Filter for IR contamination

Leonard Levy · February 25th, 2009, 01:06 AM

Chuck,

I do appreciate what ryan has been doing a great deal, but don't count Tiffen out.
I recently talked to them at a trade fair in San Francisco and they brought a whole group of new "Hot mirror" and "Hot Mirror IR ND" filters designed specifically for the Red and other video cameras. They have also been doing their homework on this one though no one ahs shown our little camera the attention that Ryan so generously has.

As for our beloved camera manufacturer....

Lenny Levy

Tip McPartland · February 25th, 2009, 02:36 AM

As an owner of an EX-3, I have noticed reddish blacks and not loved it, but being very happy with the camera otherwise, didn't ruminate too much over the problem. This thread has been extremely enlightening about the true import of the problem and about the range of present solutions and their limitations.

It has also stimulated the manufacturers (some anyway) to greater efforts and made us endusers aware of the possibility of more optimal filters in the pipeline.

Having read all the posts, I've elected not to jump the gun and buy the 486 or anything else, but to patiently wait for Schneider, Tiffen and Rosco to sort things out and give we thousands of EX owners what we need to get great results with our cameras under almost all conditions.

Thank you Chris Hurd for such a great forum, and congrats on your role in getting the ball rolling on the SI camera!

Tip McPartland

John Hamlik · February 25th, 2009, 06:21 PM

Quote:

Originally Posted by David C. Williams

The green vignetting usually happens when wide angle is used. If you zoom in somewhat as your concert shot is, you shouldn't get the green fringe. If you were setup on stage, you probably would have gotten fringing.

Here is a stage shot at full wide, to me an acceptable image in comparison to burgundy tuxes or brown and some black.

John Hamlik · February 25th, 2009, 06:23 PM

Quote:

Originally Posted by Ben Chiu

Thanks for posting the your findings, John. Can you tell us if you're using any of the picture profiles from the forum (like Bill R's)?

720/60p BBC Video setting stepped down 1/2 stop. 3200k preset for color temp. Still quite haven't figured out the best way to deal with downlighting and bald heads...

Matt San · February 26th, 2009, 03:57 PM

Quote:

Originally Posted by Derek Reich

yeah, I noticed that too. but with the wide angle vignetting that filter seems to produce, it did not seem like the ideal solution. the IRNDs without Hot Mirrors seem more versatile and effective. I've ordered one, and will post my experience when I've had a chance to use it.

which one did u order?

Leonard Levy · February 26th, 2009, 04:46 PM

What's recommended though for indoor applications that don't need ND?

Derek Reich · February 26th, 2009, 07:17 PM

Quote:

Originally Posted by Matt San

which one did u order?

Well, I had originally ordered a .9 IRND. But when I called to add some additional items to my order, I found out that Tiffen does not make that filter in a 4x4, and they would have to cut it custom which would take up to two months. (the original order was accidentally made as an IRND WITH Hot Mirror, which is apparently available)
So I canceled the order, and am in limbo at the moment. I really didn't want to wait two months for the filter.....

Tip McPartland · February 26th, 2009, 09:13 PM

Regarding shooting indoors without ND filters, I would think that another approach would be to use lights that have a reduced IR signature like solar spectrum Alzo Digital phony HMIs (actually HID) lights that run much cooler, their "higher color temperature" notwithstanding. Also solar spectrum fluorescent or better yet LED instruments would give you less IR with which to cope.

This might be one reason that Sony is selling LED on-camera lights.

Tip McPartland

Bob Grant · February 26th, 2009, 09:42 PM

Dare I say this?

I think Adam's methodology is possibly flawed. Look at his first shot, no ND filter and no IR contamination. The problem he's trying to fix is different to the one we're trying to fix.
Also his test fabric is synthetic, we know that certain fabrics don't show up the problem.

By adding ND filters he's induced a different kind of problem where the sensor and it's filters are swamped by IR as the external ND filter doesn't cut IR. We see IR problems in all of the frame including the chart. This is not the same problem that we're having. Sure it could be a problem as well if adding external NDs.

Piotr Wozniacki · February 26th, 2009, 09:56 PM

Quote:

Originally Posted by Bob Grant

Dare I say this?

