Exposing to the Right & PP

[ajohnw]

If you do buy a strong ND filter for very long exposures Shane try and get one that blocks Infra Red or also buy a UV/IR block filter to go with it.

The alternative is lots of shorter exposures loaded as layers and then averaged.

John
-

[ShaneS]

Thanks you John! I will use this advice when, probably not if , I begin to look at ND filters. Colin Southern was kind enough to educate me on averaging images in another thread and I have been playing with that as part of my research

[Polar01]

John can you state your reason or reasons for using a IR or a combo UV/IF filter combined with using a strong ND filter. I do believe that the IR blocking filter in front of the sensor is very effective. I have used 10 stop filters by both Lee and B+W, for a number of years, exposures up to an hour in length with no ill effects.

Cheers: Allan

[ajohnw]

Most cameras are sensitive to IR to a degree. This will vary from camera to camera. That's why IR pass filters can be purchased to fit on dslr's.. It seems that some ND filters do not block IR at all and the exposure times can get long enough for the sensor to pick up the IR that gets through. This will give colour casts.

This was mentioned in a thread somewhere at some point and I think some one posted a link to a site that demonstrated what was happening by taking shots with various light sources. A quick google bought this up

http://www.lifepixel.com/infrared-ph...ers-nd-filters

I'm not at all sure I would take the make sure they are glass and all will be ok as gospel. As for instance a general purpose optical glass such as BK7 lets through a fair amount of shorter wavelength IR - that's why IR block filters are multi coated. If I remember correctly - so don't quote me - plain glass that blocks IR has a greenish colour. BK7 is the sort of optical glass filters would be made of. I would also guess the dearer ND filters do block it.

John
-

[Polar01]

John so the reason to put a IR filter on the lens is to prevent any IR from reaching the sensor as the exposure time with a ND filter.
You state, "Most cameras are sensitive to IR to a degree. This will vary from camera to camera. That's why IR pass filters can be purchased to fit on dslr's.. It seems that some ND filters do not block IR at all and the exposure times can get long enough for the sensor to pick up the IR that gets through. This will give colour casts."
This seems to differ from the site you selected, "There are two types of infrared filter; infrared cut-off filters and infrared passing filters. Infrared cut-off filters prevent light in the near infrared range of the electromagnetic spectrum from reaching the sensor. Infrared passing filters allow this light to pass, but block light in the visible range of the electromagnetic spectrum from reaching the sensor.

For infrared photography, the type of filter you need is the ‘infrared passing filter’ so that near infrared light can reach the sensor, but most visible light is blocked." http://www.lifepixel.com/infrared-ph...nal-ir-filters

As to the use of ND filter your link seems to say that a ND filter will work in IR photography.

As for the colour caste when using large stop ND filters has more to do with how they are produced that IR light. My B+W ten stop has a warm caste to it, by Lee ten stopper has a cool caste, however if I take my 4-stop Lee add 2x3-stop Lee filter for a total of ten stops no colour caste, but I do get a little light leaking in because of the long exposure. The Lee big stopper has a rubber gasket around it to prevent light leaking in where as the stacked plates do not.
So I am still having a time getting my head around your statement, of an IR filter being used with a ND filter for long exposures.

This post has become somewhat off topic as the OP was "Exposing to the Right & PP" it moved to ND filters, however using a IR filter when using a ND filter on long exposures I feel is just wrong as one who does such images.

Cheers: Allan

[NorthernFocus]

Though I don't know that IR was the culprit, the first time I ever used an ND filter to take a long exposure sunrise shot, I ended up with colors that did not in any way resemble those that were visible. At the time I was quite perplexed in that the very nature of ND is to be "neutral" and reduce light passage without introducing color. I never repeated the exercise so didn't pursue it to resolution. Plus people who saw the photo thought it was great not knowing as I did that it didn't represent the actual scene. Until now it never occured to me that it may have been an issue of light outside the visual spectrum but perceivable to the camera sensor. I'd have preferred to remain blissfully ignorant....

