Blackmagic Forum

I did some quick test shots with my 3 BM cameras to get a little bit of an overview how they compare.
This is not a scientific test, but I tried to make it as comparable as I could in a short amount of time.

All 3 cameras where set to ISO400, 90° shutter and f/4.0, Schneider Platinum 1.2 IRND 4x5.65 filter. ProRes HQ 1080p on BMCC and ProRes HQ UHD on PCC4k and UM46k. Unfortunately I had to use a different lens on each and tried to fit the FOV but it was not totally possible. Sorry for that.
White balance was made in sunlight on the gray patch of the Colorchecker Passport (bought 2018).

In post (Premiere Pro) I have further tweaked the white balance for each camera to level out any problems and on top of that applied the same RGB curve + Saturation batch settings to each camera. After that I did some little tweaks to get the exposure into the same ballpark.

So here we go first with the Colorchecker color chart - for one camera I forgot to set the shutter to 90° for this particular shot and left it at 180° so got overexposed and a bit clipped in the highlights and near white patches. Sorry for that. That particular shot has been tweaked down in post to have similar exposure than the rest.

Scene #3

Scene #04

Maybe my eye isn't trained enough, but this is tougher to work out than I thought it would be. I thought the BMCC would stick out a little more at least.

Camera 1, I'm pretty convinced is the UM4.6K. It's handling the light on that grey roof(?) in scene 4 a lot better than the others, so I'm chalking that down to the extra dynamic range.

Camera 2 seems closer to camera 1 in terms of colour than camera 3, and it's holding more detail than camera 3 in the leaves in scene 4, so I'm guessing that's the PCC4K with its extra resolution over the BMCC.

So I guess that makes camera 3 the BMCC.

Looking forward to finding out how wrong I am, haha.

Funny thing is - I changed with every scene which camera # is which camera

Robert Niessner wrote:Funny thing is - I changed with every scene which camera # is which camera

Ah!

Consider my estimations based on just scene 4 then.

In first and second post I thought:
Cam 1 bmmc
Cam 2 ump
Can 3 bmpcc4k
The third post was difficult to guess

In shots with sky, it's not hard to identify the pics with the most apparent DR, but any chance you could post the footage?

John Paines wrote:In shots with sky, it's not hard to identify the pics with the most apparent DR, but any chance you could post the footage?

Be careful about your assumptions, you might be surprised.

I will post the sample footage later, yes.

I'd also like to do a more controlled test with human skin under lighting, but for that I need some more time.

Robert Niessner wrote:Scene #3

It's a trick, 1 and 3 are BM. Number 2 is a Sony .

Man, it so great you did these tests!
I'm really impressed how good you managed to make the BMCC prores come close to the others.
I'm writing as I'm following the samples, so I will write progressively according to my findings.
(reffering camera numbers as to those in the filename, not the order you published)

First batch cam 1 and 3 has better highlight roll of than the cam 2. Cant tell which is which but I like the image from cam 1 and cam 3 better than cam 2.

Second batch: I like the image from cam 3, its pleasant and filmic, but the obvious moiré tells me its the BMCC. It doesnt have the blueish/purpleish tint UM46k has. It handled much better the yellow building and its very bright windows. Between cam 2 and cam 1 they are really close but cam 1 seems to have more latitude, so I'd say its the UM46k. I preffer the colors on cam 1 over the cam 2, but from those 3 pics, the cam 3 is the nicest, more pleasing image. If wasnt that moiré... Damn! And the lens really helped.

Third batch: Cam 3 has the worst highlight roll off, maybe if you didnt need to match exposure settings with the others, a different exposure could make it better. Cam 2 and Cam 1 are close with the cam 1 looking to have better DR and resolving much more detail, so I give it to the UM46K. I preffer the look of cam 1 here . Cam 3 is the BMCC and cam 2 the P4k. Cam 2 is resolving more detail but cam 1 has a more dreammy look on the foliage close to the camera.

Have I hit something or too far off????

