Mysterium-X dynamic range...

[Jannard]

Neil Smith and his crew from Hollywood DI supervised an over/under shoot to confirm the dynamic range of the new Mysterium-X sensor. Charts were lit in a properly controlled light booth at D50. An engineering build was used with a slight calibration issue. Results clearly showed 13 stops with more room available given a proper calibration. We will re-shoot and post results as soon as the calibration issue is resolved. We fully expect a meaningful increase once this is done.

The noise floor of the new sensor is pretty incredible. I'll let those that have seem it post their opinion since we showed plenty of examples tonight at the new RED Studios Hollywood.

Jim

[Shawn Nelson]

Wow! What a good night for Red. Congrats Jim! You sure looked happy in the pictures posted.

Looking forward to the final word, I can't wait for this sensor!

Having watched my current project (shot on Build 21) go through the Quantel Pablo I was blown away by how clean it was, I can only imagine what M-X will be like, looking forward to it big time.

[Alex Mack]

This must be from the successful raising of the ISO recently.
If this is due to low noise floor, then the shadows are where the DR has increased. (highlights over expose point still same)

Given this dynamic range, then the question must be what ISO to shoot at to get the most out of it with the best color accuracy...

[Ryan S]

13 Stops. I think we can safely call that Magic. Excellent work, Red team!

[Neil_Hdi]

Originally Posted by Jannard

Neil Smith and his crew from Hollywood DI supervised an over/under shoot to confirm the dynamic range of the new Mysterium-X sensor. Charts were lit in a properly controlled light booth at D50. An engineering build was used with a slight calibration issue. Results clearly showed 13 stops with more room available given a proper calibration. We will re-shoot and post results as soon as the calibration issue is resolved. We fully expect a meaningful increase once this is done.

The noise floor of the new sensor is pretty incredible. I'll let those that have seem it post their opinion since we showed plenty of examples tonight at the new RED Studios Hollywood.

Jim

Jim,

Thank you for putting on a great day for all of us .... that new 4k Sony projector with 4k M-X footage on that huge screen was totally stunning .... now we understand why you've been beating the 4k drum so loud for these last three years .... high quality 4k projection makes all the difference ... glad Sony finally caught up with the 4k acquisition team ;-)

The M-X Dynamic Range test we ran on Thursday night was incredibly revealing .... the lack of noise at 3 stops plus under was great news for us RED post boys ... the graceful fall off at 3 stops over was sweet .... this new M-X sensor combined with Graeme's FLUT color science is in someways more revolutionary than the original RED One footage we graded three years ago.

I had a sense today that we were not just making digital cinema news ... we're making digital cinema history .... buying Ren Mar Studios and setting up shop in the heart of Hollywood is a strategic move of significant proportions .... your sensor program combined with the new Color Science and Redcine-X is a killer combination .... RED Studios Hollywood combined with the M-X upgrade program is a streak of genius .... well done to you and your team of hard working REDsters for delivering a memorable, inspiring and enjoyable day.

Neil

P.S. I hope your back and feet and legs are not aching as much as mine .... jeezz that was a long day, mate.

[Gunleik Groven]

Fantastic!
More DR and less noise is just the Doctors order.
Congratulations. So looking foreward to working with some MX footage.

Couple of questions.
Was this with a MX'ed R1 or with an Epic proto?
Is the MX'ed R1 still recording a lin image or a log'ified one?
(Sorry Graeme for my unprecise terminology...)

Reason for asking?
Back in the day - you all know when - we were all explained that each extra bit doubled the precission and added a theoretical stop to what could be recorded, given that the signalchain and sensor allowed for that.

MX being 14 bit explains a lot of the new Epic redcodes. Redcode 144'ish would make a lot of sense for a 14 bit redcode lin image with the ballparkish same compression as a 12 bit lin redcode 36 Mysterium image, just from roughly adding the numbers and not adding Graemes magic sauces and leaving the pixies home for the new year.

But an MX'ed R1 does not all of a sudden become a 14 bit lin camera, without more or less replacing the whole interiours of the R1 AND record to higher datarate Redcodes than the current ones.

If I am not too off here... Which I may well be...

So. The question is more or less:

What would make sense - maybe - would be to apply a log curve on the data after the AD and record that log data to 12 bit on the MX'ed R1 and that the Epic records 14 bit lin.

The alternative would still leave the MX'ed R1 with less than 12 stops.

Or am I off?

Anyway

I cannot wait to play with the new footy!


G

[Tanner Stauss]

This is all very cool. The coolness and surprises are very routine by this point, but nonetheless, exciting.

I must have missed something, what is this "RED Studios Hollywood" being mentioned?

[CJ Roy]

Jim-

Posting the results would be really beneficial for us. Thanks for tonight, great job.

-CJ

[Tai Wah Lim]

After watching footages and playing around with the RED One with M-X - certainly more than 13 stops - Lim

[jonnycom]

Can you post the Leo or R8 footage for the people who weren't there? It was amazing!