Welcome to our community

Be a part of something great, join today!

Register 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 :)

Varicam LT vs Red Weapon Dynamic Range testing

I have a sneaking suspicion that even if we get the files to see for ourselves, and if they reveal that the Scarlet W clips before the Varicam and the noise in the shadows is more than the Varicam, that none of you will be convinced. But hey, I could be wrong!
 
luigivaltulini said:
exactly Kenneth .. Weapon is better in the shadows and ALexa in highlight at 800 iso , just set Weapon 1280/1600
and all be the same or better for Weapon.
So do you think Varicam LT would do better than ALexa and Weapon or Scarlet-w? I doubt! I want to see before judging the original files, then we can discuss.

Brian, please, let us can download the original files?
Click to expand...

Like I said, man. Rating the sensor at a higher ISO just redistributes the dynamic range. It just means the shadows would get noisier more quickly. You wouldn't gain any dynamic range. Maybe the highlights would clip at the same time as the Varicam, but the shadows would go dark more quickly.

The Alexa has boasted better range than Dragon for years; I don't see how it's that impossible for another camera company to figure it out.
 
Props for doing the test, since I'm in the middle of one, I now know how involved these things are and how they cost time and money like any shoot.

When looking at a test it's important to understand the aims of the test before critiquing it - there's no point in bashing a test for what it's not.

A couple of comments I'd like to make:

because the curves of V-log and redlogfilm are different it can be an aggravating factor for visual comparison.

As others have pointed out, RED debayers differently according the FLUT (floating point look up table) which to my understanding is a colorimetric transofrm to optimise the debayer for a given exposure choice/rating.

LOG was never meant to be a display format - nothing wrong with it, it's just not how we're meant to see it, so I feel that if the purpose of the test is a visual impression, we should focus on the display-referred colour science.

ISO is just another variable in the debayer equation - an interpretation of the underlying sensor data - it's a 'grading' of the RAW data if you will.

I wish that redlogfilm would be adjusted to have a flatter curve like the other manufactueres are using, but in the end it's down to the colourist to decide where everything ends up landing.
 
Last edited:
Kenneth C Merrill said:
Like I said, man. Rating the sensor at a higher ISO just redistributes the dynamic range. It just means the shadows would get noisier more quickly. You wouldn't gain any dynamic range. Maybe the highlights would clip at the same time as the Varicam, but the shadows would go dark more quickly.

The Alexa has boasted better range than Dragon for years; I don't see how it's that impossible for another camera company to figure it out.
Click to expand...

Never said the opposite for EPIC DRAGON, man.
I only said that Weapon Dragon is the same dynamic of 'Alexa or maybe better, now with Epic Dragon was not so you're right, so I have a little doubt ..
I never had time to compare my Scarlet-w with my ALexa, maybe I get the same results as Brian.

http://www.reduser.net/forum/showthread.php?141166-WEapon-Alexa-my-thoughts

So wait for Brian files.
 
luigivaltulini said:
Never said the opposite for EPIC DRAGON, man.
I only said that Weapon Dragon is the same dynamic of 'Alexa or maybe better, now with Epic Dragon was not so you're right, so I have a little doubt ..
I never had time to compare my Scarlet-w with my ALexa, maybe I get the same results as Brian.

http://www.reduser.net/forum/showthread.php?141166-WEapon-Alexa-my-thoughts

So wait for Brian files.
Click to expand...

Thanks for linking that other thread! Your test confirms what I've seen elsewhere about the Weapon. I'm hoping the Red Raven can keep up at least for the most part, with those results :)

So if Brian's methodology pans out, would that suggest that the Varicam could have even more range than the Alexa? That would be crazy. Of course, that's just luma levels. Part of the reason I really want to see Brian's images graded to "proper" exposure is to see how much color information is retained. The Alexa holds its color almost perfectly throughout exposure range, and from what I've seen, so does the Weapon. How does the Varicam handle?

Honestly the Varicam LT looks amazing. I wish the motion looked better. To my eye, it's super video-y, but that's just me. More or less dynamic range, the Red image looks more cinematic I think.
 
