Filter for IR contamination
 

I have read a lot of posts here and elsewhere regarding the IR contamination with the EX. I have personally experienced the IR contamination several times in a big way.... so to alleviate the problem and be prepared for similar situations in the future.... I am interested in purchasing an IR cut filter to correct the problem when it appears again (in 4X4 for matte box use). But which filter? I have read that the Schneider 750 is the one to buy, then I've read it won't work on an EX. I have read the 486 is going to solve the problem with an EX, then I've read that was developed for digital still cameras, and isn't the best option for video cameras. I haven't heard much about the 'Hot Mirror' filters (I think Tiffen and/or Format) and how successful they may be.

Anyone with any suggestions/experience with any or all of these filters? I'm also concerned about vignetting with some of these filters at wide angle. Any input would be greatly appreciated!

I can tell you that the 750 didn't do anything for my EX3, but the 486 does with some green vignetting that can be at least partially corrected in post. I was talking to one of the engineers at Schneider and he encojraged me to keep in touch with me so I could help with testing solutions, but he has since ignored all attempts at communication.

I hold little hope of anything being done in the near future and have decided on using a 486 as a compromise where the IR problem is too offensive to ignore.

I too have tried to clear this thing with Schneider and they have instructed me to buy the 486 instead of the Tru-Cut 750 since this does not work with the EX series but was meant to work with the RED camera. I've just received the 486 2 days ago and as soon as I've put it on I immediately saw the green vignetting at the extreme wide setting.

It seems that I have to avoid going to the full extreme wide and any slight cast will have to be corrected in post - a pain yes BUT much better than the aweful ugly contamination that I have encountered with this camera. Any filming of black fabrics under halogen light is a nightmare with this camera.

It seems that there isn't much choice and since the EX series are the only cameras that exhibit such severe contamination, I'm not sure whether companies will bent backwards to find a solution to a problem which apparently is not always problematic to all EX users.

Thanks, Brian, Les-
That's what I had read in both of your previous posts. Interesting that Schneider is now recommending the 486 over the 750? One post I had read by Ryan Avery he was suggesting just the opposite. The problem for me is that I rarely do any of my own editing, my clients take the media and run. So correcting in post is not really an option for me, and not something I want to have to explain to a client who will most likely be less than thrilled about it. A bit of a conundrum, I guess. Has anyone seen the contamination on objects other than fabrics?
Both times I noticed a rather severe amount of contamination happened under daylight conditions. One was outdoors in the Nevada desert just before sunset. The light was of course very low, so it was quite warm (a nice look, I didn't re-balance for the low light) and a woman we were interviewing was wearing a black sweater. The interview continued past sundown, and looking at the clip later, you can actually see the sweater turn from a deep magenta back to black as the sun dipped under the horizon. The producer was understanding (and fortunately that part of the interview wasn't going to be used anyway) but needless to say this is not something you want to have to explain to a client. The other experience was under HMI lighting with ambient daylight as fill coming in from a window. At least that incident remained constant.
This is frustrating, and obviously there does not appear to be a solution? I remember when tight pinstripes were our biggest worry in fashion on camera......

It's a shame one has to pay for "green vignetting." I shot a concert under theatrical lighting at a major venue, some of their tuxedoes came out black, while others came out brown (what's that tell you?). I'd rather have that than the green vignette!

Besides, the client didn't bat an eye at the differences in tuxedoes.

Once the folks at Schneider (or elsewhere) are able to offer a clean fix, then I'll invest in the filter.

I agree with you, Jay. I'd rather let people think someone was wearing a different color sweater than deal with the green vignetting. Besides, who's going to know if everything else is correct? Even my lav windsock turned the same color.... it looked like I had a perfectly matched windsock for every occasion!

I agree with you, Jay. I'd rather let people think someone was wearing a different color sweater than deal with the green vignetting. Besides, who's going to know if everything else is correct? Even my lav windsock turned the same color.... it looked like I had a perfectly matched windsock for every occasion!
Maybe one day we'll have an effective solution. I'm certainly not going to drop $200-$400 for a filter which only partially works.

This issue seems like something the companies don't want to talk about as they do not have an answer.

How do you correct for "green vignetting" anyway?

A correction for the outside of the frame would throw the inside off!

My question, Tim, would be what are they doing to get it? I'm not aware of any other filter (still or motion) that causes such an issue, and I've been making images for 40 years.

if you are not zooming too much with the filter, you can try to get a still from a white wall at full wide. this will give you a mask you can use to correct the default.
this will probably require to convert the green to magenta or orange and use some "multiply" feature in color correction.
Since i got the filter and same problem, i will soon try to find the best way to do that.
if you zoom , you can try to track some point in the video and use it to apply a zoom on the mask too , so the correction follows the zoom. I doubt it would be perfectly ok , but it could help enough to forget it.

