ND Filters & White Balance

[HoppyUK]

Re-read what they have to say. Its not about the white balance or fooling the camera at all. Its about how the camera deals with the varying sensitivity of the different colour pixels on the sensor. The RAW data isnt truly raw, its the cameras interpretation of how to amplify the seperate colour channels off of the sensor in order to produce a balanced image.

I have read that quote from Singh Ray before, but re-reading what they have to say, this is how I interpret it.

Originally Posted by Gandhi
This is from the Singh Ray website:

Q: I'm getting a color cast with my Neutral Density filter. What's the problem?
A: When using a Neutral Density filter (graduated or solid) you are generally causing your camera to make a longer exposure than normal, sometimes much longer. The software in some digital cameras uses response curves that adjust the RGB sensors sensitivity based on the exposure time, which may introduce more red into the image. Sort of a digital form of reciprocity failure.

This appears to be talking about auto white balance, and implies that one of the parameters used by manufacturers is exposure time, ie when exposure time is long, the camera assumes daylight is low (eg sunset) or artifical (tungsten light). If this was the case, then the camera would add blue as part of the AWB profile, not red which is the predominant shade induced with most ND filters.

A: In other instances, blocking visible light and a longer exposure allows infrared light to "build up" on the red sensors leading to a red or purple-ish cast. If your filter looks neutral but you see the cast in your images, try taking some test shots using the filter with a different brand of camera, or with a film camera.

Digital cameras have strong, sharp-cut IR and UV filters over the sensor, so there shouldn't be a problem. This explanation also implies that ND filters only affect the visible spectrum, and do not cut IR at all, which I think is unlikely, but possible. But really, if it was some kind of IR build up as described then the colour cast would be less with shorter exposures, and worse with longer ones. With an ND grad, it would be quite invisible. This is not the case in practise.

A: If you still see the cast, and you're within the warranty period, contact Singh-Ray to have your filter evaluated and possibly repaired/replaced.

If this is a camera characteristic, then replacing the filter will do nothing. Unless it is replaced with a filter from a more accurate batch.

I have not tested a lot of different ND filters, but the overwhelming experience of users is that they get a colour shift, and that this varies with different brands, and also between apparently identical filters from the same manufacturer. It just seems to me that a much more likely explanation is that many neutral density filters are in fact not neutral.

Not that it's a big deal as you can correct it in post processing, or (after a few experiments) it shouldn't be too difficult to adjust the white balance manually before shooting.

 

[buckas]

Q : is changing white balance "on set" worth it if you always shoot RAW? i.e. can change it afterwards

 

[pxl8]

A simple experiment should help solve the mystery.

Take the same shot with the same camera/lens/etc but with 2 different brands of ND filter. With the WB set the same for both if the saturation increases for different hues with different filters I would conclude it's the filter that is introducing the colour, not the camera using some kind of "auto detect the use of a ND" mode (which does seem unlikely).

 

[Byker28i]

I borrowed a hitech 8x ND filter and shot the sky on sunday afternoon about 3pm. Reapeated moments after with the cokin nd 8x. Cokin filter produced pink/purple skies, hitech didn't

 

[pxl8]

Don't suppose you took a 3rd shot without a filter did you?

 

[DiddyDave]

Don't suppose you took a 3rd shot without a filter did you?

What? Use a 'control'? The basis of all scientific experiments?

What's the point!



DD

 

[pxl8]

DD you are a bad bad man

 

[Byker28i]

Don't suppose you took a 3rd shot without a filter did you?

LOL - yes, but of course it was different shutter speed so I didn't mention it.

 

[pxl8]

So, you have 2 shots with different filters and a 3rd without.

The main question of the thread seems to be does (and if so why) an ND filter increase saturation.

Did you notice an increase in satuation with the hitec filter over the shot with no filter?

 

[Byker28i]

Didn't really look for that if I'm honest as I was comparing colour cast. The hitech gave me a blue sky, the cokin gave me a pink/purple sky.

