Filter for IR contamination

[Leonard Levy]

Piotr,
I don't know - I assume both. All I can tell you is what I have, technically the problem has been solved and they know there is a market out there.

[Max Allen]

Tiffen,

You should definitely release this product not only as a screw on but also as a 4x5.65 so I can purchase both! And please don't overprice it just because we really, really need it! Thanks!

Of course I believe it is Sony's responsibility to address this issue in the first place. It is a bad issue as bad as issues can get if you ask me. Accurate black reproduction is not a luxury to settle without but a fundamental necessity in any camera labeled "professional".

[Gabriel Soares]

Some more tests with the prototype far red filter from Tiffen;

ProVideo Coalition.com: Stunning Good Looks by Art Adams | Cinematography

[Max Allen]

Hi, some thoughts after reading the article...

Originally Posted by from the article

It could be that the dyes in the DSC chart are more far red-neutral than those in the MacBeth chart.

I would rely on the DSC dyes over the Macbeth dyes any day of the week. Those Macbeth charts are prone to flare errors, reduced dynamic range and they're not tested for time limits on the chemical stability of the dyes.

This implies that Sony has built some sort of far red filtration into the ND, although why they weren’t able to take care of all of it is a bit of a mystery. It could be that they are emphasizing the camera’s ability to reproduce a wide variety of reds at the expense of ensuring that every black material seen by the camera remains black.

If they did do this why don't we see any far red reduction when the NDs are engaged. Red reproduction at the expense of blacks IMO is highly unlikely. We know before you get to any colors the blacks have to be locked in place. Although perhaps not in the case of far red, as a matter of practice bad blacks will affect every color out of the gate so you would think this was noticed when the blacks were being calibrated. Not having noticed this when testing the camera is very befuddling to me. It is camera engineering 101.

Also I just wanted to mention that the kill of saturation in the colors is too much and I think simply not acceptable. Not to knock Tiffen at all, at least they have stepped up to address this. But to say we should have a fully working camera here, (Sony), and shouldn't let the pain of unsolved far red contamination now make us accept a partial solution or a solution that in turn creates another problem. Now we have to gain the colors in camera or post which may create noise. Just my 2 1/2 cents.

If Tiffen succeeds at producing this filter without having us give up saturation or something else I suggest that Sony pick up the tab for Tiffen and send a filter to every EX1 owner free of charge.

[Leonard Levy]

Max,
I don't understand what you mean by "the kill in the saturation of colors". My observation was that the new filter left the colors just like before after a minor re white balance.

[Brian Cassar]

After reading the article and seeing the with and without Tiffen filter photos, I feel that with the filter all the colours have a reduction in their punch - it seems that the contrast is lost. I'm not sure whether I'm using the right choice of words - but although I hate the IR contamination, I do not wish to lose any punch from the camera. The 486, although producing the undesirable green tint, leaves the contrast and colour punch the same.

[Simon Wyndham]

As the article points out, the reduction in colour could be due to those colours having that element of red that the filter is removing in them.

The EX paint functions are extensive enough that you can very easily dial in a compensation saturation or matrix setting if it bothers anyone. Putting a filter in front of the camera is always going to have trade off's. In this case I'd rather have proper blacks and a slight reduction in saturation that is easily compensated for than maroon blacks that are nigh on impossible to correct in post without a lot of secondary CC and masking etc.

[Ron Wilk]

I've noticed that several posters have added the "slim" qualifier to their suggestion for use of the B&W 486 IR blocker on the EX3. Is there a physical or vignetting issue with the standard version when used by itself (no stacking)?

Thank you in advance.

[Leonard Levy]

Its very dangerous to judge too much from photos on a web site. These tests were done very fast and then they are digitized on a web site. To my eye working with Art I saw no difference of consequence. ( Its possible we didn't always even re- white because the difference was subtle.)

Go back to Art's first article about the Tiffen filter where he takes a shot of a DSC chart with and without the Tiffen prototype filter, (both white balanced). Then he puts them on a vectorscope in post and imports the 2 shots to photoshop where he lays them on top of each other. (I think that was the process) The points lined up perfectly indicating no color change at all.

