ACS Weapon test

[Anton Shavlik]

I know it's not really important but for the DP who shared the video, it is OLPF not OPLF.

 

[Tom Gleeson]

Yes the Weapon gets the same Dragon sensor as the Epic but based on no hard evidence I suspect the sensor may have had some tweaks. How much the sensor is involved or the DSP in improved performance we will probably never know unless Red interjects but ultimately it is irrelevant. The only fact worth knowing is that the Weapon is producing a small but significant noise improvement.

 

[Björn Benckert]

There is probably good reason why red calls the sensor dragon and not unicorn... most likely it is still 2012 sensor tech, the Chinese year of the dragon.

 

[Brett Harrison]

Yes the Weapon gets the same Dragon sensor as the Epic but based on no hard evidence I suspect the sensor may have had some tweaks. How much the sensor is involved or the DSP in improved performance we will probably never know unless Red interjects but ultimately it is irrelevant. The only fact worth knowing is that the Weapon is producing a small but significant noise improvement.

Good point Tom: there could be a Unicorn sensor in there for all we know but what we can know is how the CF Weapon differs from the Epic Dragon.

 

[Brett Harrison]

There is probably good reason why red calls the sensor dragon and not unicorn... most likely it is still 2012 sensor tech, the Chinese year of the dragon.

Hehe yep Björn, I fully agree it's the Dragon sensor, just using colourful rhetoric.

 

[Björn Benckert]

Hehe yep Björn, I fully agree it's the Dragon sensor, just using colourful rhetoric.

That was fully understood

Feels a bit odd to wait for delivery of a 2012 sensor though. So I hope the 8k unicorn comes soon.

 

[Elsie N]

That was fully understood

Feels a bit odd to wait for delivery of a 2012 sensor though. So I hope the 8k unicorn comes soon.

That 2012 sensor may just turn out to be hard to top for the time being. And it's a matter for some debate, but I think it still leads the pack.

But yeah, 8k Unicorn... or Snowy Owl, or Pterodactyl or... whatever it's called...

 

[Tom Gleeson]

Brett,

The Weapon is seeing further into the dark than the Alexa that is in no doubt. Where it gets tough is making the call at what point does the noise levels make that extra picture information valueless? Thats hard to scientifically quantify and is more up to the individual shooter/projects aesthetic.

 

[David Battistella]

I dunno, rate it at 250/320 ASA, hold your highlights and I'm pretty sure your noise profile would be quite different.

Actually, I wouldn't even call what you guys did a test.

I would call it a "straight out of the box comparison" of which there are many.

Testing is more like what Phil does, where he takes the sensor and rates it through a range of ISO so you can see what it does. Then you could do that with the "reference" camera as well. And then that could be a basis point for knowledgable DP's to do their own project specific testing.

You post this from the ASC website as a professional organisation of image makers for an entire country and then make statements like, "i suspect they tweaked the sensor on Weapon"? While it is widely known that the Weapon electronics offer an improvement the machines carry the same sensor tech.

Specifics actually do matter when you are testing and when you represent an organisation of industry professionals.

Just my worthless .02 cents

David

 

[Brett Harrison]

That's a fair point Tom, which I guess is what makes for so many slightly different evaluations of dynamic range.

The point I am making is that, whatever the 'correct' ratio of signal to noise we label (edit lowest usable stop, you can see that there's at least a stop differential (probably 2) in the shadow response, which I believe is merely down to the bias imparted by the ARRI LogC to Rec709 transform (relative to RedGamma4).

Brett,

The Weapon is seeing further into the dark than the Alexa that is in no doubt. Where it gets tough is making the call at what point does the noise levels make that extra picture information valueless? Thats hard to scientifically quantify and is more up to the individual shooter/projects aesthetic.

 

[Kemalettin Sert]

I dunno, rate it at 250/320 ASA, hold your highlights and I'm pretty sure your noise profile would be quite different.

Actually, I wouldn't even call what you guys did a test.

I would call it a "straight out of the box comparison" of which there are many.

Testing is more like what Phil does, where he takes the sensor and rates it through a range of ISO so you can see what it does. Then you could do that with the "reference" camera as well. And then that could be a basis point for knowledgable DP's to do their own project specific testing.

You post this from the ASC website as a professional organisation of image makers for an entire country and then make statements like, "i suspect they tweaked the sensor on Weapon"? While it is widely known that the Weapon electronics offer an improvement the machines carry the same sensor tech.

Specifics actually do matter when you are testing and when you represent an organisation of industry professionals.

Just my worthless .02 cents

David

I, myself really liked this test and its more valuable than Phil's tests.Could you show me where Phil tested Dragon/Weapons against any other camera? Not a single one.We are all own Red cameras in this forum but not Alexa or F65 etc.Thats what we are wondering how Red holds up against these cameras.