I've got a little time tonight so I was planning on trying a custom white balance to see if this would help. Is it best to use a white piece of card or 18% gray (inside of a lowepro bag I think is that)

 

[pxl8]

For the purpose of the test any wb will do, as long as it's the same for all 3 shots it will reveal any cast from filters (compared to the unfiltered shot).

The theory that an ND filter increases saturation can also be tested - assuming the hitec filter is neutral then the image should (compared to unfiltered shot) show increased saturation.

If the filter isn't neutral then the perception of increased saturation could just be from the introduction colour by the filter.

The other possibility is that the ND is, for whatever reason, causing under-exposure which would also give the impression of increased saturation.

 

[Byker28i]

I tried to submit a review to Warehouse Express on this kit and emailed them asking why it hadn't been added, if they just added positive reviews. I just wrote that the kit produced magenta cast with digital cameras (as confirmed to me by their support person).

SO I was a little surprised to read an email to me today saying that:

"On receipt of your review, we tested the filter kit ourselves with some of our expert photographers as this is one of our most popular kits. We could find no fault with the kits and therefore decided that the review was not an accurate reflection of the product."


Which is really interesting because WE's support person blamed my digital camera (not a problem with film) and using them during sunrise/sunset when reds were more prevalent.
This was also kinda confirmed by Cokin as well, it's an issue with certain cameras/cmos sensors etc.

So I've emailed them back asking what their tests were so I can repeat in case I'm doing something wrong.

I'd be surprised though considering goggling 'cokin ND magenta' gives a huge number of hits and it's been discussed heavily on many forums.

 

[HoppyUK]

I am not aware of any reason why a truly neutral ND filter should produce a colour shift with digital of any sort, or how it could possibly increase saturation (provided exposure levels remain exactly the same).

If someone has any theories I'd really like to hear them. Futhermore, if we could get together a decent selection of ND filters sent to me, I will compare them all directly and post the results. I'll also point them at something bright, as you would normally do with many ND filter shots, and show how uncoated acrylic flares easily when provoked

But personally, there is nothing to prove: many NDs produce a colour cast because they are simply not neutral. If they are not neutral this could, in theory, increase/decrease saturation of some colours but my guess is that is much more likely due to differences in exposure; I don't suppose that an ND8 or whatever is any closer to being exactly 3 stops than it is to being dead neutral.

Furthermore, since the colour shifts are not huge and well within the scope of custom white balance or correction in post processing, I'm not sure I see much point. The ND filter I'm interesting in testing is the Hoya ND X400, which is nine stops, made of glass glass, and multicoated. The multicoating is I think more important than any slight shift in colour, if there is one, since the latter can be sorted quite easily but flare cannot.

If anybody has a Hoya X400 in any size, I'd love to try that, and post the results vs the B+W 10 stopper. (Sadly the X400 is special order only, takes over eight weeks at full RRP of £124 in 77mm. B&H stock them in the US, but shipping alone is $45 extra.)

 

[EdinburghGary]

I am not aware of any reason why a truly neutral ND filter should produce a colour shift with digital of any sort, or how it could possibly increase saturation (provided exposure levels remain exactly the same).

Not sure if I follow what you mean, but I can say for certainty using LEE filters, that two correctly exposed shots, one with an ND and therefore longer exposure, and one without, will be completely different in terms of saturation. Both are exposed fine, but one is rich and deep, the other flat and washed out. The longer the exposure, the more so the effect.

Gary.

 

[HoppyUK]

Not sure if I follow what you mean, but I can say for certainty using LEE filters, that two correctly exposed shots, one with an ND and therefore longer exposure, and one without, will be completely different in terms of saturation. Both are exposed fine, but one is rich and deep, the other flat and washed out. The longer the exposure, the more so the effect.

Gary.

I hear what you say Gary and have read your posts. But I cannot explain why that should be so. Unless you were shooting film, in which case reciprocity failure does all sorts of weired things - to exposure, colour, contrast, the lot.

