Nd filters versus stopping down

[tonyowen]

A question of curiosity only.

Is there any advantage in using a ND filter rather than stopping down a lens, or vice versa?

For instance using a ND8 filter on a f5.6 lens. Does that change the lens to the equivalent of a f11 lens, or does it merely allow a longer exposure without changing the optical properties of that f5.6 lens.

I’m wondering, for example, of the depth of field of a f11 lens versus that of a f5.6 lens with ND8 filter attached. Assuming no structural differences in lens type etc.

Also, for the sake of this query I’m ignoring the extra glass element between the subject and the recording media; which could affect the quality of the recorded image.

Regards
Tony

[xkaes]

A ND filter does not change the DOF. It just cuts down the light getting to the film. So how much DOF you want needs to be determined first. At the same time, you need to determine what speed you want to shoot at. An ND will impact either of these -- or both.

Instead of opening the lens to compensate, you can use a slower shutter speed and preserve the DOF.

One problem is that different manufactures use different designations for their ND filters, for example ND4, 0.6ND, ND2X, etc. They do this just to muddy the waters.

[Bob Salomon]

A ND filter does not change the DOF. It just cuts down the light getting to the film. So how much DOF you want needs to be determined first. At the same time, you need to determine what speed you want to shoot at. An ND will impact either of these -- or both.

Instead of opening the lens to compensate, you can use a slower shutter speed and preserve the DOF.

One problem is that different manufactures use different designations for their ND filters, for example ND4, 0.6ND, ND2X, etc. They do this just to muddy the waters.

No they don’t. 0.3 is 1 stop of density. 0.6 is 2 stops, 0.9 is 3 stops.
Some companies mark their filters in density values.
Others mark them in the filter factor. 2x.

[xkaes]

No they don’t. 0.3 is 1 stop of density. 0.6 is 2 stops, 0.9 is 3 stops.
Some companies mark their filters in density values.
Others mark them in the filter factor. 2x.

Exactly my point -- it confuses many shutterbugs.

[consummate_fritterer]

If it confuses folks then they can simply carry conversion charts or mark each filter in whichever way they find non-confusing.

[Jim Jones]

Tony, using a ND filter doesn't change the optical performance of the lens. Instead, it permits (or demands) longer exposure time. Some photographers find this useful in photographing star trails, moving water, or eliminating pedestrians in urban landscapes. It may be better than stopping a lens down so far that diffraction limits image sharpness.

[xkaes]

If it confuses folks then they can simply carry conversion charts or mark each filter in whichever way they find non-confusing.

Exactly my point -- it confuses folks.

[Bob Salomon]

Exactly my point -- it confuses folks.
Then those folks should learn proper terminology for their hobby. Lots of books explain it in simple language. So do many brochures from filter manufacturers, not private label filter suppliers, except for Hoya. I was Product Manager and Sales Manager for B+W and then Heliopan and I still find Hoya brochures confusing!
And that is also after many years as a photographer and a graduate of the USAF photo school!

[Dan Fromm]

Apologies for presenting a small format example.

Tony, I used to shoot flowers, insects and such at magnifications from 1:6 to 0.88:1 with my Nikon FM2n. Kodachrome 25, 1/250 shutter speed, flash illumination, typical effective aperture f/22 or f/32. Live unconstrained fish in small aquariums, too, same settings. All of the exposure came from flash, none from ambient. That's what I wanted and that's what forced shooting at such small apertures.

The results were marginal for publication, couldn't be printed as large as I'd have liked because of loss of sharpness to diffraction.

I stopped doing all this when Kodachrome processing went away. The apertures needed to let me overpower ambient with ISO 100 film were far too small.

I still have Ektachrome in the freezer and Fuji is still selling 35mm color reversal film so I've been thinking about shooting flowers again. It may be time for me to get a 3- or 4-stop ND filter or two. I find the idea of such dense filters repugnant, but its the only way to overpower ambient with flash and shot at reasonably large effective apertures.