I think Adam's methodology is possibly flawed. Look at his first shot, no ND filter and no IR contamination. The problem he's trying to fix is different to the one we're trying to fix.
Also his test fabric is synthetic, we know that certain fabrics don't show up the problem.

By adding ND filters he's induced a different kind of problem where the sensor and it's filters are swamped by IR as the external ND filter doesn't cut IR. We see IR problems in all of the frame including the chart. This is not the same problem that we're having. Sure it could be a problem as well if adding external NDs.

This is exactly what I think.

1. I don't have any serious problems in sunlight, only tungsten which implies lowlight - thus using ND is out of question

2. I never have everything contaminated, just some specific, black or dark blue materials

3. I have found that instead of using 3200K WB (typical for tungsten), it's enough to dial WB down to some 2800-3000K and the problem is gone, with only a little overall impact on the remaining colours - usually fixable in post.

4. Re: using LED, or another cold light, instead of tungsten: the rendition of people's faces it creates is a much more serious problem that the tungsten IR contamination of their tuxedos!

Bob Grant · February 26th, 2009, 10:17 PM

Quote:

Originally Posted by Piotr Wozniacki

This is exactly what I think.

1. I don't have any serious problems in sunlight, only tungsten which implies lowlight - thus using ND is out of question

2. I never have everything contaminated, just some specific, black or dark blue materials

3. I have found that instead of using 3200K WB (typical for tungsten), it's enough to dial WB down to some 2800-3000K and the problem is gone, with only a little overall impact on the remaining colours - usually fixable in post.

4. Re: using LED, or another cold light, instead of tungsten: the rendition of people's faces it creates is a much more serious problem that the tungsten IR contamination of their tuxedos!

I'd have to disagree about 4). We use nothing but daylight balanced fluro, LED or HMI lights and everyone of our users are very happy over how good color rendition is. That's on top of the reduced heat and power requirements. So happy in fact we've just doubled the number of such lights we have on offer, despite the state of the economy we cannot meet demand at times. These lights have been used with just about every video camera out there and to shoot 35mm, 16mm and stills.

I'd also point out that I haven't seen a tungsten lamp in a local television studio down here in a long time. I am of course talking about light sources designed for television and film. Certainly there's no shortage of cheap "daylight" lights with poor CRI that can indeed give bad skin tone.

Piotr Wozniacki · February 26th, 2009, 10:32 PM

Bob,

I only meant a situation where an on-camera fill light somebody mentioned above is LED, and the rest of lighting is a typical, home tungsten - this is when balancing the WB is extremely difficult, and skin suffers most (or you get everything else reddish if you set your WB high).

For a fully controlled lighting environment, you're right of course.

Bob Grant · February 26th, 2009, 11:56 PM

Fair enough, You know I could tell you the solution to that problem but I know it's probably our of your price range.

I just took the 486 off my camera and tried shooting various black fabrics in full sun. No sign of IR problems that I could see. From all the screenshots that I've seen of IR problems under sunlight they're all taken when the sun is very low in the sky. I suspect at those times of day the shorter wavelengths are more filtered by the atmosphere.

Dean Sensui · February 27th, 2009, 01:53 AM

Based on Art Adams' article, it seems that the problem isn't IR. Otherwise the filters designed to cut wavelengths longer than 700 nanometers would have the desired effect. They didn't.

Instead, he seems to have discovered that a filter which attenuates wavelengths longer than 680 nanometers appears to work well.

B&H does have a Tiffen filter that might work. But it's still an ND filter that cuts a full stop: Tiffen | Neutral Density 0.3 (ND) Infrared Glass | 45650IRND3

I can test to see if an LED light helps reduce the problem. I have a set of Cool Lights and the spectral distribution of the light has a significant dropoff after 610 nanometers. Maybe I can do a comparison of a few black objects shot under full CTB tungsten and the daylight LED.