 

[EdinburghGary]

I hear what you say Gary and have read your posts. But I cannot explain why that should be so. Unless you were shooting film, in which case reciprocity failure does all sorts of weired things - to exposure, colour, contrast, the lot.

I don't get it either. Perhaps it's because light is changing? Maybe 10 seconds if light for example, has so many different colours in it. when combined, it creates a much richer effect? I am just guessing, it makes no sense to me either.

The difference is astounding and enough to warrant working out why...

Gary.

 

[pxl8]

Not sure if I follow what you mean, but I can say for certainty using LEE filters, that two correctly exposed shots, one with an ND and therefore longer exposure, and one without, will be completely different in terms of saturation. Both are exposed fine, but one is rich and deep, the other flat and washed out. The longer the exposure, the more so the effect.

Gary.

Gary can you stick a couple of raw files online somewhere so I can look at the difference?

 

[EdinburghGary]

Gary can you stick a couple of raw files online somewhere so I can look at the difference?

Yes leave it with me, if I can't find the examples I will take more. Due to go out tomorrow night for some dusk shots.

Gary.

 

[HoppyUK]

I don't get it either. Perhaps it's because light is changing? Maybe 10 seconds if light for example, has so many different colours in it. when combined, it creates a much richer effect? I am just guessing, it makes no sense to me either.

The difference is astounding and enough to warrant working out why...

Gary.

Gary, was it film? If so, then that's almost certainly the answer.

If it was digital, in the absence of any sound theory which says saturation should be increased, I am forced to conlude that something else is happening. Even a slight exposure shift can make a big difference, depending where it falls on the sensitivity curve. And if the light pops out from behind a cloud between different shots, even if the exposure is correctly adjusted, the image will appear more constrasty and saturated.

 

[pxl8]

If you shoot some more tests include a shot of a light grey object, with and without the ND - that will make it easy to check for any colour cast from the filter.

 

[Gandhi]

I have read that quote from Singh Ray before, but re-reading what they have to say, this is how I interpret it.

This appears to be talking about auto white balance, and implies that one of the parameters used by manufacturers is exposure time, ie when exposure time is long, the camera assumes daylight is low (eg sunset) or artifical (tungsten light). If this was the case, then the camera would add blue as part of the AWB profile, not red which is the predominant shade induced with most ND filters.

Nope, it nevers mentions white balance. Or even suggests it. You are reading between lines that arent there! Different photosites have different sensitivities and responses over extended periods of time and the camera applies curves to these BEFORE they are dumped out as RAW data OR any white balncing applied. If a RAW file was truly raw and you saw what the sensor did, it certainly wouldnt make any sense due to the fact there are twice as many green photosites as there are red or blue! ALL CAMERAS APPLY PROCESSING TO RAW DATA BEFORE IT IS PRESENTED AS A RAW FILE!!!!!

Digital cameras have strong, sharp-cut IR and UV filters over the sensor, so there shouldn't be a problem. This explanation also implies that ND filters only affect the visible spectrum, and do not cut IR at all, which I think is unlikely, but possible. But really, if it was some kind of IR build up as described then the colour cast would be less with shorter exposures, and worse with longer ones. With an ND grad, it would be quite invisible. This is not the case in practise.

Its not about the colour cast on everyday short exposures, its about the shift and increased saturation you get with longer exposures, typically those created by using ND filters. You can have the most neutral filter in the world and it will still create a colour shift over long exposures.

If you think about how dense optically an R72 (infrared) filter is, and that lets through only IR, then it is blindingly obvious that a normal ND or ND grad will let ample IR through. Despite the IR filters installed on sensors, we can still do infrared photography, so they cant be cutting ALL the infrared!

If this is a camera characteristic, then replacing the filter will do nothing. Unless it is replaced with a filter from a more accurate batch.

