Low Contrast filters

[spasmoid]

I am getting a Lee Filter system for my wide-angle lens (woohoo!)

I am wondering if anyone has experiences/judgements to share about the Lee low-contrast filter or any of the Tiffen low-contrast filter. I am particularly intrigued by the Tiffen Ultra Contrast filter if anyone has any experience with that. It is very expensive ($267) compared to the Lee one ($107), plus I have to especially get 4mm guides for my Lee holder.

I would love to find out before-hand if it is worth it!

[Murray Foote]

To me the first question would be Why? - What do you hope for it to do that you can't do in Lightroom or Photoshop?

I can see a low contrast filter might have some advantage in extending dynamic range probably at some cost in resolution & ISO and maybe there might be some use for portraits. ... But I would also expect that you could do the same things in LR and PS with much greater control. Also, the dynamic range of a RAW file is quite large so that HDR is often unnecessary unless for special effects.

I find it difficult to imagine why you'd want a high contrast filter when increasing contrast is so easy in LR and PS and as for the low contrast option, you also have the alternative of local adjustments.

Regards,
Murray

[ATJ]

To me the first question would be Why? - What do you hope for it to do that you can't do in Lightroom or Photoshop?

Maybe he's going to shoot video...

[spasmoid]

To me the first question would be Why? - What do you hope for it to do that you can't do in Lightroom or Photoshop?

I can see a low contrast filter might have some advantage in extending dynamic range probably at some cost in resolution & ISO and maybe there might be some use for portraits. ... But I would also expect that you could do the same things in LR and PS with much greater control. Also, the dynamic range of a RAW file is quite large so that HDR is often unnecessary unless for special effects.

I find it difficult to imagine why you'd want a high contrast filter when increasing contrast is so easy in LR and PS and as for the low contrast option, you also have the alternative of local adjustments.

Regards,
Murray

Hi Murry, some good questions. Allow me to expand.

There are many ways indeed to produce low-contrast result in post-production with software - however, all post-production involves "mashing" data (I also mean data in a perceptual visual sense). There is always a departure from image integrity (scientifically speaking). In the case of low-contrast post-production enhancements, the associated amplification of the values in the shadows shows up the weaknesses of the capture device (without going into detail, I'm sure you know what I mean). The outcome is a less natural looking "effect" which can sometimes be just plain ugly and requires too much work to make compensations for.

Mashing the incoming light before capture-time means that we are not as exposed to the limitations of the capture device (electronic gain artifacts (noise in the shadows)), due to amplification in software, when trying to achieve the desired effect (low contrast in this case).

Going the other way is, as you pointed out, easy too (high contrast). This effect is the opposite of extracting more visual detail (and associated capture device limitations detail). It can be done safely without amplifying ugly artifacts. Increasing contrast is in fact throwing out visual data (or at least "squashing" it in the case of an S curve).

I imagine that a low-contrast filter will only have its uses in certain lighting conditions depending on the desired effect. I suspect, a lot of the time I will be clipping shadow to raise (restore) some of the black floor, but I will have better shadow detail without noise. Other times I won't have a requirement for much solid black to be in the image - particularly for some minimalism artwork.

As with all my equipment, I will use it for ANY kind of photography to make the image bend to my will. Right now, my needs are for landscape. I don't need to explain how out of range outdoors lighting can get for DSLR capture devices. Also, I will use it for outdoors portraits on bright sunny days. Often times the time of the day for the photo shoot is dictated to by my client's schedule. Needless to say, it will be good for weddings - allowing me to under-expose for the wedding dress without losing all shadow details (often the groom's suit/shoes) - sometimes the Bride's black hair. I already shoot with highlight priority turned on (forcing me to ISO 200).

As for the dynamic range of RAW, well of course everything is relative, and it is much better than 8bit/channel, however it's dynamic range is not big enough to accommodate across the board detail for every possible shooting situation here on earth - even if it was "heaps of" bits per channel. Relatively speaking, it could be said that 14bit/channel is woefully low-res.

ATJ, I absolutely plan on using it for video as well

As it turns out, I have decided to go with the Lee filter. I have no idea how it compares to the Tiffen. Unfortunately, I don't have any shops nearby (or in Queensland at all for that matter) where I can walk in and ask to test it out. You have to just buy it and spend some time with it first. It's not even easy to find any reviews online where they have hi-res samples to compare in a variety of test conditions. This is why I was hoping that someone here has already gone through all that and has some experience to share.

