BMCC6K Facts and Findings

A short film by Shugo Akatsuka (Tokyo, Japan)
Shot on Blackmagic Cinema Camera 6K with DZOFilm PAVO 55mm T2.1 (2x anamorphic)
Edited with Phantom LUTs (Vision)

Wow.... those DZO are really great looking.

BMPCC 6K Full Frame/Pyxis vs. Kinefinity Mavo LF MKII test | Blazar Remus Anamorphics (Omega Brand Productions)

I knew Cam A. was BMCC6K from the noise

I knew Cam A. was BMCC6K from the noise

But...Cam B was the BMCC 6K

I watched it thinking A was the BMCC because of your comment.
I kept thinking "the image is much nicer than cam B. Wow, it's even much better in low light? Wait, is cam A really the BMCC 6K?!".

Obviously we have no control over the post pipeline but judging from this video, I like the Mavo a lot better.

Obviously we have no control over the post pipeline....

Or, for that matter, the production pipeline. There's no stopping youtubbers from posting this stuff, but why does anyone take it seriously?

Yes but what production differences in this case?

There's no stopping youtubbers from posting this stuff, but why does anyone take it seriously?

Well for starters the Tuber in question has actually shot on both cams. Have you used either? Why should we value your opinion over his?

Good Luck

What opinion on the cameras did I offer, and what opinion about them am I urging you to accept?

Or is it your view that no matter how a "test" is conducted -- you're not privy to the details -- it's reliable?

Every youtubber earns your respect if he's shot with the two cameras he claims to have shot with?

What opinion on the cameras did I offer, and what opinion about them am I urging you to accept?

It wasn't your opinion of the cameras, of which neither of you had incidentally, rather your opinions of Tubers. Why should we accept your assertions? Especially when posting nonsense like this.

Every youtubber earns your respect if he's shot with the two cameras he claims to have shot with?

Good Luck

So pointing out that these comparisons are of dubious value (putting it politely), there being no access to how they were actually performed, is "nonsense"? As is responding to the claims for this youtubber you advanced ("well for starters the Tuber in question has actually shot on both cams"), which is evidently supposed to be a good reason, in your view, to take the "comparison" seriously?

There used to be fairly decent camera testing available. Geoff Boyle, Adam Wilt, etc. Of course, they never went for in these these "shoot-out" style tests, since they're all but impossible to perform reliably or with common baselines.

But I gather you know better, Howard. Good luck(?)

I approach all camera comparisons with a healthy dose of skepticism. What I don't do is categorically dismiss them or entertain conspiracy theories.

Cheers.

That’s his signature John.

Also, the majority of your posts seem to be overly harsh, critical, and a bummer to read.

Forums are supposed to be a place to geek out and escape from the negativity that can be pervasive in our lives.

Discourse is welcome, negativity is a drag.

Just my .02 cents

Also, the majority of your posts seem to be overly harsh, critical, and a bummer to read.

Forums are supposed to be a place to geek out and escape from the negativity that can be pervasive in our lives.

And here I thought forums, particularly an industry-run forum like this one, are supposed to be places for reliable information.

Granted, if you view BMD forums mainly as social platforms, any objection to the "shoot-out" youtubes tirelessly posted here will be unwelcome.

How's that for a bummer?

I knew Cam A. was BMCC6K from the noise

But...Cam B was the BMCC 6K

I watched it thinking A was the BMCC because of your comment.
I kept thinking "the image is much nicer than cam B. Wow, it's even much better in low light? Wait, is cam A really the BMCC 6K?!".

Obviously we have no control over the post pipeline but judging from this video, I like the Mavo a lot better.

My bad I meant Cam B. A part from low-light performance, BMCC6K performs well; it likes light as we all do

That’s his signature John.

Also, the majority of your posts seem to be overly harsh, critical, and a bummer to read.

Forums are supposed to be a place to geek out and escape from the negativity that can be pervasive in our lives.

Discourse is welcome, negativity is a drag.

Just my .02 cents

I'm not familiar with John's post but what did he say that was wrong here? A lot of shoot-outs like the one posted are questionable especially when they're presented as-is with no other information about how it was shot because we can't judge if the test conditions were flawed.

For example, the first shot appears to show more vignetting and shallower depth of field in the BMD shot than the Mavo shot which suggests the lens is open wider. Why? What other effect does that have on the sharpness of the subject? How was it exposed? If it was exposed for middle grey, what ISO was used? How was it post-processed?