 

[Phil Holland]

I, myself really liked this test and its more valuable than Phil's tests.Could you show me where Phil tested Dragon/Weapons against any other camera? Not a single one.We are all own Red cameras in this forum but not Alexa or F65 etc.Thats what we are wondering how Red holds up against these cameras.

I've been mostly restraining myself from posting such things, but honestly since Radiant Images and Dave Stump's test I didn't feel the need to do much more of that.

But perhaps I should just go ahead and post everything at this point.

The reality is if you don't test the various OLPFs and ISO Ratings you don't get a total picture. There is an answer for noise tolerance and especially when performing A to B comparisons.

Something fun to try. Grab a Weapon. Shoot ISO 400, 800, and 1600 with all 3 OLPFs. Examine sensitivity, noise, Dynamic Range, DR above and below metered/measured 18% Gray, Highlight-Roll Off, etc.

There's a bigger story to tell and if you've seen a few BTS shots from a couple of RED features you've likely noticed folks in the ISO 800-1280 range. I've been writing about that for years now, but it's sort of a magic spot on the Dragon sensor. Those DPs are seeing what I'm seeing for the most part.

These cameras are all about how you use them. More or less, my tests are more of an investigation with a bit of control to explore what the camera can do. My hopes are to show that accurately so others can achieve predictable results with their cameras.

If we want to talk about usability, practicality, and opinion I have a lot to say about various cameras. Some are good for me, some are "meh". I don't like crapping on people's parties if I can avoid it, but I have the hard DR measurements from every current digital cinema camera which have all been tested the same way.

However, I will say this, if there was a camera more suitable for my creative goals I would own it and be shooting with it. Weapon is the right choice for the work I do for a great deal of reasons. 6K, eventually 8K, in my opinion very good color, size, weight, customization, rigging, etc....

Varicam 35 for instance, decent camera good image, but Weapon is a better system in so many ways. I can't fathom a job where I'd choose it over Weapon.

The ACS is a useful comparison for what is shown. There's just a bigger story to be told if looking deeper.

 

[sam karr]

I dunno, rate it at 250/320 ASA, hold your highlights and I'm pretty sure your noise profile would be quite different.

Actually, I wouldn't even call what you guys did a test.

I would call it a "straight out of the box comparison" of which there are many.

Testing is more like what Phil does, where he takes the sensor and rates it through a range of ISO so you can see what it does. Then you could do that with the "reference" camera as well. And then that could be a basis point for knowledgable DP's to do their own project specific testing.

You post this from the ASC website as a professional organisation of image makers for an entire country and then make statements like, "i suspect they tweaked the sensor on Weapon"? While it is widely known that the Weapon electronics offer an improvement the machines carry the same sensor tech.

Specifics actually do matter when you are testing and when you represent an organisation of industry professionals.

Just my worthless .02 cents

David

I watched the test and enjoyed it. Basically they were saying that you can buy a Weapon safely, it's on par with the Alexa and is one of the few best cameras in the world. It made me want a Weapon. And then I read your comment....making me feel like if they said negative things on the camera. Come on, they just spent 25mn telling how good is the camera, the guy even owns it. Why posting some counterproductive comment like this one?

 

[David Battistella]

Sam,

All I am saying it that it is a straight up comparrison at 1 ISO with 1 OLPF in a "camera system" that has many useful ranges and capabilities. Personally, I don't care if anyone bashes the camera or loves the camera. I know what I know about this DRAGON sensor in the EPIC body with three different OLPF's and the lenses I use. That I have "tested" myself.

There are a lot of people who shit on Reduser and specifically "Redusers" and want "more serious" discussion about cinematography on the forum.

Then an organization which represents cinematographers, get the little things wrong in the thread, LIKE OLPF (optical low pass filter, which is essentially a critical component to the image path as it is the last piece of glass that filters the light spectrum prior to hitting the image sensor) instead of OPLF. Or, in the written part of the forum and not the video posted, make "presumptions" about the sensor and make factually incorrect statements about the Dragon sensors in Weapon bodies compared to EPIC bodies.

We know that there are improvements in onboard image signal processing but a professional cinematography organisations should not really be suggesting that the Weapon carries a different sensor. It is false (unless someone from RED confirms that the sensor in my camera is inferior).

The "test" is what it is. It is a comparison of manufactures recommended settings. There could be Alexa guys out there right now saying, "yeaj but if I rate the ALEXA sensor at ISO XXX then I will dig images out of the low end.

And if it sounds like I am being a bit tougher or unusually harsh, well it is because those organisations who carry the torch for an industry have to be held to a slightly higher standard.

So, it's not about good or bad reviews.

David

 

[David Battistella]

The ACS is a useful comparison for what is shown. There's just a bigger story to be told if looking deeper.

This is essentially the crux of what I am getting at and what I would consider "testing" over "comparing".

David

 

[Brett Harrison]

David, Tom's comments on RU aren't The Australian Cinematography Society - let's just keep it accurate.

