Filters: Traditional or Digital?

[rpcrowe]

I still have a plethora of various types of filters for my lenses which I accumulated when I shot film. I keep these filters with the expectation that SOMEDAY I will play with them a little when I am totally bored and can't figure out anything else to do. I can't see selling them because it would be more trouble than it woud be worth and they really don't take up much storage room...

One thing I would like to do is experiment with the various soft-focus filters I accumulated in film days against post processing and against my 135mm f/2.8 soft focus lens. BTW: I did not use the soft-focus lens during film days. I bought it recently on a whim because I got it at such a good price (less than $USD 100 on eBay) and because, after I play with it, I can always sell it for the price I paid and perhaps even make a little profit.

I normally carry only two types of filters: CPL and protection. I don't like GND filters because they simply don't fit my style of shooting. I have honestly not taken the trouble to work with the GND filter enough to make it a viable tool. IMO, it is dificult (for me) to work with and I absolutely hate the results of an incorrectly used GND.

CPL filters are a tool I will often use in my photography of the Southwesten United States. The reduction or elimination of reflections allows saturation of the colors of the various rock formations and foliage. I even used the CPL effectively in the Hoh River Rain Forest on Washington State's Olympic Peninsula. The reduction of reflections from the saturated foliage helped many shots. BTW: the CPL will often darken the sky sufficiently that a GND is not needed. I don't do a LOT of wide angle shooting in landscape work so I have no problems with the uneven polarization which often occurs when combining UWA lenses and CPL filters. BTW: I never use CPL filters when shooting panos...

Since, the Desert Southwest photography is frequently subject to a lot of blowing dust and other grit; I will often protect my camera/lenses with a UV filter and a OPTECH Rainsleeve secured with a couple of rubber bands.

[Viana]

I am still hoping that those of you with filters will take a few photos with and without of the same subject, post the RAW/SOOC non-filtered photo so that I/we can work on it with digi filters. Then, you can put up your filtered photo and we can compare.

Anyone?
_____________

@ Richard

The soft focus thing is one of my favorites. One does not always have time to optimally adjust the camera, especially in tourist areas or doing street photography. Digi filters really work well on those.

[Colin Southern]

I am still hoping that those of you with filters will take a few photos with and without of the same subject, post the RAW/SOOC non-filtered photo so that I/we can work on it with digi filters.

I could do, but I can tell you now that you'd struggle, because I shoot a lot into the light, and you'd have highlights that were blown by a full 3 stops on the images shot without the GND, because I'll be exposing to keep the shadows away from the noise floor.

[Viana]

I could do, but I can tell you now that you'd struggle, because I shoot a lot into the light, and you'd have highlights that were blown by a full 3 stops on the images shot without the GND, because I'll be exposing to keep the shadows away from the noise floor.

Awww, common Colin, at least let me try. I like struggle and challenges. No pain, no gain!

Perhaps you will convince me that I need to use a traditional filter or two?

[nickjohnson]

Viana, At the risk of causing difficulty I think you're missing the point.
As Colin has pointed out, the appropriate use of GND physical filters allows the recording – in a single exposure – of a dynamic range that would otherwise not be possible.
This has nothing to do with digital or film. It is a simple matter of the recording device not having a dynamic range that can encompass all the light values in a scene.
Digital filtering in post production is by definition completely useless in this regard since it cannot operate on data that is not captured.
Your previously posted suggestion of combining bracketed exposures has limitations. In practice few large dynamic range subjects are totally static. So there are usually ghosting artefacts that render the results unsatisfactory. Further, combining multiple frames taken at different exposure values often gives unnatural tonal values – the process brakes the Munsel Colour Model.

[Viana]

Viana, At the risk of causing difficulty I think you're missing the point...

