ACS Weapon test

[Tom Gleeson]

I worked with the ACS Technical Committee and FXPHD and we have done an "Introductory" look at Red’s Carbon Fiber Weapon although much of the information will be familiar to regular Reduser readers. We also did some dynamic range tests to see how the tweaks on the Dragon sensor and upgraded DSP in the Weapon improved DR performance over Epic Dragon. We overexposed and underexposed our image in increments and then graded back to nominal exposure to determine where it begins to break. Alexa was included as its dynamic range is generally considered gold standard.


Our tests like others such as Phil Holland who have done tests with the Weapon have shown there are small but real improvements in the Weapons Dragon sensor.


Our tests did not provoke any CMOS smear and also demonstrated again the major gains in colour over the previous MX chip.

http://www.cinematographer.org.au/cms/page.asp?ID=21969


We decided to measure the comparative dynamic range of the cameras using a "real world" human face rather than charts. Where a camera fails especially on the bottom end is not an easy call but our results saw Weapon dynamic range improved over the Dragon by at least half a stop and just short one stop on the Alexa. This is a genuinely impressive performance placing the Weapon Dragon at the top end of cinema camera performance. We were impressed by the performance of the Weapon and the Alexa and how similar they are in some regards. Considering how old the design of the Alexa chip is, it remains an impressive achievement.

We couldn't cover every aspect of the Weapon as there so much good stuff in such a small package. Would be delighted to hear any feedback.

 

[Stacey Spears]

Thank you for sharing Tom.

What is the test chart with the 6 colors in a circle that is at the start of the video behind the model? I just started watching, saw the chart and wanted to know where I can get one.

 

[Tom Gleeson]

Stacey,

I think it was called the "Serious Chart" Unsure where it came from as it was supplied by FXGuide.

 

[Mikael Lubtchansky]

Thanks Tom and Mike for the very well done camera test and was happy to bump into #foolcontrol quick demo around 20min 40s into it

 

[Tom Gleeson]

Its very easy to love Fool Control. I suspect other camera manufacturers if their smart will follow Red's lead and allow similar access to their cameras.

 

[Brett Harrison]

The fraught condition of dynamic range ratings.

Thanks Tom and Mike, that is a great look at the camera along side the reference - perhaps not the reference for ever.

 

[Mikael Lubtchansky]

Its very easy to love Fool Control. I suspect other camera manufacturers if their smart will follow Red's lead and allow similar access to their cameras.

Thanks Tom, fwiw, upcoming free update will bring some more WEAPON / DSMC2 specific features to foolcontrol.

 

[luigivaltulini]

Thanks Tom, much appreciated our effort
I would be curious to see with the skin tone OLPF

thanks for your time devoted to testing.
nb: We can have some frame of test, where to download?

[/URL][/IMG]

 

[Kemalettin Sert]

Thanks Tom, much appreciated our effort
I would be curious to see with the skin tone OLPF

thanks for your time devoted to testing.
nb: We can have some frame of test, where to download?

[/URL][/IMG]

why skintone? He already used best olpf.Do you think with skintone olpf weapon could gain 1 stop more on highlights? Nope

 

[luigivaltulini]

why skintone? He already used best olpf.Do you think with skintone olpf weapon could gain 1 stop more on highlights? Nope

subjective! I always prefer the skin tone.
Especially if it fired at between 250-500 ISO.
I'm not very interested in having below.
And since Weapon be cleaner under, I would have liked to see the difference!
Obviously Skin tone would give better highlight same ISO and aperture.
But then again, I like the test was just to have another reference to compare.

 

[David Battistella]

One thing I can appreciate is shadow detail. With MX we always had to have concerns about highlight retention and that can also be considered on the dragon, but you could always dig images out of the shadows and the lower end of the exposure. With the STH the highlight retention is better. Controlling highlights is one thing but digging out nice images from the shadow area is also important.

On the other hand I do tend to see a lot of shadow detail crushed in "mainstream" films so I get why highlights are favoured. Either way, all of these cameras make gorgeous images in the right hands.

David

 

[Nick Morrison]

Thanks for the test Tom!

 

[Patrick Tresch]

why skintone? He already used best olpf.Do you think with skintone olpf weapon could gain 1 stop more on highlights? Nope

It could be 1/4 stop. But who cares, it's allready a Formula1 car.

 

[Tom Gleeson]

Luigi,

Would have loved to shot some tests with STH OLPF but we decided to narrow our testing to specifically Dynamic range. I have heard some strange claims About the difference in Dynamic range capabilities of different OLPFs. Different OLPFs may change the exposure difference between the RGB channels slightly changing DR but the main difference is they change the camera's sensitivity to a given exposure. As you point out at a given exposure with the same iris and ASA the STH will give you an extra stop in the highlights but the trade off is a stop less in the shadows. I too love the STH OLPF but the differences are subtle.

I have spent most of my holidays working on this test so will take a few days off but if there is more interest I can look into making the originals available but that's a lot of data and the underexposure tests need more than a frame grab so as you can see the grain move.

