DR ..a good discussion!!

[rand thompson]

For Those who will see my "untraditional" Luminance curve and say "What the fuck Rand!!", here are the same images graded with a more traditional "S-Curve" and .RMD file.

"Untraditional Luminance Curve"

ISO 800

"Traditional S-Curve"

ISO 1600

Dropbox Grade with traditional "S-Curve"


ISO was raised from ISo 800 to ISO 1600, you may even prefer ISO 2000, to get an approximate image with the traditional "S-Curve"

https://www.dropbox.com/s/41qp4pqzeunmrgb/woman on boat standardd s-curve 2.RMD?dl=0

 

[luigivaltulini]

[/url]34118453_2130430700526135_8118903900855074816_n by luigi valtulini, su Flickr[/IMG]

now open the Sony Venice demo with firmware beta 2.0.
jumped out of the box while my girlfriend slept ehehe
Monstro and Venice (set on 2500 iso base) .
4 clip MONSTRO 8k 2:1/ VEnice 6k 3:2, 2 at T8 and 2 at T1.5.
At first sight Monstro has 1 extra stop in the highlights, they are also softer. Noise we are very close, I think almost similar.
The Venice is a beautiful movie cameras I love the internal ND, easy and intuitive menu. Damn bitch as much as I would like of the internal Nd on RED ...

Tomorrow or later take night shots and test on xyla DR.
Well take a look and discuss

https://www.dropbox.com/s/qgtptqmg734qhhf/sonvistared.zip?dl=0

 

[rand thompson]

Luigi,


Here is one of the .R3ds graded. A Little noise reduction can help.




Original

[/url]A001_C021_0111EZ by rand thompson, on Flickr[/IMG]



Graded

[/url]A001_C021_0111EZ by rand thompson, on Flickr[/IMG]




With Noise Reduction

[/url]monstro noise reduction_1.3.2 by rand thompson, on Flickr[/IMG]

 

[rand thompson]

Luigi,


I am not too familiar with Sony's Venice and what Luts to use so I just graded it raw in Davinci Resolve. I tried to grade the Underexposed Venice image in a similar way in Resolve as I did the Underexposed Red image in Redcine-X.



Original

[/url]Venice Original_1.1.2 by rand thompson, on Flickr[/IMG]



Graded

[/url]venice_1.1.1 by rand thompson, on Flickr[/IMG]


With more contrast and noise reduction

[/url]with noise reduction 2_1.1.5 by rand thompson, on Flickr[/IMG]

 

[rand thompson]

Here is a slight regrade of the better illuminated .R3D image.



Original

[/url]A001_C023_0111RF by rand thompson, on Flickr[/IMG]



Regrade

[/url]A001_C023_0111RF by rand thompson, on Flickr[/IMG]

 

[rand thompson]

Luigi,


Here is the Sony Venice Clip with the blown highlights. There is nothing you can do but keep from blowing out the highlights further with your grade.




Original

[/url]venice blown highlight original_1.4.2 by rand thompson, on Flickr[/IMG]





Graded

[/url]blown out venice grade_1.4.2 by rand thompson, on Flickr[/IMG]

 

[Aaron Lochert]

Rand, here you go: http://www.hingsberg.com/2018/05/sony-venice-rec709-venice-3d-lut/

Thanks for these, Luigi.

All I've done here is use nothing outside of the manufacturer's own color pipeline, which was to process each one to their respective wide gamut color and gamma curve. Then I added each manufactuere's own REC.709 output LUT on a single node. Everything else was done in the RAW tab as you can see in the settings below. On the underexposed clips that were pushed, a touch of chroma noise reduction was added to tame nasty colors, but no luminance noise reduction was done.

Once matched for brightness, Monstro appears to win on highlights clearly, Venice at first glance wins on underexposure for color reasons. Will be interesting to see the Xyla charts. I assume these were the same exposure settings for the two cameras?


Venice:

4K: https://farm2.staticflickr.com/1723/27655137607_36921874d2_o.png

4K: https://farm2.staticflickr.com/1739/42525649491_c07abbc6d6_o.png



Monstro (IPP2 Low Con/Soft Rolloff):

4K: https://farm2.staticflickr.com/1731/28652249948_107aa12ea5_o.png

4K: https://farm2.staticflickr.com/1735/28652250578_69506b9b01_o.png



Scopes:

(if viewed at 100% size, the screenshots on the right will be at 1:1 pixels for 4K delivery.)

The interesting thing is that the green "Helium shadows" exist on Monstro as well, if cranked enough.

