RED MX Latitude Tests

[Sanjin Jukic]

Great test that everybody should read:

Next Stop: The Last Stop! RED MX Latitude Tests>>>

By Art Adams & Adam Wilt @ ProVideo Coalition | 07/30

Looking forward to those two guys that can soon make Alexa Latitude test soon.

Then we could compare those two tests to get an idea...

Adam made a tunnel out of black showcard.

 

[oliver imfeld]

no link !
cheers, o

 

[Sanjin Jukic]

no link !
cheers, o

Oliver,

sorry,

now is corrected!

 

[Alexander Ibrahim]

I like the terminology they use.

The camera has 11 stops of "paycheck" latitude and an additional 3-4 stops of "gravy."

So, you can bet on 11 stops, but you might be able to pull something out of the shadows for another 4 stops, for a total of up to 15 stops.

It gets noisy after 11 stops, so I agree that I'd stick to using 11 or 12 stops for most projects.

Interesting and relatively well done test. I'd like to see the test repeated for more cameras.

 

[Thomas Wright]

This thing here is confusing me: We had fellas come in and test MX and M and they found a sort of colour difference. Besides that, I closed the iris to check where the waveforms were equal for both camera's RAW outputs and there was less than one stop difference and the RAW and colourspace files looked most alike when the camera was set around 320.

One fella thinks the MX is not a chip but a board/software upgrade that just deals with the noise better. I didn't get to play with the MX long but from what I saw there is a grain of reason in this logic.

Anyone care to comment on this conspiracy theory?

 

[David Birdy]

My 2 cents...

The new MX sensor is most likely a new Cmos sensor that utilizes smaller Nano engineering components to yield less noise.

The only conspiracy I know Red to be involved with is: Conspiring to offer new tools to produce media at a reasonable price point.

Dave

 

[sander kamp]

This thing here is confusing me: We had fellas come in and test MX and M and they found a sort of colour difference. Besides that, I closed the iris to check where the waveforms were equal for both camera's RAW outputs and there was less than one stop difference and the RAW and colourspace files looked most alike when the camera was set around 320.

One fella thinks the MX is not a chip but a board/software upgrade that just deals with the noise better. I didn't get to play with the MX long but from what I saw there is a grain of reason in this logic.

Anyone care to comment on this conspiracy theory?

That they look similar at 320 ISO makes sense because where the MX sensor is better than the M sensor is mostly in the shadows. There is less noise in the MX shadows which means you can underexpose one or two stops and up the gain in post (or in the preview on the camera) which means you can effectively treat it as a 800 ISO camera.

The chip and/or OLPF are definitely different. When you look into a M camera you see a red OLPF/sensor, when you look into a MX camera you see a gray OLPF/sensor.

 

[Roberto Lequeux]

Nice article. Thanks for sharing SJ, and thanks to Art and Adam obviously.

I'd look at it as ~10 paycheck stops plus ~2 to roll off highlights even softer by ETTR and applying curves to taste, and consider gravy only ~1.5 of the bottom three since with noise reduction and down conversion from 4.5k and 5k you are bound to get ~1 more usable stop off of all that rich Thanksgiving gravy.

 

[Hrvoje Simic]

I'd like to see the test results after WB and out of Redcine X.

 

[Paul Leeming]

My only queston would be - why did they use Camera709 and not RedColor/RedGamma combo which is stated by Red as being the optimal solution to maximise dynamic range and colour accuracy?

Otherwise it seems like a decent test.

Paul