Same thing with the darker shots. The MAVO appears darker and less contrasty than the BMD. How much less noticeable would the BMD's noise be if it's contrast and brigthness were matched with the MAVO? Would matching the MAVO to the BMD bring out any noise? Since the MAVO doesn't shoot RAW, are their any signs of noise reduction being done or is the MAVO's sensor just that much better in low-light?

The shot with the DZO lens also shows that's it's open wider on the BMD shot.

A lot of shoot-outs like the one posted are questionable especially when they're presented as-is with no other information about how it was shot because we can't judge if the test conditions were flawed.

So pointing out that these comparisons are of dubious value (putting it politely), there being no access to how they were actually performed, is "nonsense"?

C'mon guys, you need to put in a minimum of effort. They do present it. From the YT info:

We had fun and it was our first true testing experience and we are not experts. We did our best to get the settings almost the same with native REC 709. There's some weird Youtube compression going on with the Kinefinity lowlight shots as they are a lot better not on Youtube, unfortunate!

Outside settings:
Blazar Remus Lenses : 45mm, 65mm, 100mm
T4.0 W/ Tilta Mirage VND Mattebox
ISO 800
WB 5800

Indoor settings:
Blazar Remus Lenses : 45mm, 65mm, 100mm
Open at T2.0 for 45mm & 65mm
Open at T2.8 100mm
ISO 3200
WB 2900

Again we are not experts but we managed to get the color as close as possible before applying luts

C'mon guys, you need to put in a minimum of effort. They do present it. From the YT info:

There are so many unknown variables, and known variables for which they haven't compensated (production and post-production both), that nothing can be concluded from what they've done. They themselves admit that they're inexperienced in the matter, that they're mainly "having fun" and that the youtube is not actually representative of the footage.

And that's fine, as far as it goes: "fun". The same applies to any number of other "shoot-outs" routinely posted here, which suffer from the same defects. But saying so now amounts to a "conspiracy theory" against youtubbers or disrupts all the happiness and goodwill here?

Are we serious here -- or not? Are there any standards of discourse, at all? Or is everything Facebook?

There are so many unknown variables, and known variables for which they haven't compensated (production and post-production both), that nothing can be concluded from what they've done.

They fixed enough variables to be able to draw conclusions. The unknowns I see are the color correction to get both cams to the same colors and the temperature and exposure shifts that would occur when filming an uncontrolled scene outside. They used the native 709 transfers, so they didn't grade, like add contrast ect. So a comparison of 2 different cams and also their 709 transfers.

Would I rely on something like this to chose between cameras for a project, of course not. I'm also the last one to defend YouTubers, and this comparison is as flawed as most, but I can conclude things from it. To say that nothing can be concluded is overexaggerated.

I would have picked Cam B as the P6K FF even without seeing the lowlight shot. Giveaway is the warm tone.

"I always use an FX6, the FX3, you know, modern cameras with AF and all the bells and whistles that come with them, but there always has been something about Blackmagic, right? you've heard all the other people talk about it, 'oh it's the image quality'...I have been a Resolve user for a long time and, my God, they were right!" (Sony Fanboy)

"I always use an FX6, the FX3, you know, modern cameras with AF and all the bells and whistles that come with them, but there always has been something about Blackmagic, right? you've heard all the other people talk about it, 'oh it's the image quality'...I have been a Resolve user for a long time and, my God, they were right!" (Sony Fanboy)

That Tokina footage looks pretty awful; at first I thought it was just veiling flare but I think there's haze in at least one of the elements, probably from oil offgassing from the aperture blades (the haze looks oily). The Minolta shots are better, but overall I don't get the magic here. And I like using vintage lenses.

Well, what do you expect from a YouTuber, who can't even get his horizon straight?
Yes, that Tokina has a massive issue. I had oil from the blades on some old Soviet glass and had them serviced professionally. This looks similar as they did before cleaning.

Well, what do you expect from a YouTuber, who can't even get his horizon straight?

When he said, "It made me feel..." I added "Seasick" in my mind.

I'm guilty of being a flare fan boy here.

