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 MX IMAGES APPEAR BETTER THAN SCARLET-X

Alexandre Desjardins said:
According to Graeme from RED, the sensor is the same, only the processors are slower... And I shot as well on RED One MX and on my Scarlet X, not side by side testing, but I don't "feel" that the Scarlet is noisier.
Click to expand...
My scarlet has tons of noise and grain at 800 ISO. I've black shaded 100s of times with no fix or result. At one point I was in touch with someone at red, sent samples and never heard back. I even see the black clipping bar filling up in daylight.
 
Jeffrey Loewe said:
My scarlet has tons of noise and grain at 800 ISO. I've black shaded 100s of times with no fix or result. At one point I was in touch with someone at red, sent samples and never heard back. I even see the black clipping bar filling up in daylight.
Click to expand...
Jeffrey, that sounds like a major problem. Can you send over an R3D? Curious to see what your shots are looking like.
 
I've been doing quite a few tests with my Scarlet since it arrived and great imagery is all about proper exposure. At 320 ISO and with proper raw exposure/lighting the image is flawless at every frame rate an size. Even at 2K/60 the image has little to no noise in the darkest parts of the image. At 800 ISO its very similar, a little more noise, but just a tiny bit and visible mainly in the cropped sensor at 2K.

The image has to be exposed right. An under exposed shot sees the noise raise considerably. I've had luck rectifying this with Denoiser, but its obviously not optimal.

I would recommend the camera be rated and lit at ISO 320 with the knowledge that one would have the ability to take it to 800 or so in a pinch.
 
Steve Sherrick said:
One must take into account all kinds of factors with this discussion. On paper, all of these cameras are running the same sensor but as we all know are being processed differently internally. I'd also argue that no two cameras within the same family are identical. This can become evident in multicamera shoots. You have the glass in front of the camera, as well as OLPFs which all contribute to a camera's personality. I think when you start comparing cameras from two different families the gap widens even further. In the end, the key is to get to know YOUR camera, identify its "personality" and maximize its performance. A lot of times we're talking about subtleties here, so it's not always night and day differences. When Epic is brought into the conversation then yes, often times we are talking big differences.
Click to expand...
While I agree with the notion, that people should know their tools, I can't see how people will be able to learn quirks of THEIR cameras, if many of those people often use rental cameras.
On the point of "perceived noise issue" I do have a personal feeling. Yes, it is "perceived" as well, due to the fact, that so far no one was able to produce side by side images confirming the noise issue. So far it seems to me, that both Epic and Scarlet produce seemingly little more grainy images, than MX. But until we see a visual proof of that issue, it is just that- a matter of perception.
 
Much of the r1mx vs epic noise issue has to do with the 320 vs 800 native iso. I'm a firm believer that epic is still a 320 sensor. Only noise I've seen so far is when the footage is shot at 800.
 
jake blackstone said:
While I agree with the notion, that people should know their tools, I can't see how people will be able to learn quirks of THEIR cameras, if many of those people often use rental cameras.
On the point of "perceived noise issue" I do have a personal feeling. Yes, it is "perceived" as well, due to the fact, that so far no one was able to produce side by side images confirming the noise issue. So far it seems to me, that both Epic and Scarlet produce seemingly little more grainy images, than MX. But until we see a visual proof of that issue, it is just that- a matter of perception.
Click to expand...
Actually, I've profiled rental cameras as well. Although, you'll never get to know the rental cameras as well as your own, at the very least it's important to do a thorough checkout to make sure there is nothing completely out of character happening with the camera you have been assigned. Unfortunately I've noticed some jobs are cutting back on the camera checkout process, which is of course a mistake but kind of a sign of the times. On bigger jobs, with producers who get it, we're still given appropriate checkout procedures. On jobs where the cameras will be used over a long stretch and you have some prep days, you can indeed learn a lot about a camera, even a rental one.

I acknowledge though that there will be jobs where you are handed a camera and told to start rolling. In that case, you do the best you can with what you've got. And most of the time it turns out okay. It's just those times when it doesn't that you wish you had a chance to catch the problem earlier.
 
Sid Idris said:
I'm a firm believer that epic is still a 320 sensor.
Click to expand...