What RED needs to figure out and what Sony and Arri both clearly already have, is how to keep the shadows neutral as the ISO is increased. The Sony Venice footage when underexposed and pushed looks relatively the same, but there is more noise and less color fidelity (as one would expect). But the colors that do remain are the same. As seen in the Vectorscope, the general shape and location of the white blobs are the same from shot to shot. On Monstro, the blobs severely skew towards green and yellow. We can also see how the shadows on Monstro end up clipping below zero. On Venice, all shadow information remains above the 0 line.

You might have noticed I didn't change the white balance on the underexposed Monstro footage. This is because the tinted shadows is non uniform. If I tried to dial out the green with a global tint, I'd be making my whites magenta in the process.

So to tackle this, you need to do something with a little more precision, which I'm now quite practiced in as a Helium owner.

4K: https://farm2.staticflickr.com/1757/41803339974_bfbbe1f9a1_o.png

And now it appears that Monstro looks maybe a touch cleaner than Venice because the grain is smaller. RED is doing themselves a disservice by not fixing this. I know, I know. "Raw is raw". But I don't think people really wan't "raw." We want the flexibility that comes with raw, we don't want the fallbacks. Expecting too much? Maybe. But the Venice gives nicely balanced image even when pushed a few stops. Seems to be RED's MO. "Here's this piece of technology that we know is kickass. But you're going to need to figure out how to polish it."

Definitely interesting to see. Venice looks quite lovely - I like the way those saturated reds are looking. Maybe a little more forgiving to the cinematographer who doesn't want to do a whole lot of post work.

And now, an actual grade:

4K: https://farm2.staticflickr.com/1744/41804558024_0006bd5d28_o.png

 

[rand thompson]

Aaron,


Thanks for these test and a link for the Sony luts. I noticed some of the things you pointed out in your post and also the Purple/Pink/Red noise in the shadows which appears in your images as well.

 

[rand thompson]

Aaron,


I just didn't like those S-log Transform luts in Resolve. So I regraded the underexposed image with a Custom Curve, 58 Sat, -0.5 Log Shadow and 2000 ISO in the Raw tab and a 38 Better/Small noise reduction.

[/url]underexposed corrected_1.1.1copy by rand thompson, on Flickr[/IMG]

 

[luigivaltulini]

Rand, here you go: http://www.hingsberg.com/2018/05/sony-venice-rec709-venice-3d-lut/

Thanks for these, Luigi.

All I've done here is use nothing outside of the manufacturer's own color pipeline, which was to process each one to their respective wide gamut color and gamma curve. Then I added each manufactuere's own REC.709 output LUT on a single node. Everything else was done in the RAW tab as you can see in the settings below. On the underexposed clips that were pushed, a touch of chroma noise reduction was added to tame nasty colors, but no luminance noise reduction was done.

Once matched for brightness, Monstro appears to win on highlights clearly, Venice at first glance wins on underexposure for color reasons. Will be interesting to see the Xyla charts. I assume these were the same exposure settings for the two cameras?


Venice:


4K: https://farm2.staticflickr.com/1723/27655137607_36921874d2_o.png



4K: https://farm2.staticflickr.com/1739/42525649491_c07abbc6d6_o.png



Monstro (IPP2 Low Con/Soft Rolloff):


4K: https://farm2.staticflickr.com/1731/28652249948_107aa12ea5_o.png


4K: https://farm2.staticflickr.com/1735/28652250578_69506b9b01_o.png



Scopes:

(if viewed at 100% size, the screenshots on the right will be at 1:1 pixels for 4K delivery.)



The interesting thing is that the green "Helium shadows" exist on Monstro as well, if cranked enough.

What RED needs to figure out and what Sony and Arri both clearly already have, is how to keep the shadows neutral as the ISO is increased. The Sony Venice footage when underexposed and pushed looks relatively the same, but there is more noise and less color fidelity (as one would expect). But the colors that do remain are the same. As seen in the Vectorscope, the general shape and location of the white blobs are the same from shot to shot. On Monstro, the blobs severely skew towards green and yellow. We can also see how the shadows on Monstro end up clipping below zero. On Venice, all shadow information remains above the 0 line.

You might have noticed I didn't change the white balance on the underexposed Monstro footage. This is because the tinted shadows is non uniform. If I tried to dial out the green with a global tint, I'd be making my whites magenta in the process.

So to tackle this, you need to do something with a little more precision, which I'm now quite practiced in as a Helium owner.


