EX3's Black Balance, and the IR contamination problem
 

Has anyone with the EX3 checked whether it still has the EX1's infamous IR contamination problem? Since the colour most affected is black, and the Auto BLK Balance function has been added to the EX3 menu - could it be the way Sony's addressed it?

In other words: when light changes (e.g. like from daylight to tungsten), can you get true blacks without any IR cut filters but simply re-balancing the Black?

If so, we could hope the new fw for EX1 will provide it, as well...

I can't see how an auto black balance can fix the problems caused by IR contamination anymore than they can be fixed in post.
It'd be interesting to know though what this function does. Some cameras have a set black point function that's used to calibrate out any offsets in the sensors and A/D converters. As I've found out the hard way use it incorrectly and you can have very serious problems with your image.

I agree. But still it's possible Sony wanted to alleviate the problem without changes to hardware...

Speaking of the Black Balance, it can actually make multi-camera shooting with EX1 and EX3 difficult, unless they add it to EX1 as well.

As black balancing is performed with the iris closed and thus no light coming through the optics I don't see how it can compensate for raised IR levels.

I didn't know it's the "blind" black balancing, Alister.

I have a B+W 750 filter arriving in the next few days and plan at least to test the IR contamination issue.

Please post your results. It's a shame for Sony (and Schneider ???); I've already bought and tested 3 filters (a single-threaded 486, a double-threaded 486, and a 489); while they're basically doing what they're supposed to do - I'm not satisfied with either.

Nevertheless, I am still ready to invest in a 4x4" 750 filter to use with a matte box (hopefully the Redrock Micro one), if only it works as advertised while not creating a greenish vignette!

This problem cannot be corrected with any software or firmware upgrades.

IR blocking filters installed directly in front of your camera's sensor are designed to work at all angles of view from the lens. As the angle of view increases, the level of cut off required to keep away green casts on your images is much higher. The EX1, EX3, and RED Camera all have weak IR filters (cutting off around 800nm) due to the wide angle or possible wide angle nature of the lenses attached. This means that in certain shooting situations you will get IR light contamination because the IR filter is not strong enough. If you cut IR light too strongly, you will get green casts. Our previously manufactured 486 filter cuts at 680nm which is too strong for wide angle lenses on video cameras. This is why our True-Cut IR 750 filter cuts off at 750nm. This is the happy medium between avoiding IR light contamination while simultaneously avoiding green cast caused by greater angles of incidence.

Ryan Avery
Schneider Optics

I've been testing the IR 750 and in one instance shooting a black fabric fanny-pouch in bright sunlight there was a slight but definitely perceptible difference between the IR750 and no filter. I've been waiting for some sunny days to test again, but we've just been having nothing but clouds and rain here for the past several days.

Ryan, will the IR 750 filter be available in circular threaded: 77mm, 52mm and other diameters?
If so, when can we expect to purchase them?

As Alister says, it's done with the iris closed. Black balancing can be thought of as zeroing out any residual current from the pixel sites when exposed to a true black condition. The inherent thermal noise from the sensors can change over time and with different operating temperatures so a black balance becomes necessary.

It's also worth noting again that black balancing does not cure stuck pixels. A stuck pixel is a shorted photosite and will produce maximum current even in total darkness. That's why pixel masking must be done, instructing the camera to ignore the offending pixel and approximate its output by looking at adjacent photosites.

-gb-

Thanks for the explanaion, Greg! It's a pity that EX1 doesn't have this function...

It most likely does. On the EX1 it can be done automatically, possibly that's what the camera is doing at power up or when you change certain settings.
On the EX3 where you could fit prime lenses it needs manual intervention. Doing a black balance without the iris closed or a lens cap on can lead to a "What just happened to my camera??" moment or two. I've done this once on a camera, until it finally dawns on you what you've done it's a bit of a worry!

Is infrared light polarized?

Can circular polarizer help eliminating IR contamination?

The EX1 does have black balance. I believe it's in the maintenance menu.

Yes it it there, but isn't active (grayed out) - at least with the 1.02 fw I still have.

Robert,

We can make almost any common filter size of the True-Cut IR 750. We have a few 77mm sizes in-stock and will likely make those a regular limited stock item. The other sizes will be special order with an average 3 to 4 weeks lead time. You may contact your favorite dvinfo.net sponsor dealer, tell them the filter and size you want, and we will work together to get you a price.

Ryan Avery
Schneider Optics

Ned, have you tested the IR750 under tungstun?

Polarization isn't a property of special wavelengths. Most surfaces reflect diffuse, so they destroy any polarization, if there was any. Transparent stuff like glass is reflecting purely linear-polarized, if the light hits it with the brewster-angle of incidence.

Dominik, then do I understand correctly that Polarizers block light more or less evenly throughout both visible and IR spectrum, and therefore do not work as IR blockers?

Not yet.

High priority this week is IR750 testing.

Ned

That's correct.

Ditto that.

Ryan Avery
Schneider Optics

Hi,
FYI I have a new EX3 with firmware 1.04-0084 and yes the IR problem is still there Sony really sucks on this one.
I have tried black balance, white balance and black balance again and it has not fixed the problem. I have just ordered a Tiffen 4 x 4 IR (IR680) due in 3 weeks to try to fix the problem. Most of my work is under QI lights. I had to do a interview yesterday and decided to not to use the Blonds and Redhead instead I use some new high intensity 2300K LED lights for the shoot. The images were great no IR contamination black were solid and crisp.
That’s great for close up work my problem is I do a lot of stage and musical productions that I have no control over lighting.
I would also be very interested in any tests using IR filters if some can post results.

Thanks
.

Please report how your 680 is doing, preferably post some grabs!

TIA,

Piotr

my ir problem dissapeared when i used the b&w 486 filter.

Dano

Same here.

Also, the amount of IR which is allowed to hit the sensors is determined by a fixed hi-pass filter which is permanently attached to the CMOS sensor assembly inside the camera. I wouldn't have expected a firmware update to alter that. Many CMOS-based cameras exhibit this behavior. If you want to see an extreme example, Google "Leica M8 IR contamination" - it will make you feel better about your EX1/EX3. (If "Sony sucks on this" then so does RED, Canon, ...)

I frequently shoot under hot lights and the 486 filter lets me control how far down the spectrum to go.

Same here as well.