Mysterium X sensor is ISO 320. 800 is Red Co. preferred ISO and is recommended to use to protect the highlights, among a few other things.
 
please Jeffrey could U send me a PM with some r3d attachement about your problem? I got no problem at all with my Scarlet, everything superclean, little noise only if underexposed, and I'd like to figure out what's going on with yours.

Jeffrey Loewe said:
My scarlet has tons of noise and grain at 800 ISO. I've black shaded 100s of times with no fix or result. At one point I was in touch with someone at red, sent samples and never heard back. I even see the black clipping bar filling up in daylight.
Click to expand...
 
Karl Krings said:
hey guys, i think if there is no reliable test from a laboratory you better forget talking about all this. i think most of you using these cams the wrong way...try to make movies with intelligent stories ...and remember "tech. quality of picture is not all". basta :D - and sorry for my hard words but there are more important issues to talk about! thanks
Click to expand...

This might surprise you but I'm actually not a screenwriter so more often than not, my time is better spent perfecting the image than writing dialog.
 
If there is any noise according to my experience the best solution is

The Foundry's Furnace Core De-Noise Plug-in and is more powerful than Neat Video.

To illustrate this have a look at the latest two examples of my test with Scarlet-X.

Furnace Core De-Noise was applied at the night shots of the first example at YouTube.

Shot on Scarlet-X with Leica R lenses.

Shot on Scarlet-X with Leica R lenses.
 
Sanjin Jukic said:
If there is any noise according to my experience the best solution is

The Foundry's Furnace Core De-Noise Plug-in and is more powerful than Neat Video.
Click to expand...
Evaluating noise with a trans-coded YouTube video isn't that useful.

Full size images would be best with before/after.

Why exactly do you feel that plug-in is better than NeatVideo? Do you have any side by size comparisons done by someone who knows how to setup both plug-ins?
 
Paul Schneider said:
I've been doing quite a few tests with my Scarlet since it arrived and great imagery is all about proper exposure. At 320 ISO and with proper raw exposure/lighting the image is flawless at every frame rate an size. Even at 2K/60 the image has little to no noise in the darkest parts of the image. At 800 ISO its very similar, a little more noise, but just a tiny bit and visible mainly in the cropped sensor at 2K.

The image has to be exposed right. An under exposed shot sees the noise raise considerably. I've had luck rectifying this with Denoiser, but its obviously not optimal.

I would recommend the camera be rated and lit at ISO 320 with the knowledge that one would have the ability to take it to 800 or so in a pinch.
Click to expand...

Neat Video will help you a lot more than DeNoiser. best $100 you will ever spend on gear.
 
What is all this talk about 320 ASA... The Scarlet/Epic/ RED MX are all Native 800 ASA.. I use a meter for almost all my shooting and have rated these cameras at 800 ASA with perfect results.. I have shot Features, TV and many other projects
always keeping in mind 800 ASA.. Now that does not mean I dont rate it at other ASAs but what that does is change the latitude of the camera. For instance when you rate at 320 ASA you are going to clip your highlights sooner... that may be ok for you but
you are not getting the true Dynamic range of the camera... This has been discussed with several leading DITs that I work with like Brook Willard, Chris Cavanaugh, Nate Kalushner.
What you are seeing is better Blacks but with sacrifice for the highlights.. With this being said this may be the look you are after and 320 ASA works for you but the Scarlet/Epic/ RED MX are Native 800 ASA and gives you the most Dynamic range exposed there.

Dana Gonzales
Director of Photography
www.danagonzales.com
 
I think alot of the complaining about noise in the Scarlet-X comes from people who are not familiar with how to expose digital sensors correctly. And the whole native ISO thing is confusing people. It's really easy actually, and has been pointed out before. Red recommend using 800 ISO as a "standard" so that when you are out shooting you will automatcally protect your highlights, because the image on the monitor then seems a little over exposed compared to what the sensor is actually recording. If you set the ISO to 320 on your scarlet-x and then go to Settings-Display-Tools and turn on RAW you will see a very similarly exposed image, this means the monitor shows you an image that is very close to what the camera is actually recording.