As a footnote, here is an example of me employing software to manipulate contrast. Its low contrast except for the bottom end where the black floor is lifted. All if it was achieved in lightroom.
http://terencekearns.com/before-after/carl_and_haley2.png

[surenj]

How about overexposing by +1 then pulling back the highlights. This will save your shadows.

As a footnote, here is an example of me employing software to manipulate contrast. Its low contrast except for the bottom end where the black floor is lifted. All if it was achieved in lightroom.
http://terencekearns.com/before-after/c ... haley2.png

This looks like the look from "davehill" actions I have seen for LR.

[Murray Foote]

I'm just thinking about responding to the earlier post but I'll make a comment on this one first.

It's not clear what overexposing by +1 stop means. There is no accurate way to see the dynamic range of the RAW image through the JGP preview and the histogram associated with it. There is certainly more highlight exposure available than you can see and the correct exposure uses up every whit of it and not an iota more. Exposing by one stop might simply be a correct exposure but there is no easy or accurate way to know.

Regards,
Murray

[spasmoid]

How about overexposing by +1 then pulling back the highlights. This will save your shadows.

Could do, but this is contingent upon dynamic range of the scene being small enough for the camera (file) to have the sensitivity/resolution headroom to pull it off. It's all hackage anyway, and all unnecessary if.... I use a low contrast filter
The principle is not unlike your sound engineer choosing to use a compressor to pre-process the audio before capturing it to the recording device (or file).

This looks like the look from "davehill" actions I have seen for LR.

I'll have to look it up sometime.
I went on a user-presets download binge one time, but i never use them. I always roll my own look, then save it as a preset - often only calling on that preset from the one particular photo session. Each session gets a dedicated sub-style.
Check out the rest of the sub-styles from the shoot.

[Murray Foote]

In the 80s I used to shoot live music on slide film, push it 2.5 stops and then print on Cibachrome. Not a lot of latitude in the slide film to start with and reduced by pushing it but still possible to get good results. Similarly I used to take landscapes on Velvia 50 on 5x4. Potentially greater image quality than a D3x with the right scanner and technique if you have the need to print huge. Dynamic range was undoubtedly a constraint in taking images but it didn't usually seem to be a problem at the time.

These days I have a D3 (though I may well also go back to 6x17) and the dynamic range is huge. So's the light sensitivity. Being forced to go to 200ISO is one thing; I was at an event last friday night where the light was so low I had to go to 9,000ISO and the lenses were searching for focus but still got good images. Different kind of images to yours, though and different requirement for shadow detail.

My camera preview is sRGB, my monitors are sRGB and aRGB and my printer is closer to aRGB than sRGB but a quite different shape to that. On the other hand, the RAW image captured by my camera is close to L*A*B - and the relevant dynamic ranges would be equivalent in each case. (For those reading who don't know what that means, the captured RAW image from the camera approximates what the human eye can see (actually it's constrained not to go too far beyond it) whereas everything else is much smaller. Likewise with the tonal range.)

Surenj's post is a good reminder that the elusive perfect exposure is a better aid to image quality than either post-processing (of a single image anyway) or use of a filter.

Terence, I don't accept your presumption that post-processing to expand dynamic range has to compromise shadow detail or produce an ugly effect however often that might happen with people who either don't know what they are doing or have "different" aesthetic sensibilities. There is significant leeway in a single RAW file and these days it's really quite easy to take separate exposures for the landscape background and for the couple and to combine them (and I'm not really thinking HDR). I think it's often possible to expand shadow detail in post-processing without compromising it whereas I suspect a filter inevitably compromises detail to some extent though it may or may not be apparent.

Your two images actually illustrate the leeway in the file. The way I see it, the attraction of the second image is that you can see their eyes. I like the idea of a differential combination of the two in Photoshop (not Lightroom in this case). I really like the sky, the sea to the left and maybe the rocks at back right in the first image. The sand and grass could come back a bit from the second to the first, then maybe a bit more colour in the couple while keeping the eyes as they are. Then balancing everything for coherence. I think it might be possible to get a dreamy effect for the couple with more ambiguity as to an overall realistic or ethereal effect. Of course, my imagination, not yours, but all presumably from a single image.

I'm not trying to say you shouldn't try out the filter. Everything's legitimate if it works and for that matter I'd argue that creative imagination is more important than technique (because the second will flow from the first but not vice-versa). I'd be interested to hear how much difference you find it makes in particular circumstances and as compared to the revealed potential of optimal post-processing.