 

[Phil Holland]

interesting perspective and I enjoy the back and forth discussion a lot. There's a lot I can agree with on those fronts.

One thing that is extremely time consuming to test is to check across the entire usable ISO range of over exposure and underexposure versus noise character. Examining the effect of ISO 250-2000 in particular on all three OLPFs is important and effects the results of tests like these. Good to do this with all of the OLPFs as well as that does effect the Highlight Roll-Off to Clip as well:




Another thing I didn't hear brought up is format size and resolution, which is understandably not the focus for everybody in the world. Arri Open Gate on the Alexa is around 7.5 megapixels at 3.4K versus 19.4 megapixels at 6K. If considering the S35 format you'd be comparing 2.8K versus 5K essentially as well. Makes a difference for some.

Totally agree about testing. It's long, labor intensive, and detail oriented work. I certainly need time between tests to shoot pretty things to get me back to normal.

 

[luigivaltulini]

Luigi,

Would have loved to shot some tests with STH OLPF but we decided to narrow our testing to specifically Dynamic range. I have heard some strange claims About the difference in Dynamic range capabilities of different OLPFs. Different OLPFs may change the exposure difference between the RGB channels slightly changing DR but the main difference is they change the camera's sensitivity to a given exposure. As you point out at a given exposure with the same iris and ASA the STH will give you an extra stop in the highlights but the trade off is a stop less in the shadows. I too love the STH OLPF but the differences are subtle.

I have spent most of my holidays working on this test so will take a few days off but if there is more interest I can look into making the originals available but that's a lot of data and the underexposure tests need more than a frame grab so as you can see the grain move.

Sure Tom,
It was not a criticism respect on your test. I'm happy that people like you do this, and understand the difficulties.
I shot with my Alexa now , without color correction and WB, put a bit of denoise on all fast and tried to level image +- .. fast.
As I said I can extract slightly more with Alexa, but not much more.
I can see something like that in my tests of about 1/2 -- 3/4 stops difference with Epic Dragon vs Alexa.
In a real situation now I am sure that no one sees the difference, now confirmed , you and Phil, that Weapon be better than Dragon about 1/2 stop

Now it would be nice to do a Lut 3D to match the cameras eheheh very difficult thing
I am anxiously awaiting the arrival of my Weapon

Thanks again Tom

little fast test.

[/URL][/IMG]

 

[David Mullen ASC]

Isn't it "Weapon Vs. Epic" not "Weapon Vs. Dragon"? The Weapon uses the Dragon sensor.

 

[Tom Gleeson]

David,

Yes you are correct about the naming but although the Weapon and Epic Dragon share the same sensor I understand the Dragon sensor being put in the Weapon has been tweaked with improved CMOS smear and noise floor. Also the new DSP is a major influence on the Dragon chips performance in the Weapon. I do not know which is most responsible for the incremental but clear improvements in the Weapons DR over the Epic Dragon.

No worries Luigi I did not take your input as a criticism. It's totally plausible your tests would vary somewhat as there are so many variables. I am more than happy to admit that any test has serious limitations and in our case our results are specific to the shot we photographed and its specific tonal values and colors. We did not add any Denise or sharpening in our tests. We also struggled with determining the base line for the Weapon so decided to go with 800asa on the Alexa and 800asa on the Weapon but maybe 500 or 640asa may be better for the Weapon with the Standard OLPF? I just haven't had enough time with the STD OLPF to decide how I will rate it. While the choice of ASA doesn't change your total DR it will influence how far each direction you can go before it craps out.

 

[David Mullen ASC]

Trouble with "Weapon Vs. Dragon" is that someone is going to go to a rental house after seeing the test and say "Don't give me that Dragon sensor camera... I want the Weapon." Should be "Weapon Dragon Vs. Epic Dragon".

 

[luigivaltulini]

David,

Yes you are correct about the naming but although the Weapon and Epic Dragon share the same sensor I understand the Dragon sensor being put in the Weapon has been tweaked with improved CMOS smear and noise floor. Also the new DSP is a major influence on the Dragon chips performance in the Weapon. I do not know which is most responsible for the incremental but clear improvements in the Weapons DR over the Epic Dragon.

No worries Luigi I did not take your input as a criticism. It's totally plausible your tests would vary somewhat as there are so many variables. I am more than happy to admit that any test has serious limitations and in our case our results are specific to the shot we photographed and its specific tonal values and colors. We did not add any Denise or sharpening in our tests. We also struggled with determining the base line for the Weapon so decided to go with 800asa on the Alexa and 800asa on the Weapon but maybe 500 or 640asa may be better for the Weapon with the Standard OLPF? I just haven't had enough time with the STD OLPF to decide how I will rate it. While the choice of ASA doesn't change your total DR it will influence how far each direction you can go before it craps out.

Tom certainly, discuss, this is positive and should always be. Share feelings and experiences. This all to grow

Trouble with "Weapon Vs. Dragon" is that someone is going to go to a rental house after seeing the test and say "Don't give me that Dragon sensor camera... I want the Weapon." Should be "Weapon Dragon Vs. Epic Dragon".

David , my habit and still call Epic heheh