Er, no. I have gotten the point. I simply do not find it necessarily valid.
According to this CiC article:

“Graduated neutral density (GND) or "grad" filters are an essential tool for capturing scenes with a broad dynamic range. They're also perhaps a hidden secret of successful landscape photographers. While they have been used for over a hundred years, nowadays the grad ND effect can also be applied digitally, either during RAW development or in subsequent photo editing…

…The most versatile strength is perhaps the 2-stop variety; anything weaker is often too subtle, and anything stronger may appear unrealistic. In either case, it’s often not difficult to reproduce the results of a 1 or 3-stop GND filter in post-processing… https://www.cambridgeincolour.com/tu...ty-filters.htm

The article goes on to cover the plusses and minuses of both physical and digital filters. The implication is that digi-filters are more flexible and that the effect of physical filters are more difficult to fix in PSP if there is a screw-up such as with the gradient line. Then there is the issue of extra glass, smudges, scratches, etc.

That is precisely what I have been saying. I am not against using physical filters. Used to use them all the time when shooting film. These days, I find digital more flexible.

Again: Put up some photos and let’s adjust them and see if one is superior to the other or it’s all a wash.

[Colin Southern]

Again: Put up some photos and let’s adjust them and see if one is superior to the other or it’s all a wash.

I shoot something just for you later on in the week (ie when it stops raining). If nothing else, I'm looking forward to seeing how you recover the detail when all three channels are blown by around 3 stops over the portions of the image that my physical GND will be covering (and when I say blown, I mean blown blown - not "looks blown on in-camera JPEG conversion, but actually recoverable in RAW").

[Viana]

I shoot something just for you later on in the week (ie when it stops raining). If nothing else, I'm looking forward to seeing how you recover the detail when all three channels are blown by around 3 stops over the portions of the image that my physical GND will be covering (and when I say blown, I mean blown blown - not "looks blown on in-camera JPEG conversion, but actually recoverable in RAW").

Why not take several and not just one with all 3 channels blown? I also want to see how the GND filter does on such a photo, with the same settings of course.

You could also do some 3 shot bracketing and we can do some dynamic range recovery via HDR blending.

[nickjohnson]

Er, no. I have gotten the point. I simply do not find it necessarily valid.
According to this CiC article:

“Graduated neutral density (GND) or "grad" filters are an essential tool for capturing scenes with a broad dynamic range. They're also perhaps a hidden secret of successful landscape photographers. While they have been used for over a hundred years, nowadays the grad ND effect can also be applied digitally, either during RAW development or in subsequent photo editing…

…The most versatile strength is perhaps the 2-stop variety; anything weaker is often too subtle, and anything stronger may appear unrealistic. In either case, it’s often not difficult to reproduce the results of a 1 or 3-stop GND filter in post-processing… https://www.cambridgeincolour.com/tu...ty-filters.htm

The article goes on to cover the plusses and minuses of both physical and digital filters. The implication is that digi-filters are more flexible and that the effect of physical filters are more difficult to fix in PSP if there is a screw-up such as with the gradient line. Then there is the issue of extra glass, smudges, scratches, etc.

That is precisely what I have been saying. I am not against using physical filters. Used to use them all the time when shooting film. These days, I find digital more flexible.

Again: Put up some photos and let’s adjust them and see if one is superior to the other or it’s all a wash.

Well - I see that Colin has volunteered to feed you something to chew on. By the same token that you don't necessary agree with my point – I don't necessarily agree with the CiC tutorial you quote. Whilst your waiting for Colin you might like to see if you can do a digital filter that duplicates a 3 stop reverse GND – you may find your going to need it. By which I mean works on a raw file blown on all three channels by 3 stops.

[rpcrowe]

Viana...

You posted...

"The soft focus thing is one of my favorites. One does not always have time to optimally adjust the camera, especially in tourist areas or doing street photography. Digi filters really work well on those."

IMO, there is a difference between "soft-focus" and "out of focus". In the soft-focus effect, primarily the highlights seem to bleed into the shadows.

In the days of film using a soft-focus filter on the camera would achieve this while using a softening filter on the enlarger when making a print would have the shadows bleed into the highlights a totally different effect.

Photographers also used nylon pantyhose material or smeared petroleum jelly over UV filters (some poor souls tried it overthe lens - bad move) or painted UV filters with clear nail polish to achieve soft-focus look.

