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 :)

real native iso of dragon?

M

Magnus Dennison

Member
Joined
Apr 30, 2012
Messages
16
Reaction score
0
Points
0
Hi All,
I have a scarlet MX and often shoot at 320iso as I find it to be considerably less noisy than 800iso (I accept a reduction of dynamic range for this cleaner image).

I'm wondering what the Dragon sensors cleanest iso is... and if 2000iso is where you get the most dynamic range?

Thanks!
 
It won't be clean at 2000 iso but yes if you can live with some noise you'll have the best DR results. I still shoot at 800 iso as I would with the Epic-x.

To answer the REAL native iso question, this is a tough thing. RGB channels don't even have the same native iso, and I'm not yet speaking about wich color temp you would shoot. But if you use 800iso and it suites your taste name it 800.

Pat
 
The popular consensus is if Dragon doesn't give you what you expect then it means: because user error son.
 
I believe maximum dynamic range with middle gray at the center of the exposure curve is at 800 - though there is room above and below that to move around without significant loss of DR.

Graeme expounded up this thoroughly in another thread.
 
Magnus Dennison said:
Hi All,
I have a scarlet MX and often shoot at 320iso as I find it to be considerably less noisy than 800iso (I accept a reduction of dynamic range for this cleaner image).

I'm wondering what the Dragon sensors cleanest iso is... and if 2000iso is where you get the most dynamic range?

Thanks!
Click to expand...

The accurate answers to the questions are:

Dragon is cleanest when rating/exposing for ISO 250.
Dragon has more Dynamic Range at every ISO rating compared to Mysterium-X because there is more total available Dynamic Range. You do "see more of" the Dynamic Range at ISO 2000. However, with Dragon and Mysteroum-X it's all about how that DR gets mapped for your particular ISO and eventual gamma or look.
ISO 800 is a nice middle-ish point where you have approximately equal tonal range above and below middle gray.
 
Thanks a lot everyone, very useful insight. I'm very much considering a Scarlet Dragon upgrade..
 
The reality is with new firmware that has been announced you probably won't be able to use this as a guide by the time the camera arrives. The new firmware will change whatever current rules we adhere to on rating the ISO. Also do you use denoise in your redcine/resolve/etc exports? That also changes the perceived grain levels.. I think its far less cut and dry then people make the issue out to be, every workflow is a snowflake and I would test it. You may very well find that the denoise feature and the bit rate of your encodes affects the perceived noise levels. With so many flavors (and corresponding bit rates) of prores and dnxhd its easy enough to run some encode tests and play with settings. There are more variables to worry about then which ISO you expose for in camera that is making the issue very cut and dry when you don't factor in the other variables I attempted to list off...
 
Hi Brian, I totally agree that users should test their gear and develop the way that they want to work. I've had my scarlet for over 2 years now and after trying every workflow that's possible (and some that aren't) I have settled upon my favourite. I like to get the cleanest+purest image out of the camera that doesn't require any de-noising or sharpening. I use lights/NDs/Grads to manage dynamic range like I had to with the lesser cameras I owned previously. It may not be the way everyone works, but that's fine with me.

I' not suggesting that ISO is the most important factor in my workflow but I would say it matters a lot. Enough to warrant a discussion about it in isolation.

The reason I asked this question is it seems Red defines their 'Native ISOs' based on a compromise between dynamic range and noise level. What I wanted to know is if the Dragon sensor's 2000iso is a similar compromise. I think others have confirmed that it probably is. So that helps me decide whether to buy the upgrade or not.
 
They said 2000 was the native ISO previously, but that was before the new OLPF/dragon color (and subsequent red channel noise issues). I am not sure what is the optimal answer given these new variables, the current popular answer is rate it at 320 or so is what most people say, but I think if your doing an upgrade and not a new purchase surely this issue will be solved by the time your camera arrives, and the theory then might very well be different when it comes to ISO choice. Also what redcode do you use for compression? I notice that the more lossy the redode the more the gain/grain pops up as well...
 
elisey gladnikov said:
So, native ISO is 250?
Click to expand...

Depends on the OLPF.

