Camera shootout 2011

[Eric Z]

Here's some things I wanted to highlight when the RED One MX was compared to the Alexa:


Click the image for the large version.

To me it seems that the RED One MX is better even here - in the highlights test.
The only thing that's going in favor of the Alexa is color saturation, but that could be just the grading.

 

[Paul Duvilla]

I noticed the light on the couch missing too. It was there for every other shot except Alexa. And also the shadows playing off the arm of the couch. Anybody have a clue???

 

[Hans von Sonntag]

Here's some things I wanted to highlight when the RED One MX was compared to the Alexa:


Click the image for the large version.

To me it seems that the RED One MX is better even here - in the highlights test.
The only thing that's going in favor of the Alexa is color saturation, but that could be just the grading.

This side by side example is colour corrected. Someone altered the original files. Not good for a qualified comparison. However, it shows that both cameras are in the same ballpark regarding the intended mood of the this scene.

Hans

 

[Eric Z]

Hans,

This is a frame grab directly from the video on Zacuto's website. I didn't alter anything except adding my comments on it.

 

[Hans von Sonntag]

Hans,

This is a frame grab directly from the video on Zacuto's website. I didn't alter anything except adding my comments on it.

Haha. Eric, I never thought that you touched the picture. The test crew did. Such tests are more or less meaningless, IMHO.

Hans

 

[Eki Halkka]

As an aside, I think one reason why some DPs may have a diminished opinion of dynamic range on the MX sensor is that availing yourself of the maximum DR requires you to dig very deep into the shadows, which is pretty counterintuitive for any DP who cut his teeth on film negative.

This is my impression too...

 

[Peter Dmitriyev]

That's a good pic. I noticed those things too and allow me to add a few of my own as well

notice the highlights on everything including the woman, her hair, arm, the shirt etc. Everything is under control with the MX but is blowing out on the Alexa yet they expect us to believe this camera on the left has 3 stops higher DR than the one on the right. If I was ignorant and just looked at this footage I'd say the camera on the right has far more DR.

 

[Peter Dmitriyev]

All I can find is this comparison between Epic and Alexa:
http://reduser.net/forum/showthread.php?47529-EPIC-and-Alexa...

I'm seeing 13 to 13.5 stops on the Epic there. I would find it hard to believe that a camera that uses the same exact sensor has 2 stops or more dynamic range over its predecessor (Red MX) which is why it makes more sense that MX has 12.5 to 13 stops or more, not 11 and change like Zacuto's test found.

 

[Martin Stevens]

Not to mention that the test scene is not particularly high contrast.

 

[Sanjin Jukic]

Now it's a right time that I jump in this discussion.

First thanks to to Steve Weiss from Zacuto who made possible and produced this great camera shootout test that involved so many respectful professionals from the industry…

I would like to comment mostly on RED MX, Alexa and Kodak film stock 5219.

For my point of view the winner is RED MX that I was already taking here almost a year and half ago during the time of Alexa introduction.

If you would like to know more about it then search this forum…

So let's talk about the facts that can be found on the first part of the comprehensive great Camera Shootout 2011 tests.

Dynamic range of Alexa, REd MX and Kodak 5219?

Click on the each image to see it at the full size of 1080p, all taken from the Zacuto's GCSO 2011 Vimeo video.


Latitude stops of all cameras involved in Zacuto's GCSO 2011.



Alexa has exposure latitude degree about 12.6 T stops of dynamic range values.

Somebody made a mistake and wrote that Alexa has exposure latitude degree about 14.1 T stops of dynamic range values.

That's obviously not a true just because at the picture above from Arri DRTC can be seen only about 12.6 T stops.

It is a picture taken from ARRI Dynamic Range Test Chart that was supervised by Michael Bravin.

Maybe somebody whated to say that Alexa has 14.1 T stops of Potentially Acceptable Extractable Latitude but that totally differs from 12.6 T stops of Useable Extractable Latitude.



RED MX has exposure latitude degree about 11.9 T stops of dynamic range values.



Kodak 5219 has exposure latitude degree about 14.5 T stops of dynamic range values.

That means this Great Camera Shootout 2011 told us something we all knew from before and Jim Jannard stated here at the forum that Kodak stock 5219 is still unbeatable by DR,

also he said that Alexa has about 0.5 to 1.0 T stops more DR than RED MX that depends on how you could do your test.

Let's see some of the real world high and low key comparison images between Kodak 5219, Alexa and RED MX taken from this great comparative test.

LOW KEY comparisons:





HIGH KEY comparisons:





Looking forward to see two other parts of the test during this summer.

 

[Roberto Lequeux]

It isn't the end of the world, but it is a shame that the lighting is different on a comparison.