[Mike Buckley]

I'd have preferred to remain blissfully ignorant....

Your only solution in all such situations is to go into vehement denial.

[Manfred M]

Carefully said; I will have to agree with John's view on this one. All filters have very specific response curves and are by no means "perfect" in terms of which wavelengths they cut off. This means that the camera manufacturer has to make a decision regarding the amount of IR and red wavelengths will be let through at the longer wavelengths side of the red part of the spectrum. The boundary between red and IR is 400nm.

If you look at the response curve from the Kenko / Hoya UV / IR cutoff filter (see the link below), you will see that some of the red (around 680 nm to 700nm) is cut off to some degree and some near IR (up to 720 nm) is transmitted. If you look at the other end of the visible light spectrum we see a similar issue at th UV end of the spectrum. While UV is defined as wavelengths that are shorter than 380 nm, and the filter shown does block these rays, but does not transmit the short violet end of the spectrum at 100% (top of the curve from 380 nm to 680 nm).

About 5% of the IR at frequencies between 700nm and 720nm gets through, after that, effectively none does. There is a cost, though, about half of the red light between 680 nm and 700 nm does not get through.

http://www.kenkotokinausa.com/hoya/p...lters/uvircut/

What does this mean to the photographer; adding an identical filter will remove a bit more IR (effectively doubling the amount of IR that does not get through), but the same thing happens to the red light.

This is where my knowledge stops, as I have no idea as to how the image is effected. In theory these effects could be reduced using camera firmware / processing.

[pnodrog]

You also need to take into account that unfiltered silicon sensors have their peak sensitivity at about 750 to 900nm so this makes it even more complicated...

PDF]Paper: Spectral Response of Silicon Image Sensors - Aphesa

www.aphesa.com/downloads/download2.php?id=1

It seems using a IR filter is prudent advice for any very long exposure and in particular when attenuating visible light with a ND filter.

P.S. I don't have one and have survived.

[Manfred M]

You also need to take into account that unfiltered silicon sensors have their peak sensitivity at about 750 to 900nm so this makes it even more complicated...

PDF]Paper: Spectral Response of Silicon Image Sensors - Aphesa

www.aphesa.com/downloads/download2.php?id=1

It seems using a IR filter is prudent advice for any very long exposure and in particular when attenuating visible light with a ND filter.

P.S. I don't have one and have survived.

I'm not sure if that's necessarily a correct assumption. The Hoya data suggests that there is virtually no IR transmission beyond 740nm. I would assume that the built in IR filter would likely have comparable performance. I could see this kind of filter being used on cameras that have been converted to IR / UV photography and someone wanted to use it as a "regular" camera again.

[pnodrog]

The sensitivity raises throughout the visual spectrum and at 700nm the sensor is about twice as sensitive as it is at 450nm. Of course the various filters including bayer and also RAW conversion will be designed to take into account most of the variations in response. Just pointing out it may not be as simple as just looking at filter performance but using a IR filter seems logical and may be of benefit.

I have a few ND filters and some graduated ones both ND and tobacco somewhere but I prefer to avoid filters as much as possible. It is nearly all multiple or differing exposures followed by PP since I stopped using film.

[ajohnw]

John so the reason to put a IR filter on the lens is to prevent any IR from reaching the sensor as the exposure time with a ND filter.
You state, "Most cameras are sensitive to IR to a degree. This will vary from camera to camera. That's why IR pass filters can be purchased to fit on dslr's.. It seems that some ND filters do not block IR at all and the exposure times can get long enough for the sensor to pick up the IR that gets through. This will give colour casts."
This seems to differ from the site you selected, "There are two types of infrared filter; infrared cut-off filters and infrared passing filters. Infrared cut-off filters prevent light in the near infrared range of the electromagnetic spectrum from reaching the sensor. Infrared passing filters allow this light to pass, but block light in the visible range of the electromagnetic spectrum from reaching the sensor.

