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 :)

RED DRAGON Scores Highest DXOMark Score ever.

Brook Willard said:
Total number of visible stops versus total number of usable stops will yield a much more dramatic difference between the MX and Dragon sensors. It's not just about how many stops we can see in a single frame, it's about how usable the darker stops are given their noise levels. The Dragon really excels in this regard.
Click to expand...

Spot on.
Then add usability of the top stops, colorprecission and saturation in the full range.
And throw in some extra pixles as a bonus. :)

"in-between" I will post a few camera-comparisons here (still pretty jammed in real-work, though, so it will come as time allows):
http://blog.gunleik.com/2014/03/03/epic-and-dragon-comparison-part-1-overexposure/

In this first part, the highlights handling (not only "where it clips" difference) becomes pretty visible... :)

And Epic is accepted for stock-delivery for stills.
 
Last edited:
Brian Boyer said:
I see what you're getting at but, seeing how RED specifically markets the "stills" part of their camera system, I'd argue DxO DID test Dragon against RED's competitors - Nikon, Canon, Phase One, Sony, etc.

It's the overall best stills camera they've ever tested. Maybe Alexa would score higher than Dragon in some areas but its resolution (and size) make it unsuitable for a lot of serious print work. The F65 is also too big for any serious print work.

This test demonstrates that Dragon's images are in the same league as the best stills cameras in the world and blaming the camera is going to make people look less competent and should be increasingly less tolerated as an excuse for poor imagery.
Click to expand...

2K is not enough for most stills work, so Alexa would not work.
 
Congratulations, Red.
Jim, Jarred, Graeme and all the rest - Job well done.

Now to get a bit more technical:
Considering that an EPIC-M RED DRAGON GUNNER COLLECTION costs $42,360, it's a very good deal.
Just add a lens, black-shade the sensor and you're ready to shoot.

I don't care that Red employs some frame-combining technics. It's the result that counts.

I've seen frame-combining in Sony mirrorless cameras, but nothing that matches the speed of sensor read-out of RED EPIC Dragon.
Those ASICs are amazing.

I'm sure other manufacturers (Sony, Canon, Nikon, Fuji) would employ frame-combining to a RAW still, once they get the electronics fast enough to do so.
It wouldn't be considered a "pure" RAW file, but with such amazing results - should we even care?
 
Brook Willard said:
Total number of visible stops versus total number of usable stops will yield a much more dramatic difference between the MX and Dragon sensors. It's not just about how many stops we can see in a single frame, it's about how usable the darker stops are given their noise levels. The Dragon really excels in this regard.
Click to expand...

Preaching to the converted mate :)
 
Congrats Jim, Jarred and all the RED team. Can't wait for my DRAGON.
 
I saw the news this morning and it was a pleasant surprise. Yes, the DSLRs are cheaper, but the MFD cameras are more expensive (as noted above). And the RED whipped 'em both.


But a caveat about DxO scores. Leicas tends to score lower than you'd expect. Check out the M9's score for example. Yet the people who actually look at the finished product will not accept a DSLR over an M9. And if you think the D800 is a match for the S2... forget it. It's great, but it's not equal. So why do the Leicas score low, but are revered by users? Noise at high ISO, I imagine. The M9 has 'only' 12 stops of DR - but slide film has about half that, and yet for many, Kodachrome is the gold standard for image tonality, colour and depth.


What would Kodachrome 64 - or EXR 5298 - score on DxO?
 
That Jim Jannard devil! I remember he mention D800 sensor in one of his latest post. And now he beat it, and from the video vs still territory!!!

That is absolute INSANE!

Congrats Jim, you really delivered some awesome things!
 
The next big step is to balance RGB beam out of the sensor with upcoming firmware upgrades.

And then we are there with Dragon at the top of the mountain!!!!!

RGB signal out of the sensor should be balanced in a way that the peak points of each color channel (RGB) make a horizontal line (as possible)!!!!

Right now the line is something about 45˚ to Y axis and it should be that all three colours have the same level of vertical response .

For example RED signal has value of 1 and also other two GREEN and BLUE should have the same level (1) on Y axis.

In that case it much easier to de-mosaicing or debayering algorithm to work properly, OLPF getting less issues and overall colour gamut is much wider.

L_E19.tmp.png

DXO Mark Test: Red Epic Dragon sensor color response. The 3 channels are normalized with the highest sensitivity measured
 
Jean Maurizio said:
That Jim Jannard devil! I remember he mention D800 sensor in one of his latest post. And now he beat it, and from the video vs still territory!!!

That is absolute INSANE!

Congrats Jim, you really delivered some awesome things!
Click to expand...

These numbers hardly say anything about the picture quality. D800e photos still look better than motion stills from Dragon.
Hopefully Red will announce a 645 cam @ nab, that would really kick ass. :D
 
Jarred Land said:
The Dragon rose above all still cameras tested before it, including the best full frame and medium format cameras.

QUOTE]

And Red can be !

However, ranking last in low-light performance. Due to the choice of more, smaller photosites ? Please Red, hear us. EFP/ENG users can simply not fully - if ever - benefit from the best sensor in the world if the light is too scarce to shoot with acceptable noise...
Click to expand...
 
Dunkan Wolf said:
These numbers hardly say anything about the picture quality. D800e photos still look better than motion stills from Dragon.
Hopefully Red will announce a 645 cam @ nab, that would really kick ass. :D
Click to expand...

DXO don't take in consideration the resolution, and a D800 still is twice the Dragon. Red accomplished its goal anyway, they broke the chart scoring a 101, that is a serious thing.
Regarding a medium format Red, it would be awesome, but maybe (and i say maybe given what i see, but i don't have direct knowledge about this matter) they would encounter a media problem. New mini mags record 6k at maximum 1/5 (or 1/4). With twice the resolution the compression, especially for 100FPS would be much higher than 13/1. Maybe it's not the time yet for such a monster, but who knows, RED is always some steps ahead ;)
 
It is not unlikely that a medium format camera would require different media... :)
THAT is probably the least of challenges. At least a lesser than a lot of others.

100 FPS @ medium format. I didn't even think that thought...
(Which probably means Jim and Jarred & co are all preparing for it)

I would love a stills FF 4:3 sensor. But that is maybe considered a compromise... :)
 
+1 on 4:3 FF, so anamorphic 2x can be used, or 1.33 for 16/9 :)
 
Denis Buhot said:
However, ranking last in low-light performance. Due to the choice of more, smaller photosites ? Please Red, hear us. EFP/ENG users can simply not fully - if ever - benefit from the best sensor in the world if the light is too scarce to shoot with acceptable noise...
Click to expand...

Let's correct this right now. Dragon is far from last in low light performance. They have tested the sensors in hundreds of cameras; Dragon is 10th on that score, while coming in #1 in every other category.
 
correct, sorry

correct, sorry

D Fuller said:
Let's correct this right now. Dragon is far from last in low light performance. They have tested the sensors in hundreds of cameras; Dragon is 10th on that score, while coming in #1 in every other category.
Click to expand...

too quick reading... thank you for correcting me. Anyway, still some truth in my point :-)
 
You must log in or register to reply here.
Back
Top