Brett Harrison said:
Props for doing the test, since I'm in the middle of one, I now know how involved these things are and how they cost time and money like any shoot.

When looking at a test it's important to understand the aims of the test before critiquing it - there's no point in bashing a test for what it's not.

A couple of comments I'd like to make:

because the curves of V-log and redlogfilm are different it can be an aggravating factor for visual comparison.

As others have pointed out, RED debayers differently according the FLUT (floating point look up table) which to my understanding is a colorimetric transofrm to optimise the debayer for a given exposure choice/rating.

LOG was never meant to be a display format - nothing wrong with it, it's just not how we're meant to see it, so I feel that if the purpose of the test is a visual impression, we should focus on the display-referred colour science.

ISO is just another variable in the debayer equation - an interpretation of the underlying sensor data - it's a 'grading' of the RAW data if you will.

I wish that redlogfilm would be adjusted to have a flatter curve like the other manufactueres are using, but in the end it's down to the colourist to decide where everything ends up landing.
Click to expand...

Great insight Brett, thanks for the information. It seems it's always the display format that gets everyone up in arms. What do you think would have been the best way to display this for comparison?


luigivaltulini said:
Brian, please, let us can download the original files?
Click to expand...

luigivaltulini said:
So wait for Brian files.
Click to expand...

I'll try my best, it's a good 26 GBs though.
 
With the RED you are of course comparing a curve applied to RAW to a curve that is baked in with compression on the Panasonic - So you aren't comparing the roll off of the sensors, just the roll off from REDLOG Film to V-Log. Without knowing the differences between these curves or going in and grading the results to stress test what parts of the image are actually clipped, you can't actually tell if clipping is occurring on the sensor at the same or different points.
 
Brian Nguyen said:
Great insight Brett, thanks for the information. It seems it's always the display format that gets everyone up in arms. What do you think would have been the best way to display this for comparison?
Click to expand...

In the case of RED vs Panasonic perhaps this is the best it gets unless you can have access to the linear sensor data from the camera? If you can, then it's worth putting them together under the same curve using ACES - not always possible. That or have a colourist manually match them to their best ability allowing them to adjust RAW settings.
 
The Devils in the detail when testing cameras but I would not be surprised if the Varicam, Alexa and Weapon sensors are within a half stop of each other.


When I purchased my Weapon I seriously considered the Varicam. The DR is excellent on it and I have also always been a fan of Panasonic colour science choices. Don McAlpine ACS ASC recently tested the camera against a Red Epic Dragon and an Alexa and he chose the Varicam for his latest movie. But as I have stated before arguing over camera specifications is of limited value as the top end cameras are offering similiar performances albeit each has its strengths. While some cameras may outperform the Weapon in one area its best to look at the big picture.


With the choice of four serious contenders in the market here are my thoughts as an owner operator and why I chose the Weapon.


1/ For all the bad mouthing and prejudice that goes on about Red generally Arri and Red are in the minds of most Producers the go to cameras. Panasonic Varicam presently is an unknown to most players and it will have to win back its reputation over time.


2/ At the time of my decision there was no stated upgrade path or any similiar process from Panasonic. Arri while it does offer upgrades there is never a clear roadmap and that's scary when blowing over 100k.


3/ For all its growing pains the R3D format is now well understood within the post world and remains a simple, amazing, flexible and efficient codec.


3/ Like many camera manufacturers Panasonic stuck to the standard S35 sensor size. The promise of a Full Frame sensor was a major factor in my decision. While I doubt that I would use the full width of the sensor for every project I find that ability to choose very compelling. In a competitive world I can "soon" offer something unique.


4/ Both the Alexa SXT and full size VariCam 35 are big and heavy cameras compared to the Weapon. While on bigger jobs this is probably an advantage on much of my work I prefer a lighter and therefore more flexible camera system. I know there is the Alexa Mini but will it be upgradeable to the new Arri chip and if so what cost? The Varicam LT has since been released and although it has some performance limitations compared to its big brother it looks good.