Replacing a filter may help, especially if there has been degradation of the dye over time, or a fault with the dye process in the first place. They state this quite clearly and suggest that you test it empirically using another camera or some film. Otherwise you are screwed, as singh ray politely explain.

Case in point. I have gone from digital to film. I uses to get colour cast using my Lee filters on my D70, D200 and D300. I do not get those casts with any of the colour film I use now. Go figure. Must be the cameras.

 

[Chaz Photos]

I have got a 7dayshop ND8 so I set my camera up on tripod and took two photos out my kitchen window on with out and one with
Put both in to Photoshop on top of each other in layers
Colour sample the lamppost
I had a grey lamppost in shot so I took a reading of it
With out filter 243-242-233
With ND8 233-236-229
It has a much cooler look to it BUT I can not see a cast as such
This is for my filter from 7Dayshop

 

[HoppyUK]

Ghandi, I am only trying to make sense of a strange explanation from Singh Ray. Sadly, I cannot, and I don't believe it. Nothing you say here has changed my view that many ND filters are simply not neutral. The evidence is overwhelming, it explains everything, and I really don't see any reason to look further.

Nope, it nevers mentions white balance. Or even suggests it. You are reading between lines that arent there! Different photosites have different sensitivities and responses over extended periods of time and the camera applies curves to these BEFORE they are dumped out as RAW data OR any white balncing applied. If a RAW file was truly raw and you saw what the sensor did, it certainly wouldnt make any sense due to the fact there are twice as many green photosites as there are red or blue! ALL CAMERAS APPLY PROCESSING TO RAW DATA BEFORE IT IS PRESENTED AS A RAW FILE!!!!!

It might not say white balance, but that was my best guess at what was being described. You appear to be describing reciprocity failure, and digital does not suffer from this.

Its not about the colour cast on everyday short exposures, its about the shift and increased saturation you get with longer exposures, typically those created by using ND filters. You can have the most neutral filter in the world and it will still create a colour shift over long exposures.

If you think about how dense optically an R72 (infrared) filter is, and that lets through only IR, then it is blindingly obvious that a normal ND or ND grad will let ample IR through. Despite the IR filters installed on sensors, we can still do infrared photography, so they cant be cutting ALL the infrared!

I don't accept that long exposures produce a significant colour shift with digital (or even a noticeable one). That is a film effect, due to reciprocity failure.

I don't understand what you are saying about IR light. There is no IR within the visible spectrum, so of course a full IR-only filter will appear absolutely black to the naked eye. Digital cameras are notoriously useless at producing IR images, which is why Canon produced a special version of the 20D to do it, and if you want to shoot digital IR you need to have the IR filter removed from the sensor (which is easily done by specialist firms).

Replacing a filter may help, especially if there has been degradation of the dye over time, or a fault with the dye process in the first place. They state this quite clearly and suggest that you test it empirically using another camera or some film. Otherwise you are screwed, as singh ray politely explain.

Case in point. I have gone from digital to film. I uses to get colour cast using my Lee filters on my D70, D200 and D300. I do not get those casts with any of the colour film I use now. Go figure. Must be the cameras.

Frankly, the only reason I can see why replacing an ND filter might improve things (within the warranty period, as stated) is because the original was not neutral in the first place. I can well understand that there is batch variation with these.

I can't explain why you had problems with digital and not with film, but I am not convinced by your reasons either Film only produces more variables, not less.

 

[buckas]

I tried to submit a review to Warehouse Express on this kit and emailed them asking why it hadn't been added, if they just added positive reviews. I just wrote that the kit produced magenta cast with digital cameras (as confirmed to me by their support person).

SO I was a little surprised to read an email to me today saying that:

"On receipt of your review, we tested the filter kit ourselves with some of our expert photographers as this is one of our most popular kits. We could find no fault with the kits and therefore decided that the review was not an accurate reflection of the product."


Which is really interesting because WE's support person blamed my digital camera (not a problem with film) and using them during sunrise/sunset when reds were more prevalent.
This was also kinda confirmed by Cokin as well, it's an issue with certain cameras/cmos sensors etc.