Regards,
Murray

[spasmoid]

I hear what you say, but just because RAW's capacity to store dynamic range is similar to what the human eye can see, doesn't mean that it can retain the desired richness in detail in both the highlights and the shadows when the capture device is faced with a scene where the range goes to extremes. You need to remember that we survey a scene differently when we are present at the scene, than we do through a 2-dimensional photographic representation of that scene. In real life, we make adjustments with our irus depending on which way our head is turned. Other adjustments are also made on a chemical level (cones and rods) as well as psycho-visually in our brain.
A huge function of the 2D photographic representation is the interpretation to convey the interpreters ideas of the original scene. To this end, they may take an any means necessary approach to secure their very own (and subjective) interpretation.
Specifically in this case, I may choose to forgo some image replication integrity subtracted by a filter, in order to arraign details in the shadows along my camera and it's RAW format to wholly accommodate not only the shadow detail more comfortably, but also now the highlight detail due to the fact that it has been compressed before it enters my recorder. Furthermore, I have changed the profile/distribution of the detail - depending on how that particular low-contrast filter functions.
You must remember that the outcome of this change will take on different signifigance in different situations. I don't accept that one can intellectually theorize that there can't be a signifigant improvement in the quality of shadow detail redistributing the spread of the shadow detail through a filter. A lot will depend on the successful design of the filter, but there are so many factors, that the only viable way to get good information about how much it will help, is to try it out. Hence, asking for experiences of other people who may already have tried it out.
I don't have much money to experiment with, so I am going for the Lee system. The annoying thing is that I can't get my hands on the lee low-contrast filter until July. There are a lot of people using the low contrast filters for various reasons in film. I like the idea of not having to spend too much time munjing and image in post in order to do something that can be achieved through the trivial act of mounting a filter on to the lens.
Many experienced digital darkroom practitioner-photographers will vouch for the fact that it is too difficult to achieve the same quality effect in post, than it is starting with the filter.
I will still subject my work to post, but starting off with a data set which more resembles my desired interpretation means less munging in post - and less work. Munging in post is destroying data. I do it all the time, but the only improvement in detail is simulated. Besides, when you have a different starting point, you end point is likely to be different in this case.

Put simply, if you have a scene where the highlights aren't so high, and the shadows aren't so low, there would not be any point in using a contrast reducer. But if the shadow detail is really far apart from the desired highlight detail, then a low-contrast filter drags them within range of the RAW file - which is not infinite. The range of the human eye at any given iris size is also not infinite - so it is not useful as a benchmark for what needs to be recorded by the camera.

[spasmoid]

I hear what you say, but just because RAW's capacity to store dynamic range is similar to what the human eye can see, doesn't mean that it can retain the desired richness in detail in both the highlights and the shadows when the capture device is faced with a scene where the range goes to extremes. You need to remember that we survey a scene differently when we are present at the scene, than we do through a 2-dimensional photographic representation of that scene. In real life, we make adjustments with our irus depending on which way our head is turned. Other adjustments are also made on a chemical level (cones and rods) as well as psycho-visually in our brain.
A huge function of the 2D photographic representation is the interpretation to convey the interpreters ideas of the original scene. To this end, they may take an any means necessary approach to secure their very own (and subjective) interpretation.
Specifically in this case, I may choose to forgo some image replication integrity subtracted by a filter, in order to arraign details in the shadows along my camera and it's RAW format to wholly accommodate not only the shadow detail more comfortably, but also now the highlight detail due to the fact that it has been compressed before it enters my recorder. Furthermore, I have changed the profile/distribution of the detail - depending on how that particular low-contrast filter functions.
You must remember that the outcome of this change will take on different signifigance in different situations. I don't accept that one can intellectually theorize that there can't be a signifigant improvement in the quality of shadow detail redistributing the spread of the shadow detail through a filter. A lot will depend on the successful design of the filter, but there are so many factors, that the only viable way to get good information about how much it will help, is to try it out. Hence, asking for experiences of other people who may already have tried it out.
I don't have much money to experiment with, so I am going for the Lee system. The annoying thing is that I can't get my hands on the lee low-contrast filter until July. There are a lot of people using the low contrast filters for various reasons in film. I like the idea of not having to spend too much time munjing and image in post in order to do something that can be achieved through the trivial act of mounting a filter on to the lens.
Many experienced digital darkroom practitioner-photographers will vouch for the fact that it is too difficult to achieve the same quality effect in post, than it is starting with the filter.
I will still subject my work to post, but starting off with a data set which more resembles my desired interpretation means less munging in post - and less work. Munging in post is destroying data. I do it all the time, but the only improvement in detail is simulated. Besides, when you have a different starting point, you end point is likely to be different in this case.

