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 :)

Dragon in the UK

Awesome news John....THANKS.

People will save a lot of money by not having to buy IR filters.
 
Hey John, Thanks for the info. How is apparent noise vs. MX? Would love to see some charts of profiling Dragon noise up to 6400ISO if you have the time. Really interested to see if 2000ISO is truly "clean." Curious also how Dragon performs at 250ISO in a daylight situation to reduce ND on set.
 
I need to run a few things past Jarred... see whats good to share. I have in front of me a set of shots of a regular day exterior (no pretty girls I'm afraid) exposed with middle grey at the same point in the image, at everywhere from 250ISO to 12800ISO. On that image I'd happily use the range 400 to 4000 for HD delivery and 400 to 2000 for 4K delivery. I prefer between 800 and 1600 for its combination of sufficient headroom and superclean shadow... but with Dragon this is all much more down to taste than technicality. There could easily be an aesthetic argument for ISOs beyond 4000.

Low light is a different story, and one that I hope to cover shortly. Jamming tests around real work :)
 
250 works, but the rolloff to a strong highlight is that bit rapid for my taste... and you're more likely to have to correct the tone of any blown areas too.
 
John Marchant said:
Something less pretty, but very interesting. This should be self explanatory with any luck - you're looking at a tungsten lamp split into a spectrum with a custom prism rig I built for this purpose. IR contamination is the stripe to the right of the red...

Dragon kicks ass on this front. Comparable IR contamination to Epic-X only arrives by the time you have 7 stops of regular ND added, AND the grade pushed a couple more stops... So you can put all the ND you own up front, miss your exposure by a country mile, and still only have as much IR as the old camera.

Short version: IR is not an issue.

I'll run this under true daylight too, but from experience a tungsten lamp only four feet from camera tends to be worst case for this test.

dragonirtungsten.jpg


By the way, don't stress about grain in the image - this is a crop from a tiny tiny area of the sensor.
Click to expand...

Comporable to teh Epic-X under what conditions? With zero ND do you still see IR on current MX using this test? I am also unfamiliar with this test - ins't IR very heat-dependant (hot day gives more)?
 
John,

Thank you so much for sharing your tests. As far as new Color science Graeme is of course creating new color science for the Dragon. It is a completely new sensor with new boards and he tweaking away. It is pretty damn good but it will continue to change as Graeme's mad scientist (and I mean that in the best possible way) genius works its magic.

But as a few of have said, "out of the box" the Dragon looks pretty cool......

That's all,

Cheers,

Peter
 
Thanks John

Running through my testmaterial here, too... ;)

And can just confirm all you say.

I think the 800-1600 range keeps everything "in control" if you have a good image there, post will be good independently of how you whitebalance.

If you push it beyond that on set...
Dragon is rarely "bad" in anything I have seen, but light and AD is still physics despite "new magic pixies" at work.

Get a good image, and it will only get better

Cheers

G
 
Robert Ruffo New said:
Comporable to teh Epic-X under what conditions? With zero ND do you still see IR on current MX using this test? I am also unfamiliar with this test - ins't IR very heat-dependant (hot day gives more)?
Click to expand...

I posted some spectral response test a while back comparing X to Mono. They're still online here
As you can see, there's IR pollution on X with zero ND. Worlds apart from Dragon.

And IR as it concerns us here isn't heat dependent as such. Its near-infrared that causes the issue, and that's still pretty much light in the way we normally understand it. Far-IR, as sensed by a thermal camera is a very different beast, and doesn't play with the optics we use for visible light... nor does it interfere with the sensor. (otherwise the sensor's own temperature would dramatically 'expose' the image)

The observation that IR and heat are related is circumstantial in so far as it is associated with bright light - bright sun and bright artificial light means heat, most of the time... but you could observe IR pollution in bright arctic sun, for example.
 
Peter Lyons Collister said:
John,

Thank you so much for sharing your tests. As far as new Color science Graeme is of course creating new color science for the Dragon. It is a completely new sensor with new boards and he tweaking away. It is pretty damn good but it will continue to change as Graeme's mad scientist (and I mean that in the best possible way) genius works its magic.