GEAR:
BMCC6K FF + Leica R 50mm Summicron / NISI VND
RS4 PRO / TILTA RING / EASYRIG VARIO 5 STABIL + EASYTILT
DJI FOCUS PRO & DJI Transmission

Some artifacts happens when you do cross dissolve with color grade and using the gyro stabilisation.

Don't know how to have a good workflow on sequence shot like that when you need to shift contrast and grade on some part of the image while maintaining a GYRO stabilisation...

Didn't had time to investigate further as this one had to be release very soon after being shot.

If anyone had better results in timebased grade and gyrostab, please tell me !

I'm bummed I missed the BMCC6K deal they had going on! Does anyone have any scoop on whether the deal will come back? I'm in the market for this camera and am willing to wait. Could another sale be coming around Amazon's Prime day?

For fun, I made a short video for a friend. The more videos I take, the happier I’m with BMCC6K.

Your comments are welcome

Well, what do you expect from a YouTuber, who can't even get his horizon straight?

This personal attack is totally uncalled for.

Like everybody else, I also have to endure attacks when I make my mistakes public.
This time, I was not the only critic, if you look at the next response…

Like everybody else, I also have to endure attacks when I make my mistakes public.
This time, I was not the only critic, if you look at the next response…

I think there is nothing wrong with criticizing the work but I do not think it is appropriate to criticize the person.

I'm sorry if it came across like that.
Yes, I should have made it more clear that it was the clip that disappointed me.

I mentioned something in my feature request topic for adding Ursa 12K-styled crop selection but this topic gets more eyes so I wanted to mention it here to get more opinions on it.

The dimensions of Super 35 are 24.89 x 18.66mm and the 6K FF and PYXIS sensor is 36 x 24mm. Since the full resolution of the sensor is 6064 x 4040, that would mean a Super 35-sized area should be 4192 x 3144, not 4096 x 3072 as it is on these cameras. Would anybody be opposed to S35 and 4K DCI crops getting replaced with 4192 x 3144 and 4192 x 2224 ones? They're only about 5% higher resolution but it would make better use of an S35 image circle and the frame rates you can shoot with them should be the same. I don't see any good reason to adhere to shooting at a standard delivery resolution like 4096 x 2160 when it's adding more of a crop then needed and it doesn't make it any less suited to delivery in 4K DCI.

I'll admit the idea does look more attractive in the contest of a system of crop/resolution selection that he Ursa 12K uses, but I feel like it has merit on it's own. I got into my detail about all of this in my topic.

I suggest reading this and see if it changes your mind: http://endcrawl.com/blog/2048x1152-is-a-total-crock/

I suggest reading this and see if it changes your mind: http://endcrawl.com/blog/2048x1152-is-a-total-crock/

That article is really more about the finishing process and selecting the appropriate resolution for your output master. It's not as applicable when it comes to selecting a resolution for your original camera negative, and it's even less applicable when you start considering a raw OCN's capture specs. It is worth being reminded how detrimental fractional scaling can be in certain contexts though.

Would anybody be opposed to S35 and 4K DCI crops getting replaced with 4192 x 3144 and 4192 x 2224 ones? They're only about 5% higher resolution but it would make better use of an S35 image circle and the frame rates you can shoot with them should be the same. I don't see any good reason to adhere to shooting at a standard delivery resolution like 4096 x 2160 when it's adding more of a crop then needed and it doesn't make it any less suited to delivery in 4K DCI.

You're asking if users want more options when they acquire their OCN? I mean... Sure? I guess? I just think it's possible to get so drunk on spilling ink with napkin math that camera design seems simpler from an end user's perspective than it actually is. There are so many bottlenecks in this process that it's rarely a question of just sliding around the variables.

Joe, agree with the prohibition for delivering in 1152 lines of resolution. I always capture 2048 x 1152 and process in Resolve with Centre Crop no Resizing to deliver one of the mentioned standard aspect ratios without creating artifacts. The extra captured pixels are for reframing which I rely upon.

As for the observation that shooting 4096 will result in using a physical area of the sensor smaller than traditional 4K film capture, I’m fine with whatever BMD wants to define as 4K such as recording 4096 in an area physically smaller than 24.9mm. Once you get over the outrage of BMD marketing 4K Super 35 in a camera that is recording the equivalent of the Canon photography APS-C physical format because there’s no international convention describing 4K Super 35 for digital cameras, I think they’re being reasonable anchoring 4K to 4096 photosites rather than the size of the active sensor/window.

