Meaning of 16.5 stops of Dynamic Range

[Christopher Probst]

At this point, I feel nit picking a stop or so in dynamic range is really unnecessary. BOTH cameras capture more DR than any display medium can exhibit. And BOTH cameras have such a broad DR that they really expand exposure options to cinematographers. Clearly Dragon offers a marked improvement in DR over the MX sensor. And I was REALLY happy with the dynamic range of the MX chip. Dragon also has improved the color response over the MX as well, both in saturation as well as in highlight color neutrality and roll-off.

Now, I was suprised by my first Dragon/MX torture test to see highlights clipping magenta on the MX, simply because the way I expose digital, I NEVER clip my highlights! Dragon has made great strides in this area specifically and I'm thrilled with that. And to me, Dragon's DR is very close to ALEXA, but without the "gauzy" feel of the Alexa camera. That is the big difference to me. I LIKE having the option of a sharp, clean image that if I want to make bloomy and gauzy, I CAN, but I'm not forced with it for every job.

Does Alexa still have perhaps a bit more dynamic range? Maybe... I really don't fucking care when I'm getting WELL OVER TWICE the resolution in 1/3 the size, weight, cost! Whatever you choose to believe the Dragon's dyname range to be, it's well beyond any necessary capture range for a cinematographer with half a sense about lighting/exposure. AND, the Dragon sensor is still in its infancy. I'm sure we will see improvements to its performace as this sensor matures and Red squeezes every drop of performance juice out of it that's possible.

When a new sensor comes out, of course there is need to test it and determine its performance specs. I certainly ran as fast as I could to quantify and compare it to my MX Epic so I had an understanding of how to utilize it. I have now been shooting exclusively on Dragons since October and shot my first Dragon material last JULY, so I'm feeling very comfortable with how I want to make the camera perform. The way I shoot it may be very different than each of you, but that's something that each of you has to grapple with. I NEVER take a camera at stock manufacturer recommendations. I test them to my shooting style and see what works best for me and the results I want to achive.

Go forth and do likewise gents.

 

[Gunleik Groven]

the forum won't let me attach the alexa sekonic profile

Did you manage?

let us not make the conspiracies larger than they are...

You could upload them to some server and post them with the embedd button. Really not that hard.

By now, though, the new OLPF and calibration is in place.

And not "postulate in your head"-ware.

Best

G

 

[luigivaltulini]

MArk Wuerthner,
These examples we've seen them so many times and we also discussed a lot, is boring cut and paste.
Make you a new "test Sekonic DR" between Dragon (new OLPF) and Alexa and show us the difference, so it would be more instructive. So we can discuss something new.

 

[Niko Pueringer]

But how can red claim 16+ stops when 14 stops alexa shows better DR?

The Alexa doesn't have 14 stops. It has around 18 stops when fully measured out. Of that, about 14 stops are 'useable'. The Dragon has around 16.5 stops when fully measured out, with a 'useable' amount that's a bit less.

People need to stop comparing two different types of DR measurement.

 

[Zeb B]

Does Alexa still have perhaps a bit more dynamic range? Maybe... I really don't fucking care when I'm getting WELL OVER TWICE the resolution in 1/3 the size, weight, cost! Whatever you choose to believe the Dragon's dyname range to be, it's well beyond any necessary capture range for a cinematographer with half a sense about lighting/exposure. AND, the Dragon sensor is still in its infancy. I'm sure we will see improvements to its performace as this sensor matures and Red squeezes every drop of performance juice out of it that's possible.

Word. Now let's move on

 

[Dunkan Wolf]

The Alexa doesn't have 14 stops. It has around 18 stops when fully measured out. Of that, about 14 stops are 'useable'. The Dragon has around 16.5 stops when fully measured out, with a 'useable' amount that's a bit less.

People need to stop comparing two different types of DR measurement.

I don't think most people know there are different ways of measuring DR, i sure didn't. Perhaps red should only advertise with usable DR, to avoid confusion.

 

[Mark Wuerthner]

Did you manage?

let us not make the conspiracies larger than they are...

You could upload them to some server and post them with the embedd button. Really not that hard.

By now, though, the new OLPF and calibration is in place.

And not "postulate in your head"-ware.

Best

G

Gunleik, I did manage but I still couldn't get the grahic on. I figured there was a limit on how many? Anyway, here it is. Its the same stuff from previous posts that other people put up. My point here is not to propose "conspiracy" theories but to just present the evidence that's out there and let people make their own assessments, as opposed to relying on other people's assessment and then taking issue with their assessment. Make your own opinion based on the evidence put before you.

