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 :)

New RED Color Science - Test Pilot Feedback Thread

Sanjin, awesome examples, thanks!

What's even more amazing is in your candle example, the new build seems to have cleaned up artifacting on the flame itself!

Graeme, using Sanjin's candle example (full 4K versions) would it be possible at some future point to clean up the black dots around the specular highlights on the bottle to the right of the flame To my eye that's the last "video-like" artifact left in the image, everything else (like the candle flames smooth highlight rolloff) is pretty much film territory in quality now!!!

Thanks all.

Paul
 
Graeme, you're a bloody genius.
Amazing work.

When I came back to this thread, didn't think I would see something so amazing as that comparison right there..
 
Quick test with the new build and new RA.:thumbsup:

320ISO_build10v3.72_2.8f.jpg

320ISO_build10v3.72_5.6f_.jpg

320ISO_build10v3.72_5.6f_1000.jpg
 
I would like to be one of the "special ones".

When can we get this?

I've seen enough candy to be convinced that it tastes good!!
 
One more clarification. This new science is not a rewrite of the rec 709 and redspace gammas, but a new debayering algorithm? Just want to make sure because I usually always work in RGB/Red Log. If i usually do my CC in Color with Redcode native Log and Transfer then my new best workflow would be printing a flat image from RA and doing CC from that, until the SDK is finished?
 
May have missed it, but does the better monitoring on set also mean we won't see clipped highlights when viewing 320 that are clearly not clipped in raw? Going back and forth is started to get old. Many people I talk to stay in raw most of the time, but I don't think thats the best solution.
 
Michael Ragen said:
May have missed it, but does the better monitoring on set also mean we won't see clipped highlights when viewing 320 that are clearly not clipped in raw? Going back and forth is started to get old. Many people I talk to stay in raw most of the time, but I don't think thats the best solution.
Click to expand...

They asked recently about making the traffic lights pull from he RAW info regardless of what color space you use on set, which would fix the issue letting you set exposure correctly and keep your clients in the loop. It seems the new camera build will be sure to have that and possibly other awesome treats regarding exposure in the mix.
 
And now about a light flare correction "issue" between RA 3.60 & RA 3.72B.

Also to mention that a light flare "issue" is "mostly" connected with the lens coating and having or not floating aspherical elements for the optical corrections.

But have a look at the following examples (BTW there are all shot handheld on R1):


RA_comparison5.jpg

Shot on RED1 with Leica Noctilux-M 50mm f/0.95 @ f/0.95 wide open, RA comparison 1 @ ISO 500.


RA_comparison6.jpg

Shot on RED1 with Canon 50mm FDL f/1.2 @ f/1.2 wide open, RA comparison 2 @ ISO 2000.


RA_comparison7.jpg

Shot on RED1 with Canon 50mm FDL f/1.2 @ f/1.2 wide open, RA comparison 2 @ ISO 2000.


RA_comparison8.jpg

Shot on RED1 with Leica Noctilux-M 50mm f/0.95 @ f/0.95 wide open, RA comparison 1 @ ISO 2000.


4K RA 3.60 example 1>>>

4K RA 3.72B example 1>>>

4K RA 3.60 example 2>>>

4K RA 3.72B example 2>>>

4K RA 3.60 example 3>>>

4K RA 3.72B example 3>>>

4K RA 3.60 example 4>>>

4K RA 3.72B example 4>>>
 
Is the sensor protection thing where the full blown sections are dark "fixed"? Looking at the center of your projector shot it seems there may be a bit of that going on but it is gone on the new version.

Those 2000 ISO shots look great, wonder what they would look like after a bit of denoising though they seem a bit too noisy for my personal taste. Sanjin, do you think you could post the same shot of the van at 1280 ISO please?
 
Roberto Lequeux said:
Is the sensor protection thing where the full blown sections are dark "fixed"? Looking at the center of your projector shot it seems there may be a bit of that going on but it is gone on the new version.

Those 2000 ISO shots look great, wonder what they would look like after a bit of denoising though they seem a bit too noisy for my personal taste. Sanjin, do you think you could post the same shot of the van at 1280 ISO please?
Click to expand...


Roberto,

per your request the van shot @ ISO 1280 in RA 3.60 and RA 3.72B:


RA_comparison6-ISO1280.jpg

Shot on RED1 with Canon 50mm FDL f/1.2 @ f/1.2 wide open, RA comparison 2 @ ISO 1280.



4K RA 3.60 example 2>>>

4K RA 3.72B example 2>>>




So ladies and gentlemen the camera firmware build 20 and RA newest upgrade is on and from now do it yourself all the tests!!!!!!!!!!!!!
 
Hi Sanjin,

Very good and different test with the projector image.
I processed a couple of old files and everytime BANG the noise is down and the dynamic range is higher. Dramatic...

Totally incredible isn't it :) :) :)

Antoine
 
Thank you Sanjin! I pulled it up in Photoshop to test Neat Image after Evin started his thread talking about his use of NeatVideo. I resized to 2k, applied a little de-noising with minor sharpening then moved the bottom end four points over with curves. My guess is that at 2k there would be VERY little noise noticeable and a little better tweaking might give far better results. I wonder what this would look like in motion!

(Below is a 1:1 2k)

This sort of exposure seems almost perfect for my purposes but I can't wait to see what Mysterium-X manages to add to something like this! Can you imagine 1280 ISO without denoising being mandatory, or close to 2000 ISO with denoising being acceptable for 1080?!! Am I pushing it?
 
Tim Fassnacht said:
is it just me or does it have to do with the image warming up a bit?
Click to expand...

Yes it is, but you could apply a CC in post. That not a big deal when before we used to export at 4,200K (or so) and make it 3,200K in CC after that.
 
Roberto Lequeux said:
Thank you Sanjin! I pulled it up in Photoshop to test Neat Image after Evin started his thread talking about his use of NeatVideo. I resized to 2k, applied a little de-noising with minor sharpening then moved the bottom end four points over with curves. My guess is that at 2k there would be VERY little noise noticeable and a little better tweaking might give far better results. I wonder what this would look like in motion!
Click to expand...

Good job!
I'm liking the new default-warm look, Kodak esque?

:couch:
 
I agree the images look nicer, but...

I'm finding the images are simply just warmer than they used to be. Warmer as in, the blue channel is attenuated.

When I adjust the two versions of RA to give the same white balance and exposure, the difference seems to be just that the blue channel is slightly less noisy now, which is a step forward to be sure, but not the drastic difference you see here where people are just letting the default white balance fall where it may:

RA2.png


RA1.png
 
You must log in or register to reply here.
Back
Top