I posted charts comparing the Cool Lights LED to full daylight here:
http://www.dvinfo.net/conf/photon-ma...ml#post1018808

Bob Grant · February 27th, 2009, 04:32 AM

I think I can confirm that the issue is not caused by typical IR wavelengths. I pointed a remote control at my EX1 and with the 486 filter on saw nothing. With the 486 off I could just see the LED blink but it was quite faint. My other camera with the internal IR cut filter out of the path (Nighshoot) the IR LED in the remote lights up very brightly. Also I noted that the EX1 showed the IR LED as white so it's getting in through all the RGB filters on the photodetectors roughly the same amount.
I don't know the wavelength of these IR LEDs, they typically seem to be around 800nm to 900nm.
So I think what Adam Wilt has confirmed is what Ryan Avery said some time ago. The red filters on the sensor are still letting wavelengths just below visible red through. That might have a deliberate move by Sony to get more sensitivity out of the sensor. Cutting it right off at the edge of the visible spectrum many have also attenuated some of the visible spectrum as well.

So really I think we're back to square one. We need a filter with a very sharp cutoff at 680nm out to at least 720nm. Any wavelength longer than that and the camera is already filtering that out.
Now the Tiffen filter according to Adam is using a complimentary color filter and it has to match the ND filter, that's all good but doesn't give us a filter we can use to cure the problem with the camera at all.
I'd also suggest that the 486 is not a typical "hot glass" filter. We have a ND 1.5 and ND 1.8 hot glass filter and the coating on the front looks just like a mirror, I could use it to check my makeup.

To put it simply , Adam has confirmed the nature of the problem, we still do not have a solution. Shooting stage shows with a ND1 filter on the camera even if it cured the problem is not an option, period.

February 25th, 2009, 02:36 AM	#62
Tip McPartland Major Player Join Date: Nov 2005 Location: Crestline, California Posts: 351	This is what makes this forum great... As an owner of an EX-3, I have noticed reddish blacks and not loved it, but being very happy with the camera otherwise, didn't ruminate too much over the problem. This thread has been extremely enlightening about the true import of the problem and about the range of present solutions and their limitations. It has also stimulated the manufacturers (some anyway) to greater efforts and made us endusers aware of the possibility of more optimal filters in the pipeline. Having read all the posts, I've elected not to jump the gun and buy the 486 or anything else, but to patiently wait for Schneider, Tiffen and Rosco to sort things out and give we thousands of EX owners what we need to get great results with our cameras under almost all conditions. Thank you Chris Hurd for such a great forum, and congrats on your role in getting the ball rolling on the SI camera! Tip McPartland

February 26th, 2009, 09:13 PM	#68
Tip McPartland Major Player Join Date: Nov 2005 Location: Crestline, California Posts: 351	Another way: reduce IR Regarding shooting indoors without ND filters, I would think that another approach would be to use lights that have a reduced IR signature like solar spectrum Alzo Digital phony HMIs (actually HID) lights that run much cooler, their "higher color temperature" notwithstanding. Also solar spectrum fluorescent or better yet LED instruments would give you less IR with which to cope. This might be one reason that Sony is selling LED on-camera lights. Tip McPartland

February 26th, 2009, 10:32 PM	#72
Piotr Wozniacki Inner Circle Join Date: Aug 2006 Location: Poland Posts: 4,086	Bob, I only meant a situation where an on-camera fill light somebody mentioned above is LED, and the rest of lighting is a typical, home tungsten - this is when balancing the WB is extremely difficult, and skin suffers most (or you get everything else reddish if you set your WB high). For a fully controlled lighting environment, you're right of course. __________________ Sony PXW-FS7 \| DaVinci Resolve Studio; Magix Vegas Pro; i7-5960X CPU; 64 GB RAM; 2x GTX 1080 8GB GPU; Decklink 4K Extreme 12G; 4x 3TB WD Black in RAID 0; 1TB M.2 NVMe cache drive