So I've emailed them back asking what their tests were so I can repeat in case I'm doing something wrong.

I'd be surprised though considering goggling 'cokin ND magenta' gives a huge number of hits and it's been discussed heavily on many forums.

sounds like a fix to me - "expert" endorses it, turns out to be sub-standard, real factual opinions hushed up as don't want their reputation tarnished etc

only hear what they want to hear?

 

[Gandhi]

Ghandi, I am only trying to make sense of a strange explanation from Singh Ray. Sadly, I cannot, and I don't believe it. Nothing you say here has changed my view that many ND filters are simply not neutral. The evidence is overwhelming, it explains everything, and I really don't see any reason to look further.

If all filters were not neutral everybody would be experiencing colour casts all the time.

It might not say white balance, but that was my best guess at what was being described. You appear to be describing reciprocity failure, and digital does not suffer from this.

I know that digital doesnt suffer from reciprocity failure. I'm not going to explain it all again.

I don't accept that long exposures produce a significant colour shift with digital (or even a noticeable one). That is a film effect, due to reciprocity failure.

I think Gary, Mike, Woodsy, Myself and many other digital photographers would dis-agree with you there.

I don't understand what you are saying about IR light. There is no IR within the visible spectrum, so of course a full IR-only filter will appear absolutely black to the naked eye. Digital cameras are notoriously useless at producing IR images, which is why Canon produced a special version of the 20D to do it, and if you want to shoot digital IR you need to have the IR filter removed from the sensor (which is easily done by specialist firms).

I'm not an idiot, so please dont patronise me. If an R72 optically opaque filter cannot block infrared then a 3 stop ND wont either, is what i'm trying to say.

Frankly, the only reason I can see why replacing an ND filter might improve things (within the warranty period, as stated) is because the original was not neutral in the first place. I can well understand that there is batch variation with these.

Thats exactly what i said. But it may not fix the issue completely or at all!

:bang:

Right, I have explained it in as simple terms as I possibly can.

Dont make me break out the Physics!

 

[pxl8]

Dont make me break out the Physics!

I for one would love to know why longer exposures mean increased saturation on a digital sensor and if the effect is different with different sensors. If that means breaking out the physics then so be it

Note I said longer exposure as it would seem sensible to remove any doubts about a filter being neutral or not from the equation.

 

[EdinburghGary]

OK I think I may have a handle on it, and it may be as I said earlier.

http://singhray.blogspot.com/2008/03/terrific-triple-threat.html
"the 5-stop ND filter creates long exposure times which has two benefits; first it shows painterly blur in any object moving in the frame (swirling tides, running water, blowing clouds) and second it builds up colour intensity especially at sunrise and sunset by allowing colour to wash over the land and sky over time."


This is a perfect description as to what I experienced. So I am going to presume that a long exposure in times of sunrise / sunset, allow the colours to mix and multiply so to speak, creating a combination of lots of lovely warm colours. It would therefore seem sensible to assume it's NOT filter related, except with regards to using filters to decrease shutter speed?

I am going to do 8 tests.

1: 10 stop at sunrise / sunset
2: No filter at same sunrise / sunset

3: 10 stop in household scene - bulb lit
4: No filter...

5: 10 stop at night on TV - with lots of scene changes
6: No filter...

7: 10 stop outdoors during the day
8: No filter...

I expect that the only differences will exist on the sunset / sunrise shots, and the TV shots. I expect to see much deeper colours.

Gary.

 

[pxl8]

Gary, the idea that exposures of 30+ secs where the light in a scene can change/move by a visible amount makes a lot of sense and I expect the results would be very different from an exposure of 1/250s for example.

At the same time Ghandi seems to be suggesting there's some underlying physics with the sensor at play here which sounds very interesting too - maybe one sensor is better at saturating colours than another?

