Can you guess which camera was used, Alexa, Red, Sony, etc.?

G

Gillvane

Veteran
All 4k cameras.
Shots from all these cameras back to back in a random order, and you have to guess which is which, or simply decide which one you like best, least.



Answers are here:

http://nofilmschool.com/2014/08/put-eyes-test-can-tell-difference-todays-top-cinema-cameras


 
Last edited:
Rapidly converging to "near enough as makes no difference." I didn't bother to make exact guesses, only noted those by which I was less impressed in the graded sample. Those turned out to be the BM4K, 1DC, 5D3, and BMPCC w/o booster... which is odd, because I thought I really liked BM camera images in general... and I know I've liked 1DC and 5D3 images in other works... so yeah.
 
I wish they didn't reveal the cams so soon. It would've been better if they allowed a week where people could go on record with their guesses:)

I couldn't guess which cam was which, honestly. Plus the framing differed, so f.ex. it wouldn't be fair to judge which camera handled the big white blowout on the left upper corner better.

Regardless, I made a list on paper and put down which I liked and which I didn't. I liked the Alexa best - but the framing was really close on the woman, so can't say how fair it was, plus I hated the next one immediately the Kineraw Mini, so maybe the Alexa gained by contrast. I wasn't impressed by Dragon, but there were some focus problems on that sample. Wasn't a fan of the Sony F55, but liked the BMPC 4K and liked the BPCC even better (which is bizarre), and liked the Sony FS700, so go figure. I also liked the GH4, which surprised me a little given the much better cameras it was competing against.

Honestly, all of them are "good enough" in this test - maybe under more challenging conditions, their various strengths and weaknesses would come out, but in this static test, well, all were "good enough".
 
mcbob said:
Rapidly converging to "near enough as makes no difference." I didn't bother to make exact guesses, only noted those by which I was less impressed in the graded sample. Those turned out to be the BM4K, 1DC, 5D3, and BMPCC w/o booster... which is odd, because I thought I really liked BM camera images in general... and I know I've liked 1DC and 5D3 images in other works... so yeah.
Click to expand...
Yes, it seems to me any of these cameras make an acceptable image. I think it's going to start being more about features and cost. Red can make 6k, 8k, 10k, etc., but I believe we have reached the eyeball limit. Our eyeballs can't see more detail so more "k" is kind of pointless now.
 
The problem with these kinds of tests is that the lighting is so controlled. The dynamic range could have been five stops, for all we know. Try outside at high noon.

Cameras are so very close, especially after grading. I wince when people talk about highlight roll-off, color science, and mojo. Record it to ProRes or raw, and I can get you whatever mojo you want.

The important differences these days are: global shutter, sensitivity, dynamic range, and ergonomics.
 
Last edited:
mcbob said:
Rapidly converging to "near enough as makes no difference." I didn't bother to make exact guesses, only noted those by which I was less impressed in the graded sample. Those turned out to be the BM4K, 1DC, 5D3, and BMPCC w/o booster... which is odd, because I thought I really liked BM camera images in general... and I know I've liked 1DC and 5D3 images in other works... so yeah.
Click to expand...

I can only speak for the 1Dc footage since I know that camera very well: their "ungraded footage" is not Canon Log as it comes out of the camera. Not only did they mess up the white balance, but the Log footage is no where near that level of saturation. Somebody messed with the settings and they made a mistake simple as that. It happens.

The 1Dc should look like the C500, but with a little less resolved detail. They are both shooting Canon Log. To bad they included the footage anyway. For the less informed, this video now represents what the 1Dc looks like.
 
AndreeOnline said:
I can only speak for the 1Dc footage since I know that camera very well: their "ungraded footage" is not Canon Log as it comes out of the camera. Not only did they mess up the white balance, but the Log footage is no where near that level of saturation. Somebody messed with the settings and they made a mistake simple as that. It happens.

The 1Dc should look like the C500, but with a little less resolved detail. They are both shooting Canon Log. To bad they included the footage anyway. For the less informed, this video now represents what the 1Dc looks like.
Click to expand...
I rather liked the 1DC in this test. But yeah, they could have done a bit better with it.
 
James0b57 said:
I rather liked the 1DC in this test. But yeah, they could have done a bit better with it.
Click to expand...

I am specifically concerned by the way the red... was it a pillow or something... to the left broke under grading. I don't ever see that. I also feel the "non graded image" was way too saturated for default Log. But some under exposure and "wrong" (compared to the others) WB won't help there...

Anyway. Doesn't really matter. Those who NEED to know what a camera can or can't do always find good information.
 
SPOILER ALERT!

This post shows cameras used in order of appearance.


I love Alexa as much as the next guy, but my observations were:

1. Alexa
Liked the image, but didn't really vibe with the red hue on the lipstick and the circular object in the background. Not knowing what the scene looks like in real life, I cannot really say if this is desired or not. And after seeing all of the other cameras, I realized it was more the color of the lipstick and BG object that I didn't like.

2. Kine Raw
Moire, but was most likely a scaling issue. Still, the moire dominated my opinion, so I didn't look at the other aspects. Seemed to be slightly more noisy in the shadows? After a second look, it seemed nice, but I still might prefer an Epic/Scarlet MX.