5/ Red have always offered an amazing cost to performance ratio advantage over its competitors although the Weapon CF saw most of that advantage wither it was still competitive. (As a side note the Varicam record media is more expensive than Reds Media.)


6/ Varicam's clean 5000asa is its most powerful difference. That's an approximate two stop advantage over my Weapon with the LLO Olpf. While most of my work would not require it that extra two stops would undoubtably be very handy on an occasion and for some natural light shooters a big win. Red please take note.
 
What we can not tell with this test is whether the clipping is happening because of the dc2/redlogfilm or if it actually clipped data on the sensor. At an 800 ISO rating I have found that it's about a stop and a half of overexposure protection. So what might "look" clipped is not actually clipped at all.

brian,

Do you recall at which point the stop lights came on? That is really tha only time any of the channels are actually clipping and when you are actually loosing data.

David
 
Kenneth C Merrill said:
Thanks for linking that other thread! Your test confirms what I've seen elsewhere about the Weapon. I'm hoping the Red Raven can keep up at least for the most part, with those results :)
Click to expand...

eh eh eh well depends ! I prefer Weapon or i prefer ALExa mmm ......
I choose the one that gives me more chance now in post production.
if I look at this still, clipping and noise, it is difficult to deny that Weapon is worse than ALexa.
Well play around with these in DaVinci. Cropped 2:35WS. no WB,no denoise, only have leveled the signal.

https://www.dropbox.com/s/wxd2a7lsu8u5r1d/alexaweap3.zip?dl=0


Kenneth C Merrill said:
So if Brian's methodology pans out, would that suggest that the Varicam could have even more range than the Alexa? That would be crazy.

Honestly the Varicam LT looks amazing. I wish the motion looked better. To my eye, it's super video-y, but that's just me. More or less dynamic range, the Red image looks more cinematic I think.
Click to expand...

Yep is possible.


Brian Nguyen said:
I'll try my best, it's a good 26 GBs though.
Click to expand...

Magic :) Thanks Brian ;)


Tom Gleeson said:
The Devils in the detail when testing cameras but I would not be surprised if the Varicam, Alexa and Weapon sensors are within a half stop of each other.


When I purchased my Weapon I seriously considered the Varicam. The DR is excellent on it and I have also always been a fan of Panasonic colour science choices. Don McAlpine ACS ASC recently tested the camera against a Red Epic Dragon and an Alexa and he chose the Varicam for his latest movie. But as I have stated before arguing over camera specifications is of limited value as the top end cameras are offering similiar performances albeit each has its strengths. While some cameras may outperform the Weapon in one area its best to look at the big picture.


With the choice of four serious contenders in the market here are my thoughts as an owner operator and why I chose the Weapon.


1/ For all the bad mouthing and prejudice that goes on about Red generally Arri and Red are in the minds of most Producers the go to cameras. Panasonic Varicam presently is an unknown to most players and it will have to win back its reputation over time.


2/ At the time of my decision there was no stated upgrade path or any similiar process from Panasonic. Arri while it does offer upgrades there is never a clear roadmap and that's scary when blowing over 100k.


3/ For all its growing pains the R3D format is now well understood within the post world and remains a simple, amazing, flexible and efficient codec.


3/ Like many camera manufacturers Panasonic stuck to the standard S35 sensor size. The promise of a Full Frame sensor was a major factor in my decision. While I doubt that I would use the full width of the sensor for every project I find that ability to choose very compelling. In a competitive world I can "soon" offer something unique.


4/ Both the Alexa SXT and full size VariCam 35 are big and heavy cameras compared to the Weapon. While on bigger jobs this is probably an advantage on much of my work I prefer a lighter and therefore more flexible camera system. I know there is the Alexa Mini but will it be upgradeable to the new Arri chip and if so what cost? The Varicam LT has since been released and although it has some performance limitations compared to its big brother it looks good.


5/ Red have always offered an amazing cost to performance ratio advantage over its competitors although the Weapon CF saw most of that advantage wither it was still competitive. (As a side note the Varicam record media is more expensive than Reds Media.)