 

[EdBray]

I have two types of filter Kood and Cokin, the Cokin filters definately have a magenta cast compared to the Kood versions, this can be seen with the naked eye never mind on the sensor.

Straight jpeg out of camera, white balance corrected from paper before placing the filters. Cokins top, Koods bottom!

I realise this does not have anything to do with the long exposure change in colour just the difference between makes of filter.

 

[Woodsy]

I don't accept that long exposures produce a significant colour shift with digital (or even a noticeable one). That is a film effect, due to reciprocity failure.

It has been tried and tested more times than I care to think about. That is like us saying, can you prove that this is a film effect? Can you provide sound scientific reasoning as to why you see this effect on film? And don't just quote reciprocity failure, because, like you, we all want convincing?

I don't understand what you are saying about IR light. There is no IR within the visible spectrum, so of course a full IR-only filter will appear absolutely black to the naked eye. Digital cameras are notoriously useless at producing IR images, which is why Canon produced a special version of the 20D to do it, and if you want to shoot digital IR you need to have the IR filter removed from the sensor (which is easily done by specialist firms).

You see, this is where your understanding of how the EM spectrum, and indeed how it is detected breaks down. People will say, "yes, we can see 400-700nm of the EM spectrum, and therefore that constitutes the "visible" part of the spectrum". Well, frankly this is utter crap. It is entirely possible that one person might be able to see 375nm for example, which is, "technially" in the UV.

Regarding filters, please bare in mind that there is not such thing as a perfect filter, IE, there is nothing in optics which will have a perfectly flat pass band, and then a delta function-esq stop to the stop region. It should also be noted that every kind of optical filter will have an extinction ratio: a measure of how effective a filter is. Even the best linear polarisers available to optical physicists etc will have an extinction ratio of around 1:10^4. that means that for every 10^4 photons incident, one will not be affected by the filter, and these are polarisers that costs in the thousands each. I would not like to think of the ratio of a £50 filter made in mass production. So, the UV + IR filters on the sensor and the external filters we buy are never perfect.

I shot IR on my D50 perfectly well, and it required no IR filter to be removed from the sensor. Entirely plausible considering my above argument.

Your comment about there being no IR within the visible spectrum is also flawed, as the red and NIR parts of the spectrum can most certainly interfere.

It should also be understood that the efficienty of an optical filter, in this case an ND, falls off at the edges of the "visible" spectrum, and it is also entirely plausible, even with this point made, that the blue end is more highly attenuated than red, giving rise to a red cast.

Frankly, the only reason I can see why replacing an ND filter might improve things (within the warranty period, as stated) is because the original was not neutral in the first place. I can well understand that there is batch variation with these.

I can't explain why you had problems with digital and not with film, but I am not convinced by your reasons either Film only produces more variables, not less.

Now. Let's assume, for a moment, that there is such thing is a perfect optical-pass filter, ie one that attenuates the "visible" spectrum over a flat curve, and does not either side. If you then consider the percentages of red green and blue coming from a source that appears red to you (a vivid sunset for example). by adding said perfect filter, you are attenuating those percentages by the same amount. Now, if the incident light has been attenuated to the extent where the non predominant colours are below the noise level of the sensor (do not mistake this with image noise, as that is nearly always caused by heat from the active sensor, and amplification noise due to ISO (the former of which will certainly give a tiny boost to the red end - after all, IR is associated towards the red end)), then the red will appear more vivid in the resulting image. This then becomes true to any area of the image where a strong ratio of a single colour orginates from. Therefore it is totally plausible that a solid ND filter will act to make any photo more saturated.

Now, I hear you say, "ahh yes, but like you say, there is no such thing as a perfect filter". Indeed. However, the more expensive filters will have a flatter (though not ideally flat) attenuation curve. This supports the argument that ND filters will make colours more vivid. The difference in shape of the attuenation curve at the red and blue end (compared to the middle) support the existance of a colour cast in cheaper filters. And this is all possible at the same white balance setting and the same colour temperature.