 

[Nick Morrison]

the forum won't let me attach the alexa sekonic profile so you'll have to take my word on that one. Both the CML tests and the Sekonic meter tests show the same thing. Dragon has +6.5 above mid gray, -7 below (at that point you're into the noise floor. There are others on here that will argue that there's more below but my "opinion" is that in any realistic grade there will be nothing there.) Alexa has +7.5 above and -7 below. So, with this info, you can come to your own conclusion. It may not match a lot of other forum member's opinions but so what. There's a lot of agenda going on in this forum. Hell, that's what makes it so fun to read!:redface:

As many have said, those tests were old OLPF, old color science.

 

[Paris Remillard]

I'll start by reiterating that I don't care how much DR the Alexa and the Dragon have. I'm not nitpicking or saying that the Dragon doesn't have enough DR for me. It's got more than enough. I've rarely wished the MX had more. I'm only talking about how we label the DR.

And people keep responding to the question of the discrepancy in these numbers by saying, "Who cares? Dragon has enough DR and is smaller and cheaper, etc...! Isn't that good enough?" But that's not the point of this discussion. I know all of those things to be true which is why I own an Epic. And I'm sure Dragon does have enough DR and is great. The point is trying to come to a place where these numbers are consistent across different cameras and manufacturers. Even if testing methods vary, those varied tests can be performed on all cameras and across manufacturers lines allowing one to view results based on the same frame of reference.

I'm just trying to temper expectation and get everyone on the same page in terms of what we're seeing and how we label it. If we're not speaking the same language then it's very difficult to communicate.

And I understand that these numbers are a slippery topic and can be interpreted many ways and there are many other factors that go into the overall image chain, etc...

But just as with any situation, you'll earn more respect by under claiming and over achieving than by over claiming and under achieving.

Even if those achievements are very impressive, as is the case with Dragon. It will be diminished in the eyes of the observers if they expected more. And this 16.5+ number makes people expect more because of what has been seen with cameras that claim 14.

That's all. I'm not trying to piss people off. Sorry if it's ruffled any feathers.

 

[Matt Ryan]

Gunleik, I did manage but I still couldn't get the grahic on. I figured there was a limit on how many? Anyway, here it is. Its the same stuff from previous posts that other people put up. My point here is not to propose "conspiracy" theories but to just present the evidence that's out there and let people make their own assessments, as opposed to relying on other people's assessment and then taking issue with their assessment. Make your own opinion based on the evidence put before you.

That chart is a stop incorrect. The middle should not have a break like that so if we remove it and properly set it up, it should read 17.9 stops total most likely. Those reading should be linear so any gap and offset like that means it is inaccurate. Either way, total DR goes to Alexa, but I'd say useable goes to Dragon. I'd put the usable of Alexa around 14-14.4 and the usable Dragon at 14-14.8.

This is with the Old OLPF so I'd expect with the new that the highlights are equal. I know which one I'd choose...

 

[Paris Remillard]

Luigi, if I seem insistent it's probably because several times after I've posted images and numbers and quotes collected from various tests and threads, someone will then post something that contradicts that information or ignores or dismisses it without any foundation in fact or data. Just opinions or claims. So I've reiterated and reposted only to try to keep things based in empirical data and real examples rather than ideas.

I don't mean it to seem insistent or aggressive. I'm just trying to stay unbiased.

And as I've said, I very much appreciate the tests and information that everyone posts on this site. It's very valuable as is their time which was used to share their findings. So I hope not to seem ungrateful for that.

I'm just trying to get everyone speaking the same language. Which I think ultimately would save a lot of explanation on our part and confusion on the part of non-redusers. And also help improve red's image outside of this forum.

 

[Gunleik Groven]

Paul.
Relax.
For what it is worth: I did not plan to get the Dragon.
That is until I had the chance to play with it.

Have you used it?

 

[luigivaltulini]

Luigi, if I seem insistent it's probably because several times after I've posted images and numbers and quotes collected from various tests and threads, someone will then post something that contradicts that information or ignores or dismisses it without any foundation in fact or data. Just opinions or claims. So I've reiterated and reposted only to try to keep things based in empirical data and real examples rather than ideas.

I don't mean it to seem insistent or aggressive. I'm just trying to stay unbiased.

And as I've said, I very much appreciate the tests and information that everyone posts on this site. It's very valuable as is their time which was used to share their findings. So I hope not to seem ungrateful for that.

I'm just trying to get everyone speaking the same language. Which I think ultimately would save a lot of explanation on our part and confusion on the part of non-redusers. And also help improve red's image outside of this forum.