After all, a century ago when I was coding a K in random access memory meant 1024 bytes, so 4K means 4096 units of something, not a physical distance. I can still rage over the fate of values when a K of a disk storage media began to mean 1000 bytes and not 1024; never in my day, but, the logic and truth of “my day” is long gone, replaced by the intentional lies of marketeers.

I suggest reading this and see if it changes your mind: http://endcrawl.com/blog/2048x1152-is-a-total-crock/

By that logic, BMD made a bad decision choosing 4096 x 2308 for it's 4K 16:9 crop instead of UHD on the Ursa Cine. I think they made the right decision though. It's allowing the camera to shoot with a wider field of view and higher resolution which allows for oversampling and/or cropping/motion stabilization.

Maybe I'd be convinced if the article showed any examples with actual footage but they didn't. It makes complete sense with credits because they're all sharp edges with a lot of contrast but it makes less sense with recorded video. Recorded video has some level of noise and is already resolution limited via de-mosaicking, in camera compression, and lens optics. Using credits as an example of this is like saying that JPEG was a bad choice for compressing photos just because it compressed clip art badly.

I wouldn't even know how to test this in a way that insures that resolution is the only variable. If I shoot at 5% higher resolution then I intend to downsample to then yea, there's going to be some fine details that I saw before downsampling that I won't be able to see as clearly after downsampling... because I lowered the resolution. However in this case I also got 5% wider of a field of view and the noise is every so slightly finer and I can also still crop to keep that detail if I don't need data around the edges. To recreate the same shot with a 4K DCI crop, I'd have to increase the field of view somehow by 5% and I'd lose some of those finer details just because the same amount of photons is being collected by fewer pixels making those details more prone to being lost from de-mosaicking, the lens optics, or compression. I also would no longer have the option to crop any more without further diminishing the resolution and the noise will be slightly less fine.

We're also talking about down-sampling to 4K instead of 2K so any issues are going to be less apparent anyway.

As for the observation that shooting 4096 will result in using a physical area of the sensor smaller than traditional 4K film capture, I’m fine with whatever BMD wants to define as 4K such as recording 4096 in an area physically smaller than 24.9mm. Once you get over the outrage of BMD marketing 4K Super 35 in a camera that is recording the equivalent of the Canon photography APS-C physical format because there’s no international convention describing 4K Super 35 for digital cameras, I think they’re being reasonable anchoring 4K to 4096 photosites rather than the size of the active sensor/window.

It's not outrage over anything. I get that the term S35 is loosely defined when it comes to image sensors but the entire purpose of the S35 crop is so that you can use lenses made for the image circle of S35 film. That's why the S35 crop is 4:3 as well, because S35 film is 4:3. So if you have the resolution and the sensor size, why not maximize what you can get out of those lenses. BMD could have made the S16 crop 2048 x 1080 since it's close enough but they didn't. They chose 2112 x 1184 which is the resolution closest to physical width of Super 16 film.

After all, a century ago when I was coding a K in random access memory meant 1024 bytes, so 4K means 4096 units of something, not a physical distance. I can still rage over the fate of values when a K of a disk storage media began to mean 1000 bytes and not 1024; never in my day, but, the logic and truth of “my day” is long gone, replaced by the intentional lies of marketeers.

I'm not sure how this is relevant to this discussion though because we are talking about physical distance in this case. Generally the only reason to have crops on these sensors is to enable higher frame rates so if they were just picking crops to gradually sacrifice spatial resolution for temporal resolution. If that's all they were doing then they could have picked 4760x 2520 to get 60 fps and maybe 3072 x 1632 to get to 80fps as a stop gap to 1080p. That's not why the S35 and S16 crops exist though, they're chosen for their physical size.

To your point about S35 digital sensors, the Cine 12K's S35 crop seems to be based off the Mini 12K's sensor size (~27.03mm wide) instead of S35 film and it has no 4:3 crop, only 3:2. If that can be reliably covered by S35 lenses then they should have gone with 4560 x 3040 and 4560 x 2416 crops for the 6K FF/PYXIS.

What I was suggesting would be a bit more conservative... but also now I'm noticing that a 4608 x 2592 crop should fit comfortably into a S35 image circle.

[... but also now I'm noticing that a 4608 x 2592 crop should fit comfortably into a S35 image circle.

