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 :)

RED Weapon - Dynamic Range and Latitude Examined

Hey Phil -

Thanks as always - amazing.

How exactly did you expose for this chart? (Sorry if it was explained already).

Did you happen to bracket the chart with the lens by any chance?
 
  • Thread starter
  • Moderator
  • #62
Mark Wilkinson said:
Hey Phil -

Thanks as always - amazing.

How exactly did you expose for this chart? (Sorry if it was explained already).

Did you happen to bracket the chart with the lens by any chance?
Click to expand...

Thanks Mark, it's explained in the first post. These charts represent total captured stops of dynamic range, meaning all chips are not clipping in R, G, B channels. And there is something very important that must be done when conducting a test like this, you need a lens that exhibits no light loss variation across the field due to vignetting at given apertures as that can hide both the clipped and possible shadow detail.

I did test out a few things at various apertures, no bracketing however. One thing you can and should do is using the included shutters to eliminate possible light bleed/flare/glare of individual patches on the chart.


Anton Shavlik said:
It also seems like the final ISO is a bigger jump in exposure than expected.
Click to expand...

ISO 12800 is the wildcard as always. At that point you're pretty much into the thick of the noise floor. Personally, even if I was to go to the extremes, I wouldn't work beyond ISO 6400 if you could avoid it. And truly at that point I'd also recommend the LLO or STD OLPFs.
 
Thanks, Phil - I did get that in the first post. I was trying to figure out if you metered it or just eyeballed the histogram.

I think I'm mostly curious about how the exposure tolerance shifts as middle gray is placed in the curve.

A test for another day :)
 
Hey Phil... truly appreciate the effort you put in & not to miss the tons of hours... it's pays to follow your test, as we can prep better, based on your findings...
Sure, some interesting pov & questions from fellow redusers just make the whole game a lot more exciting...
Once again... Thanks Phil!
 
This is truly phenomenal work - testing, analyzing, and posting results like these takes a tremendous amount of time and effort. Thank you for sharing these, Phil.

My takeaway is that Weapon, with the STH OLPF is giving 15 very clean, useable stops up to ISO 2000. Add the LLO and it will easily add another stop of latitude, possibly more. I'd love to see this same series repeated with the LLO on a Weapon (...not requesting this, btw, just saying I'd love to see it!)

16 squeaky clean stops is pretty mind-blowing stuff. Certainly stops of exposure beyond these parameters still fall in the useable category - but past the 15th chip, we enter the world where everyone argues over whether these are useable stops.

15-16 stops, depending on the OLPF, which are beyond debate and above reproach is mind-blowing enough for me.

From Build 12 RED ONE to here, Weapon, in an 8-year time frame...It is almost double the latitude of the original camera! I can't help but be impressed.
 
  • Thread starter
  • Moderator
  • #68
Just got home, pulled an all nighter, eating leftover Thai food.


Patrick Tresch said:
I do repeat myself but : was this test done with frame averaging?
Click to expand...

Jake Wilganowski said:
No that would be silly
Click to expand...

What would have been silly is Frame Summing the results :) Frame Averaging allows for slightly cleaner patches that are easier to read.

Do not worry, I did both and examined both. And if you're really curious I also did some tests averaging up to 120 frames because I was curious.

But to be thorough and transparent here's an Averaged and No Averaged result at ISO 2000, REDlogFilm so you can see the difference.

Frame Averaged of 16 frames @ ISO 2000, RLF:




And No Frame Averaging @ ISO 2000, RLF:





And yes, your eyes are not deceiving you. The No Frame Averaging image reveals better separation of steps 19, 20, and 21 due to the increased noise across the frame while the Frame Averaged version produces a more solid "patch".


While I'm typing away, I've done lots of other tests during the last week as well. Various capture resolutions downsampled to a 4K finish, experimentation with sharpening and scaling methods, as well as some more 3D LUT tests.

Here's some nice patches I don't mind sharing. These are from all 3 OLPFs captured at 6K FF resolution, downsampled to 4K, and cropped in at 1080p (100% of the downsampled 6K>4K image in this case). These are the tonal patches on the color chart at every ISO rating within the Recommended ISO Range:











This is essentially a look at all of the patches all at once. This is useful to see in motion and you can see that at the end of the video I posted last week.