 

[Snapper73]

It has been tried and tested more times than I care to think about.

Yes I've tried it and I get different colours when using ND filters on long exposures. From memory I think the Cokin one's gave a stronger one but also a horrible colour cast.

A very quick test shot

2 shots with camera set to Aperture Priority at F10 and auto white Balance

First shot no filters, second shot Hitech ND Grad for 30 sec's (thats what the camera thought the shutter should do to get the correct picture)

looks different to me

 

[pxl8]

by adding said perfect filter, you are attenuating those percentages by the same amount. Now, if the incident light has been attenuated to the extent where the non predominant colours are below the noise level of the sensor (do not mistake this with image noise, as that is nearly always caused by heat from the active sensor, and amplification noise due to ISO (the former of which will certainly give a tiny boost to the red end - after all, IR is associated towards the red end)), then the red will appear more vivid in the resulting image. This then becomes true to any area of the image where a strong ratio of a single colour orginates from. Therefore it is totally plausible that a solid ND filter will act to make any photo more saturated.

Sorry, I still don't get it. Your basically saying the ND makes the colour (light)darker and therefore the satuation increases because the light no longer exceeds some "noise level" of the sensor? No matter how you get there (ND, shutter speed, or aperture) if the colour stops exceeding this noise level shouldn't the saturation always increase? What is special about the filter that makes this happen but not in other cases?

I'm happy to believe but first I need to understand what's happening and at the moment it all sounds very much like magic

 

[Woodsy]

Not quite, I'm not saying the filter makes all the difference. I simply mean that it is the only aspect of all the methods by which you can alter the exposure that can make a difference to the light entering the sensor. Changing the aperture, of course, simply lessens the amount of light passing through the lens per unit time. Indeed, changing the exposure time simply determines for how long said light is allowed to enter the sensor for. An ND filter however does alter the ratios of the wavelengths passing through to the sensor compared to without it. It might be that greens and yellows do not suffer from this effect as much, and indeed are attenuated closer to the amount stated by the filter (1, 2, 3 stops etc). After all, it does seem that people often state that the reds and blues of an image appear much more vivid than before... does green and yellow get mentioned as much? I don't know.

In that last part, I was simply trying to get at the principle of, if the signals of the non dominant colours in the scene are near to the lower cut off point of the signal, ie, the amount of photons of the less dominant colours than is needed for the site to register their presence, then they wont show up.

Now, please don't take that last part "too" seriously, I was mearly trying to make a point about how the sensor works. What matters is that it is possible for a filter to alter the light passing through it (given by extinction ratio / attenuation curve), and give both a slightly higher saturation, and/or a colour cast.

 

[pxl8]

So is that a posh way of saying the filter changes the colour of the light hitting the lens?

 

[HoppyUK]

Ghandi, Woodsy, I have already said how an imperfect ND filter can not only induce a colour cast, but can also alter contrast/saturation in the same way that any coloured filter will do to respective parts of the spectrum.

I also agree that no filter is perfect and things overlap at the extremes particularly. Different cameras will no doubt react slightly differently to IR.

But I have not dwelt on these points because I don't believe them to be significant. I just believe that a simple explanation to the colour cast induced with an ND filter is that it is simply not nearly as neutral as it should be in the visible spectrum, let alone the IR.

Why should it be any more complicated than that? I am more interested in what is causing the problem here, and rather less what it might be if we had to explore the extremes of scientific possibility with a proven neutral filter and still got problems.

Right now I am of the view that if people had use of a reasonably neutral ND filter, the problems being reported would disappear. At the moment, that is a hypothetical situation. We need to distinguish between an ND filter problem - is it at least reasonable neutral or is it not? - and one created just by long exposures which is only incidental to the ND filter issue. We have not yet done these things.

Also, this particular aspect of the debate was brought about by the quote from Singh Ray where they say that once you've tried one of their filters on various cameras and concluded that it was indeed not neutral, then they will replace it with one that was. In other words, the explanation from Singh Ray is barely plausible techno babble which you might expect from a customer complaints department.