Not when the person happens to be the groom and the groom knows that he has specifically hired a black outfit....

...and yes that's my question as well: How does one correct the green vignetting? I was thinking of creating a sort of a mask with colour correction for the vignet only. Does this makes sense?

...and by the way Derek, it was Ryan Avery himself who way back in November had told me in an e-mail NOT to buy the 486 as it is not good for the EX cameras and now in January this year he told me NOT to buy the Tru-Cut 750 as this is not good for the EX but that I should buy the 486..!!!!! I have only seen this IR contamination on certain black fabrics - black plastic or metal does not produce such an effect.

I have this gut feeling that no-one will correct this issue - except that future cameras will be produced in such a way to minimise it. In our case we have to lump it.

It is disappointing that this issue isn't taken care of much more at source (ie built into the camera).

This issue irritated me on the EX DVD I made. I am in a couple of shots wearing a black jacket and a black fleece underneath, but the fleece is brown. This isn't easily correctable in post because if I bring the fleece closer towards black my jacket will turn bluer.

This is clearly not just an issue with the EX. It affects a lot of other cameras as well. Just a shame that the problem doesn't look like being solved any time soon.

After a couple of incidents with black fabric turning into ugly brown, I did bite the bullet and bought this 486 filter for EX1.

Hot mirrors I tried did not work; but this filter does very well.

Note that this is a slim version and it fits under the stock lens shade of EX1 easily.

I figured, since this filter works mostly by *reflecting* IR light, I can't put any other optical elements in front of it anyway, so why would I need a front groove.

Also, luckily, I have not run into any perceptible green vignetting yet, so can't attest to that.

I don't know as I do not own the camera. I am looking at purchasing one but I shoot a lot of live events and all are lit with the type of lighting that brings out this red colored black issue.

My understanding is that this only happens on the wide end of the lens or if you are using a WA adaptor.

Maybe the light is defracted at a weird angle when the lens is full wide and it escapes the full effect of the filter? Or distorts it?

In the end, between this and the roling shutter the EX-1 does have some compromise built into it. And if one of these compromises happen to be on your job this camera does not hold that much value IMHO.

I talked to Schneider as well and they and Abel Cine are recommending the 486 filter for every EX-1/3.

Sad that Sony does not step up and do something as these are "pro" models.

Actually Sony does have built-in IR filter, but apparently it is ineffective in certain situations.

I keep my 486 on the lens at all times.

Has anybody seen the 486 in a 4X4? The 77mm will be useless with a matte box on. (if you plan on using any filters in the matte box, that is)

No, that's not the case. It happens regardless of the focal length and without any adaptors.

The rolling shutter is an urban legend. It's not the camera, it's the user.

Agreed! Surely they saw this issue in pre-production models!

Really?! Ouch..! From the little testing that I did I only saw it in th extreme wide setting of the stock lens. So can it happen even in the telephoto setting?

...and how are you correcting it in post, if I may ask?

is the IR contamination only a problem with tungsten/halogen based lighting sources or have their been any other nasty suprises?

We're talking about the blacks looking brown, right?

Insofar as correcting in post, I haven't. You can't really. So far I've been lucky. The concert I shot was lit with theatrical lighting, which ranged from white to blue to green to magenta to red--mixed. So the audience presumes it's just the colored lights.

Problem was, as I mentioned, some of the tuxedoes looked brown and some looked black! Also, as someone mentioned, if you try to remove the red cast in the black, you add green or blue to the other areas. So pick your poison!

There may be some very sophisticated (expensive) applications out there that could correct it, but I certainly can't afford them.

I have seen it both outside under daylight (although it was very warm daylight, just before sunset) and controlled HMI (5600 balanced lighting) with some ambient daylight coming in from windows. I would estimate the sunset light temp would have been around 4400-4500K, and I remember the HMI lighting was about 4600K (realizing that the number the camera gives is rather subjective) but these two values are pretty close. As I described in my earlier post, when the sun dipped below the horizon in the sunset scenario, the IR contamination disappeared before my eyes. That color temp (when the sun was no longer visible) was probably around 7000K. I have not done much incandescent light shooting with this camera, yet, so I can't speak to that.... but I know others have seen the contamination under incandescent light around 3200K.

That said, I just remembered something a Sony tech I spoke with told me, which I haven't tested yet. He asked a very good question when I described my experience with the IR contamination, he asked if I was using a custom picture profile, or the standard one? I hadn't thought of that.... I was on a custom profile. He did not infer that a custom profile was causing the issue, but wondered if adjusted gamma and black might have made the contamination appear worse? I think it's a valid question, at least I will switch to the standard profile next time I have a contamination issue, and just see if anything looks different? Worth a shot, anyway..... which leads to the next thought: I wonder if a PP could be developed specifically for IR contamination? I was using Doug Jensen's PP from the Vortex Media DVD before anyone asks. I'm not saying for a moment that this PP is causing to the problem, but it's valid to see if anything other than the standard set makes a difference. Sony probably never bench tested the cameras in anything other than their vanilla standard setting.
Hey, all you PP developers out there? Here's your chance! Knock yourselves out.....

