quick exposure test emulating dragon with epic

[Ashley Stracke]

I'm aware that the sensor is entirely different and with that comes lots of improvements that can't really be simulated at all. But is there a "ballpark" setting I could put the Epic on in a low-light situation to get a sense of how much better of an exposure I would be getting if I was shooting with Dragon instead of Mysterium-X ? Basically, what I'd like to know is, can I set up a very low light scene on an Epic at ISO1000 and record a few seconds, and then change the shutter to a 24th (or lower) instead of a 48th, and therefore see a fairly good 'ballpark' of what the Dragon might be getting exposed in the same kind of light ?

 

[Noah Yuan-Vogel]

Probably not... Seems unlikely the difference is as simple as one stop more sensitivity.

 

[Mike P.]

Using HDRx +3 on MX you can get an approximation of the dynamic range of Dragon, but you best keep it limited to still images since motion artifacts would be a-plenty. Conversely, Dragon is ISO2000, which, in low-light, would look pretty damn grainy on MX (since your sensor would be comparatively starved, whereas Dragon's would still be within its "noise floor cleanliness range").

Honestly though, why not grab some of Phil's .r3ds from the "It's just Paint" thread. Play around with those to get an idea of the DR (although not low-light).

 

[Detlev Eller]

... the feeling of the image and its response to its environment is different ... so much so to not being expressed in math or comperative data.

It's a quantum leap that has to be experienced. no substitute ... ;-)

detlev

p.s.
"wouw!" was the word most used around the dragon - deep into the night, i should add ...

 

[Brett Harrison]

... the feeling of the image and its response to its environment is different ... so much so to not being expressed in math or comperative data.

It's a quantum leap that has to be experienced. no substitute ... ;-)

detlev

p.s.
"wouw!" was the word most used around the dragon - deep into the night, i should add ...

No other comment has made me as excited as this one. "deep into the night". That's where it goes on for me in the grading system, pushing and pulling it around for hours. I look forward to that moment with Dragon.

 

[Phil Holland]

This one is a tricky to quantify at the moment.

If you look at the straight up sensitivity improvement you're looking at an overall increase in performance and sensitivity by about 1.3 stops. Essentially what ISO 320-800 is to Mysterium-X, ISO 800-2000 is to Dragon. Mysterium-X can be taken up to ISO 2000 under certain conditions. I think we'll see Dragon go up to ISO 3200 under those similar conditions. But typically we'll working in those optimal ranges.

That's not exactly the whole story though. Interestingly enough, if you pull down Dragon material to say ISO 320 and sometimes further, it holds up significantly better than Mysterium-X material. In my little shoot a couple weeks ago I pulled one shot down to ISO 100 and it holds color accuracy nicely. Also, I added a fairly interesting grade to one shot that I pushed to ISO 2000 and purposefully underexposed (the CU on the mirror make up thing). The lights I pushed up to around ISO 4000 via an S curve while keeping the midtones and absolute highlights in the ISO 2000 range. It had a bit of texture to it, but was totally fine for a 4K finish IMO.

However, the sensor itself also has more dynamic range which effects what is captured. It's listed as 16+ stops. This could be viewed as an additional 1.5-2-ish stops in dynamic range on either side of your exposed midtones compared to the 13 stops found within Mysterium-X.

So in theory, potentially about 2.8-3.3 stops worth of improvement in low light sensitivity if you think of it this way. In terms of highlight protection though I'd say that's an easy 2 stops up there as well, which is why I think it's rated 16+. I personally think it's closer to 17, maybe a bit more in total. However, without the new color science being released yet, it's hard to know exactly how much juice is in there.

Remember, this in itself will completely rely on how you expose for what you want to see. You could expose to the right or left and have more room to play with depending on a given frame and lighting scheme. It's basically a lot more room to play with when exposing, which provides more power during capture and in post.

That's all guesstimate and theory. Not hard numbers. I will say this, you'll instantly notice how much more Dragon can see if you've been shooting Mysterium-X for some time now. It's a pretty immediate "feeling" thing. Noticeable in the dark and in the light.