February 27th, 2009, 01:53 AM	#74
Dean Sensui Inner Circle Join Date: Oct 2001 Location: Honolulu, HI Posts: 2,054	Based on Art Adams' article, it seems that the problem isn't IR. Otherwise the filters designed to cut wavelengths longer than 700 nanometers would have the desired effect. They didn't. Instead, he seems to have discovered that a filter which attenuates wavelengths longer than 680 nanometers appears to work well. B&H does have a Tiffen filter that might work. But it's still an ND filter that cuts a full stop: Tiffen \| Neutral Density 0.3 (ND) Infrared Glass \| 45650IRND3 I can test to see if an LED light helps reduce the problem. I have a set of Cool Lights and the spectral distribution of the light has a significant dropoff after 610 nanometers. Maybe I can do a comparison of a few black objects shot under full CTB tungsten and the daylight LED. I posted charts comparing the Cool Lights LED to full daylight here: http://www.dvinfo.net/conf/photon-ma...ml#post1018808 __________________ Dean Sensui Exec Producer, Hawaii Goes Fishing

February 25th, 2009, 01:06 AM	#61
Leonard Levy Trustee Join Date: Feb 2006 Location: USA Posts: 1,684	Chuck, I do appreciate what ryan has been doing a great deal, but don't count Tiffen out. I recently talked to them at a trade fair in San Francisco and they brought a whole group of new "Hot mirror" and "Hot Mirror IR ND" filters designed specifically for the Red and other video cameras. They have also been doing their homework on this one though no one ahs shown our little camera the attention that Ryan so generously has. As for our beloved camera manufacturer.... Lenny Levy

February 26th, 2009, 04:46 PM	#66
Leonard Levy Trustee Join Date: Feb 2006 Location: USA Posts: 1,684	What's recommended though for indoor applications that don't need ND?

February 26th, 2009, 09:42 PM	#69
Bob Grant Trustee Join Date: Nov 2005 Location: Sydney Australia Posts: 1,570	Dare I say this? I think Adam's methodology is possibly flawed. Look at his first shot, no ND filter and no IR contamination. The problem he's trying to fix is different to the one we're trying to fix. Also his test fabric is synthetic, we know that certain fabrics don't show up the problem. By adding ND filters he's induced a different kind of problem where the sensor and it's filters are swamped by IR as the external ND filter doesn't cut IR. We see IR problems in all of the frame including the chart. This is not the same problem that we're having. Sure it could be a problem as well if adding external NDs.

February 26th, 2009, 11:56 PM	#73
Bob Grant Trustee Join Date: Nov 2005 Location: Sydney Australia Posts: 1,570	Fair enough, You know I could tell you the solution to that problem but I know it's probably our of your price range. I just took the 486 off my camera and tried shooting various black fabrics in full sun. No sign of IR problems that I could see. From all the screenshots that I've seen of IR problems under sunlight they're all taken when the sun is very low in the sky. I suspect at those times of day the shorter wavelengths are more filtered by the atmosphere.

February 27th, 2009, 04:32 AM	#75
Bob Grant Trustee Join Date: Nov 2005 Location: Sydney Australia Posts: 1,570	I think I can confirm that the issue is not caused by typical IR wavelengths. I pointed a remote control at my EX1 and with the 486 filter on saw nothing. With the 486 off I could just see the LED blink but it was quite faint. My other camera with the internal IR cut filter out of the path (Nighshoot) the IR LED in the remote lights up very brightly. Also I noted that the EX1 showed the IR LED as white so it's getting in through all the RGB filters on the photodetectors roughly the same amount. I don't know the wavelength of these IR LEDs, they typically seem to be around 800nm to 900nm. So I think what Adam Wilt has confirmed is what Ryan Avery said some time ago. The red filters on the sensor are still letting wavelengths just below visible red through. That might have a deliberate move by Sony to get more sensitivity out of the sensor. Cutting it right off at the edge of the visible spectrum many have also attenuated some of the visible spectrum as well. So really I think we're back to square one. We need a filter with a very sharp cutoff at 680nm out to at least 720nm. Any wavelength longer than that and the camera is already filtering that out. Now the Tiffen filter according to Adam is using a complimentary color filter and it has to match the ND filter, that's all good but doesn't give us a filter we can use to cure the problem with the camera at all. I'd also suggest that the 486 is not a typical "hot glass" filter. We have a ND 1.5 and ND 1.8 hot glass filter and the coating on the front looks just like a mirror, I could use it to check my makeup. To put it simply , Adam has confirmed the nature of the problem, we still do not have a solution. Shooting stage shows with a ND1 filter on the camera even if it cured the problem is not an option, period.