Jump to content

L.A.R.T. under NDA


Stephen Williams

Recommended Posts

  • Replies 72
  • Created
  • Last Reply

Top Posters In This Topic

  • Premium Member

It was independent of RED, but it was organized by Steve Gibby under his company Gibcor, so if he wants to assert some control over how the information is disseminated, that's his prerogative. All will be revealed eventually, just in some sort of semi-organized way.

 

There were also some elements of new equipment being used that were under some wraps. It was interesting for me to momentarily step into that inner core of RED users and look under the hood, so to speak.

 

I can actually say though that my experience trying out the camera was pretty positive for the most part. Exposing it was not as tricky as I thought it might be -- there are a number of methods to double-check a basic meter reading, and it's all pretty fast and intuitive. For example, you can quickly punch up a false color image or a histogram to look for any problems with information getting clipped.

Link to comment
Share on other sites

  • Premium Member
It was independent of RED, but it was organized by Steve Gibby under his company Gibcor, so if he wants to assert some control over how the information is disseminated, that's his prerogative.

 

Hi David,

 

It all seems so strange to me, either the camera performs as it should or it doesn't.

 

Having seen Mac's thread on 'issues' I am rather surprised that this has not been previously reported by Steve Gibby or other camera owners.

 

It looks like Vision 3 will come to Switzerland before a Red camera.

 

Stephen

Link to comment
Share on other sites

  • Premium Member
All this reminds me of the Russian election, also held this weekend.

 

Hi Max,

 

I guess the Russian NDA would state 'Good News only may be reported on RED matters'.

 

NOTE This is supposed to be a Joke, sometimes there is very little humor in this part of the forum.

 

Stephen

Link to comment
Share on other sites

Interesting issues thread on Reduser. I'm surprised that you'd have to colour correct from 5000k to 3200k on the RED. Anytime I'm shooting at low light levels with gain in it's almost always at 3200K and that's when you don't want possible additional noise from the colour correction.

 

However, early days yet with the RED.

Link to comment
Share on other sites

  • Premium Member

The sensor has a native balance of around 5000K. This is not that unusual -- I've heard that the Viper sensor is similar.

 

You just shoot under 3200K light and set that balance on the camera, which is stored as metadata, so you can open up the file already corrected if you want. The only question is whether there is increased noise from this, just as it is an issue with the Viper or many other digital cameras. I don't think using blue filters or HMI lighting for night interiors makes much sense, so it's just something you'll have to accept about many digital cameras. But when doing chromakey work, it may make sense to work under 5500K lighting rather than 3200K lighting, just for less noise in the blue channel.

Link to comment
Share on other sites

All digital sensors -- CMOS, CCD or whatever -- are blue-starved, making them roughly daylight balanced. When you use a broadcast camera that is nominally tungsten balanced it is because the camera manufacturer has skewwed the red/gree/blue balance to do so. This is why many cameras -- Panasonics especially - have the option to electronically set the base balance to daylight instead of tungsten. It not only plays to the sensors stregnths and has less noise in the blue channel (where the noise is always the worst because of this issue) but it also increases sensitivity in daylight settings over shooting tungsten balanced with the 85B correction filter dialed in.

 

Anyway, all electronic sensors are daylight due to the laws of physics. The Phantoms are around 4600 degrees or so, basically the same as RED.

Link to comment
Share on other sites

All digital sensors -- CMOS, CCD or whatever -- are blue-starved, making them roughly daylight balanced. When you use a broadcast camera that is nominally tungsten balanced it is because the camera manufacturer has skewwed the red/gree/blue balance to do so. This is why many cameras -- Panasonics especially - have the option to electronically set the base balance to daylight instead of tungsten. It not only plays to the sensors stregnths and has less noise in the blue channel (where the noise is always the worst because of this issue) but it also increases sensitivity in daylight settings over shooting tungsten balanced with the 85B correction filter dialed in.

 

Anyway, all electronic sensors are daylight due to the laws of physics. The Phantoms are around 4600 degrees or so, basically the same as RED.

 

I'm pretty sure the dvx100 is 3200K balanced; I think Panasonic says so and it's noticeably cleaner (at equal levels of gain) under tungsten rather than daylight. That said, it shows more grain in blue than in other colors and it is also by far the noisiest channel.

 

What is the chemical difference between a daylight and tungsten stock? Why not make a sensor that has more blue photosites? 2 blue 1 green 1 red?

Link to comment
Share on other sites

  • Premium Member
I'm pretty sure the dvx100 is 3200K balanced; I think Panasonic says so and it's noticeably cleaner (at equal levels of gain) under tungsten rather than daylight. That said, it shows more grain in blue than in other colors and it is also by far the noisiest channel.

 

What is the chemical difference between a daylight and tungsten stock? Why not make a sensor that has more blue photosites? 2 blue 1 green 1 red?

