Filter for IR contamination

[David C. Williams]

If that's the filter with the dye only, no dichroic layers, that's great news because that is the perfect filter for the EX. Cuts all IR and no vignette. I think the process is branded ColorCore.

These http://www.tiffen.com/pr_infrared_filters.html

[Leonard Levy]

Ed,
This sounds like very good news.

Do you have a name for that filter without ND.
Also did you talk with them about the value of making round filters for it Every EX-1 owner would buy one.
Usually they can make a round filter if you ask them, but it will cost.

Who did you talk to at Tiffen?

[Ed Kukla]

Harold E. (Harry) Hofmann
Senior Technical & Customer Service Rep.
Tiffen Acquistion LLC
hhofmann@tiffen.com
631-273-2500 ext. 1405

[Derek Reich]

Originally Posted by Ed Kukla

David
I was on the phone today with Tiffen. You can get the filter WITHOUT ND. It is somewhat confusing as they mostly make the IR WITH ND for Red cameras.

Interesting. When I was on the phone to Tiffen, they told me this was not available. They also told me that any 4x4 IRND would have to be custom-cut and would take at least 4 weeks!

I also do not agree with the post which said hot mirrors do nothing for the EX. I recently took delivery of a Tru-Cut 680, and it seems to work quite nicely so far! I know the 750 does not work with the EX, but the 680/486 seems to do the trick. I have not even noticed the vignetting yet, so that does not seem to be a big issue.

I would be wary of the Tiffen without ND built in, as it seems using external ND's (including the ones in the camera) may very well affect the performance of the IR correcting capacity, since the amount of dye is based on the level of ND used. If no ND is needed, then it's probably okay.... but what do you do when using this filter outdoors?

[David C. Williams]

Your correct, the 680 works, albeit with wide angle green vignette. I should have specified the 720 and 750. The built in EX hot mirror seems to work from @ 700nm onward.

The Tiffen IR ND range have the dye component balanced to not cut far red, but to reduce it in proportion to visible light. A normal ND 0.9 reduces visible light to 12.5%, but your still getting 100% of far red getting to the sensor. It then becomes predominant, and you get problems. The IR ND 0.9 dye reduces far red to the same proportion as visible light, to 12.5%

That's why I thought they didn't make them in that fashion as IR only. I hope is possible.

[Derek Reich]

But what about the amount of dye being used in proportion to the level of ND being used? If a ColorCore filter is used without ND added, does this mean no external ND (camera NDs or other external NDs) can be used?
This is the way I understand it.... and why I decided against the Tiffen IRND filters. I simply wasn't going to pay $250-$300 for EACH level of ND I might need to have the proper amount of dye in proportion. I already have an expensive set of 4x4 ND filters. Personally, I'd rather save $$$ and deal with the possibility of some minor vignetting with the True-Cut 680 than have to buy a whole new set of NDs. I don't see how the ColorCore can be used successfully outdoors on a bright day.
If the ColorCore without ND can be used with various levels of external NDs, (including grad NDs) I'd be more interested...... anyone know if this is possible?

[Piotr Wozniacki]

Derek,

Since you're the first one here to have the 680 filter, it'd benefit all of us if you posted some screen grabs showing its effectiveness. Before ordering, I for one would like to be sure it's as effective as its screw-on sibling, the 486.


Thanks in advance !

[Bob Grant]

I've been following this thread for some time and am left with the feeling this issue will never be solved to everyones satisfaction.
I believe the 486 will be as good as it gets. The solution requires an extremely sharp cut at a specific wavelngth. This can not be achieved with a dye based filter. The 486 is multiple layers of a metallic coating which probably act like a stack of HiQ notch filters to effect a band cut filter.
I did a very crude test using two IR LEDs and our HC5 in nightshoot. One of the LEDs is around 1100nm and is invisible to the eye. The other would be around 900nm and is just visible. Both light up the HC5 nicely although the longer wavelength one is better. The 486 totally kills the shorter wavelength IR but not so much of the longer wavelength IR LED.
The IR filter in the HC5 (Nightshoot Off) kills both.
This was a very crude test but it certainly doesn't disprove my belief that the 486 is a very complex filter and that to achieve the same effect in any other way would be impossible. It is not a hot glass filter. We do have several extreme ND hot glass filters and from the outside they are like looking into a mirror, same as welding goggles. These filters are designed to stop heat (long wavelength IR) that could cook a sensor or eyeballs or at the very least cause green shifts in the image.

[Piotr Wozniacki]

I'm not sure what you mean by that, Bob - is it that the 680 isn't going to work as effectively as the 486?

I did have the 486 but sold it, since I'm using the 35mm adaptor and need a 4x5.65" version for my matte box.