It was an elegant test and convinced me.

[Serena Steuart]

In that article Art did say "I’d be curious to see which version of the chart you prefer: saturated reds, or desaturated reds". And for those seeing the problem as an unforgivable flaw: "It could be that (Sony)are emphasizing the camera’s ability to reproduce a wide variety of reds at the expense of ensuring that every black material seen by the camera remains black".

Generally the cameras do record black correctly, but there are some fabrics that cause problems. In a production it is possible to avoid using those fabrics, but not easy in less controlled work (e.g. weddings and interviews). My memory of the first (I think) posting about this problem, in which a talent's blonde hair had recorded with a red tinge, was alarming. The new Tiffen filter looks like the answer to such issues, but I'm beginning to think that Leonard has a point in suggesting this is not IR contamination. In fact Art has confirmation from Sony that the camera's hot filter cuts everything longer than 720nM and here we're talking of "visible far red" as being 780nM, so that alone says we're dealing with visible red "contamination".

[Leonard Levy]

Serena,

I think your basically right but without our own spectral measuring devices its hard to know exactly what's what, likewise a bit of fuzziness about what the limit of visible red is.

I would be quite surprised if Sony consciously chose to give us better red rendition at the expense of our black materials.

My completely wild guess is that until the Red and the Ex-1 started seeing this stuff nobody had much noticed it before. I was stunned to see the problem with my naked eye while looking through professional grade ND's . It would surprise me if the manufacturers knew it would be an issue but didn't care. However I think that people in general would notice black material in tungstun situations. Actually i want to test the old workhorse D-30 next week with Tiffen ND's (only because I have them) to see if the issue has been with us for longer than we realize. If anybody else can do that at home let us know.

On the other hand I'm also curious about digital still photographers - You'd think professional fashion photographers would scream about this problem. How have they escaped it. Are the chips that different?

[Serena Steuart]

Very likely the black fabric issue wasn't noticed in development because it wasn't included in tests; not surprising. In early reviews the different rendition was noticed but not considered significant and for most subjects there is no problem. Generally the visible bandwidth is taken to be 380 to 750 nM, perhaps extending to 780 nM. Still photographers refer to IR as being 700 to 1000 nM, so there is an area of possible overlap. Individual eyesight varies as well. However if Sony say that nothing longer than 720 nM gets through to the sensors it is reasonable to suggest that little or no IR is contaminating the image. I presume Tiffen have measured the spectral sensitivities of the cameras (easy to do when one has the lab gear) and have designed their new filters accordingly. I would have thought they would happily release that data, and normally they provide transmission spectral characteristics with every filter they sell. That you could see by eye the fabric problem tells me that it falls within the visible bandwidth. I would expect with the development of digital cinema Sony would be working to broaden the spectral sensitivity of cameras, so once we start thinking about the process I'm not surprised that previously unnoticed problems are discovered. If they had designed their hot filters to cut into visible red (say, start attenuating at 680 nM) we wouldn't have seen the non-black black issue. This may well be the reason your fashion photographers haven't seen it.

[Giroud Francois]

probably photographers are using lot of light , more than most videographers, because they are using flashes. flashes produces very few IR and probably not so much in the red.

[Ron Wilk]

Check out the Lecia M8 forum ... it is replete with IR issues that require the use of a Leica supplied IR blocking filter. So photographers, at least those using Leica M8 series of digital cameras, are experiencing the same issues with black, mostly synthetic, fibers.

[Serena Steuart]

Originally Posted by Ron Wilk

Check out the Lecia M8 forum ... it is replete with IR issues that require the use of a Leica supplied IR blocking filter. So photographers, at least those using Leica M8 series of digital cameras, are experiencing the same issues with black, mostly synthetic, fibers.

It appears that the M8 can be used for IR photography (the subject of several threads) so one would need to use IR control filters (pass or block) in normal work. But certainly they have encountered the same problem with fabrics.