Sony VENICE and RED Monstro Anamorphic in 4K HLG HDR right out from cameras

[rand thompson]

Sony VENICE and RED Monstro Anamorphic in 4K HLG HDR right out from cameras




By Sittipong Kongtong

I had a test with 2 high end camera, Sony VENICE and RED Monstro in anmorphic mode. I used 2x anamorphic lens; Cooke, Master Anamorphic, Kowa, LOMO. All lens has beuatiful color for both cameras.

I tested at Life Boat, a camera and lens rental house in Bangkok.
I export this file in 4K 25p DCI Scope DNxHQX 12bit for Youtube.
You can see it on any OLED HLG HDR devices.

 

[Christoffer Glans]

Why is the Monstro so unbalanced in its tint?

 

[Björn Benckert]

Yes, when white walls are green you are doing it wrong some how. I have a difficulty to believe both cameras are at equal kelvins and tint. And even if so. Why make a test where one has the colors clearly pushed out of balance.

 

[rand thompson]

From looking at the clapboard he presented for each camera at the beginning of the shoots, he did use a White Balance of 5500 for each camera. It seems he is unaware of the Tint required to better white balance the monstro or chose deliberately not to use it for this test.


Edit,


The Venice seems more blue than it probably should as well. So I'm sure there are some Sony Venice owners who will probably take issue with this test as well. So I believe neither camera is truly "White" Balanced.

 

[Björn Benckert]

From looking at the clapboard he presented for each camera at the beginning of the shoots, he did use a White Balance of 5500 for each camera. It seems he is unaware of the Tint required to better white balance the monstro or chose deliberately not to use it for this test.


Edit,


The Venice seems more blue than it probably should as well. So I'm sure there are some Sony Venice owners who will probably take issue with this test as well. So I believe neither camera is truly "White" Balanced.

I find cameras to be just as monitors, they can drift over time and / or loose their zero for calibration. But and as the cameras captures raw a gray card balancing would be a good start I think. But possibly they did not compare colors but other things.

 

[rand thompson]

Bjorn,


Yeah, They probably just thought it was easier for them to just set the Kelvin the Same and just go from there. They do say that no grading was done to the images.

here is my attempt to show how the Venice and the Monstro both could have been corrected to atleast fix the White Balance of each in this test. And yeah! I know they are jpeg images that I graded you can tell by the shitty banding in the image.


Venice

Monstro

 

[Chris McKechnie]

Gosh....I really really wish RED would focus on color accuracy as opposed to resolution. I feel that the MX & Dragon sensors were much more accurate than any of the latest RED tech in terms of color. Productions care about the post workflow, and amount of work to get the image looking great, and with RED, it just takes more to get it there, where it didn't used to be that way with my tried and trusted MX sensor. That thing is a beast in terms of color, reliability, etc. I still love RED, but I think they need to hit it out of park with a DSMC 3 model that shoots 4K, but is more focused on producing amazing color as opposed to just increasing the K's. Just my opinion as I've seen RED progress/change through the years.

 

[Nick Morrison]

From looking at the clapboard he presented for each camera at the beginning of the shoots, he did use a White Balance of 5500 for each camera. It seems he is unaware of the Tint required to better white balance the monstro or chose deliberately not to use it for this test.


Edit,


The Venice seems more blue than it probably should as well. So I'm sure there are some Sony Venice owners who will probably take issue with this test as well. So I believe neither camera is truly "White" Balanced.

Yeah properly white balancing with offsets in camera is a discipline that rarely happens - esp in cam tests.

Looks like they are just eyeballing it

 

[Nick Morrison]

Why is the Monstro so unbalanced in its tint?

Nothing to do with the camera, they clearly eye-ball balanced to the Sony and plugged in the same values to the Monstro.

Diff cameras almost never match...so...

 

[Patrick Tresch]

Gosh....I really really wish RED would focus on color accuracy as opposed to resolution. I feel that the MX & Dragon sensors were much more accurate than any of the latest RED tech in terms of color. Productions care about the post workflow, and amount of work to get the image looking great, and with RED, it just takes more to get it there, where it didn't used to be that way with my tried and trusted MX sensor. That thing is a beast in terms of color, reliability, etc. I still love RED, but I think they need to hit it out of park with a DSMC 3 model that shoots 4K, but is more focused on producing amazing color as opposed to just increasing the K's. Just my opinion as I've seen RED progress/change through the years.

Don’t take this as a color test. I never had a color issue with Red (appart from some yellow peaks you can easily turn down). I love the 6k resolution my Dragon gives me. I would never trade it for a 3.2k lowres bayer sensor.

 

[David J. Buchanan]

Still, it's a shame to see the color from the Monstro so off... I know there are a couple of people who believe it's accuracy but it's too obvious. Good thing grading takes care of it on the Monstro, for the most part.

 

[Nick Morrison]

Gosh....I really really wish RED would focus on color accuracy as opposed to resolution. I feel that the MX & Dragon sensors were much more accurate than any of the latest RED tech in terms of color. Productions care about the post workflow, and amount of work to get the image looking great, and with RED, it just takes more to get it there, where it didn't used to be that way with my tried and trusted MX sensor. That thing is a beast in terms of color, reliability, etc. I still love RED, but I think they need to hit it out of park with a DSMC 3 model that shoots 4K, but is more focused on producing amazing color as opposed to just increasing the K's. Just my opinion as I've seen RED progress/change through the years.

Chris this test has nothing to do with RED's color accuracy - this crew is clearly not white balancing (at all).

Sorry, you're being pretty unfair here.

If you pointed an Alexa at this slate with similar settings it could be green too (I know of a very talented DP that always dials out the green in Alexa - did it on an Amira shoot with us).