Time to go build a rig and wireless thing.
 
More greatness from Phil.

Love it.
 
Phil, have you had any chance to play with a MG Weapon? Because of the differences in internal circuitry, I wonder if it will perform up to the same levels (dynamic range, noise, elimination of CMOS smear, etc.) as the CF Weapon?
 
  • Thread starter
  • Moderator
  • #74
Tom Greenberg said:
Phil, have you had any chance to play with a MG Weapon? Because of the differences in internal circuitry, I wonder if it will perform up to the same levels (dynamic range, noise, elimination of CMOS smear, etc.) as the CF Weapon?
Click to expand...

Yes. Two of them actually:



They appear to be the same when it comes to Dynamic Range, Noise, and the CMOS Smear improvements.
 
Phil Holland said:
Yes. Two of them actually. They appear to be the same when it comes to Dynamic Range, Noise, and the CMOS Smear improvements.
Click to expand...
That is excellent news for those of us waiting for MG Weapon!! As always, you are a wealth of information, Phil!
 
Phil Holland said:
It is exactly that. Difficult to quantify/describe.

Here's what I can say. The images are slightly cleaner, so the noise floor is a smidge lower. There appears to be more highlight retention for sure as well. That lamp image shows a clear 1 stop advantage at the top end. So it feels like things have been tuned differently across the board. What's really telling is the ISO 2000+ samples. There's a lot of DR in there. For me in practical terms it's around a 1.5 stop "difference" overall perhaps. I rarely shoot beyond ISO 1280 for my typical shoots, but there's more meat now and where the meat is what is actually important.

Slight difference for some, noticeable more when you start taking the implications of things like high speed shooting into account. There's also the CMOS Smear improvement, that can't be overlooked. It's a big difference.

Right now I'm very, very happy with where the camera is, it's already revealed that in the tests and the handful of shoots so far. Weapon is just doing better slightly in pretty much every way. Feels a little bit like a huge body upgrade with a slight sensor upgrade on top, if you want to classify it that way.
Click to expand...
Phil, I'm testing my new Weapon and it's much better then I was expecting ... which is kind of confusing me. I'm trying to take what I know from the MX and dragon, I was expecting it to be this simple learning curve but I'm finding it's just way better. For example I really like the contrast on the weapon, not sure why it's different, there's just all these little subtle things I wasn't expecting and I don't have enough cinema film background to really understand these differences through testing. If you could give some more insights on this would be great. Also it's not so much "spec" type of things, but it just seems better for practical shooting. It almost feels to me like a new sensor, even though RED says it's the same dragon sensor. I re-read the link you put at the begining of this thread where Jarred explaina the weapon technology, in that I think he hints all over the place about this ... but how it effects practical shooting is where I need some more insights.

P.S. would be nice if your name had that "skip forward to next message" feature like some other RED folks have on thier threads, there's a bunch of stuff you've said but it's kind of hard reading through every page.
 
  • Thread starter
  • Moderator
  • #77
PatrickFaith said:
Phil, I'm testing my new Weapon and it's much better then I was expecting ... which is kind of confusing me. I'm trying to take what I know from the MX and dragon, I was expecting it to be this simple learning curve but I'm finding it's just way better. For example I really like the contrast on the weapon, not sure why it's different, there's just all these little subtle things I wasn't expecting and I don't have enough cinema film background to really understand these differences through testing. If you could give some more insights on this would be great. Also it's not so much "spec" type of things, but it just seems better for practical shooting. It almost feels to me like a new sensor, even though RED says it's the same dragon sensor. I re-read the link you put at the begining of this thread where Jarred explaina the weapon technology, in that I think he hints all over the place about this ... but how it effects practical shooting is where I need some more insights.

P.S. would be nice if your name had that "skip forward to next message" feature like some other RED folks have on thier threads, there's a bunch of stuff you've said but it's kind of hard reading through every page.
Click to expand...

Heh. There's no magic "go to next Phil post button yet", but I try to do my best to hit the main points or at least link to them on the first post when it comes to tests.