Put simply, if you have a scene where the highlights aren't so high, and the shadows aren't so low, there would not be any point in using a contrast reducer. But if the shadow detail is really far apart from the desired highlight detail, then a low-contrast filter drags them within range of the RAW file - which is not infinite. The range of the human eye at any given iris size is also not infinite - so it is not useful as a benchmark for what needs to be recorded by the camera.

To use an example, if I could take a photo like the one here of a scene at sunset taken with a telephoto lens, and through some amazing new technology, I could get all the details in the shadow (which are currently silhouetted) and the detail of the surface of the sun - solar flares et al, then that, to me, would be way cool. It doesn't matter that the human eye is not capable of seeing that in real life, what matters is that I now have that detail at my disposal for generating my 2D representation via the post-production process - bringing everything (or at least what I chose) into the range of my output device. That's what matters, and that is the purpose of dynamic range compression pre-capture time in order to fit it into a recording device/format that is necessarily limited in dynamic range.

[surenj]

Very interesting conversation.

With a contrast reducing filter I think the RAW file would already be degraded even before any post applied. [As with ANY filter]

I agree that post degrades image data as well.

But in the end, one has to ask, does the degradation matter in (or can be seen) in the final print/presentation?

I would be keen for a comparison when you eventually get your hands on this filter.

Video is a different story.

[Matt. K]

A touch of refined fill-flash....wait for a cloudy day....some deft work in Photoshop (copy image/desaturate/invert/slight Gaussian blur and blending mode to soft light or overlay. Then reduce opacity to suit). These all work for me most of the time. If they fail then a blast of hot breath on the front lens element will turn any image into instant low contrast art. $260 dollars is more than my first Nikon SLR film camera cost. You would have to want it bad to pay that much.

[spasmoid]

A touch of refined fill-flash....wait for a cloudy day....some deft work in Photoshop (copy image/desaturate/invert/slight Gaussian blur and blending mode to soft light or overlay. Then reduce opacity to suit). These all work for me most of the time. If they fail then a blast of hot breath on the front lens element will turn any image into instant low contrast art. $260 dollars is more than my first Nikon SLR film camera cost. You would have to want it bad to pay that much.

 LOL@Matt

[Murray Foote]

I don't accept that one can intellectually theorize that there can't be a signifigant improvement in the quality of shadow detail redistributing the spread of the shadow detail through a filter.

I wasn't saying that if that's what you mean. I'm sceptical and not particularly tempted for myself in the absence of compelling empirical evidence but that's another matter.

Munging in post is destroying data.

However I don't think this is true at all (and munging in a post such as this one could well destroy nuances of meaning). As most people are well aware, the amount of retained detail declines exponentially from the highlight end to the shadow end of a RAW image. Only part of the RAW image is ever going to be presented in any realisation of the data. Therefore if you post-process to enhance the dynamic range by revealing more of an optimally exposed highlight area you are enhancing the presented data rather than destroying it. Likewise with combining exposures and potentially with revealing any different parts of the RAW data.

To use an example, if I could take a photo like the one here of a scene at sunset taken with a telephoto lens, and through some amazing new technology, I could get all the details in the shadow (which are currently silhouetted) and the detail of the surface of the sun - solar flares et al, then that, to me, would be way cool.

Well you could if you really wanted to with staid existing technology. It's a fairly static shot so with the right camera you could autobracket for shadow detail and flare on the sand. Then you'd need some combinations of probably radical neutral density filters to get detail in the sun, which wouldn't need to stay in the same place. The only minor problem would be combining all the exposures which would require great technical and aesthetic skill to have any chance of pulling it off. A little bit like what some people used to do with elaborate pin registration systems and multiple enlargers in the fume room.

[spasmoid]

I do appreciate all the input from you guys, esp Murray who does seem to put a lot of thought into his posts, however I really would like to know if anyone has any experience to share with low-contrast filters or if they know another stills photographer who does.
I have done plenty of multiple exposure combining with get more dynamic range captured - both for HDR, *and* for layers combinations by using layer masks. I've been there, but I don't even use a tripod when I shoot live events so assumptions about shooting conditions are not helpful.

If anyone here genuinely wants to learn anything about low-contrast filter usage, I found that Ken had some useful comments in his filters article.