But as a few of have said, "out of the box" the Dragon looks pretty cool......

That's all,

Cheers,

Peter
Click to expand...

No problem Peter. I hope techy stuff like this can be valuable to us all in making informed decisions on how best to shoot. Lots of 'from scratch' learning to do, even though its arguably an evolution of something we understood... some critical elements are different to a revolutionary degree.
 
John Marchant said:
No problem Peter. I hope techy stuff like this can be valuable to us all in making informed decisions on how best to shoot. Lots of 'from scratch' learning to do, even though its arguably an evolution of something we understood... some critical elements are different to a revolutionary degree.
Click to expand...



100% agree!


What you post is VERY valuable to all of us John!

Thanks for plowing the grounds!
 
John Marchant said:
I posted some spectral response test a while back comparing X to Mono. They're still online here
As you can see, there's IR pollution on X with zero ND. Worlds apart from Dragon.

And IR as it concerns us here isn't heat dependent as such. Its near-infrared that causes the issue, and that's still pretty much light in the way we normally understand it. Far-IR, as sensed by a thermal camera is a very different beast, and doesn't play with the optics we use for visible light... nor does it interfere with the sensor. (otherwise the sensor's own temperature would dramatically 'expose' the image)

The observation that IR and heat are related is circumstantial in so far as it is associated with bright light - bright sun and bright artificial light means heat, most of the time... but you could observe IR pollution in bright arctic sun, for example.
Click to expand...

That has not been my experience. I could show you a series of shots under a bright, Atlanta sun where the fabrics start out Ok and then turn magenta as the day progresses. Same Epic X, same lens, same f-stop, same (1.2) IRND, and very little actual light change (11 AM - 12:30 PM) The difference really was how long the actors had been out under that sun. Extras that arrived later have the same normal fabrics at first, but they then turn magenta as they heat up. In fact you could tell how long an actor wearing black had been out in the sun by the level of magenta of their black synthetic fabrics.

Shooting in Montreal in the winter - no real IR issues, ever, even at noon - but at sunset on a very hot day - lots of IR unless you use Formatt products.

So will all due respect (and that is a lot of respect) direct observation tells me different.

That said I'm sure Dragon fares better than Epic-X, and this is most welcome news.
 
Here's the daylight version of the same chart. Better even than under tungsten.

I added a mega-pushed version of the IR pass spectra, just to illustrate that you could shoot real IR images for day exteriors if that's your thing. You need ten to twelve stops of overexposure compared to regular imaging to make a deep-red / near IR image. Of particular note is that the 720nm range IR pass filter is actually totally blocked in the very near IR - all the image it does make is in the deep red visible range.

dragonirdaylight.jpg


Note: These are aligned less precisely than the tungsten ones only because this is done with direct sunlight, and I have to chase the sun across the sky to do it ;)
 
Robert Ruffo New said:
That has not been my experience. I could show you a series of shots under a bright, Atlanta sun where the fabrics start out Ok and then turn magenta as the day progresses. Same Epic X, same lens, same f-stop, same (1.2) IRND, and very little actual light change (11 AM - 12:30 PM) The difference really was how long the actors had been out under that sun. Extras that arrived later have the same normal fabrics at first, but they then turn magenta as they heat up. In fact you could tell how long an actor wearing black had been out in the sun by the level of magenta of their black synthetic fabrics.

Shooting in Montreal in the winter - no real IR issues, ever, even at noon - but at sunset on a very hot day - lots of IR unless you use Formatt products.

So will all due respect (and that is a lot of respect) direct observation tells me different.

That said I'm sure Dragon fares better than Epic-X, and this is most welcome news.
Click to expand...

I'll test this asap on an IR converted camera here... I have IR converted Epics that will show exactly what's going on. I have a black bag I often use for IR comparison - its half manmade, half natural fibre and shows in IR as a black and white bag ;)

Maybe I'll put it in the hot sun, and then hit a section with cold spray... see what happens :) Always willing to be proven wrong in the name of learning.
 
You must log in or register to reply here.
Back
Top