I am sure that these effects can be duplicated in Photoshop but, I have not seen it done quite effectively.

There are several filter add-on programs for Photoshop. I have experimented with the Tiffen filter program and I do not think it is a good substitute for a soft-focus filter. I have the NIK Color Efex program but, have not experimented with their filters in post processing...

[Viana]

Whilst your waiting for Colin you might like to see if you can do a digital filter that duplicates a 3 stop reverse GND – you may find your going to need it. By which I mean works on a raw file blown on all three channels by 3 stops.

As far as I know, GND filters are not very good at blending 3 tonal ranges. For that, you need a custom filter or a bracketed/blended, tone mapped digital HDR in which the bracketed photos of the most bright exposure should show the darkest parts of the image “noise free” while the darkest image should show no blown highlights, for a proper final blending.

[Colin Southern]

You could also do some 3 shot bracketing and we can do some dynamic range recovery via HDR blending.

Look, nobody is saying that you can't do a lot of these things digitally; as I said before, I use digital techniques myself - a lot. Heck - I think I use on average 2 or 3 Digital GNDs on just about every location portrait that I shoot - BUT - like every tool, they're just not perfect for every job. Let's take your example of combining multiple shots to simulate a ND filter - it works - it absolutely works - no question about it - BUT - if I'm going to pop up to the source of the Riwaka and capture some flowing water, it's far Far FAR easier to whack on my Vari-ND - spin it around to an 8-Stop attenuation and take a 30 secound exposure than it is to take 256 shots totalling around 6GB and merge them (even with the automation). Again, with landscape, it's far Far FAR easier to take a single shot with a physical GND to compress the dynamic range into a single exposure with good low-tone quality than it is to try and shoot multiple exposures to cover the dynamic range - combine them into an HDR composite - and then massage the data back to what it would have been in the first place (especially when the light is changing fast and I'm already shooting multi-minute exposures WITH significant cloud and/or water movement). Same goes for IR; yes you can simulate the result, but you can't simulate the atmospheric clarity you get in the IR region. With CP filters; yes, you can do a better job of skies digitally - but you can't remove glare from water to reveal something underwater that a CP would have picked up without any effort what so ever.

At this stage, I think I'm going to leave you to it with this thread -- it's starting to consume far too much time, and to be honest, I suspect that very little is going to convince you otherwise anyway. If digital manipulation works for you 100% of the time then great - it's obviously something that suits your shooting style. But keep in mind also, that others may well have shooting styles that are very different to yours; in my case I'm often shooting scenes that range from several minutes all the way up to around an hour - often in changing light conditions - and for these kinds of shoots it's just plain easier to use a physical filter.

[Viana]

Viana...IMO, there is a difference between "soft-focus" and "out of focus". In the soft-focus effect, primarily the highlights seem to bleed into the shadows...

...Photographers also used nylon pantyhose material or smeared petroleum jelly over UV filters (some poor souls tried it overthe lens - bad move) or painted UV filters with clear nail polish to achieve soft-focus look.

I am sure that these effects can be duplicated in Photoshop but, I have not seen it done quite effectively....

I love soft focus for portraits and some landscapes. I don’t think it’s out of focus, because the edges stay crisp.

Yes, I remember using the stocking over the lens. In fact, I have a couple of framed photos around here where I used that technique. As for petroleum jelly, hair-spray or any other gunkies— never did get into that for obvious reasons, like cleaning glass. You can also use a star filter to diffuse if you are careful about the light source. All those old movie star “dreamy” photos used soft focus filters. Today, I prefer to use a digi-soft focus, because of the usual—control. You can have a little or a lot or nothing at all. Can’t change the “soft” focus to more or none if you use a physical filter.

Photoshop is not always the best solution. There are other programs, specifically those for portraits, as better options. One thing that is important, IMHO is to not soften the eyes and often the lips—depending.

