Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Red MX blown away by f35, d-21, alexa in DR test?

Peter Dmitriyev said:
as I understand it dynamic range is all about photosite size. (well other things play a role but that's the most important) as the size of the photosites determines the light gathering capabilities and such. Red's Photosites are around 5 microns, Alexa is about 8 or more from what I remember.
The bigger your photosites are the less of them you can have obviously as the size of the sensor can only fit so much of them. So, since photosites comprise pixels, that means the bigger your photosites the less pixels you have and thus the less resolution. So in short, Arri sacrifices resolution for dynamic range. Red has super resolution and thus less dynamic range because its photosites are smaller.
Click to expand...
That can be kind of a rule of thumb, but not the sole determinant in dynamic range. For example: the Dragon will have smaller photo sites than the MX chip, but have 2 stops higher dynamic range. In ARRI's case, I'm sure the larger pixels help - as well as their dual-gain method. They use analog gain as opposed to RED's digital gain, and apply two levels of gain to the image - recombining the two streams for the final image. The same trick doesn't work on RED (processing out different ISOs and recombining them), and I believe it is due to the digital vs. analog gain. At least that's my understanding.
 
Kemalettin Sert said:
i know this guy and he's totally on drugs..he shot tvc with Epic rated @ 160 ISO too.He believes Red cameras has default iso 160..
Click to expand...

Thats weird, I shot a tvc and rated the Epic at 20,000ISO :banghead:

This guy is nuts! 160? Where does he even come up with that? Why would someone hire him to shoot with a camera he has no clue about??
 
Kevin Marshall said:
That can be kind of a rule of thumb, but not the sole determinant in dynamic range. For example: the Dragon will have smaller photo sites than the MX chip, but have 2 stops higher dynamic range. In ARRI's case, I'm sure the larger pixels help - as well as their dual-gain method. They use analog gain as opposed to RED's digital gain, and apply two levels of gain to the image - recombining the two streams for the final image. The same trick doesn't work on RED (processing out different ISOs and recombining them), and I believe it is due to the digital vs. analog gain. At least that's my understanding.
Click to expand...

That's very intriguing, and more than I knew before. It sounds almost - almost - as if Arri are doing something analogous to HDRx, at an earlier stage in the processing pipeline, and outputting something vaguely resembling a 'baked in' HDRx grade?

The photosite size makes total sense; see Nikon - D3s versus D3x.

Mike
 
Worth a read but inadequate overall. The fact that RED does have a "Learn" page is a good start but much more detail and indepth education on "processing REDCODE RAW correctly" is needed.

There is a RED 101 video, 102 etc... but there is a lot of contradictory information and it's 3rd party, not from the source. The guy in these videos says "Expose to the Left," I believe Ted said the same. Then checking REDUSER, you have a group "ETTR." The phrase is common enough to get it's own abbreviation. Expose to the right, give the sensor as much light as possible without clipping the right goal post and just before getting stop lights.

With contradictory information out there, and others like Mark Toia saying "it's easy," his kid does it, there it is no wonder people have difficulty, I certainly have.
The post process is where I have learned maybe the most. This shot and this setting worked, and this one over here is a pain to correct. Trial and error for the most part.

So let's be done with it once and for all, no user groups, no 3rd party videos. Please, Jim make a video series showing exactly how to make your camera shine.

Matt Ryan said:
Here's how to achieve proper Exposure:
http://www.red.com/learn/red-101/exposure-with-red-cameras
http://www.red.com/learn/red-101/red-camera-exposure-tools
Click to expand...
 
Sorry... photosite (pixel) size used to be one of the most important pieces for DR but has now taken a back seat to pixel design.

The MX sensor has more DR than the original M sensor... and the pixels are the same size. The new Dragon sensor has mind-blowing DR and the pixels are slightly smaller than either the M or MX.

The new native DR King will be crowned later this year. And that will be by anyone's testing methodology. Geez... and it is an upgrade!