Tiffen won an Oscar (or maybe it was an Emmy) for these filters. They figured out a way to make this filter so that it did not add a foggy look or reduce resolution while adding flare. They demonstrated this using a waveform monitor display for an image of two flat dark and bright areas. Fog filters would have let the bright area leak into the dark area, looking like fog. This looks like a slope on the waveform monitor. Tiffen's ultra-con filter instead simply lightens the dark area instead of looking like the light area is leaking into it. The waveform monitor was flat for each area showing the lack of fog. I realize waveform monitors aren't known outside of professional video.

Used properly it saves the need to make a double exposure for pre-exposure for contrast reduction and control. See Ansel Adams' book "The Negative" chapter on the zone system for this technique.

Used properly the Ultra Contrast (flare adding) filter will throw just enough light into shadows too dark to capture and give them contrast that otherwise would be a black blob. Again, ignore what you read on the internet and read from Ansel Adams.

This filter is much easier to use than making a double exposure through the special flashing device Ansel used. If you have a Mamiya 6 or 7 you need these filters if you normally use pre-exposure since the Mamiyas can't make double exposures.

If you really do your homework and like to push film then these flare filters actually can increase film speed!

[digitor]

... Tiffen's ultra-con filter .....

The Ultra-Con filter- a perfect description for both the low & high contrast versions!

Maybe with film, the low contrast filter has a similar effect to pre-fogging, which would increase the overall sensitivity somewhat (well not really, rather it just moves the whole exposure up the curve a bit) but with digital?

I'll be interested to see your results!

Cheers

[spasmoid]

Yeah, the idea (all these theories) really intrigues me. There is no substitute for field studies in context.

I have decided not to get the Lee low-contrast filter, but this means I will have to wait longer until I can afford the Tiffen. The good news is that my foundation kit Lee system comes with the bits I need to modify it to accommodate the 4mm think Tiffen filter.

If anyone can find some decent reviews on this equipment in the context of still, digital photography, I would be most grateful for a reference. Otherwise, weeks from now, I will have to write one myself.

[Murray Foote]

The problem with Ken Rockwell is that frequently he's an airhead and therefore you can't necessarily believe a thing he says.

I'd be interested to hear what you find. It's obviously designed for film, where contrast can be an issue, especially where there is scanning involved. Whether it has a useful effect in digital remains to be seen. I don't think a better looking JPG would count for much. If it has a benefit it would have to be if it were able to compress more of the image into the low-noise highlight end of the RAW image for a significant benefit net of filter loss.

My guess would be that even if there is a positive effect, it would be much less than a more expensive camera with a better dynamic range.

[spasmoid]

In terms of dynamic range capabilities of DSLR cameras, I found this website useful
http://www.dxomark.com/

I had made up my mind to switch to Nikon (due to Canon's continuing trend for pixel cramming) and then I read this article
http://www.dxomark.com/index.php/eng/Insights/More-pixels-offsets-noise!
which made me lighten up about it a whole lot.
No, this is not a trigger for a religious brand war...

IQ at the thin end of the wedge, which is what DSLR technology intrinsically is, will always be a complex issue exposed to variances in interpretations and people's personal experience and values.

As stated before, I am hoping for a munging of the scene intensities to squeeze more image data into my camera which has a finite dynamic range - on one frame grab. When you get more detail at one end, you can shift your exposure a bit in order to achieve whatever end - so what affects the shadows, affects how much highlights you expose for.

We'll see what happens in real life. It should be fun.

[Murray Foote]

When it talks about more pixels offset noise, that's only at the same pixel pitch, though, as it says. I think you'll find that increasingly as ISO increases, pixel size is much more important which is why the Nikon D3s has such high sensitivity and low noise.

I was thinking more of this site and for example ....

[spasmoid]

Yup, it would be nice to have a D3. Unfortunately, I'm doing photography for a living, so I can't afford such professional equipment

The good news is, I got my Lee system this morning. It comes with some extra 2mm guides and some longer screws. After the unwrapping ceremony, I am initially impressed with the build quality and design. So might be

I'll have a read of those articles. Thanks!

ps. I meant to say something about the Ken Rockwell thing before. I have never met the guy so I'm not given to any opinions about the man's intellectual capacity, but the information he presents in this instance is consistent with what I have read elsewhere and it makes sense. I'm not in the habit of throwing the baby out with the bath water. At the end of the day, I will be putting my money where his (and some others) mouth is and placing a bet based on perceived odds. I have to say, as it turns out (despite some brief worry about pixel cramming), I haven't done too badly in the betting stakes with regards to equipment purchases on my severely limited budget - and I am willing to bet AU$267 on a Tiffen Ultra Contrast filter
Hopefully my next significant purchase will be the EF 400L f/5.6