I'm seeing 13 to 13.5 stops on the Epic there. I would find it hard to believe that a camera that uses the same exact sensor has 2 stops or more dynamic range over its predecessor (Red MX) which is why it makes more sense that MX has 12.5 to 13 stops or more, not 11 and change like Zacuto's test found.

I see 12.5 solid usable stops. More to be dug out with good grading. Add some careful noise reduction and you should be able to get a good ~13.5 for 2k. 0.02

 

[David Diperstein]

The lamp also seems to lack a horizontal line through the center in the low key comparison specifically in the Alexa frame. What's up with that?

 

[Roberto Lequeux]

Actually, the lamp itself was moved on the Alexa frame... lol, don'cha love conspiracy theories?

 

[Peter Dmitriyev]

The lamp also seems to lack a horizontal line through the center in the low key comparison specifically in the Alexa frame. What's up with that?

If you look at the legs of the lamp on which it is standing on the table, they are completely in different position in relation to how the legs are situated in the MX and Film frames. This means that for some bizarre reason things were moved around during the Alexa shoot of that scene.

This adds to the strangeness/mystery because in the high contrast window scene there is also evidence that the Alexa had things moved around. For instance the missing highlight on the couch that is present in all the other cameras but not in the alexa. Why did Alexa get its scenes shot differently both times? There seems to be something going on here

 

[Mike Van Laanen]

Thanks Sanjin for the post. Is it just my eye or does there seem to be a stark difference between the Arri Alexa image and that of the Red MX and the Kodak 5219. This is more of a question of the "look" of the image rather than the DR discussion in this thread. Sorry for a brief hijack of the thread. When I watched my initial Red One images the first thought that popped in my head was how much the Red looked just like a film image - not digital. Do you find that this is true and why does it?

 

[Martin Stevens]

Thanks Sanjin for the post. Is it just my eye or does there seem to be a stark difference between the Arri Alexa image and that of the Red MX and the Kodak 5219. This is more of a question of the "look" of the image rather than the DR discussion in this thread. Sorry for a brief hijack of the thread. When I watched my initial Red One images the first thought that popped in my head was how much the Red looked just like a film image - not digital. Do you find that this is true and why does it?

I believe this is due to the Red having zero sharpening (?), film type gamma and a wide, film-like dynamic range.

 

[Peter Dmitriyev]

I believe this is due to the Red having zero sharpening (?), film type gamma and a wide, film-like dynamic range.

And how about the biggest thing: Red is the only camera on the market that comes close to the resolution of film which is far above and beyond all other cameras including Alexa, F35, etc, etc

 

[Sanjin Jukic]

Thanks Sanjin for the post. Is it just my eye or does there seem to be a stark difference between the Arri Alexa image and that of the Red MX and the Kodak 5219. This is more of a question of the "look" of the image rather than the DR discussion in this thread. Sorry for a brief hijack of the thread. When I watched my initial Red One images the first thought that popped in my head was how much the Red looked just like a film image - not digital. Do you find that this is true and why does it?

Mike,

certainly that RED MX looks like a film if you exposure correct that means properly and try not to clip in the highlights.

If you want to dig more in shadows then is the best to use on RED1 MX fast lens like a Leica Noctilux M 50mm f/0.95 with that you can see in available light situation

at about -9.5 T stops of under exposure latitude that is something not possible to achieve with a fast film stock or any other digital cinema camera like an Alexa or similar.

 

[Martin Stevens]

And how about the biggest thing: Red is the only camera on the market that comes close to the resolution of film which is far above and beyond all other cameras including Alexa, F35, etc, etc

Yes, but film/video/digital cinema does not have to have high resolution to look like film.
Even the images from a VHS copy of "Blade Runner" still look like film.

 

[Peter Dmitriyev]

Yes, but film/video/digital cinema does not have to have high resolution to look like film.
Even the images from a VHS copy of "Blade Runner" still look like film.

but you're mistaking something. A VHS of Blade Runner was shot on native resolution much higher (let's say around 4k film resolution) and then subsequently down rezzed through the various DI methods to result in that final VHS resolution which itself might be 640x480 or whatever. But as we all know something shot on much higher resolution but then down sampled, looks far better than something shot at that native resolution. This is why Red footage shot at 4k down sampled to 2k still looks much sharper than let's say Alexa or DSLR footage that is near 2k.

Similarly a 640x480 VHS version of Blade Runner originally shot/scanned at 4k 35mm film, will look far better than a 1980x1080 DSLR footage down sampled to that same 640x480 so even if you're watching it on a crappy down sample like VHS, it is still looking far better if it was shot at a high original resolution