Jay, I was speaking to the green cast from the 486 filter which is only supposed to appear when at full wide or using a WA adaptor.

And the rolling shutter effect that bother me is the strobe handling with half exposed frames which is not a myth.

Okay, sorry! Yes, that is correct.

From the tests I've seen online from other videographers, it's user error. This would also include my personal experience with the EX3. I've recording people being photographed with a strobe and there were no half exposed frames. What else can I say?

I've been trying to sort out the IR issue for quite some time.

While endemic to CMOS cameras, IR contamination can even occur with CCD cameras. I shot Steve Job's keynote at MacWorld 2008 (wow, the last Jobs keynote ever) on an HVX200. It really looked great, even down to Steve's brown turtleneck. Yes, IR contamination under tungsten light using CCD's.

I've tested the Schneider 750 but really can't see any difference, mainly because I just have not been able to re-create a contaminated scene.

I've got an email into my contact now at Schneider just to see what they are recommending these days. I have not tested the 486 yet.

Rolling shutter on CMOS cameras is real. The question is the extent to which it affects your shot. If I were shooting strobes in a club, I'd probably shy away from a CMOS camera.

But for what I do (and I confess to being more of a tech type and writer than full-time shooter), the EX cameras produce amazing images.

Jay, it was my confusion. I was referring to the green vignetting - I thought that you saw the green vignetting at the telephoto setting - that's why I was shocked. But now I realized that you were referring to the IR contamination.

Also I was asking how the green vignetting (and not the IR contamination) can be arranged in post.

And as for Dereck's question, no I always use the original factory settings (no pp's no alteration whatsoever of the picture at all) and yet I saw the abysmal IR contamination on particular fabrics only. I could be shooting in halogen and one piece of black fabric appears black and another black piece of different material appearing nauseatingly brown.

Jay I do not want to side track this thread but what could be the user error in the rolling shutter effect? I film a lot of footage where there are a lot of flashes going off - it's a matter of how the light falls on the subject and ultimately on the sensor. But I've seen loads of half frames and I cannot see what I'm doing wrong. I'm doing the same what I used to do with the CCD cameras - in the CCD cameras I use to get 1 whole frame white washed and now I get half a frame. No big deal really but it's the camera's fault not mine, in my opinion.

Brian, I replied via e-mail. Thanks!

We have 3 EX1s in our fleet and 2 EX3s. All are fitted with 486 slim filters. These cameras are used by many people.
Prior to fitting the 486 filters some had noticed the IR contamination problem. Since fitting the filters no one has had that problem and no one has mentioned the green vignetting problem. I've not noticed it myself or if it was there it's not bad enough to need correction to my eyes.

If you do need to correct it, it would be easy enough to correct. Duplicate the track, apply color correction and a circular mask to the upper track. If you're a Vegas user the Cookie Cutter FX with a circular cutout and feathered edge should do the trick.

The rolling shutter issue is not an urban myth. It's is well documented, understood and affects many cameras. What is an urban myth is how much the issue affects most of us. it might be a problem for those few matching CGI elements in post in fast moving scenes. The half flash frames should be very easy to fix if they are a visual nuisance but the odd one or two I've noticed I just left.

If you're looking for a reason not to buy any camera you can always find one or two :)
Neither of these are good reasons in my opinion.

Good to know about the filters Bob.

IF the rolling shutter was ever more than anecdotal, and that's a BIG if, it isn't now. The one thing I will agree with Bob on is if you're looking for a reason not to buy any camera...

The IR contamination goes further than being anecdotal. However, taking everything else into consideration, not the least of which is the image quailty, it's well worth that small inconvenience, in my opinion.

Ignoring shortcomings of the particular camera seems like disservice to people who are trying to learn about strong points as well as weaknesses of the device.

IR contamination: certain black surfaces turn brown under *any* lighting conditions - I've seen it under the natural sunlight; tungsten; fluos...

Rolling shutter: what is this discussion about? Of course there is rolling shutter issue with EX1. It manifests itself in skewed vertical lines on very fast pans; as well as in the fact that photo flashes are exposed only partially in the frame. Go frame-by-frame in your NLE to see for yourself. If you can't see it, you are not looking close enough.

Now, does it matter for the average viewer?

Blacks turning into brown - yes it matters. That's why I always have my slim 486 filter on my EX1.

Rolling shutter issues - not so much. Fast pans are blurred anyway, so it's hard for the viewer to discern the skewed lines at normal playback speed. Same with flashes; hard to see that one frame is only partially exposed at normal projection speeds...