 

That's sort of a contradiction, that the blue channel is noisiest yet the picture is less noisy in tungsten lighting when the blue channel is working the hardest.

 

Most of your sharpness and luminence values comes from the green photosites usually; besides, I'm not sure adding more blue photosites necessarily makes the blue channel less noisy. It may add more resolution to blue detail but not necessarily more signal.

 

In film stocks, a daylight balanced film has a slower-speed blue layer whereas it is faster (grainier) on tungsten stocks to compensate for the lower levels of blue wavelengths in 3200K lighting.

 

I suppose you could ND gel the green and red photosites on a sensor to make it more (relatively) blue-sensitive in tungsten light, but then it would effectively be much slower than the typical 320 ASA of most digital cameras. You might as well leave the sensor alone and just put a blue filter on the camera.

Link to comment
Share on other sites

  • Premium Member
Anyway, all electronic sensors are daylight due to the laws of physics. The Phantoms are around 4600 degrees or so, basically the same as RED.

 

Hi Mitch,

 

So the laws of Physics apply to RED, what happened to the revolution?

 

Stephen

Link to comment
Share on other sites

That's sort of a contradiction, that the blue channel is noisiest yet the picture is less noisy in tungsten lighting when the blue channel is working the hardest.

 

Most of your sharpness and luminence values comes from the green photosites usually; besides, I'm not sure adding more blue photosites necessarily makes the blue channel less noisy. It may add more resolution to blue detail but not necessarily more signal.

 

In film stocks, a daylight balanced film has a slower-speed blue layer whereas it is faster (grainier) on tungsten stocks to compensate for the lower levels of blue wavelengths in 3200K lighting.

 

I suppose you could ND gel the green and red photosites on a sensor to make it more (relatively) blue-sensitive in tungsten light, but then it would effectively be much slower than the typical 320 ASA of most digital cameras. You might as well leave the sensor alone and just put a blue filter on the camera.

 

I've tested the dvx under 6000K lighting with an 85b filter versus with daylight white balance (and compensated for exposure, plus took IRE readings) and gained it up and it definitely is cleaner at 3200K; that said it's almost unnoticeable. I agree that it doesn't make much sense, though, but I honestly think it has a 3200K native sensor.

 

That makes sense about different stocks. Would 500D (if it existed) theoretically be cleaner than 500T is or is the preference for daylight unique to digital photography?

 

EDIT: Further contradicting what I wrote about the dvx, I shot something this summer with an HVX and we didn't have enough light to gel the 1k's blue at night so we gelled the practicals with 1/2 CTO instead and turned the digital white balance way down in color temperature. All I can say is I hope most of that footage gets cut. It was so bad.

Edited by Matthew Wauhkonen
Link to comment
Share on other sites

All digital sensors -- CMOS, CCD or whatever -- are blue-starved, making them roughly daylight balanced. When you use a broadcast camera that is nominally tungsten balanced it is because the camera manufacturer has skewwed the red/gree/blue balance to do so. This is why many cameras -- Panasonics especially - have the option to electronically set the base balance to daylight instead of tungsten. It not only plays to the sensors stregnths and has less noise in the blue channel (where the noise is always the worst because of this issue) but it also increases sensitivity in daylight settings over shooting tungsten balanced with the 85B correction filter dialed in.

 

If it's the physics that's the nature of the beast.

 

I suppose that means the reason for getting increased blue channel noise when you do an electronic daylight white balance (without the correction filter) on say a F900 is you're applying a "double correction", since the naturally blue balanced sensors have been skewed.

Link to comment
Share on other sites

If it's the physics that's the nature of the beast.

 

I suppose that means the reason for getting increased blue channel noise when you do an electronic daylight white balance (without the correction filter) on say a F900 is you're applying a "double correction", since the naturally blue balanced sensors have been skewed.

Correct. That's why the Panasonic cameras have the ability to reset the base correction to daylight instead of tungsten. It is definitely cleaner in daylight balanced light this way.

 

The DVX100 is skewed to a base 3200 tungsten balance, as are all video cameras. You just don't have a way to get past this. That's because the signal from the sensors is going through a DSP and being highly compressed from there. With a raw sensor output it is quite obvious that all of these sensors are around 5000K.

 

This is why the blue channel is always where you'll see noise first.

 

If the Mysterium sensor had arrived with a true tungsten balance than that would have really been something unique.

Link to comment
Share on other sites

  • Premium Member
That makes sense about different stocks. Would 500D (if it existed) theoretically be cleaner than 500T is or is the preference for daylight unique to digital photography?

 

A 250D stock would have less grainy blues than a 250T stock, such as Eterna 250D vs. Eterna 250T.

 

The only 500D stock is Reala 500D, and it's an older emulsion that is graininer all-around than Eterna 500T.