For infrared photography, the type of filter you need is the ‘infrared passing filter’ so that near infrared light can reach the sensor, but most visible light is blocked." http://www.lifepixel.com/infrared-ph...nal-ir-filters

As to the use of ND filter your link seems to say that a ND filter will work in IR photography.

As for the colour caste when using large stop ND filters has more to do with how they are produced that IR light. My B+W ten stop has a warm caste to it, by Lee ten stopper has a cool caste, however if I take my 4-stop Lee add 2x3-stop Lee filter for a total of ten stops no colour caste, but I do get a little light leaking in because of the long exposure. The Lee big stopper has a rubber gasket around it to prevent light leaking in where as the stacked plates do not.
So I am still having a time getting my head around your statement, of an IR filter being used with a ND filter for long exposures.

This post has become somewhat off topic as the OP was "Exposing to the Right & PP" it moved to ND filters, however using a IR filter when using a ND filter on long exposures I feel is just wrong as one who does such images.

Cheers: Allan

Quote from the link

Most ND filters will also work in infrared light, just make sure it’s made of glass as plastic and resin filters tend to only work in visible light and pass IR light unaffected. Just take a picture of a person with sunglasses on and you will see that in IR those sunglasses become transparent.

BUT as I said I wouldn't be at all sure that the fact that a filter is glass will definitely achieve anything.

And yes as I mentioned IR pass filters can be bought and dslr's can take IR images when they are fitted with them but the exposure times tend to be long - as they are with extreme ND filters on.

John
-

[ajohnw]

Though I don't know that IR was the culprit, the first time I ever used an ND filter to take a long exposure sunrise shot, I ended up with colors that did not in any way resemble those that were visible. At the time I was quite perplexed in that the very nature of ND is to be "neutral" and reduce light passage without introducing color. I never repeated the exercise so didn't pursue it to resolution. Plus people who saw the photo thought it was great not knowing as I did that it didn't represent the actual scene. Until now it never occured to me that it may have been an issue of light outside the visual spectrum but perceivable to the camera sensor. I'd have preferred to remain blissfully ignorant....

I'm glad the post has been of use to some one - even if it's means more expense or I would rather not know. On the other hand I am often bemused by the way threads sometimes go and would have thought that most people were aware that cameras can take IR images by fitting them with an IR pass filter even with the cameras own filter in place. It just takes rather long exposures.

Actually using older lenses can cause similar problems. Not on here but I recently had a conversation with some one who wondered why his OM lens on an adapter gave incorrect colours. When fitted with a UV/IR block filter he found it was just as good as a Zeiss lens he had. This suggests that some manufacturers may put some of the filtering in the lens, might just be the glasses used and probably is. He contacted me because he had found a US ebayer who was selling B&W UV/IR filters at very reasonable prices - seller refurbished - I don't think much of their cleaning.

John
-

[george013]

Going back to ETTR.
Is there nobody who could react on post 32?
George

[Manfred M]

And yes as I mentioned IR pass filters can be bought and dslr's can take IR images when they are fitted with them but the exposure times tend to be long - as they are with extreme ND filters on

Sort of. This is only really practical if the internal IR / UV filter sitting overtop of the sensor has been removed, otherwise no, as the amount of IR getting through to the sensor is quite small and over a very narrow frequency range.

I had looked into the cost of doing this with my semi-retired D90, but my daughter has "borrowed" it so, that is no longer being considered by me. There are several companies in the US that seem to do this modification.

[Manfred M]

I'm glad the post has been of use to some one - even if it's means more expense or I would rather not know. On the other hand I am often bemused by the way threads sometimes go and would have thought that most people were aware that cameras can take IR images by fitting them with an IR pass filter even with the cameras own filter in place. It just takes rather long exposures.

Actually using older lenses can cause similar problems. Not on here but I recently had a conversation with some one who wondered why his OM lens on an adapter gave incorrect colours. When fitted with a UV/IR block filter he found it was just as good as a Zeiss lens he had. This suggests that some manufacturers may put some of the filtering in the lens, might just be the glasses used and probably is. He contacted me because he had found a US ebayer who was selling B&W UV/IR filters at very reasonable prices - seller refurbished - I don't think much of their cleaning.