So the conclusion I have made is this: If you find youself shooting outside in the sunlight or in a high contrast situation with a lot of light, then set your ISO to 800. If you are shooting in a low light situation with low contrast, then set your ISO to 320, and light the scene to get an image that is as noise free as possible. This is the exact opposiite of what you would do if you were shooting film, and I think this is what is confusing to most people.

So in other words ISO 800 protects you from getting blown out highlights, ISO 320 protects you from noise in the low light areas. I guess the reason RED tells us to stick with ISO 800 is that blown highlights is generally a much bigger problem than a little noise in the shadows, which can be fixed with noise reduction software. There is no blown out higlight fixing software (as far as I know ;)).

A very good tool to make shure you don't get noise is the RAW noise indicator, the little bar on the left side of your histogram if this fills up with red, you will get noise. same thing the the highlight clipping indicator bar on the right side of the histogram, if the bar goes red you are overexposing. Keep the indicator bars free from red and you are golden, it's as easy as that really.

Another thing that most people dont understand is that RAW images generally seem more noisy than compressed footage because the detail has not been lost in compression and there has not performed any noise reduction. So if you want to see the same low light performance out of your Scarlet-X as the other cameras out there like the FS100, AF100 etc, you will have to apply noise reduction like they do. Only difference is when you shoot RAW, you do it in post.

Another thing; the thread starter keeps saying the RED one has different, and better performing "sensors". This is wrong. The Scarlet-X and the upraded RED Ones have the exact same Mysterium-X sensor inside, but they do have different kinds of processors. The Red One has a processor much like a CPU ( Central Processing Unit) in your computer, while the Scalet-X and the Epic have application-specific integrated circuits (ASIC's). Diffrence is that the ASIC's are customized for a particular use, rather than intended for general-purpose use, like the CPU. Using ASIC's is much more effective in terms of heat, space and cost. Hence the smaller lighter less expensive Scarlet-X. The main difference between Scalet and Epic is that the Epic has higher quality/better performing ASIC's inside, plus the circuit boards that go along with these. As a result the analog to digital conversion process is probably a bit different between the RED One and Scarlet/Epic but the compression is probably very similar (both use redcode raw).

If anything the compression in the Scarlet-X and EPIC is more efficient than in the RED One and and you will get more detailed information out of the R3D. More information = more noise. (this is pure speculation on my part as I have not done any scientific testing, nor have I ever shot with a Red One camera)
 
Last edited:
radfilm said:
What is all this talk about 320 ASA... The Scarlet/Epic/ RED MX are all Native 800 ASA.. I use a meter for almost all my shooting and have rated these cameras at 800 ASA with perfect results..
Click to expand...

ASA as we know it does not apply to digital imaging. We only use it because we have not adopted a newer and more precise way to describe what is happening. We use it because it is familiar. A sensor or camera system is not locked into a single ASA like a film stock, so why do we insist on talking about a "native" ASA?

I don't doubt you Dana, but as Steinar mentioned, if you look at the RAW feed you will notice that the RAW recorded signal is equal to 320. That may be why folks are still referring to 320 as native.

On the other hand, if the sensor is indeed recording a 320 ASA equivalent, I see no real reason to use any other setting during production. A/B 320 and other settings to see what post processing results might be, but if you are mindful of how the system works and what you need to do to expose it well then there is no reason to shoot at any ASA setting other than 320. Make all lighting and exposure descisions based directly upon 320 feedback. That is my current thinking anyway, after just a small amount of shooting at 800.

My next question is at what compression setting do higher ASA settings become moot? To rephrase; You can shoot at 12800, but at which compression setting do you cross the threshold to where the shadow details you see while shooting are not practically recoverable?

Does the viewfinder or monitoring path reflect the effects of the compression setting? I don't see it mentioned in the documentation, but my instinct says no... Anyone? Red team?
 
Last edited:
Steinar Andersen said:
I think alot of the complaining about noise in the Scarlet-X comes from people who are not familiar with how to expose digital sensors correctly. And the whole native ISO thing is confusing people.
Click to expand...

Agreed.. I was getting noiseless images out of my M sensor.

Only under exposure causes noise... and trying to retrieve information would enhance that...

PS. I shoot everything at 320. Maybe thats the secret.
 
You must log in or register to reply here.
Back
Top