4K: https://farm2.staticflickr.com/1757/41803339974_bfbbe1f9a1_o.png

And now it appears that Monstro looks maybe a touch cleaner than Venice because the grain is smaller. RED is doing themselves a disservice by not fixing this. I know, I know. "Raw is raw". But I don't think people really wan't "raw." We want the flexibility that comes with raw, we don't want the fallbacks. Expecting too much? Maybe. But the Venice gives nicely balanced image even when pushed a few stops. Seems to be RED's MO. "Here's this piece of technology that we know is kickass. But you're going to need to figure out how to polish it."

Definitely interesting to see. Venice looks quite lovely - I like the way those saturated reds are looking. Maybe a little more forgiving to the cinematographer who doesn't want to do a whole lot of post work.

And now, an actual grade:


4K: https://farm2.staticflickr.com/1744/41804558024_0006bd5d28_o.png

Just observations Aaron.
The Venice was set on 2500 iso base, I wanted to see if the background noise was equal to Monstro, but as you have noticed the high lights lose almost 2 stops.
I was hoping to see Venice cleaner, but from how it looks like Monstro is still the same or maybe better. This sistuation reminds me of Gemini vs Monstro.
In standard mode on venice, base 500 iso) the highlights are identical
Monstro tends to green, while Venice is too laden with red and Monstro in the shades tends to go green - yellow.
For the dynamic I believe that Monstro and Venice are identical, we will see xyla what he tells me in the coming days.

https://www.dropbox.com/s/dnt6tx5714q1soh/sonymon2.zip?dl=0

 

[Nicolas Chareyre]

I think that it could be interesting to see how behaves GEMINI in the shadows and his DR against MONSTRO and VENICE even if it is not full frame

 

[Michael Wahba]

I think that it could be interesting to see how behaves GEMINI in the shadows and his DR against MONSTRO and VENICE even if it is not full frame

Agree, can we do that? Thanks!!

 

[rand thompson]

Luigi,


Here is the Monstro.



Original

[/url]Luigi Monstro original_1.1.42 by rand thompson, on Flickr[/IMG]


Graded

[/url]luigi Monstro new 2 1 by rand thompson, on Flickr[/IMG]

 

[rand thompson]

Luigi,


Here is Venice.



Original

[/url]venice original_1.2.3 by rand thompson, on Flickr[/IMG]



Graded

[/url]venice full res of_1.2.2+final by rand thompson, on Flickr[/IMG]

 

[rand thompson]

I was messing around with the Monstro image and accidentally applied a Red "Low Con/Very Soft" output Transform Lut on the wrong image, the Venice image that was right next to it in the "Color" Clips. I was surprised by what I saw. So in the Raw tab for Sony, I changed the iso from 500 to 620 and set the "Shadow" to 9.00. This along with a very slight green adjustment in the green curve and a very slight luminance curve adjustment and got the image below.

[/url]Red low con very soft_1.2.4 by rand thompson, on Flickr[/IMG]

 

[Les Hillis]

What's Up Les!!! I hadn't heard from you in a while, you've probably just been in different threads than I have lately...your grades have been looking quite nice...

Thanks Rand, just keeping busy with other stuff.

Is interesting to see the Sony Venice/Monstro comparison.

Here's a creative grade I applied to both.

Monstro -




Venice -




Monstro -




Venice -




With this grade there was a bit of going backwards and forwards, but mostly it was trying to get the Monstro to match the Venice.

I do like what Sony and others are doing, but I personally still prefer what RED are doing with their sensors and colour science.



Edit - Just want to add that besides my only being able to get a rough match in terms of overall colours, with the slightly different framing and light falling on the subject, these frames aren't meant to be presenting a totally accurate comparison between the different camera's used.

 

[rand thompson]

S-Curve and 71 sat, -0.1 "lift" green

 

[rand thompson]

Thanks Rand, just keeping busy with other stuff.

Is interesting to see the Sony Venice/Monstro comparison.

Here's a creative grade I applied to both.

Monstro -




Venice -




Monstro -




Venice -




With this grade there was a bit of going backwards and forwards, but mostly it was trying to get the Monstro to match the Venice.

I do like what Sony and others are doing, but I personally still prefer what RED are doing with their sensors and colour science.

Those are some good looking grades there Les!!!

 

[Les Hillis]

Thanks Rand, and thanks Luigi again for the test frames, we're lucky to have guys like you and the others providing real world samples from these camera's.

 

[luigivaltulini]

Ok Guys lets go

Xyla DR on Monstro + Sony Venice standard/lowlight mode + Gemini standard/low . light mode .

https://www.dropbox.com/s/ey72iiypat3ofgv/Mon-son-gemi.zip?dl=0

What I'm seeing:
All the cameras are really at an incredible point of DR, there is little difference, at the first clip and clipped on the chart.
What I see but we can discuss Monstro have something more on all in general , even the background noise is very good.

Post the results and discuss