It's the same Dragon sensor but what's in front and behind it has changed with the DSMC[SUP]2[/SUP] bodies, which is why we are seeing that notable improvement. The two easy things to dig into are an more efficient optical path and the improved image processing power, but there's more going on under the hood too. It's also important to remember that this effects the encoded REDCODE RAW file in good ways, lower noise floor for instance. The slightly expanded Dynamic Range makes life a bit easier for most practical shooting situations because we just getting a bit more of it. More on both sides of the scales really. It's also a hair cleaner overall.

From a practical perspective that more or less means improved image quality across all conditions and light sources. That's the simplest way to explain it. Digging deeper is where the occasionally subtle yet sometimes giant improvements can be seen.

As for the rest of the system, Weapon itself is just wicked small and lightweight. That has rather practical benefits as well.
 
I'm really much happier with my WPN DRAGON than I expected to be.

The analogy to describe the difference between EPIC DRAGON and WPN DRAGON that's now most stuck in my mind is this: you go to The French Laundry for a nice meal and some fine wine. The sommelier offers two nice reds. The first (representing EPIC DRAGON) is spectacular, but it has some sedimentation that must be removed before enjoying. No problem...they have a filter for that. Te second (representing WPN DRAGON) is ZOMG!!! and it has no sedimentation. How convenient! Both are excellent wines, but one makes you do a little more work before you can enjoy it.
 
Phil Holland said:
Heh. There's no magic "go to next Phil post button yet", but I try to do my best to hit the main points or at least link to them on the first post when it comes to tests.

It's the same Dragon sensor but what's in front and behind it has changed with the DSMC[SUP]2[/SUP] bodies, which is why we are seeing that notable improvement. The two easy things to dig into are an more efficient optical path and the improved image processing power, but there's more going on under the hood too. It's also important to remember that this effects the encoded REDCODE RAW file in good ways, lower noise floor for instance. The slightly expanded Dynamic Range makes life a bit easier for most practical shooting situations because we just getting a bit more of it. More on both sides of the scales really. It's also a hair cleaner overall.

From a practical perspective that more or less means improved image quality across all conditions and light sources. That's the simplest way to explain it. Digging deeper is where the occasionally subtle yet sometimes giant improvements can be seen.

As for the rest of the system, Weapon itself is just wicked small and lightweight. That has rather practical benefits as well.
Click to expand...
I like that explaination how it relates to the R3D. It kind of reminds me of working on a equation that is slightly off, and it just explodes to this large "thing", but an equation that that is perfectly aligned reduces to a very small form (i.e. e=mc2).

Michael Tiemann said:
I'm really much happier with my WPN DRAGON than I expected to be.

The analogy to describe the difference between EPIC DRAGON and WPN DRAGON that's now most stuck in my mind is this: you go to The French Laundry for a nice meal and some fine wine. The sommelier offers two nice reds. The first (representing EPIC DRAGON) is spectacular, but it has some sedimentation that must be removed before enjoying. No problem...they have a filter for that. Te second (representing WPN DRAGON) is ZOMG!!! and it has no sedimentation. How convenient! Both are excellent wines, but one makes you do a little more work before you can enjoy it.
Click to expand...
Exactly. How I'm starting to think of it, I use to need 4 pixels to get 1 solid pixel in that "clean up" process that I ran in nuke. Now every pixel is clean. Also I don't have a gaffer and other stuff that most feature film guys have, so my lighting is ussually on natural light with a few reflectors - my shooting style required a lot of filtering in post. I think the raven/scarlet-w guys are in for a bigger treat then they expect because each pixel has more "value", so not having 6k isn't as important. So for different shooting styles, let's say the bottom of the wine barrel shooting style, it's the same great wine but without the sediment.
 
Last edited:
Weapon - More Dragon than Dragon. :thumbsup:

Love the colours and textures in that painting you posted an R3D frame of Patrick. Pushed it around a bit in RCX and had another one of those slightly confusing 'wow' moments myself when I found the image quality was better than I was expecting.
 
You must log in or register to reply here.
Back
Top