… If digital manipulation works for you 100% of the time then great - it's obviously something that suits your shooting style. But keep in mind also, that others may well have shooting styles that are very different to yours; in my case I'm often shooting scenes that range from several minutes all the way up to around an hour - often in changing light conditions - and for these kinds of shoots it's just plain easier to use a physical filter.

Now, Colin, where did I ever say or even imply that digital manipulation works 100% of the time? In my experience, virtually nothing works 100% of the time. That goes for a physical filter as well.

If you are standing at a scene for so long, then yes a physical filter might be a better solution, although things (vegetation, animals) do move. It’s not only the lighting changes. Then there is the issue of extra glass.

I am not about all or nothing. I prefer options. To me, and many other people, digital manipulation offers more choices “most” of the time. I just wanted someone to post some photos that could be digitally filtered for group comparison.

[Glenn NK]

Further to the "magic cloth" or MC:

Of course it's not magic, but it can be quite effective. What it can do is prevent overexposure on the parts of the sensor that would be receiving too much light.

For those that haven't followed it, the method is effective with too bright skies - sunsets over land - where the sky is overexposed and the ground is underexposed (over water, the reverse can be true).

The usage is simple, however without a ND filter, relatively long exposures are required; with short exposure times, the MC can't be moved quickly enough into the right locations. I've taken a few shots at sunset over the Pacific Ocean. At f/16 and ISO 160, many shots in AV mode were metered to 1/80th second. A MC simply isn't feasible n this situation. So it would seem that the MC method requires exposure times of several seconds or more. However, with a ND filter and low ISO, the times can be extended significantly, making the MC more practical.

There is only one issue for me - when shooting over water, I prefer to not blur the wave action so as to achieve the detail that the human eye sees.

So, one then thinks of digital filters. Well this works too, but it requires multiple images, and with fast moving surf, even rapidly fired sequential shots won't capture the action (of my images of water 1/80 second was about the slowest).

Finally, the MC has a unique advantage over the ND - a piece of cardboard can be cut to fit the profile of the horizon. I've seen soft grad ND filters used with a bright sky over dark land that was irregular - however part of the sky was overexposed because the graduated ND filter could not match the profile of the land.

Glenn

[McQ]

I found this thread a little late, but thought I'd try and chime in. Despite being a little heated at times, it really seems that everyone is largely in agreement, with perhaps a few exceptions. I think we all agree that, everything else being equal, digital filters are preferred over physical filters, and that there are still a handful of hold-out scenarios where physical filters are either unavoidable or impractical not to use. Ultimately a lot of this comes down to style and preference though. Viana: If you're seeking full resolution through a digital vs physical filter shoot-off, I suspect you'll be sorely disappointed because the ultimate winners will be personal and depend on workflow. Also, no need to escalate this; we try and keep things as friendly as possible around here, even when in disagreement. I've edited a few posts to try and maintain this goal.

Personally, I prefer to use digital filters whenever possible because these are usually much more flexible. In my mind though, the following advantages of physical filters are still insurmountable:

• GND's. Can handle moving subject matter seamlessly with continuous averaging (if needed). Also produces the end result (or something near that) in-camera, which can make it easier to know whether you've got the type of shot you're after.
• Polarizers. Although one can sometimes duplicate the effects of a physical polarizing filter in retrospect, it is impossible to fully simulate the effects a priori, since this relies on knowledge of the underlying physics of the subject matter.
• Color Filters. These are unavoidable for special scenarios where you need to filter out a monochromatic light source while still capturing all other image light with low noise (such as filtering out the effects of a sodium arc lamp street light in a night shot so that the scene's lighting consists primarily of the much dimmer moonlight).
• Soft Focus. Although one can theoretically simulate the soft filter effect digitally, this requires creating/knowing the depth map of a subject for each image, and is often impractical in a high-throughput portrait studio scenario.

I am still hoping that those of you with filters will take a few photos with and without of the same subject, post the RAW/SOOC non-filtered photo so that I/we can work on it with digi filters. Then, you can put up your filtered photo and we can compare.