Mark, my outrage is always a little tongue in cheek as occasionally I try to “lighten” the discussion rather than “enlighten.”

If 4608 x 2582 fits Super 35 image circles then that’s a good suggestion. I just haven’t checked the math as dinner is almost ready in a few minutes.

Mark, my outrage is always a little tongue in cheek as occasionally I try to “lighten” the discussion rather than “enlighten."

My apologies. Couldn't read the tone lol

If 4608 x 2582 fits Super 35 image circles then that’s a good suggestion. I just haven’t checked the math as dinner is almost ready in a few minutes.

I just re-checked the numbers and the Ursa Cine 12K's S35 crops seem to somewhat take advantage of the image circle being a circle as oppose to limiting the crop to the size of S35 film. Check them out

Code: Select all

9408 x 6264 (9K 3:2)
8688 x 4896 (9K 16:9)
9312 x 4896 (9K 17:9)
9312 x 3864 (9K 2.4:1)
7680 x 6408 (9K 6:5)


Obviously it's 16:9 and 2.4:1 crops are sub-crops of the 17:9 crop. Not exactly sure why they did that though because, if you look at the 3:2 crop, you can see it's the widest off all of them and the 6:5 crop is the tallest of all them. I think since S35 film is 4:3 and none of the non-cropped resolutions have a 4:3 crop, they just took the height of what a 4:3 crop would have been and extended it's width to 3:2. So technically it's too big but anyone who truly needs to shoot 4:3 S35 can just use guides within the 3:2 crop. Obviously the 6:5 crop has the most height because it's also taking advantage of having access to the whole image circle... like this

image_circle.png (117.47 KiB) Viewed 11127 times

So if the 6K FF/Pyxis had some Super35 and Super16 crops that took full advantage of the image circles, the crops would look like this:

Code: Select all

         3:2             16:9          17:9           2.4:1          6:5
S35   4716 x 3144     4608 x 2592   4688 x 2480    4840 x 2016   3840 x 3200
S16   2112 x 1408     2144 x 1200   2168 x 1144    2288 x  952   1690 x 1408


Some of the crops could be pushed a little further but I chose those numbers because of the frame-rates they should be able to hit. That way each of those crops would have extra utility instead of just being for S35 lenses.

So if the 6K FF/Pyxis had some Super35 and Super16 crops that took full advantage of the image circles, the crops would look like this:

Code: Select all

         3:2             16:9          17:9           2.4:1          6:5
S35   4716 x 3144     4608 x 2592   4688 x 2480    4840 x 2016   3840 x 3200
S16   2112 x 1408     2144 x 1200   2168 x 1144    2288 x  952   1690 x 1408


Super 35 lenses were designed for an image circle of 31.1mm.

16:9 4608 x 2592 on a BMCC 6K FF has an image circle of 31.47mm.

17:9 4688 x 2480 on a BMCC 6K FF has an image circle of 31.57mm.

The 4K DCI 17:9 4096 x 2160 on the BMCC 6K FF has an image circle of 27.56mm.

The classic cine lenses designed for a 1932 35mm Academy standard that is 22 mm x 16 mm has an image circle of 27.2mm. Cooke Speed Panchro Series II and III lenses made for this standard can be bought used for under $3,000 USD. Modern Cooke lenses based on the Panchro formula sell for much more.

If you crop to 4K UHD 16:9 3840 x 2160 on the BMCC 6K FF, it has an image circle of 26.23mm.

16:9 4608 x 2592 BMCC 6K FF
36 x (4608/6048) = 36 x .7619 = 27.43
27.43 x 27.43 = 752.33
24 x (2592/4032) = 24 x .6429 = 15.43
15.43 x 15.43 = 238.04
752.33 + 238.04 = √990.37 = 31.47 image circle

17:9 4688 x 2480 BMCC 6K FF
36 x (4688/6048) = 36 x .7751 = 27.9
27.9 x 27.9 = 778.68
24x(2480/4032)= 24 x .6151 = 14.76
14.76 x 14.76 = 217.91
778.68 + 217.91 = √996.59 = 31.57 image circle

4K DCI 17:9 4096x2160 BMCC 6K FF
36x(4096/6048)= 36 x .6772 = 24.38
24.38 x 24.38 = 594.43
24x(2160/4032)= 24 x .5357 = 12.86
12.86 x 12.86 = 165.31
594.43 + 165.31 = 759.74
√759.74 = 27.56 mm image circle

4K UHD 16:9 3840x2160 BMCC 6K FF
36 x (3840/6048) = 36 x .6349 = 22.86
22.86 x 22.86 = 522.45
24x(2160/4032)= 24 x .5357 = 12.86
12.86 x 12.86 = 165.31
522.45 + 165.31 = 687.76
√687.76 = 26.23 mm image circle

Super 35 lenses were designed for an image circle of 31.1mm.