Alex, can you name any camera that is 100% perfect--no short-comings whatsoever?

I've not seen any rolling shutter in the EX3--ever--the same is true for many, many others, even with the EX1. The IR issue is, evidently, correctable, so what's the problem?

I stand by assertion that the rolling shutter is a user induced problem. If it weren't, all the EX cameras would be exhibiting it, just like the IR issue.

All EX cameras DO exhibit the rolling shutter phenomenon. Along with the RED One, Nikon's D90, and many other CMOS cameras. It's more obvious on some models than others, of course, and on the EX cameras it doesn't seem to be very severe. As mentioned before, most of the time when it would occur, there is enough motion blur that the effect is not objectionable to the eye. I've certainly seen it, but it's one of those things that you really have to be looking for to notice. The EX series cameras do not have bad problems with rolling shutter issues, so I would agree that it's not something to worry about.

Apparently the EX does have some rolling shutter compensation built in, but it can't get rid of all issues. Although I have to say that in all my shooting, including fast moving rally cars etc, I have never really noticed it. I've also done a lot of work at full telephoto and not noticed any issues.

Has anybody tried using this filter on the EX1 or EX3? -

77MM STANDARD HOT MIRROR - Tiffen.Com

If so does it manages to cut the IR? Does it produce any green vignetting? And above all does the lens hood manages to attach to the filter (even though it is not of the slim type)?

Brian, I had purchased the Tiffen 77mm hot mirror filter about 6 months ago and it is totally useless on the EX1. I returned it for a refund. The Schneider 486 is the best option so far. I have a standard 77mm 486 and a 52mm 486 for my Nikon lenses which I use on my Letus.

for what it's worth....
 

So I was just preparing for a shoot today, and had my camera on my bench with some stuff laying about. I noticed an old Tiffen filer pouch (which is black) was exhibiting some IR contamination so I played around a bit. I was interested in seeing if changing the profiles made a difference, as a Sony tech once suggested to me.

Here's what I found. First of all, going to the 'standard' (no adjustment) profile made no difference at all. So there's that. But I did notice that one profile I had entered, the recently released 'BBC Film' profile minimized the effects of the contamination somewhat, or at least it seems to. There is probably several reasons for this, but the main thing I notice between the profiles is the different matrix setting. One profile, which is Doug Jensen's which I like very much and use as my regular setting, uses 'HI SAT'. This is the brighter and more saturated of the two images. The other is the BBC profile, which uses the 'CINEMA' matrix setting.

I'm open to suggestions as to why the IR seems much less obvious between the two settings? Maybe there is a chance that changing matrix settings will help when the issue arises? Thoughts, anyone? Also interesting to note that the smooth black of the Chrosziel adapter ring is uncontaminated, as is the border of the old filter pouch.

Yes, we've noticed that on a number of occasions. It seems that this is the result of a variation of fabric and/or dyes.

I've had a fair amount of experience with infrared(IR) sensing equipment. What appears "black" to the human eye is anything but black in IR wavelengths. You, simply, cannot judge the "color" of an object in the infrared regime with your eyes. It is the black anodizing on the chroziel that exhibits the "un-contamination". I don't know who coined that term, but, I find it rather laughable. Contamination, indeed!

All objects reflect light, to one extent or another. If one thinks in terms of the three primaries, RG and B, objects will reflect each color in different strengths, absorbing some frequencies more than others. Infrared is not a single "color", per se, but a range of colors, or wavelengths. THe CMOS in the EX1 responds differently to different wavelengths, as well, resulting in the uneven interpretation of RGB channels, displaying red more than other wavelengths. No single filter is going to account for the unique infrared signature of all objects that appear black to the human eye. Filter design uses specific coatings to either absorb or reflect certain wavelengths. No single filter will block all the frequencies that are in the IR range. Furthermore, attenuation of undesireable wavelengths is not a step-function, but, rather, a gradient of attenuation as the frequency moves into the IR regime.

Robert thanks for the info re:Tiffen. I found another possible candidate : Rosco US : Film/Video : Trucolor

The problem is that they do not do them in 77mm circular format. I've contacted the company and they did tell me that they are aware of the green vignetting of certain filters and that theirs do not exhibit such green cast. They didn't mention however whether they have tested it on the EX series. They even went further in offering to cut a 77mm sized filter for me but that it was up to me to mount in on a 77mm thread. (I have no idea how to do it!!)

If someone can confirm that they have used such a filter and that it works with the EX then I will consider buying a hood + filter holder (I'm not interested in matte boxes as they are too cumbersome for my run n gun work that I do).

Anybody bought the Rosco?

Great information, Brian. One would think (hope) that if Rosco has "solved" the problem, surely Schnieder couldn't be too far behind.