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 :)

Helium Dynamic range

R

Rohan B

Well-known member
Joined
Feb 18, 2011
Messages
371
Reaction score
0
Points
0
Hi guys...
i was wondering what is the dynamic range of the new 8k helium sensor compared to the dragon 6k. The spec listed during the sale was "16.5+ stops", which is the same as Dragon. But that can't be true, that all it increases is the resolution. So i would like to know what other improvements does this new sensor tech bring apart from resolution like DR COLOUR LATITUDE HIGHLIGHT ROLLOFF etc. Dragon was +2 stops over MX, does the case with Helium is same, but that would put it at 18.5 stops.....

would love to get the some reply and info on this from the Awesome Red Magicians.....................
 
Have there been any .r3ds released yet? Kinda wanted to take a peek. Also interested in low light performance... actually has anything been officially announced?
 
RED penned 16.5+ stops

notice the plus

my suspicion is this is a cloak and dagger type deal in terms of not freely giving the competition information.
Dragon in WEAPON added roughly a half-stop, Helium may or may not add another half to full stop.

Or it may truly be exactly/roughly the same as Dragon in WEAPON, but in a smaller S35 form - which is an achievement in its own
 
is helium considered a weapon/dragon upgrade? .. i figured it was just all about matching s35 glass to an 8k sensor
 
konrad grant said:
is helium considered a weapon/dragon upgrade? .. i figured it was just all about matching s35 glass to an 8k sensor
Click to expand...

It's a new sensor since manufacturing higher pixel density means that everything needs to be new. DR is the same, but we do not know about the rest. Greame however, informed that it's a new beast at least :)
 
I asked Toia about dynamic range and he said the image is so clean you can dig things out of the shadows like you couldn't before. I guess this statement means about the same DR as Dragon. But if you can dig out more from the shadows you essentially buy some more dynamic range.... but in a different way.
 
DR vs SNR per sq mm.

I speculate the Helium sensor has the same DR as Dragon that is 18+ stops.

And the helium has the same pixel SNR or 80dB.

But that the Helium SNR per sq mm has just Jumped up a notch (approx 1.7) times improvement.

Ergo the usable Helium ISO range has just jumped up > ~1.7 times would you normally find acceptable.

AJ
 
Mike P. said:
Have there been any .r3ds released yet? Kinda wanted to take a peek. Also interested in low light performance... actually has anything been officially announced?
Click to expand...

I have a feeling with the names of some of the buyers who put their orders in... we may end of seeing a sample or two on this page soon... http://www.red.com/sample-r3d-files
 
We sure will see a comparison soon. But it's always a trade off.
Amount of Pixels
Readout speed of pixels (frame rate)
Dynamic range
Sensitivity

The Helium troopers got a lot of pixels but can't
do super high frame rates no 100-120fps in 8k.
thinking about it it would be nuts if they would :)
But the Helium got apparently improved low light even
though the pixel size was reduced.
So it would be an amazing achievement to increase
the DR in any way
 
"do super high frame rates no 100-120fps in 8k."

I could be wrong - but the 80fps 8K bottleneck was a 'reading vertical pixel frame rate' limit of the Dragon architecture.

With the known requirements of the Olympics in 2020 (ie 120Hz @ 8K) - I would expect there (in the fullness of time) to be some unlocking of this potential from both Sensor and ASIC architecture within the DSMC2.

And if required tweaking of the SSD to make this possible.

AJ
 
Last edited:
I've been one of the lucky ones to be able to test the 8k HELIEM, and to answer your question in the least technical way possible, this is what Im seeing.

Even though the 6k Weapon & Dragon carry the same sensor and DR range of 16+ stops... and the Helium apparently the same. What I'm seeing straight up is more use of that 16 stop range. In other words, I'm pushing well into my blacks 1 and two stops more compared to waht I can do with the 6k. So much less noise ! So you can.
Highlights I'm not seeing a great difference, but in saying that, again when pushing the R3D file the highlight seem to hang in a bit longer before they burst.
Is this more range...? I have no idea. But what is very apparent is you can push harder on set. 3200 ISO 8k, sampled down to 4k or even 2k is like WOW !!!
 
Mark Toia said:
I've been one of the lucky ones to be able to test the 8k HELIEM, and to answer your question in the least technical way possible, this is what Im seeing.

Even though the 6k Weapon & Dragon carry the same sensor and DR range of 16+ stops... and the Helium apparently the same. What I'm seeing straight up is more use of that 16 stop range. In other words, I'm pushing well into my blacks 1 and two stops more compared to waht I can do with the 6k. So much less noise ! So you can.
Highlights I'm not seeing a great difference, but in saying that, again when pushing the R3D file the highlight seem to hang in a bit longer before they burst.
Is this more range...? I have no idea. But what is very apparent is you can push harder on set. 3200 ISO 8k, sampled down to 4k or even 2k is like WOW !!!
Click to expand...

Mark, am I understanding you correctly that Helium is the same Dragon sensor as Weapon, just a smaller pixel pitch ?
 
Mike Nagel said:
Mark, am I understanding you correctly that Helium is the same Dragon sensor as Weapon, just a smaller pixel pitch ?
Click to expand...

I'm sure there is more to it than that... but fundamentally... yes. Actually I can't even answer that, Im not a RED Sensor engineer. But the Pixels being so small are making the camera so great !
 
Mark Toia said:
I'm sure there is more to it than that... but fundamentally... yes. Actually I can't even answer that, Im not a RED Sensor engineer. But the Pixels being so small are making the camera so great !
Click to expand...

ah ok, got ya. I assume it's a completely new sensor, but just wanted to clarify your post.

Thanks and can't wait to see some of your tests of this new beast !
 
Mike Nagel said:
Remember, u aways gain a bit of DR when u downrez the original high res capture... coming down from 8K should gain a bit more...
Click to expand...

You don't gain any DR by downrezzing.

DR is captured data in a range between clipping and noise.
No DR gets enhanced after capture.

With downrezzing the apperance of DR improves due to perception of supersampled noise in the blacks and resulting usability of shadow range. The DR is the same, no additional data appears in shadows, what improves is usability of DR, as Mark mentioned. Think of it as reducing the gap between technical, mathematical DR (16.5) and actual usable one.

In short - your point referring to DR usability is correct, but the formulation isn't, which is spread around the web and if literally interpreted can be misleading.
 
Hrvoje Simic said:
You don't gain any DR by downrezzing.

DR is captured data in a range between clipping and noise.
No DR gets enhanced after capture.

With downrezzing the apperance of DR improves due to perception of supersampled noise in the blacks and resulting usability of shadow range. The DR is the same, no additional data appears in shadows, what improves is usability of DR, as Mark mentioned. Think of it as reducing the gap between technical, mathematical DR (16.5) and actual usable one.

In short - your point referring to DR usability is correct, but the formulation isn't, which is spread around the web and can be misleading.
Click to expand...

agreed, bad wording.

but the perceptual gain will/should be slightly higher with an 8K capture when downsampling to same target as lower res capture, so it will be interesting to see real world results.
 
Last edited:
You must log in or register to reply here.
Back
Top