Paris ,
I understand up to a certain point ...
It saddens me to hear that not enough 16Dr (or 14.8DXO usable) Dragon , maybe it would be nice that you can try it as soon as possible to give you an idea of how best to Alexa.
I also find it hard to figure out where the hell you can use the full dynamic range when firing , especially on a stage with lights controlled ?
In my experience now with the new esternemante OLPF is difficult to burn high light , but even if they burn a splendid roll-off .
No one here on the forum despises Alexa , indeed it is a beautiful camera.
But you can certainly continue to discuss quietly , all this helps


Mark Wuerthner ,
thanks for your SEKONIC test , in this test are showing this :

Alexa 18.9 DR total
11.9 usable

Dragon 16.3DR total old OLPF
13.6 usable

well we 're saying that Dragon has more usable dynamic range about 1.7 DR!!! ,
then why do you claim the opposite?

However, this test is done poorly, the line is not closed ! ? ! ?
Surely you've done something wrong ! !
You've taken my old place and you just pasted the result ...
I had to remove it from the forum ... it was better .. and repost the correct one.

This should make you realize how hard it is to do testing, and each has different results.



¨¨¨¨¨I redid Alexa sekonic, well here are the results.¨¨¨¨
I find a way to remake soon with the new OLPF .

[/URL][/IMG]

 

[Paris Remillard]

Luigi, I think you misinterpreted my comments. I went back and edited to make myself more clear. I don't think that Dragon doesn't have enough DR. It's got more than enough.

For some reason everyone keeps making this about a competition between Alexa and Dragon. Or about how much DR one needs or how other things are more important than DR. Which wasn't what the original poster asked.

Everyone should go back and look at the first post.

The question was just, and I'm going to paraphrase, "why does the Dragon look to have around the same DR as the Alexa and f55, which are claimed to be and assumed to be 14 stop cameras. Shouldn't it have 2.5+ stops more? If not, then what does 16.5+ stops mean?"

 

[Nick Morrison]

Shouldn't it have 2.5+ stops more? If not, then what does 16.5+ stops mean?"

Paris, what is frustrating is this has become a legal argument now, not a technical one. As far as I know, no one has actually tested an UPDATED Dragon with new OLPF and Dragon Color against an Alexa, head to head. All the previous tests are with beta Dragons.

Until we can SEE WITH OUR OWN EYES how the new Dragon stacks up, getting into a war of semantics seems fruitless, and oddly detached from reality. We are like warriors trapped in a tower throwing books at each other - look at what this document says! Look at the runes in this text!!

Lets take a deep breath. Wait for REAL WORLD TESTS, and take it from there. Otherwise we are just descending into a never ending unanswerable pit of endless conjecture and madness. No one KNOWS what RED means by 16.5 (that could be what the sensor is CAPABLE of, though yet untapped). Until then, we need to verify for ourselves in the real world what Dragon can to. Anything less is empty words and idle banter.

 

[luigivaltulini]

Luigi, I think you misinterpreted my comments. I went back and edited to make myself more clear. I don't think that Dragon doesn't have enough DR. It's got more than enough.

For some reason everyone keeps making this about a competition between Alexa and Dragon. Or about how much DR one needs or how other things are more important than DR. Which wasn't what the original poster asked.

Everyone should go back and look at the first post.

The question was just, and I'm going to paraphrase, "why does the Dragon look to have around the same DR as the Alexa and f55, which are claimed to be and assumed to be 14 stop cameras. Shouldn't it have 2.5+ stops more? If not, then what does 16.5+ stops mean?"

hehe an endless struggle ... hehe

Paris, you're saying that the f55 has the same dynamic range dell'Alexa? 14 stop ..
Do you think the same??
Let's be honest Alexa has more.
Well, now we make a comparison F55, Red DRagon, testing Dragon has about 0.5 / 1 stop in more we could say ... 14 F55, 15 Dragon (old OLPF) ...
Well now you're saying it for a long time .. Phil in his post, in my examples that the new OLPF has approximately 1/1.5 stop in more in the high lights ....
Well then let's go from 15 to 16/16.5 DR,
Another interesting thing would be to use ADD, to see what happens ..
Nobody has made ​​some real tests, why?
Why are the images that come out of this sensor are fucking beautiful, and maybe that's just me and you "argue" to try to understand where these numbers are of DR, in a real situation where visual and we would not even begun to discuss

 

[Zeb B]

I'd rather be surfing porn than arguing about DR

 

[Gunleik Groven]

I'd rather be surfing porn than arguing about DR

It is potentially a more precise and productive activity...

 

[luigivaltulini]

Ehehehehee

 

[Ramesh Jai]

It is potentially a more precise and productive activity...

More a reproductive activity..