 

It's not a "preference" for daylight, just that the blue layer is made to be slower for daylight by using smaller-sized grains, smaller grains being less effective at collecting photons than larger ones.

 

Although in truth even film emulsion has had an easier time getting exposed in daylight since the beginning -- blue wavelengths have more energy than red ones. Early b&w emulsions were blue-sensitive only.

 

We don't realize just how much blue & green light there is relative to red in reality, hence why sensors pick up those wavelengths easier.

Link to comment
Share on other sites

Since most video cameras don't allow you to change the base correction you are better off shooting with an 85b filter on the camera when in daylight, correct? Otherwise the blue channel gets gained up, then overexposed and then brought down digitally?

 

Which cameras allow you to change the base correction? Which Panasonics? Are there any cameras besides the Red and the Viper that have kept their chips native 5000k balance?

 

~Jess

Link to comment
Share on other sites

Since most video cameras don't allow you to change the base correction you are better off shooting with an 85b filter on the camera when in daylight, correct? Otherwise the blue channel gets gained up, then overexposed and then brought down digitally?

 

Which cameras allow you to change the base correction? Which Panasonics? Are there any cameras besides the Red and the Viper that have kept their chips native 5000k balance?

 

~Jess

For the first sentence, that is correct.

 

All of the little camcorders switch between 3200 & 5600 electronically, which simply means they are balancing the relative levels of red, green & blue. In daylight they don't need to bring up the level of blue very much, hence the lower noise. All the professional Panasonics will do this as well. Hey Phil Rhodes, if you're reading this, did your old industrial 200 camera do it as well? I bet it did.

 

Any camera that functions in a RAW mode will put out the native balance of the sensors because they are simply recording that direct output without treating the image through a DSP. That would include Dalsa Origin, Silicon Imaging, Phantom, etc.

 

It is the difference between a raw camera head and a DSP engineered camera.

Link to comment
Share on other sites

All of the little camcorders switch between 3200 & 5600 electronically, which simply means they are balancing the relative levels of red, green & blue. In daylight they don't need to bring up the level of blue very much, hence the lower noise.

So if shooting outside are you better off not using a filter and instead setting the camera to 5600? It seems like if the camera were setup correctly this would be the case, but from your previous statement it sounds like there is a problem with doing this.....

 

Is the DSP applying gain to the blue channel regardless of the white balance setting?

 

When shooting outside with no filter is it gaining up the blue and then having to bring it down digitally, or does it simply turn down the amount of gain on the blue channel when switching from 3200k to 5600k?

 

~Jess

Link to comment
Share on other sites

So if shooting outside are you better off not using a filter and instead setting the camera to 5600? It seems like if the camera were setup correctly this would be the case, but from your previous statement it sounds like there is a problem with doing this.....

 

Is the DSP applying gain to the blue channel regardless of the white balance setting?

 

When shooting outside with no filter is it gaining up the blue and then having to bring it down digitally, or does it simply turn down the amount of gain on the blue channel when switching from 3200k to 5600k?

 

~Jess

Switch the camera to 5600.

 

Stating that the DSP is applying gain is a misnomer. The DSP will adjust the relative levels of the three color channels appropriately. When set to 3200 it needs to pull more from the blue channel than when set to 5600 because that is the natural balance of the chips. But it does all of this at the start in analog RGB colorspace, and then later digitizes the information through its A-to-D converter to turn it into a digital signal (in the DVX's case, into a 4:1:1 color sampling).

 

You're overthinking this and I hope I have not confused the issue. When a camera allows you to reset the base electronic white balance from tungsten to daylight, this means that it is doing it at the earliest possible electronic point and this is a good thing to get the most from the chips.

Link to comment
Share on other sites

  • Premium Member
I guess I should mention that Jim confirmed that the NDA was not from red.

 

 

 

Its dissapointing to see people jump to negative conclusions before all the facts are in. Repeatedly.

 

R.

 

I wouldn't read too much into that -- it is true that the LART was done under an NDA afterall... and it's not surprising that some people thought RED had organized the test, hence the incorrect assumption.

Link to comment
Share on other sites

I wouldn't read too much into that -- it is true that the LART was done under an NDA afterall... and it's not surprising that some people thought RED had organized the test, hence the incorrect assumption.

 

Well it's the second time a mod here has started a rumour based on an incorrect negative assumption, before the facts were confirmed.

 

But sure, I wouldn't want to fall into the trap of making um... negative assumptions.

 

R.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Forum Sponsors

Visual Products

Film Gears

BOKEH RENTALS

CineLab

CINELEASE

Gamma Ray Digital Inc

Broadcast Solutions Inc

Metropolis Post

New Pro Video - New and Used Equipment

Cinematography Books and Gear



×
×
  • Create New...