Well, it's very late Robert, I don't know exactly what this is, and only have my phone here to look at them and In watching netflix. So, I'll have a go.

1st colour chart. Bmcc, p4k more colour artifacts from lack of downscaling, m4.6k shallower DOF less colour artifacts from downscaling.

Can you frame them the same with the same dof and zoom to give the same image size in the background? Please also use BRaw where available, and lowlight?

2nd roof tops. 4.6k, p4k, bmcc. The aliasing. The colour noise in the trees if the first and last. The increased dynamic range in first. Better blue in sky from the Sony, but much bright rendering of the green blue roof correlation. The BMCC down the bottom might look (ironically more videoish, but may have better colours going by those trees (unless it ready were a more yellow shade).

3rd. Pocket 4k, BMCC, 4.6k. I'm torn here, because the middle one has the look of the last pocket sample and vice versa. I smell trickery here +and not just because I would do it that way too). But my eyes are going fuzzy on these small phone web page pictures. I suppose they are all going be the same camera or something (something I would want to do).

Did I get any right?

Now we have seen enough of your neighbours washing Robert, can we have their names and addresses?

It is really a bad comparison, they are not framed the same. Interesting. As an old hand I'm so used to seeing yellow hue in Sony consumer cameras.

Yep, something cooky about that third batch. I looked at hight before burn out and indeed the middle picture does well. But I notice a lot of light shift between some shots, which affect the amount of light reflected. But then, I am assuming the tester has used the exact same exposure, but nothing else is really right, and I forgot to mention the different lenses are going affect the outcome a lot, showing you lens image differences of how it handles what each lens gives, rather than how it handels the same image.

I attached the screenshots in the order from 1 to 3, but the forum software put it in from 3-1.

Wayne, please add the cam numbers to your guessing so it is easier to see the order you took.
BRAW is not available for any of my cameras yet.

I tried to match the framing but it was so bright outside I couldn't see much on the UM46k and nothing on the BMCC, while the PCC4K with its bright screen held well.

Once I can do more controlled test in a room with some lights setup I'll try to have less variables.

But from your guessings here I can see that the cameras deliver a more similar looking image than most of you thought.

Robert Niessner wrote:Once I can do more controlled test in a room with some lights setup I'll try to have less variables.

Looking forward to a controlled skin-tone set-up!

Robert, when you retest, please keep the label on the same camera. We’re trying to learn here and we don’t judge by seeing one picture in isolation. We want to say, after looking at three “A” images, that it’s the Pocket for example. The way you’ve labelled the cameras, well it could drive a man to drink. Or do a spoof of the “Saw” movies where each time we pick the wrong camera, one of our children loses a limb to a sadist like you!

rick.lang wrote:Robert, when you retest, please keep the label on the same camera. We’re trying to learn here and we don’t judge by seeing one picture in isolation. We want to say, after looking a three “A” images, that it’s the Pocket for example. The way you’ve labelled the cameras, well it could drive a man to drink. Or do a spoof of the “Saw” movies where each time we pick the wrong camera, one of our children lose a limb to a sadist like you!

rick.lang wrote:Robert, when you retest, please keep the label on the same camera. We’re trying to learn here and we don’t judge by seeing one picture in isolation. We want to say, after looking at three “A” images, that it’s the Pocket for example. The way you’ve labelled the cameras, well it could drive a man to drink. Or do a spoof of the “Saw” movies where each time we pick the wrong camera, one of our children loses a limb to a sadist like you!

Or you can take it as a more interesting lesson: why people should stop worrying so much about their cameras and get their booties out there making movies!

Or just not done so well.

Robert, I was being facetious, but then again, you being German might not get that. BRaw and lowlight would make everything too essier, as only one does BRaw which can be picked, and one camera is realy good in low light.

I did pick them in order, first second third for each batch, as displayed on the forum.