16:9 4608 x 2592 on a BMCC 6K FF has an image circle of 31.47mm.

17:9 4688 x 2480 on a BMCC 6K FF has an image circle of 31.57mm.

The 4K DCI 17:9 4096 x 2160 on the BMCC 6K FF has an image circle of 27.56mm.

The classic cine lenses designed for a 1932 35mm Academy standard that is 22 mm x 16 mm has an image circle of 27.2mm. Cooke Speed Panchro Series II and III lenses made for this standard can be bought used for under $3,000 USD. Modern Cooke lenses based on the Panchro formula sell for much more.

If you crop to 4K UHD 16:9 3840 x 2160 on the BMCC 6K FF, it has an image circle of 26.23mm.

16:9 4608 x 2592 BMCC 6K FF
36 x (4608/6048) = 36 x .7619 = 27.43
27.43 x 27.43 = 752.33
24 x (2592/4032) = 24 x .6429 = 15.43
15.43 x 15.43 = 238.04
752.33 + 238.04 = √990.37 = 31.47 image circle

17:9 4688 x 2480 BMCC 6K FF
36 x (4688/6048) = 36 x .7751 = 27.9
27.9 x 27.9 = 778.68
24x(2480/4032)= 24 x .6151 = 14.76
14.76 x 14.76 = 217.91
778.68 + 217.91 = √996.59 = 31.57 image circle

Thank you for providing the math! Looks like I was off by a little bit however your numbers are a little off, too. The sensor's full 36 x 24mm area is actually 6064 x 4040, BMD just uses 6048 x 4032 of it though. So 4608 x 2592 would be 31.39 instead of 31.47, for example.

So the revised numbers with the image circles being rounded to 31.1mm would be along with the frame rates these crops would be able to reach:

Code: Select all

         3:2             16:9          17:9           2.4:1          6:5
S35   4716 x 3144     4568 x 2568   4632 x 2448    4840 x 2016   4024 x 3352
          50              60            60             70            46

S16   2112 x 1408     2144 x 1200   2172 x 1152    2272 x  944   1888 x 1568
         100              110           120            144           90


Filmmaker and content creator Geoff Fagien (Florida, USA), with a lot of experience with Sony mirrorless and RED Komodo OG, recently acquired a BM 6K FF and he is putting some good content out there. He is quite articulated and pedagogical providing a good service to the community, specially those starting with filmmaking and content creation.

In this comparison, you can see that the BM 6K FF has more "3D pop" than the Komodo, albeit he was using different lenses (both equivalent Sigmas though). My guess is that this better 3D pop has to do with the 6K FF having more final resolution, but specially better color separation and tonality, and better highlight roll-off (this comes more clear on the outdoor scenes with the park behind). This is not a scientific comparison but it's consistent with many other comparisons of these two raw cameras.

"Sigma 28-45 vs 18-35 1.8 ART Lens. What's The Difference? Shot on Blackmagic 6k Full Frame & RED Komodo"

In this comparison, you can see that the BM 6K FF has more "3D pop" than the Komodo, albeit he was using different lenses (both equivalent Sigmas though). My guess is that this better 3D pop has to do with the 6K FF having more final resolution, but specially better color separation and tonality, and better highlight roll-off

I don't mean to be pedantic here, but the video you linked to, shows no such thing, to me at any rate; there is only a brief couple of shots from either, and they look pretty similar to me. I still don't understand what '3D Pop' actually means and it has an urban placebo ring to me. With the right combination of skills; lenses and grading almost any camera can achieve an impressive punch, in the output these days; and other factors make your choice more important; to me anyway, that is.

I don't mean to be pedantic here, but the video you linked to, shows no such thing; there is only a brief couple of shots from either, and they look pretty similar to me. I still don't understand what '3D Pop' actually means and it has an urban placebo ring to me. With the right combination of skills; lenses and grading almost any camera can achieve an impressive punch, in the output these days; and other factors make your choice more important; to me anyway.