3. Red Dragon
Simply liked it

4. Canon C500
Liked it, but felt it had cooler colors, may have been a white balance thing, may just be the C500 look.

5. Sony F55
Didn't find any fault, but due to compression or some other factor, this one seemed to have more blockiness to it. And that distracted me from other important factors. (After looking at the image in the downloaded version, the blockiness was gone. I used to think I didn't care one way or the other for the f55 image, but seeing it amongst all of these cameras, I have a new appreciation for the way it renders. )

6. BMC 4K
Weird golden haze. Didn't like it, but could be an aesthetic for someone else.

7. Red Epic
liked it (though, i wasn't crazy about the color in her cheeks when compared directly to Alexa and Dragon. The 1DC was nice, but looked 8bit was sneaking up, and the C500 was just a touch cool, while pretty, didn't look as pleasing as the warm glow skin tones from the Alexa footage. Again, probably grading issues. )

8. Canon 1DC
Gradient Banding on the circular object in the BG. I otherwise liked it. Especially after the first watch and later comparing. It renders colors nicely, but too bad about the gradient banding. 8bit and Jello are the main drawbacks of this camera, but it has form factor, small batteries, and lowlight going for it.

9. Canon 5d3
odd haze (more a symptom of grading?)

10. BPCC
Reds were too Magenta and shirt was very green (trade mark black magic pocket, imho)

11. GH4
Nice, but low saturation and the highlights change color (see red circular object in BG goes from reddish to yellowish). After adding some saturation back to the image, I didn't like it so much. But for the price... compared to the bpcc footage it was sandwich between, it was lent a bogus sense of great color, but when later compared to leading cameras, it more obviously falters a bit. (I do wonder if the gradient color change is due to the angle of the camera to the reflected light? The FS700 exhibited the same phenom)
GH4_cameracompare.jpg
I am still tempted to get a GH4

12. BPCC
same as 10. Magenta and Green combos. Not entirely unpleasing, and a usable aesthetic for some.

13. Sony FS700
highlights change on circular object from reddish to orange (similar to the GH4). Otherwise neutral usable image.
 
Last edited:
AndreeOnline said:
I am specifically concerned by the way the red... was it a pillow or something... to the left broke under grading. I don't ever see that. I also feel the "non graded image" was way too saturated for default Log. But some under exposure and "wrong" (compared to the others) WB won't help there...

Anyway. Doesn't really matter. Those who NEED to know what a camera can or can't do always find good information.
Click to expand...
After watching the comments section, I think there was a definite 'raw' bias to the whole workflow. So, a baked in 8bit image is going to suffer under these constraints. Whereas someone familiar with a 1DC would shoot differently, compensating for the fact that changes in post could have bad effects on the footage.

The colorist graded all of the cameras to match the skin tone of the Alexa as close as possible, and then the rest of the image was what it was. So each camera had different side effects because of that. They did not give each camera a custom grade independent of their own.
 
Well, watching it in a vimeo compressed presentation doesn't show any of their strong sides. Also agree, fairly controlled environment and not the best situation to test the advantages of each sensor. However, what this proves is, for Web delivery, any of these, if properly lit, creates a beautiful picture!
 
SPZ said:
Well, watching it in a vimeo compressed presentation doesn't show any of their strong sides. Also agree, fairly controlled environment and not the best situation to test the advantages of each sensor. However, what this proves is, for Web delivery, any of these, if properly lit, creates a beautiful picture!
Click to expand...
Without disagreeing with your point above, it also shows how better cameras still look better, even when compressed and on the web, even in a blind test.
 
AndreeOnline said:
Do you know this for a fact, James? It would explain a lot.
Click to expand...
I do know that for a fact, because they have a video with the colorist explaining their methods for getting the grade they have. They matched all of the other cameras' skin tone to the Alexa. Everything else was left to be what it was basically. So the other cameras may have nice skin, but look weird in other parts.

And to further explain what I meant about the grades, is that the other cameras were no made to look in the way they might excel, but were matched to the Alexa's skin tone.

https://vimeo.com/104924908
 
James0b57 said:
I do know that for a fact, because they have a video with the colorist explaining their methods for getting the grade they have. They matched all of the other cameras' skin tone to the Alexa. Everything else was left to be what it was basically. So the other cameras may have nice skin, but look weird in other parts.
Click to expand...

You're always going to get bashed for doing these type of tests, and I feel somewhat if your process is flawed - then you go into it knowing that's going to occur. It was pretty obvious to me they'd all been grading to "one look". What this test tells us I think most of us know - know the camera you're shooting on through your own tests, then shoot with it to get best results. These tests don't incorporate that - Zacuto gave it a good shot by allowing operators that were familiar with each system.
 
Noel Evans said:
You're always going to get bashed for doing these type of tests, and I feel somewhat if your process is flawed - then you go into it knowing that's going to occur. It was pretty obvious to me they'd all been grading to "one look". What this test tells us I think most of us know - know the camera you're shooting on through your own tests, then shoot with it to get best results. These tests don't incorporate that - Zacuto gave it a good shot by allowing operators that were familiar with each system.
Click to expand...

Well said. I get to be around these cameras, but not all of them at one time. These are great supplemental camera comparisons.
 
You must log in or register to reply here.
Back
Top