 

[David Battistella]

David, Tom's comments on RU aren't The Australian Cinematography Society - let's just keep it accurate.

Ok. Thanks for that info.

David

 

[Elsie N]

I, myself really liked this test and its more valuable than Phil's tests.Could you show me where Phil tested Dragon/Weapons against any other camera? Not a single one.We are all own Red cameras in this forum but not Alexa or F65 etc.Thats what we are wondering how Red holds up against these cameras.

Surely there are other places where these tests are being performed by people who do not own either camera. That should give one an idea of the comparison (as long as you understand any shortcomings of the testing method.)

I've been mostly restraining myself from posting such things, but honestly since Radiant Images and Dave Stump's test I didn't feel the need to do much more of that.

But perhaps I should just go ahead and post everything at this point.

The reality is if you don't test the various OLPFs and ISO Ratings you don't get a total picture. There is an answer for noise tolerance and especially when performing A to B comparisons.

Something fun to try. Grab a Weapon. Shoot ISO 400, 800, and 1600 with all 3 OLPFs. Examine sensitivity, noise, Dynamic Range, DR above and below metered/measured 18% Gray, Highlight-Roll Off, etc.

There's a bigger story to tell and if you've seen a few BTS shots from a couple of RED features you've likely noticed folks in the ISO 800-1280 range. I've been writing about that for years now, but it's sort of a magic spot on the Dragon sensor. Those DPs are seeing what I'm seeing for the most part.

These cameras are all about how you use them. More or less, my tests are more of an investigation with a bit of control to explore what the camera can do. My hopes are to show that accurately so others can achieve predictable results with their cameras.

If we want to talk about usability, practicality, and opinion I have a lot to say about various cameras. Some are good for me, some are "meh". I don't like crapping on people's parties if I can avoid it, but I have the hard DR measurements from every current digital cinema camera which have all been tested the same way.

However, I will say this, if there was a camera more suitable for my creative goals I would own it and be shooting with it. Weapon is the right choice for the work I do for a great deal of reasons. 6K, eventually 8K, in my opinion very good color, size, weight, customization, rigging, etc....

Varicam 35 for instance, decent camera good image, but Weapon is a better system in so many ways. I can't fathom a job where I'd choose it over Weapon.

The ACS is a useful comparison for what is shown. There's just a bigger story to be told if looking deeper.

This being REDuser, I think breaking down the differences between OLPFs and settings is an invaluable test for those of us (the majority) who do not own Alexa cameras. Those who are subject to renting an Alexa vs Weapon are going to do their own tests anyway.

 

[Michael Tiemann]

Surely there are other places where these tests are being performed by people who do not own either camera. That should give one an idea of the comparison (as long as you understand any shortcomings of the testing method.) [...]

Disagree. Some of the out-and-out dumbest tests I've seen have been from people whose testing criteria are so arbitrary and obtuse that of course they'd never own the camera (or, in some cases, the lenses they put on the camera). What I have appreciated about Phil Holland's tests is that he not only looks at what happens at the extremes, but he also reports details that are meaningful in the middle. I've decided to ignore results from folks that look only at extremes and don't ground those findings with equally rigorous "normal" results. Speaking platitudes about "of course this is an extreme and you'd never do this normally" doesn't cut it. A valid test should cover the entire spectrum with equal care and diligence. There should be as much effort put into how good things can look as there is into how bad things can look. (Actually, much more effort should be put into the former than the latter, because it's always easier to do something wrong than it is to do it correctly.)

 

[David Battistella]

I am all for what you are saying and I am all for creating a set of exposure strategies. Of understanding how the sensor performs in different conditions and applying the chosen (creative) and tried strategy.

There are MANY ways to expose something and MANY factors (story, mood, tone, texture, style, etc) that go into an exposure decisions. So comparative tests, to me at least, have become quite banal. Many of the tests Phil posts , I do myself along with other ones.

I don't bother posting because he is more in the thick of it where he lives and his service to the community is so top notch and precise most other tests would end up seeming repetitive.

But you are right on here Michael, we should be talking about how to make things look great. That's kinda the point, isn't it?

David

Disagree. Some of the out-and-out dumbest tests I've seen have been from people whose testing criteria are so arbitrary and obtuse that of course they'd never own the camera (or, in some cases, the lenses they put on the camera). What I have appreciated about Phil Holland's tests is that he not only looks at what happens at the extremes, but he also reports details that are meaningful in the middle. I've decided to ignore results from folks that look only at extremes and don't ground those findings with equally rigorous "normal" results. Speaking platitudes about "of course this is an extreme and you'd never do this normally" doesn't cut it. A valid test should cover the entire spectrum with equal care and diligence. There should be as much effort put into how good things can look as there is into how bad things can look. (Actually, much more effort should be put into the former than the latter, because it's always easier to do something wrong than it is to do it correctly.)