Please don't patronise me by threatening the heavy physics just to prove an esoteric point (which I already metioned ahead of you) when we have not yet examined the obvious solution. Why are people getting strong colour casts with exposures that are barely outside the normal range? A second or three? No need to invoke heavy IR theory here.

The other issue is the apparently increased colour saturation noted with ND filters, and very long exposures (tens of seconds). Taking into account everything else that has been said, I am inclined to believe that this is not an ND filter creation, but is down to exposure levels not being quite the same A-vs-B, and extreme duration allowing subject movement to affect the image.

(And Gary has not yet said if his earlier observations were based on film images, or digital, unless I've missed something.)

 

[HoppyUK]

So is that a posh way of saying the filter changes the colour of the light hitting the lens?

Yes.

Which is what some folks have been saying all along.

 

[Woodsy]

But I have not dwelt on these points because I don't believe them to be significant. I just believe that a simple explanation to the colour cast induced with an ND filter is that it is simply not nearly as neutral as it should be in the visible spectrum, let alone the IR.

But they are significant! And that is the whole point of the thread! The filters do make a difference, it is noticeable and makes a significant change to the images produced when the filters are used!

Fair enough, this might well be due to the fact that the filters are not perfect, in fact no where near perfect. But as a perfect filter is not available, or even one that comes close to being perfect, I can only think that you are arguing a moot point. The fact that we don't have near perfect ND filters will probably be a mystery for a long time, at least until someone comes up with a new material with a flat attenuation curve.

(And Gary has not yet said if his earlier observations were based on film images, or digital, unless I've missed something.)

He is discussing Digital I believe, as I was there when we both discussed it in Scotland. We were both shooting digital at the time iirc.

Appologies if it sounded threatening, that was by no means the intent

 

[Snapper73]

He is discussing Digital I believe, as I was there when we both discussed it in Scotland. We were both shooting digital at the time iirc.

Appologies if it sounded threatening, that was by no means the intent

By the way did you and Gary shoot any medium format stuff when up in Scotland

 

[AliB]

I like the physics and they make perfect sense to me. Used to use a lot of bandpass fluorescence in microscopy so I can just about follow your gist.

I did have an errant thought creep in too regarding the quantity as well as the quality of light. At an exposure of 1/250 you see the light that is available within that short time frame. Using long exposures the light has more of a chance to reflect and refract within a scene and perhaps by using long exposures this is where the "Colour washing" comes in especially if the blues and greens are being held back or not recorded but the reds are...............

Ponders...........

 

[HoppyUK]

But they are significant! And that is the whole point of the thread! The filters do make a difference, it is noticeable and makes a significant change to the images produced when the filters are used!

Fair enough, this might well be due to the fact that the filters are not perfect, in fact no where near perfect. But as a perfect filter is not available, or even one that comes close to being perfect, I can only think that you are arguing a moot point. The fact that we don't have near perfect ND filters will probably be a mystery for a long time, at least until someone comes up with a new material with a flat attenuation curve.

I am not arguing a moot point at all. That is you and Ghandi, prompted by the misleading pseudo-science put forward by Mr Ray Singh. Don't drag me into that one.

As far as I can see, the points raised in this thread concern colour casts introduced by imperfect ND filters. You have explained this, and I think that in the main we agree on that. Then there is the matter of changes to saturation/contrast which I think are most likely related to other things, not directly the filter. The filter has yet to be eliminated from the latter point, but when it is, I believe the matter will be cleared up.

Of course filters are not perfect. That is the root of the colour cast problem, which has been my position from the very start. I don't want to talk about IR at all as I don't think it has any bearing on the situation.

Appologies if it sounded threatening, that was by no means the intent

Edit: No worries at all I missed your edit in my reply above, but I think it was Ghandi who was threatening (in the nicest possible way) to bring out the big guns of phyisics. And he would be missing the point completely IMO.