6/ Varicam's clean 5000asa is its most powerful difference. That's an approximate two stop advantage over my Weapon with the LLO Olpf. While most of my work would not require it that extra two stops would undoubtably be very handy on an occasion and for some natural light shooters a big win. Red please take note.
Click to expand...

Tom always well balanced your post, I like :)
 
Last edited:
I think tests like these can be fun and done for fun... I don't think we need to test cameras like this anymore because almost any high end camera out there performs well behind the needs for most kind of imagery.

Tests I like to see is about production, what is the time to set up a camera, move a camera, rig a camera? How does they perform in deep cold and extreme heat? When rigging a car, how do they function then, are they flexible? What is the actual post production time from shoot to edit to finished master of a few shots? What's the resolution and what kind of outputs can you do?

Any experienced cinematographer will make any camera deliver superb images and you can't use two cameras in the same way at the same exposure. You use a camera as that camera is supposed to be used during a certain light, but you will probably always get good images out of any camera if you do it like so.

I just think there are too many comparisons between sensors and too few between the actual cameras.
 
Scarlet-W vs Varicam LT Dynamic Range Test

Scarlet-W vs Varicam LT Dynamic Range Test

oops
 
I think that each DP might want something different from a camera, sensor, and lens. For me, shooting dramatic feature films my first requirement is skin tone and how actor's faces are rendered. 2nd and 3rd for me are DR and then ISO sensitivity.

For other DPs the needs and requirements are different.

Yes one camera/sensor/lens setup that hits on all cylinders would be awesome but the night is young!
 
Peter Lyons Collister said:
I think that each DP might want something different from a camera, sensor, and lens. For me, shooting dramatic feature films my first requirement is skin tone and how actor's faces are rendered. 2nd and 3rd for me are DR and then ISO sensitivity.

For other DPs the needs and requirements are different.

Yes one camera/sensor/lens setup that hits on all cylinders would be awesome but the night is young!
Click to expand...

Definetely, but wouldn't you also agree that almost all high end cameras now hit DR and ISO sensitivity pretty high as well? For me, the whole "which has the best image"-discussion has hit an end or at least went into a blur. Some time ago there was a great difference between different cameras but nowadays, skin tones, DR and ISO feels maybe not the same, but enough from almost any camera. It's hard to NOT get the image right if you work with great minds, whatever camera you are using. I can realize my vision though most of them. I prefer Weapon because of the proxies and resolution, but last short film I directed was shot on Alexa and it looks wonderful because I had a great DP, even though we shot it at a lower resolution than I would prefer.

Problem is that most people seem to think about what the camera can do for them, not what they can do with the camera. That's why I own my own Weapon, I could rent one, but I want to experiment with this thing into the tiniest details in order to achieve perfection. Hard to do that if you only have your hands on one during each job.
 
Since highlight performance is way more important to me than noise-free shadows, I'm more interested in the results of the STH OLPF overexposure test.

The DSMC² body style improved upon the noise floor over the Epic/Scarlet body by 1/2 stop or so and I was fine with STH in those. I know I'm in the minority on that one.

The STD OLPF gives up a little overexposure latitude and gains a little shadow detail compared to the STH as part of an effort to be more things to more people.

Yes, as it appears with the STD OLPF in this test, it can be argued the Varicam LT might have a wider overall DR because it does better on both ends.

However, I expect a 4K S35 sensor to be cleaner and more sensitive at the low end than a 6K sensor because it presumably has bigger photosites. So, the most salient test results for me are those using Dragon's best highlight performance option - STH.

I can excuse the only marginally poorer shadow performance (when exposed correctly) of the STH OLPF than most 4K-or-less competitors given Dragon's higher count/smaller photosites and the physics of it all.

If pressed, there are the other two OLPFs from which to choose.

Anything beyond that is why they invented noise reduction. I believe people see it a lot more than they think they do.
 
Last edited:
You must log in or register to reply here.
Back
Top