"simple and ridiculous" dragon vs Mx test..

[Phil Holland]

If you're curious, while it's hard to tell exactly without precise measurement, I'm seeing around 16.3 or so stops of DR. I won't measure/publish anything hard until the new OLPF arrives though. Luigi's images showcase where a bit of that lands in both the highlights and shadows. Something tells me that there's more than 16.3 stops in this particular shot even.

Significantly more highlight information retained. A bit more sensitive and more tonal variety/density in the shadows. That's the long and the short of it.

Chroma separation is just freaking impressive on Dragon, richer/purer even. Nice dense tonal variety too.

Now think of that extra DR in combination with a bit of HDRx in say a 18-20 stop scene like this one.

 

[Phil Holland]

Oh. One thing I should mention, for those who have experience working with Linear color. Make sure to check that out. I saw Gunleik did a bit. There's actually some interesting color information in the whole image when not wrangled into RC3/RG3/RLF.

 

[luigivaltulini]

ok buddy, Gunleik, MAtt, Denis
then how to proceed? Tomorrow I still have a bit of time.
give me perfect directions.

 

[paul schefz]

iso 2000 is 1 plus 1/3 stop from iso 800....so yes 1600 would have been easier...either stop down on the lens (the extra depth probably would not have made much difference in this case) or change the shutter speed....

regardless...thanks again for the test...the difference for me in this is that i actually prefer the epic...cleaner, less noise, i would never shoot this much detail at 10:1 so i am surprised at the clarity/detail....and all this with a slightly underexposed shot....
this test does not show the obvious DR advantage of dragon....
i am actually wondering if a color checker (towards the camera) or person in the frame would have made me like the dragon image better...

 

[Scott C]

Forgive me if i'm missing something dumb, but are the dots I'm seeing in the dragon image on the purple/blue squares hot pixels? If so I assume a blackshade would correct for that?

1:1 crop of the colorchart

Again in no way a controlled test, but a good indication of the different "feel" of the two cameras.

 

[Gunleik Groven]

ok buddy, Gunleik, MAtt, Denis
then how to proceed? Tomorrow I still have a bit of time.
give me perfect directions.

Hahaha

I will try.

You need an environment with controlled lights.

Actually I would prefer you to stop down by shutter, so that DOF dos not come into play as a variable on noise, but I realize that shutterspeed could be an issue too. I would accept that risk for 2 stops and otherwise image consistency... I do when I shoot.

For the consistency, it would be nice to shoot with the same compression. 5:1.

You do NOT need to change ISO in-camera to do the test., You can, but it is not necessary.

Shoot a "fat" within-range high-contrast scene with the Epic at 5.6 1/50
Nice with a Mac Beth in a darker part of the image, but not lower than you can see the colors in the Epic.

Swap cam and lens

Shoot the same scene with the same settings on the Dragon.
Change shutter to 1/100 shoot
Change shutter to 1/125 shoot
Change shutter to 1/200 shoot

To be able to control for shutter noise, it would be nice to have these three Dragon exposures too:

Shoot T8 1/50
Shoot T8+ 1/50
Shoot T 11 1/50

That way we can control for noise introduced by shutter changes.

THANKS!

 

[D Fuller]

It's the same thing as if you shot both at 800ISO with same exposure and pushed Dragon to 2000ISO in post. Same outcome since it's RAW.

This seems quite basic, but I think I see this misunderstanding of ISO fairly frequently in the discussions of Dragon. The ISO setting on a Red camera is "just metadata", but an ISO number has real meaning for exposure.

ISO is a standard reference for measuring exposure; it is a measure of the assumed sensitivity of the imaging device (sensor or film stock). The fact that for Red cameras, the sensitivity setting (ISO or FLUT) is applied in post (or in the monitoring chain) doesn't matter, ISO is still the working assumption about how sensitive the device is when you calculate exposure.

To expose a gray card to a given level at ISO 800 requires twice as much light to hit the sensor (+1 stop) as is required at ISO 1600.

 

[Gunleik Groven]

Mr Fuller!


I could embrace you!

hahaha

Actuall the comments I have seen from RED all point in the direction that the 1/125 exposure should be "as usefull" as the 1/50", given that the reference Epic Exposure at 1/50/800 is NOT underexposed.