The Skintone OLPF is rated 320. Great for studio and outdoors (where the lower ISO is very CONVENIENT)

The Lowlight OLPF is rated 800 ISO and looks like it can shoot clean at significantly higher ISOs, which should be VERY useful for lighting challenged scenarios! The KNICK was shot as 1280 I believe, and looks GREAT (on the first version of this OLPF, looks like it's gotten even better since then!).

It also appears like RED will be making a "Standard" OLPF, rated 640 (ie in-between both OLPFs above).
 
Excited for the dragon standard OLPF but also holding out foolish hope that they'll fix the red orbs on the low light olpf and come out with a sort of a... low light v2 olpf
 
Stephen Franchek said:
...also holding out foolish hope that they'll fix the red orbs on the low light olpf and come out with a sort of a... low light v2 olpf
Click to expand...

They already did that, no? The Low Light Optimized OLPF is the updated version. The orb issue is fixed.
 
Brandon J.F. said:
They already did that, no? The Low Light Optimized OLPF is the updated version. The orb issue is fixed.
Click to expand...

Definitely not fixed on my interchangeable OLPF with LLO. Just pan across a bright object like the sun and at some point the red orbs grid shows up. If its fixed, but I got an un-fixed LLO installed (this was just a couple of weeks ago), I'm gonna be super pissed. Just sayin'.
 
Nikhil Kamkolkar said:
Definitely not fixed on my interchangeable OLPF with LLO. Just pan across a bright object like the sun and at some point the red orbs grid shows up. If its fixed, but I got an un-fixed LLO installed (this was just a couple of weeks ago), I'm gonna be super pissed. Just sayin'.
Click to expand...

What stop were you shooting at?
 
Brandon J.F. said:
They already did that, no? The Low Light Optimized OLPF is the updated version. The orb issue is fixed.
Click to expand...

Nikhil Kamkolkar said:
Definitely not fixed on my interchangeable OLPF with LLO. Just pan across a bright object like the sun and at some point the red orbs grid shows up. If its fixed, but I got an un-fixed LLO installed (this was just a couple of weeks ago), I'm gonna be super pissed. Just sayin'.
Click to expand...
I think you guys are talking about two separate issues.
According to Phil's tests, the red orbs around highlights were visibly improved on the LLO (vs "V1 OLPF") but not eliminated completely, and the dot grid pattern still exists on the LLO.
 
Nikhil Kamkolkar said:
Definitely not fixed on my interchangeable OLPF with LLO. Just pan across a bright object like the sun and at some point the red orbs grid shows up. If its fixed, but I got an un-fixed LLO installed (this was just a couple of weeks ago), I'm gonna be super pissed. Just sayin'.
Click to expand...

Not 100% certain this "red grid" is the same as the "orb" issue...as it sounds a lot like the red dot sensor pattern we see occasionally on MX. The "orb" issue was the red "halos" around bright sources yes?

Often with the red dot pattern, stopping down would help it go away. Was that the case with this grid pattern?

thanks!
 
Nikhil Kamkolkar said:
Definitely not fixed on my interchangeable OLPF with LLO. Just pan across a bright object like the sun and at some point the red orbs grid shows up. If its fixed, but I got an un-fixed LLO installed (this was just a couple of weeks ago), I'm gonna be super pissed. Just sayin'.
Click to expand...

What's your F stop? is it above F11? Like F16 or above or something? Are you using ND? Can you post a sample?



Battistella
 
Yes - seems some confusion here... red orb is a red/purple 'orb' around light sources (that can bleed into/around the image)... the red grid is caused by stopping down and shooting at narrow stops (rather than using ND for outdoor shootings) and isn't LLO related (the MX did it).

On the subject of the Knick, was the whole thing shot at 1250? Any minks to more info, would be intrigued to read more.
 
Nikhil Kamkolkar said:
Definitely not fixed on my interchangeable OLPF with LLO. Just pan across a bright object like the sun and at some point the red orbs grid shows up. If its fixed, but I got an un-fixed LLO installed (this was just a couple of weeks ago), I'm gonna be super pissed. Just sayin'.
Click to expand...

As some here have said this is a separate issue you won't see this in less you're shooting a very high f-stop
 
You must log in or register to reply here.
Back
Top