSTOMER ;(

July 27, 2012

0
Reply
RebelPhoton

For the price, and what you get, this camera looks great. I'm not gonna be too nit picky about it.

July 27, 2012

0
Reply

Nasty moire in the back of the chair at 0:22. Disappointing.

And you don't blow the chalk off the cue. Amateur move.

July 27, 2012

0
Reply
Graham

Her blowing the chalk is a 'cinematic' move... stop moaning lol.

July 27, 2012

0
Reply
VillageBoi

Yeah it's there. But honestly I rarely see it. In this case, if you think about it, it's amazing that it's showing up. That is a pretty wide shot and you've got some very fine mesh at quite a distance.

It will never be nearly as bad as say a 5D because the root cause is quite different.

jb

July 28, 2012

-1
Reply
John Brawley

The moire in the chair have nothing to do with the camera, at least in this shot. I would guess it’s some “real” moire that you would also see with your eyes because the chair has two layers of slightly transparent fabric (basically a very small grid) stretched around it which is causing moire.

Your eye will always see moire patterns when two grids are held one behind another and are a little tilted. Moire in camera appears also because of two grid layers, the one being filmed, like a brick wall or some fine fabric (you won’t see moire here with you eye, only with some cameras) and the other being the pattern of pixels on the sensor of the camera itself.

So correct me if I’m wrong, but this as I think is moire filmed by the camera that you will also be seeing with your eyes. So there is nothing to worry. Or is there? ;)

There are some colour shifts in 0:21-0:23 on the chair, I guess this is some camera issue and an effect you wouldn't see with your own eyes.

What do you think?

July 28, 2012

0
Reply

There is in installation at MOCA that actually uses "natural" moire as an intended effect. Its a bunch of wires stretched from edge of the ceiling to the other, and you walk down the hallway and look up and watch the lines go crazy, like some bad 5d footage.

July 28, 2012

0
Reply
john jeffreys

Thanks so very much John Brawley for posting this video and all of the other stuff you've done concerning the BMC. I (and others) really appreciate it

July 28, 2012

0
Reply
DeaconWhite

Pages