I agree completely. "3D pop" is one of those terms that starting to frustrate me more and more because it means so little. It's not quite as bad as people describing things as "cinematic" or "digital looking" though. I've seen people compare two cameras or even two lenses and say one looks "more digital" and the other looks "more organic". What's that supposed to mean? Both cameras are digital and no lens optics are digital. Just now I was scrolling through YouTube and saw a video about "cinematic settings" on a camera and every accessory is supposed to make your shots more "cinematic". It's all so annoying.

I agree completely. "3D pop" is one of those terms that starting to frustrate me more and more because it means so little.

I agree, but it's an easy concept to illustrate: it's simply a feeling of three-dimensionality, of the subject (or a portion of the subject) appearing to "pop" out of the screen as in this example below (especially her hands). It's due more to lighting than anything else, but depth of field also plays a role. I've seen arguments for images on full frame having more "3D pop" than equivalent images on Super 35/APS-C due to the shallower depth of field that you get with the same framing on a full-frame camera, but I've seen plenty of very 3D-looking images on Super 35/APS-C or even Micro Four-Thirds as well. I think it's mostly an effect of lighting.

Threading the needle by Brad Hurley, on Flickr

For some good "3D pop" on Super 35, check out this footage by Note Suwanchote on the Ursa 12K, which also confirms that it's mostly about lighting (there's less pop in images with flatter lighting):

But I think the point is, "3D pop" is not a criteria for evaluating a camera.....

It amounts to a cult at this point, with favored terms which confer authority and membership but don't mean anything, or have no agreed upon meaning or don't mean what the speaker wants them to mean.

The offenses also include the likes of "highlight roll-off" and "tonal gradation" which *could* be meaningful but which aren't in context of poorly controlled camera comparisons and with unknown post practices. But they sure sound authoritative!

And speaking of "pedantic", how about this objection? Parties who actually don't actually know anything about cinema using the word "cinematic"....?

But I think the point is, "3D pop" is not a criteria for evaluating a camera.....

Of course it's not, it's not even a criterion for evaluating lenses. But I don't think that means "3d pop" needs to be banished as a term, because it's an effect that anyone can see with their own eyes. Like many things, its causes are misunderstood because everyone's instinctual response seems to be, "wow, that's amazing, what [lens or camera] did you use to achieve that effect?" As if you could achieve that effect simply by buying that camera and/or lens.

"Cinematic" is much more vague and impossible to pin down, and is a useless term.

Brad, you illustrate 'it' well; but as I say, and you concede, it can be achieved quite easily across a range of cameras; given the right mojo, each time. Ultimately therefore it's a pretty meaningless term when applied to a specific camera; and I would warrant especially this one. I have seen some pretty impressive images from this iteration of BMD sensors, but so then with it's forbears too. It falls into the esoteric, for me, like the 'special' quality that FF imbues, you know?

Ultimately therefore it's a pretty meaningless term when applied to a specific camera;

Yes, that's my point: it's not a term that should be applied to cameras or lenses, but to images (and possibly to lighting). It's possible to demonstrate that an image has 3D pop, but most people attribute that quality to the wrong things, or at least to the ingredients that matter least. I was just arguing that the term "3D pop" on its own is valid, it just shouldn't be described as a quality of certain cameras or lenses.

The offenses also include the likes of "highlight roll-off" and "tonal gradation" which *could* be meaningful but which aren't in context of poorly controlled camera comparisons and with unknown post practices. But they sure sound authoritative!

I was gonna mention "highlight rolloff", too! I saw someone refer to one camera doing it better than another... when comparing log footage!

And speaking of "pedantic", how about this objection? Parties who actually don't actually know anything about cinema using the word "cinematic"....?

I really think it has to do with a lot of people not realizing that the things that make movies feel like movies can't be obtained by applying a LUT, using a gimbal, or using a certain type of lens.

They really need to see 28 Days Later or Pieces of April and then see other things shot on the XL1 and PD150P.

But I think the point is, "3D pop" is not a criteria for evaluating a camera.....

Of course it's not, it's not even a criterion for evaluating lenses. But I don't think that means "3d pop" needs to be banished as a term.

I agree. The concept just needs to be demystified which will cause the term to be thrown around less on Youtube.