That would also give validity to the 2000 is as useful as 800 on the Dragon.

The tests I have seen so far, have pushed the Dragon a lot more than 1,25 stops compared to the Epic.

 

[D Fuller]

Mr Fuller!


I could embrace you!

hahaha

LOL!

Well… without standards of measurement, these discussions just don't have any meaning.

 

[Matt Ryan]

This seems quite basic, but I think I see this misunderstanding of ISO fairly frequently in the discussions of Dragon. The ISO setting on a Red camera is "just metadata", but an ISO number has real meaning for exposure.

ISO is a standard reference for measuring exposure; it is a measure of the assumed sensitivity of the imaging device (sensor or film stock). The fact that for Red cameras, the sensitivity setting (ISO or FLUT) is applied in post (or in the monitoring chain) doesn't matter, ISO is still the working assumption about how sensitive the device is when you calculate exposure.

To expose a gray card to a given level at ISO 800 requires twice as much light to hit the sensor (+1 stop) as is required at ISO 1600.

I get it, I was just overcomplicating it....

 

[D Fuller]

I get it, I was just overcomplicating it....

Easy to do. Especially, it seems, when designing a test.

 

[Joel Arvidsson]

Why does mx look little hazy? Us is a mattebox on the dragon?

 

[Robert Ruffo New]

iso 2000 is 1 plus 1/3 stop from iso 800....so yes 1600 would have been easier...either stop down on the lens (the extra depth probably would not have made much difference in this case) or change the shutter speed....

regardless...thanks again for the test...the difference for me in this is that i actually prefer the epic...cleaner, less noise, i would never shoot this much detail at 10:1 so i am surprised at the clarity/detail....and all this with a slightly underexposed shot....
this test does not show the obvious DR advantage of dragon....
i am actually wondering if a color checker (towards the camera) or person in the frame would have made me like the dragon image better...

What are your thumbnails exactly of? What do you prefer in Epic? Your post is not clear.

 

[Robert Ruffo New]

Pretty crazy how Dragon holds detail and color on both ends. The MX low end is muddy and loses saturation. Pretty impressed with Dragon.

Would a simple grading curve not fix the MX image?

 

[luigivaltulini]

iso 2000 is 1 plus 1/3 stop from iso 800....so yes 1600 would have been easier...either stop down on the lens (the extra depth probably would not have made much difference in this case) or change the shutter speed....

regardless...thanks again for the test...the difference for me in this is that i actually prefer the epic...cleaner, less noise, i would never shoot this much detail at 10:1 so i am surprised at the clarity/detail....and all this with a slightly underexposed shot....
this test does not show the obvious DR advantage of dragon....
i am actually wondering if a color checker (towards the camera) or person in the frame would have made me like the dragon image better...

Paul, I'm sorry in what sense?
Epic Dragon is better to 80o iso, be cleaner.
Why prefer Epic?

 

[Matt Ryan]

Would a simple grading curve not fix the MX image?

try it haha. Doesn't work well with MX.

 

[Gunleik Groven]

I think Matt and I stumbled upon a few of the same issues...

 

[Hans von Sonntag]

Those shots are not ridiculous. Interesting and informative test. The only caveat is the flare in the MX shot.


IMO, the only meaningful way to compare pictures of different cameras is to transcode them into RGB with a similar log curve and similar exposure. I transcoded the MX and Dragon shots with 320 ASA (MX base-sensitivity according to RED), 5600 K, no curves and such applied, RC3 and RedLogFilm.

Both sensors seem to have a similar base-sensitivity. Meaning 320 ASA delivers with both sensors comparable results regarding brightness.


Spot he Dragon. Haha!

Hans

 

[Gavin Greenwalt]

Great simple test. One request if someone hasn't asked already would be an ETR test. Instead of using the extra resolution in the highlights I would be curious to see a shot where both peak the same and then show off the shadows. (Also 3:1 compression so that we can see the real noise not a de-noised image). It seems like most of the tests are matching the gray point but it would be interesting to see both ends by themselves.

 

[Gunleik Groven]

that is on my list Gavin, but that would be mid Feb, Luigi...