I now know it wasn't a good decision, but at least would like to see the effectiveness of the 680 before I pay again :)

[Derek Reich]

Originally Posted by Piotr Wozniacki

I'm not sure what you mean by that, Bob - is it that the 680 isn't going to work as effectively as the 486?

I did have the 486 but sold it, since I'm using the 35mm adaptor and need a 4x5.65" version for my matte box.

I now know it wasn't a good decision, but at least would like to see the effectiveness of the 680 before I pay again :)

Piotr, unless I misunderstood Ryan at Schneider Optics, the Tru-Cut 680 is exactly the same filter as the 486. (correct me if I'm wrong, Ryan?) The only difference being the 486 is a screw-on filter and the 680 is a 4x4. (perhaps the same designation for a 4x5.65 if they make that size) Why it has a different name is confusing, but irrelevant to the performance of the filter.

I'll post some frame grabs when I have a chance to set up a decent test, but don't expect them to be any different from previous tests done by others with the 486/680.
I agree with Bob, this filter is as good as it gets at the moment. No filter currently is 'the' perfect filter, but the 486/680 is as close as we can come.

[Piotr Wozniacki]

Michael,

Yes I have read the whole series of those articles.

Derek,

Yes I'm also assuming the 680 to be the rectangular equivalent of the 486; however I'd appreciate seeing some grabs of it in action. TIA!

[Derek Reich]

This from Art Adams:

Read the comments which follow the short cheat sheet. Interesting that Art doesn't list the 680/486 as a filter to use for the EX, also interesting that a couple of the comments mention that the Tiffen IRND is not available without the ND, but some people here have mentioned that it is. (for the record, I was told by Tiffen that it is not available without ND)

I'll post some grabs sometime in the next few days with the 680, looking forward to hearing from anyone if they are successful in obtaining a Tiffen without ND, and how it works?

[Ed Kukla]

I'm still trying to sort out the tiffen confusion. I think there are two kinds of tiffen filters. An IRND and a Hot Mirror (with or without ND). But I'm still not sure and can't get further answers back from them.
I think the Hot Mirror can be had without ND and should be a good solution but I can't yet confirm that.
And, they don't come in 77mm, only square or rectangle.

[Bob Grant]

I think I've said this before on another thread on the same subject.

I believe there's a good chance that Art is solving a different problem to the one we're trying to address. If you add an external ND to any camera and if that ND does not also cut IR you can get more IR than visible light hitting the sensor and then you can have a problem. This seems to be one reason why the extreme ND filters commonly are hot glass. If you cut 10 stops of visible light and no IR it's pretty easy to see how a problem can arise. Of course that's an extreme example but I've even read of people having issues with a HVX 200 when they gelled windows and the gells did not cut IR.

So perhaps someone should invite Art to elaborate more on this issue before anyone jumps the gun and spends serious money on a filter that may not do the job or worse introduce more problems.

I also found this photo from another thread interesting:

http://www.dvinfo.net/conf/attachmen...r-dsc_3391.jpg

I see what appears to be a deep greenish filter in the optical path of the EX3, no doubt it's the same in the EX1.

[Derek Reich]

Okay, so I promised to post some frame grabs with my Tru-Cut 680. I (hastily, as ususal) threw some black stuff on my bench and shot a couple of clips under different color temps, with and without ND, and with and without the 680.

This is not intended to be a test such as the comprehensive one conducted by Art Adams at the ProVideo Coalition. Just an everyday example of what one might expect using the 680. I have no other filters for IR with which to test anyway.

My conclusion is this. The 680 works perfectly for removing IR contamination (or, far red or whatever it is) with little side effects other than a slight color shift in the image (very slight) and a rather major green vignette at full wide. This was to be expected. Zoomed in at about 15mm on the lens, the vignette all but disappears. So the short answer to this problem (IMHO) is buy ONE filter, the Tru-Cut 680 and have it work in most of your shooting conditions, or buy SEVERAL Tiffen IRNDs (because they cannot be used with other NDs, and you have to buy a different filter for each level of ND you might need) and have them work in most of your shooting conditions. Neither is going to be a 100% solution.

Interesting to note that I put several different black items on the bench, with varying textures and structure. Only one was consistently and heavily affected in all color temps (the Skullcandy bag) and one fabric item was never affected (the gray and black shirt)
Other items were affected in different areas, like the borders on the Lectro cube cover or the media wallet. Smooth surfaces don't seem to be affected at all. (no surprise there) The reddish band on the speaker is indeed that color, that is not contamination.

So here you go, for your viewing pleasure: the highly professional stickies will indicate the status of that clip. I shot under incandescent (3400K) Fluorescent (4900K) and Daylight (HMI @ 6300K)
I did NOT re-balance when using the 680. (don't know if I should have done that or not)
The images have not been adjusted at all, they are straight frame grabs-
enjoy. There will be a second part to get the rest of the images in......