This team obviously wasn't thinking about WB at all.

Maybe there were offsets in the Monstro from a previous shoot?

Maybe the light really is green?

We always WB RED's before we shoot (off a card or chart) and never have problems. And we have a demanding colorist who doesn't tolerate cams with wrong color balance, so trust me...if there was an actual problem I'd know.

 

[Nick Morrison]

Still, it's a shame to see the color from the Monstro so off... I know there are a couple of people who believe it's accuracy but it's too obvious. Good thing grading takes care of it on the Monstro, for the most part.

Again this test is meaningless. For all we know this Monstro may have had +green offsets from a previous shoot.

If this team had WB'd off a card (or just AUTO WB'd from the touch screen) I'm 100% positive the Monstro would look normal.

This is much ado about nothing.

 

[David J. Buchanan]

Again this test is meaningless. For all we know this Monstro may have had +green offsets from a previous shoot.

If this team had WB'd off a card (or just AUTO WB'd from the touch screen) I'm 100% positive the Monstro would look normal.

This is much ado about nothing.

I've noticed green when shooting with the Monstro before and so have others who rented the camera from me. It's not a myth. It's there. I've even seen people do the - tint in camera to get rid of it as to not deal with it in post. At the end of the day, still a beast of a machine, but the color isn't phenomenal.

 

[Björn Benckert]

I've noticed green when shooting with the Monstro before and so have others who rented the camera from me. It's not a myth. It's there. I've even seen people do the - tint in camera to get rid of it as to not deal with it in post. At the end of the day, still a beast of a machine, but the color isn't phenomenal.

It shoots raw, does not matter where you do your white balance calibration, in camera or post gives same result.

Again, see caneras as monitors even the best of monitors needs to be calibrated to show an acurate image. Hitting the default button will not always get you there. So ifin daylight and you sellect k5600 you get close but only if you shoot a graycard and white balance on it you get correctly dialed in.

 

[rand thompson]

As the poster of this video noted;

"I tested at Life Boat, a camera and lens rental house in Bangkok."


So as Nick stated, who knows what settings were left from the previous renters for both the Monstro and the Venice cameras, the Venice's white balance didn't look all that hot either.

I mainly posted this video to show how the different Anamorphic lenses used in this test looked on each of the cameras. I realized after the first anamorphic comparison that someone must have screwed up the white balance for the Monstro, due to it's green cast, and simply viewed it as just an anamorphic lenses comparison and not on color rendition.

 

[Hrvoje Simic]

I mainly posted this video to show how the different Anamorphic lenses used in this test looked on each of the cameras.

Which sparked drama about raw camera colour performance on missed white balance with different toned lens and unknown light source, quality and temperature in non colour neutral scene.

While color correcting a 0.00005 MB frame grab from a video on which it says "no any grading".

 

[Aaron Lochert]

For all the wides, the ceiling lights in the background are off in all the Monstro shots but turned on in all the Venice ones. It gets more inconsistent in the closeups - sometimes on in the Venice shots, sometimes off (you can even see it turn off at 3:05). Appears to be consistently off in all the Monstro closeups but I might be wrong. If the background lights are changing, I wonder if all the ambient lights in front of the subject are also changing. Not to mention the daylight through the windows is more than likely shifting all day.

And lastly, who knows what that LED key light they're using is like. People should definitely not consider this test as a color test.

Yeah properly white balancing with offsets in camera is a discipline that rarely happens - esp in cam tests.

Looks like they are just eyeballing it

While it is upsetting that so many people do tests like this because they seem to think that keeping the WB numbers constant between two cameras is somehow being objective, I hope RED understands that THAT is the type of person they need to target. Although the tech-savvy of us understand that WB is just a non-destructive multiplier to the raw data and it doesn't matter whatever the hell the numbers actually say, there's still many people that have the compulsion for the number on screen to read zero -- if it's anywhere away from 0, they think something is inherently wrong.

This same phenomenon is why I speculate that Panavision re-rated the Monstro sensor to 1600 ISO. While nothing changed, technically, it is a resetting of the origin. Exposing at 1600 on the DXL2 = proper, where exposing at 1600 on a RED would be looked as "one stop under", which can be seen as "wrong" practice. It's semantics, really.

Now I don't know what light sources they use in house when calibrating the sensors, but I've consistently needed -7 to -15 ish for my Helium on just about everything I've done and many other shooters echo the same. So perhaps a re-calibration of the WB scale is in order for Helium users. Or perhaps RED should give us the option to zero the scale with our own offset so then we can only blame ourselves when things aren't white balanced properly

 

[Hrvoje Simic]

Exposing at 1600 on the DXL2 = proper, where exposing at 1600 on a RED would be looked as "one stop under", which can be seen as "wrong" practice. It's semantics, really.

Yeah it's cool semantics when your mid grey moves to the left so whites clips later but not so cool when A list actor's face is underexposed and compression & underexposure goblins dance around the frame on a six figure camera rental only and nine figure project.

 

[Phil Holland]

color accuracy as opposed to resolution

On a measureable level the current sensor tech is far, far more accurate than Mysterium-X.

While I'm on the line, a few things about this test as it's "straight out of camera". Something is weird here for sure in terms of white balance of both bodies. In testing the Venice recently there was nowhere near this color variance between Monstro and it.

However, I will say it is critical to have good white balance when comparing camera systems because literally no manufacturer hits each others targets in terms of color temperature, tint, and occasionally other bias related things.

Also, it's alarmingly easy to match to the Venice shot or vice versa, though really the hard truth is providing an accurate target for white balance for both systems here.