Welcome to our community

Be a part of something great, join today!

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 :)

Build 16... taste

Just getting caught up on this thread now... Very impressive.

Stepping back a bit to the ISO 500 and 640 images posted, something about the noise in those shots is somewhat alluring, seems to have a nice texture. I want to see it in motion. So I second Evin's request, even if we have a clip of a static chart.
 
This is crazy, man!

Build 16 definitely creates a unique grain. Almost fuzzy looking, having it's own distinct style. I think I'm starting to like this grain, or maybe I'm looking at this thread too long.

:w00t:
 
What's this talk about lower ISO with no penalty? Does that mean you can actually slow down the camera with the ISO, or am I just reading it wrong. I'm guessing I'm wrong considering it should just be metadata. It would be cool if I wasn't though.
 
I lit this with one Kino and soft cover and an HMI in a softbox. Lights were each about 5' from the chart.

I metered this to death and came up with a matching consensus of exactly T5.6 at ISO 400. With no NDs, I opened/closed a 1/3rd stop for every ISO change (like 320, 400, 500, 640, etc.). I used a RED 50-150mm T3 zoom, so I ran out of room at ISO 125 on the open end.

Jim
 
Jannard said:
I lit this with one Kino and soft cover and an HMI in a softbox. Lights were each about 5' from the chart.

I metered this to death and came up with a matching consensus of exactly T5.6 at ISO 400. With no NDs, I opened/closed a 1/3rd stop for every ISO change (like 320, 400, 500, 640, etc.). I used a RED 50-150mm T3 zoom, so I ran out of room at ISO 125 on the open end.

Jim
Click to expand...

Many thanks. Great to know. Love it.

David
 
Graeme Nattress said:
Basically, we have two things fighting each other - tungsten light which is blue deficient, and silicon based sensors, which are insensitive to blue light. Graeme
Click to expand...
They're not fighting each other, I'd say it was more correct to say they've joined forces to fight YOU!:biggrin:
 
SF Geek said:
What's this talk about lower ISO with no penalty? Does that mean you can actually slow down the camera with the ISO, or am I just reading it wrong. I'm guessing I'm wrong considering it should just be metadata. It would be cool if I wasn't though.
Click to expand...

Yeah, I have the same question. I would love for there to really be "no penalty" for lower ISO and just remove NDs from the equation, but that doesn't sound right if ISO is still just metadata. Is it?
 
Jannard said:
I think the most amazing thing is that the only differences between ISO 160 and ISO 2000 are the noise levels. The uniformity of color and gamma is unreal. You get to decide your noise tolerance range and make ISO decisions based on that without worrying about color shifts or levels changes.

Jim
Click to expand...
So I'll have ~5 stops over and ~6 stops under at any ISO? It won't limit me like it does in current and past builds?

Also, does the noise reduction algorithm that appears to be written into the codec [that noise is way too smooth to be without some kind of magic NR... even if it goes by a different name] change depending on the color temperature at which you process? In other words, is it aware of when I'm beating up the blue channel and will it dive in to save the day without me needing to apply additional NR?

Really nice stuff. Can't wait to see some footage.
 
If that noise isn't a fixed pattern, I can see some old guard film holdouts trying to stack a couple ND 1.8s just so they can get the image into ISO2000 for that grainy film look.

I personally love the clean image of Red and do not particular like the grain of film... unless it is serving a specific aesthetic purpose. But there are others who have a strong adverse reaction to the cleanliness of the properly exposed Red image and I can see some of them chasing the high ISO to get grain into the image.

That said, the grain at ISO2000 looks like something I would find acceptable for available light shooting where I wouldn't even think of using Build 15. It really amazes me that you guys can find stuff like this inside the camera and will be able give me, in essence, a brand new and greatly improved camera for the cost of twenty minutes of my time.

Commitment and support like this is how you won over my black little capitalist heart.
 
JarredKrakow said:
Yeah, I have the same question. I would love for there to really be "no penalty" for lower ISO and just remove NDs from the equation, but that doesn't sound right if ISO is still just metadata. Is it?
Click to expand...

Well,
Depends on how you look at it.
Changing the effect of ISO metadata means changing a process that is happening both in RedAlert etc and in the camera's viewing pipeline.
It sounds to me like the changes in Build 16 are happening before the camera compresses the image as well is in the compression of the image AS WELL AS when the image is "developed" in RedAlert. According to the metadata paradigm, the first two processes should be unaffected by your metadata choices in camera. The third process sounds like it has also been redesigned.

So, I'm guessing that when Mr. Jannard says "no penalty" he means something different than "no limitations"? Are highlights more likely to crush in 100 ISO than in 1000?

Again, guessing. Hoping to test this new build very soon. I'm sure they are itching to release it.

IBloom
 
We're not talking about metadata or when or how it is applied. We're talking about actual sensor sensitivity, like in a DSLR.
 
SF Geek said:
We're not talking about metadata or when or how it is applied. We're talking about actual sensor sensitivity, like in a DSLR.
Click to expand...
From my obsessive reading of this forums I see no evidence that the sensor sensitivity has changed. I see a large amount of evidence that algorithms are changing.

I think it is about "the effects of metadata". About processing digital images and mapping values.

I could be wrong. But I feel you are barking up the wrong tree. Again I could be wrong.

IBloom
 
That's what I'm guessing. I was just hoping that that would be the case. I'd rather slow down the sensitivity than fill up my mate box with NDs and an IR filter.
 
It appears that the images that Build 16 produces are better quality and more detailed than Build 15. Does that mean that the R3D files in 16 have more information in it, making them larger files. If the R3D files in 16 are larger, how much larger are they than Build 15?

Just want to know if I'm going to need more storage space for Build 16 projects.

Thanks.
 
Not, obviously, that I'm privy to any RED secrets, but in general if they've reduced the noise relative to Build 15, prior to the Redcode compression, that would leave some "unemployed bits" that used to be encoding noise, that are now freed up to encode the real detail. So the end result might look more detailed, without actually requiring any more storage space. Just a thought.
 
You must log in or register to reply here.
Back
Top