I want to put something together like this for a tutorial. Perhaps the biggest obstacle here though is accruing a good set of sample images.

What I'd really like to see in the next generation or two of digital cameras is more intelligent in-camera multi-exposure modes. This might include in-camera HDR, GND (through single or dual exposures), averaging to simulate an ND filter, stacking to improve depth of field, f-stop variation to improve corner sharpness, etc. Of course doing these yourself will always give more control, but these modes could have an option to save all of the comprising exposures just in case you wanted to go back and change the effect. That would be neat. Maybe some day...

[Viana]

… it really seems that everyone is largely in agreement, with perhaps a few exceptions. I think we all agree that, everything else being equal, digital filters are preferred over physical filters, and that there are still a handful of hold-out scenarios where physical filters are either unavoidable or impractical not to use.

I certainly agree with that, Sean. I know a number of photographers who continue to insist that film is the only way to go, despite the fact that film is more and more difficult to get.

Nothing in this world works 100% of the time in the same manner for all things; entropy and all that in physics on the bio front cell replication begins to break down with the result being abnormal growth e.g. disease, death and rebirth in another form.

Ultimately a lot of this comes down to style and preference though. Viana: If you're seeking full resolution through a digital vs physical filter shoot-off, I suspect you'll be sorely disappointed because the ultimate winners will be personal and depend on workflow.

Not seeking full resolution at all, Sean. Different tools for different purposes and reasons. I am not into hitting everything that is sticking up with a hammer.

Personally, I prefer to use digital filters whenever possible because these are usually much more flexible. In my mind though, the following advantages of physical filters are still insurmountable:

I completely agree and have said so. I too love the flexibility of digi-filters and can’t wait for the next generation of “smart” cameras that will give us the ability of physical filters in the box.

Anyway, thanks for chiming in on the subject. When I have a moment I intend to explore that “magic cloth.”

[rpcrowe]

In camera HDR is already in some cameras while the in-camera GND is only in our imagination. That would, IMO, be a wonderful addition.. However, the camera manufacturers seem to think that video, articulating LCD screens, gazillions of pixels and umpteen thousand ISO levels are the secret to selling cameras. Perhaps they are right! But, there are a lot of ways that cameras could improve for still photography...

[Glenn NK]

However, the camera manufacturers seem to think that video, articulating LCD screens, gazillions of pixels and umpteen thousand ISO levels are the secret to selling cameras. Perhaps they are right! But, there are a lot of ways that cameras could improve for still photography...

Sadly about the "features" you are right, and even more sadly the "features" seem to be desired. The MP "race" was a bit on the "less well thought out" side. Perhaps Canon has learned a lesson - the announcement of the 1DX with 18 MP was a bit of a surprise. There are some complaining about this, but I'm starting to suspect that my 5DII with 21 MP wasn't a great leap forward over my 30D with 8.1 MP (5/8 x 5/8 x 21 = 8.2 MP - not much better than 8.1 MP).

As for the filter argument, IMO when one limits oneself to one approach, one simply limits oneself.

Glenn

[Colin Southern]

In camera HDR is already in some cameras while the in-camera GND is only in our imagination. That would, IMO, be a wonderful addition.. However, the camera manufacturers seem to think that video, articulating LCD screens, gazillions of pixels and umpteen thousand ISO levels are the secret to selling cameras. Perhaps they are right! But, there are a lot of ways that cameras could improve for still photography...

Hi Richard,

The tide might be turning there - Canon have actually reduced the MP count in a couple of cameras now (with one being the new flagship). The new super-high (native) ISO modes are a good thing though, because this pretty much equates to an increase in dynamic range (I wouldn't be surprised is the new Canon 1Dx tops out somewhere around 14 stops of DR) - which is pretty impressive considering that - unless we compress it during post-processing (which we do anyway, but often not a lot) then we only need around 4 stops for a print and 6 stops on a monitor.

[rpcrowe]

I am sure that you could accomplish this using Photoshop but, it does give the imprression of a soft focus filter...

http://prodesigntools.com/photoshop-...n-compare.html