 

[Gunleik Groven]

Agree 100% with Phil.

What took me back except for the measureable thingsm were (as Phil says) the way it holds colors in the full range.
There is a lot more chroma info in the Dragon image than... well anything I have seen. Well documeted in Phils "paint" piece.
Skintones are "improved"
Well how do you measure that?
My guess is that it has a bit to do with the next point

Compression is vastly improved.

Which, as some have said, can be connected with the lower noise floor. I think maybe it is just improved.

Precission and "fatness" of the images is wastly cooler all over the board. Even if it did not have the added DR and resolution over the Epic, I would have been excited to upgrade.
And then there is this odd "feel" thing. It just feels right.

As to seeing into the night/dark.
Dragon still needs some contrast to make images.
So we still need to know what we do in one way or another.
On set and in post
To get a useable image.

But to this day, it is the most fun camera I have ever used.

Side - by side, I guess the carracter of the images is what would tell you which is which, more than anything you could measure, if you shot within epics fatzone.

Thus... Hard to emulate, really

 

[Ashley Stracke]

So in theory, potentially about 2.8-3.3 stops worth of improvement in low light sensitivity if you think of it this way. In terms of highlight protection though I'd say that's an easy 2 stops up there as well, which is why I think it's rated 16+. I personally think it's closer to 17, maybe a bit more in total. However, without the new color science being released yet, it's hard to know exactly how much juice is in there.

So if I'm trying to get as much of a good noiseless exposure as possible and I'm not concerned with the "quantum leap magic" that people keep mentioning, and I just want to know if the shots I recently did in low light with Epic would have come out without objectionable noise if the scenes were re-shot using a Dragon, how could I test for such a thing? Would it be to rate the Epic at 800 and shoot at a shutter speed a bit slower than 24th to get some kind of approximation? I'm just looking at the noise levels in some natural light shots we had to shoot on location without any thing else. They were shot at ISO 1600 and looked great on the Red LCD but back at the office we realized we should have been more trusting of the on screen tools and less trusting of the overall image displayed on the LCD because on our computer and plasma displays the image was a bit too dark and noisy.

I know the Dragon is an entirely different thing, with a heart and soul, and a magic feel, etc, but for basic stuff like the limits of objectionable noise in a low light shot where highlights blowing out is just fine, what kind of difference could I try to somewhat closely emulate using the Epic?

I would think for judging shadow noise I could do one exposure as normal at 48th and 1600 ISO, and one at a slow shutter speed below 24th at ISO800 to emulate what Dragon might look like in the shadow noise areas when shot at 1600 at 48th


If I'm on the right track please let me know what a more specific shutter speed might be for this ''shadow noise test''

 

[Les Dittert]

When someone is allowed to post an r3d from each sensor of the exact same scene with the same lens, then we can see what's really happening.

 

[Matt Ryan]

When someone is allowed to post an r3d from each sensor of the exact same scene with the same lens, then we can see what's really happening.

Agreed. It doesn't matter what others "think" or "feel" about their experience with Dragon. It's all about being able to compare it to something to back up their "feelings." I'm getting tired of all the comments on what it's like because that means nothing to those of us who haven't used it yet... Most of the footage I've seen I don't like because it's not my style and not the way I would shoot. Unless there is data in the form of a comparison (that isn't a stylized piece), nothing matters and "feelings" are subjective.

Ashley, you won't be able to get an accurate idea regardless of what you do. The lower noise floor on the Dragon means just that, lower noise. Not only is it cleaner into the higher ISO's, it also have more dynamic range which will also give the appearance of more noise free latitude. The closest you can get to mimicking this with MX is to shoot during the day at 800ISO with +3 HDR. Then blend the two and you'll get an idea. Just keep in mind the Dragon images will still be cleaner and the compression cleaner as well. For low light you won't be able to emulate Dragon even remotely close.