John
-

John - I'm fairly certain that you are not totally correct here; but the amount of correctness, is where the debate lies.

1. Most digital cameras have a built in IR / UV filter as part of the stack that also includes the Bayer filter and the AA filter sitting in front of the lens. The only camera that I am aware of that did not have this is the Leica M8 (CCD sensor on that camera, so making a comparison to a CMOS sensor is going to complicate any comparisions). If you've ever looked at any of the Leica M cameras, the back of the lens is very close to the film / sensor plane. This camera did have a Bayer filter, but no AA filter or UV / IR cutoff filter.

These cameras definitely had issues and required an external UV / IR cutoff filter on the lens.

http://www.luminous-landscape.com/re...solution.shtml


2. I can only assume that the built-in IR / UV filters on other cameras have performace characteristics that are similar to the Hoya one I mentioned in a previous posting.

http://www.kenkotokinausa.com/hoya/p...lters/uvircut/

The performance at the UV and IR cutoff points is not a step-function, so there will be some filtration of visible light at the near and far ends of the visible spectrum as well as some transmission of either IR, UV or both, again depending on the performance characteristics of the specific filter that has been built into the camera. I have not seen any camera manufacturer provide performance characteristics to these internal elements, so their effectiveness is anyones's guess unless they have access to data that I don't.

3. The technique you suggest is dependent on the "poor performance" of the camera's built-in IR / UV blocking filter. If the performance chart from Hoya is any indication, you are looking a transmission of around 20% of the IR at 700 nm and that drops off to almost nothing at 720 nm. Yes, mounting a visible light cutoff filter (I can't find a performance curve of one right now) will do what you suggest, but, as you say you are restricting yourself to a very narrow part of the near-IR and some very long exposures. I have not played around with this and can't comment, based on any personal experience.

[Manfred M]

Going back to ETTR.
Is there nobody who could react on post 32?
George

Possibly because no one quite understood what you are getting at. I just re-read your post and that is still my reaction, so if you could please clarify, there might be something to discuss.

[ajohnw]

Going back to ETTR.
Is there nobody who could react on post 32?
George

He talks about 10 stops and then shows 7 stops which would probably best be described as EV's anyway. I haven't added the numbers up to see of they come to 2^10.

The list does illustrate one point about cameras. The sensors are approximately linear so there are less tonal steps available in each "EV" as things get darker. It doesn't matter how these EV are defined numerically this will still happen as each step up doubles the "count". We see in a logarithmic fashion which is one of the reasons our eyes have such a wide dynamic range and also can detect tonal variations within limits over a very wide range of light levels.

If some one wants to measure the real dynamic range of a camera they need to do what dpreview seems to do. They have a calibrated series of nd filters in a strip and take shots of it. Unfortunately they mostly post jpg results.. The scales they plot run from either side of zero and zero generally coincides with a count of something over 110 plus comments that Nikon have been known to go too much over - from memory and pass as to where so maybe best to say different. It fits in with Nikon over expose comments though.

I wonder how this fits in with so called mid grey / 18% or some similar figure which would put mid grey around a count of 46 not 110. Pass.

John
-

[bnnrcn]

Hi Shane Such a long thread that I don't think I will be able to follow all of it. I didn't know you started such a thread because I have never used 'New Posts' function before

Brightness of the sky wasn't the main issue in your image. The sky was full of clouds ( well also it was a bit bright) and that's why it was competing with the lava formations at the FG. IMO a plain sky would be better in that image to put the emphasis on the lava formations.

For getting the correct exposure ,

I would try using GND filters to get a healtier exposure because in sun-rise /sun-set shots dynamic range is always high, if you expose not to lose the details in shadows, you get a too bright sky, or if you expose for sky then the land is darker in the image and you might lose some details there.