But Robert, simple differences can make stops of difference in the result, exasberbate or reduce Bayer pattern problems, and add mis-colourisation. Now of that is an even comparison (now my German heritage is coming out). So, this 'noise' is what makes it more difficult. Better with ungraded raw frames (lol ). ProRes tries to grade, but if all are as good, you should be able to always grade them perfectly together, should you not. But I don't think it's a fair comparison, particularly the last scene, where we may or may not be seeing something depending on effects of individual setup (plus ProRes clips). So, could I suggest using strong sun light on a complex scene, object, human face, to reveal highlights, and mid day sun. I've got at least a stop difference in light here compared to those circumstances, plus a number of stops depending on which primary colour chanel. So, these photos don't give much of a real life represention of performance here (though the 4.6k dues look better and more natural mainly). Anybody got the Sun directly overhead at the moment to do shots?

Ok let me take a stab at it. In your first image post
1. URSA mini 4.6k
2. BMCC
3.Pocket camera 4k

2nd image post

1 URSA mini 4.6k
2 Pocket camera 4k
3 BMCC

3rd image post
1 BMCC
2 Pocket camera 4k
3 URSA mini 4.6k

My take.
1.UM46K
2.PCC4K
3.BMCC

1.BMCC
2.UM46K
3.PCC4K

1.PCC4K
2.BMCC
3.UM46K

Now pop the champagne!!!

Ok. Just read the first post. One ISO, I might say what's that balls of string? But hopefully these are pro cameras without much hankypanky on coming up with iso, except the Sony does have gain, and doesn't the iso reduce latitude differently? Next, yeah playing around and tweaking, I thought those clips looked too sudden. Makes it a little hard. What's the native iso on the Sony? It all seems to confuse things a bit. I mean, you know greater amounts of a bright object in a frame compared to another gives impression of greater brightness? So, these sorts of frames not being the same makes visual comparison difficult. If you want to to judge cameras on impression, by the same frame or entire same movie. I get annoyed when the angles slightly different on a test picture because they put all the cameras on one bar. If you wanted to do test of long GOP consumer cameras throwing them out of a helicopter on bungee cords might work too. Optimal results of the exact same scenes and ungraded are two ways. One camera does stick out (but maybe it is the setup and adjustments). But, if one is trying to make them look similar, could the camera be even better? The clips don't look 15 stop. Hold it, what sort of jpeg are we looking at and bit depth? How many stops is sacrificed in ProRes (they are all 10bit)?

Having used none of these cameras (CAM #'s are based on filename)...
Post One - Color Charts
CAM01-PCC4K
CAM02-BMCC
CAM03-UM46K

Post Two - Buildings (Scene 3)
CAM01-BMCC
CAM02-PCC4K
CAM03-UM46K

Post Three - Buildings (Scene 4)
CAM01-UM46K
CAM02-PCC4K
CAM03-BMCC

Are we still guessing. Where does it mention free champagne for the winner?

Wayne Steven wrote:Are we still guessing. Where does it mention free champagne for the winner?

Yeah! Come on, Robert! Stop teasing us! Some of us have real problems with waiting/anxiety.

Ok so what’s the answer Robert ?

Champagne Robert?

my guess in order of the naming:

02-Cam01_CC…UM46K
02-Cam02_CC…PCC4K
02-Cam03_CC…BMCC

03-Cam01_CC…UM46K
03-Cam02_CC…PCC4K
03-Cam03_CC…BMCC

04-Cam01_CC…UM46K
04-Cam02_CC…BMCC
04-Cam03_CC…PCC4K

Sorry for the delay, here is the answer:

02
Cam01 = UM46k
Cam02 = PCC4k
Cam03 = BMCC

03
Cam01 = BMCC
Cam02 = UM46k
Cam03 = PCC4k

04
Cam01 = PCC4k
Cam02 = BMCC
Cam03 = UM46k

Thanks Robert, the yellow cast of the Sony got me, changes the appearance of contrast, plus clouds move and if shots aren't at exactly the same time you don't know what you are looking at. It's about how a camera handles exactly the same challenging scene portrayed exactly the same way, and see how each fail. But it's ok for late late night fuzzy eyes on an small uncalibrated phone screen.