Jim

Peter Dmitriyev said:
as I understand it dynamic range is all about photosite size. (well other things play a role but that's the most important) as the size of the photosites determines the light gathering capabilities and such. Red's Photosites are around 5 microns, Alexa is about 8 or more from what I remember.
The bigger your photosites are the less of them you can have obviously as the size of the sensor can only fit so much of them. So, since photosites comprise pixels, that means the bigger your photosites the less pixels you have and thus the less resolution. So in short, Arri sacrifices resolution for dynamic range. Red has super resolution and thus less dynamic range because its photosites are smaller.
Click to expand...
 
Peter Dmitriyev said:
as I understand it dynamic range is all about photosite size. (well other things play a role but that's the most important) as the size of the photosites determines the light gathering capabilities and such. Red's Photosites are around 5 microns, Alexa is about 8 or more from what I remember.
The bigger your photosites are the less of them you can have obviously as the size of the sensor can only fit so much of them. So, since photosites comprise pixels, that means the bigger your photosites the less pixels you have and thus the less resolution. So in short, Arri sacrifices resolution for dynamic range. Red has super resolution and thus less dynamic range because its photosites are smaller.
Click to expand...

No, it's not. It's about sensor size and photosites vs. "dark areas" - or in other words percentage of sensor surface devoted to photosites, regardless of their number. This is a common misconception.
 
Jannard said:
Sorry... photosite (pixel) size used to be one of the most important pieces for DR but has now taken a back seat to pixel design.

The MX sensor has more DR than the original M sensor... and the pixels are the same size. The new Dragon sensor has mind-blowing DR and the pixels are slightly smaller than either the M or MX.

The new native DR King will be crowned later this year. And that will be by anyone's testing methodology. Geez... and it is an upgrade!

Jim
Click to expand...

Victory!!!
 
Choices

Choices

First, Jim's note about photo site design is spot on. IF, you have two photo sites with identical properties and one is physically bigger then size is key - if not...

Second, putting RAW sensor data in the hands of the user gives one the opportunity to tailor a nuanced image or to destroy it.

Third, the reason there isn't a one size fits all processing path is because different exposure environments and different final image goals require different techniques - see above.

Perhaps RED would have been more widely accepted more quickly if they had done more image processing prior to letting us see it - instead they followed a mantra of giving us everything the sensor saw and letting us mold it into the RGB image we desired. FWIW I have taken that sensor data and created some pretty awful RGB images along the way, I have also created some images I really love. That said, some of the most compelling images I have seen from R3Ds were carefully fashioned by elite colorists like Leandro Marini, Bobby Maruvada, Mike Most, John Tissavary and others.

Cheers - #19
 
By the way back on topic for a moment: even on arri's official website the D21 is listed as having 11 stops. Even at its absolute most pessimistic rating the Red MX by detractors has been rated at 11 stops at the very least and as high as 13+ stops by other people. So I find it really crazy how the D21 (as well as f35) utterly blows away red mx in that test on the front page. It really proves to me that test is seriously bogus/flawed and should be taken down
 
Rob Ruffo said:
No, it's not. It's about sensor size and photosites vs. "dark areas" - or in other words percentage of sensor surface devoted to photosites, regardless of their number. This is a common misconception.
Click to expand...

So are you saying that whichever sensor crams the photosites much closer together ends up getting higher DR?
 
Greater fill factor is a worthy goal for obvious reasons but I still stand with Jim in his assertion that better photo receptors are a bigger factor. I would guess he has seen things that most of us have not to support that viewpoint. I think of it like car engines; do more cylinders mean more power no matter what the other metrics are?

Cheers - #19
 
They got 10 stops out of an Af-100! I believe that when I see it too.
 
Alexa > Sony F35 > Red Epic > Scarlet > Sony F3 > Red One > Canon 5D MK II > Panasonic HVX200
 
You must log in or register to reply here.
Back
Top