I would also try at least 2-3 different exposures for the same scene in order to be able to choose the best one when I'm back home

Good point Grahame!

Mat, just seeing the sun pop over the horizon never fails to provide me with a moment or two of excitement and awe.

L.Paul...



I guess that is what I was looking for advice on...I wanted to learn how others would deal with a situation like that both in camera and in post.

John, Greg made the comment quoted below on the point in that thread and you are right in recalling that it was also mentioned in the first early morning thread as well.



I guess I feel like the brightness of the sky (and the unfortunate cloud formation) was the issue when discussing the crop. I believe that the brightness of the sky was enhanced by ETTR but when I bring the brightness down in post the sky then looks unnatural to me

[ajohnw]

John - I'm fairly certain that you are not totally correct here; but the amount of correctness, is where the debate lies.

1. Most digital cameras have a built in IR / UV filter as part of the stack that also includes the Bayer filter and the AA filter sitting in front of the lens. The only camera that I am aware of that did not have this is the Leica M8 (CCD sensor on that camera, so making a comparison to a CMOS sensor is going to complicate any comparisions). If you've ever looked at any of the Leica M cameras, the back of the lens is very close to the film / sensor plane. This camera did have a Bayer filter, but no AA filter or UV / IR cutoff filter.

These cameras definitely had issues and required an external UV / IR cutoff filter on the lens.

http://www.luminous-landscape.com/re...solution.shtml


2. I can only assume that the built-in IR / UV filters on other cameras have performace characteristics that are similar to the Hoya one I mentioned in a previous posting.

http://www.kenkotokinausa.com/hoya/p...lters/uvircut/

The performance at the UV and IR cutoff points is not a step-function, so there will be some filtration of visible light at the near and far ends of the visible spectrum as well as some transmission of either IR, UV or both, again depending on the performance characteristics of the specific filter that has been built into the camera. I have not seen any camera manufacturer provide performance characteristics to these internal elements, so their effectiveness is anyones's guess unless they have access to data that I don't.

3. The technique you suggest is dependent on the "poor performance" of the camera's built-in IR / UV blocking filter. If the performance chart from Hoya is any indication, you are looking a transmission of around 20% of the IR at 700 nm and that drops off to almost nothing at 720 nm. Yes, mounting a visible light cutoff filter (I can't find a performance curve of one right now) will do what you suggest, but, as you say you are restricting yourself to a very narrow part of the near-IR and some very long exposures. I have not played around with this and can't comment, based on any personal experience.

It varies from camera to camera Manfred. Infa red is generally said to run from 700nm to 1000nm. A typical IR remote control will put out 940nm. One way of seeing if any of the longer stuff gets through via live view. Some of the shorter stuff in practical terms is almost bound to get through. Some ND filters cost a lot more than others as well.

There are problems with the filters themselves but not with the "coloured" glass variety except those generally don't have a sharp cut off. Multi coated filters work by using a range of different thickness of coating to produced destructive or constructive interference. This has the penalty that the working thickness of the coatings vary according to the angel rays approach it at.so a rejection filter will only work perfectly when rays hit it square on. If you have a UV/IR block filter and have looked at it at an angle you would probably wonder about using it on a fish eye lens - they reflect visible red light rather strongly at extremes angles. The other aspect is that in real terms they are not as good as Hoya imply if real figures are available. Take a look at RazorEdge here. Mels Griot now seem to have a different name. Notice that the rejection is quoted in neutral density terms, in this case 6 and that there is a fair tolerance on the actual cut off wavelength.

http://marketplace.idexop.com/store/...13_final-e.pdf

If you want the technical bits that show how difficult it is to achieve this sort of performance they are here

http://marketplace.idexop.com/store/...lancingWEB.pdf

Really it's the same as any area associated with optics - making them realistically affordable and compromise. The exotic Mels bits will be extremely expensive and will still have the rays at an angle problem. Often the really good stuff doesn't pass 100%.

John
-