[Pere Casals]

Tony, using a ND filter doesn't change the optical performance of the lens. Instead, it permits (or demands) longer exposure time. Some photographers find this useful in photographing star trails, moving water, or eliminating pedestrians in urban landscapes. It may be better than stopping a lens down so far that diffraction limits image sharpness.

This is an important factor, I'd add next:

Here we have lens optical performance measured (Line pairs per mm) by practical (not lab) gear, depending on aperture: http://www.hevanet.com/cperez/testing.html

Here we have a table showing sharpness limitation depending on aperture, at f/64 no lens can deliver more than 25 Lime pairs per mm resolution: http://www.kenrockwell.com/tech/diffraction.htm


Another factor is fall off with short focals, for example a 65mm lens in 4x5 will have an insane amount of fall off in the corners at f/5.6 and using ND4, while it can be well reduced to around 2 stops at f/11 with no filter

[Jac@stafford.net]

I appreciate the scientific optics presented because I like science,
but do large format photographers really care about diffraction?

For those who do care, just how large do you print and what is
your anticipated viewing distance?

Many thanks.

[Pere Casals]

I appreciate the scientific optics presented because I like science,
but do large format photographers really care about diffraction?

For those who do care, just how large do you print and what is
your anticipated viewing distance?

Many thanks.

Well, 4x5 lenses are typically diffraction limited by f/22. If we stop beyond that we can save money in the lens, by f/32 diffraction limits resolution of a $3000 glass to what a 1960 Symmar does (coatings apart). If we stop to f/45 we are limited to 35 lp/mm and we are in the same league than Pre WW II gear...

Does it matter ? it depends... not for Sally Mann, she can take also a great image even with the bottom of a coke bottle.

But if we put big money in glass because it is 80 Lp/mm capable but we usually end stopping at f/32, then it matters.

(This is iMHO)

[Dan Fromm]

I appreciate the scientific optics presented because I like science,
but do large format photographers really care about diffraction?

For those who do care, just how large do you print and what is
your anticipated viewing distance?

Many thanks.

Jac, if you directed that question to me, 8x12 full frame prints from 35 mm KM slides (8x enlargement) were visibly soft at 18". 4-5x would have been fine but 8x was a bit over the limit.

Whether the softness detracted from the final prints is an interesting question that I can't answer because it depends on the viewer's expectations. I didn't like it at all.

[LabRat]

Use the ND if you want longer than normal exposures and/or shallow DOF...

Steve K

[xkaes]

Use the ND if you want longer than normal exposures and/or shallow DOF...

Steve K

YOU NAILED IT.

Just figure out how much ND you have or want.

[Bruce M. Herman]

One other factor to consider is color cast. Not all ND filters are truly neutral. The editor of On Landscape, an online landscape magazine published in Britain, looked at a handful and found significant color cast in some of them. As I recall, the resin filters were the worst, and one of the best was a glass ND filter from China. This may not be a big issue for B&W photographers, but for color film users is concerning.

[LabRat]

One other factor to consider is color cast. Not all ND filters are truly neutral. The editor of On Landscape, an online landscape magazine published in Britain, looked at a handful and found significant color cast in some of them. As I recall, the resin filters were the worst, and one of the best was a glass ND filter from China. This may not be a big issue for B&W photographers, but for color film users is concerning.

Very true!!!

It's like wearing sunglasses... We don't notice the color cast or shifts because our brains adapt, but what do you see on a lightbox or while printing/scanning???

Steve K

[Pfsor]

Tony, using a ND filter doesn't change the optical performance of the lens. Instead, it permits (or demands) longer exposure time. Some photographers find this useful in photographing star trails, ...

Interesting. I strongly suspect you're not speaking out of your experience. Could you kindly explain how a ND filter can be useful in photographing star trails?? Thanks in advance.

[Jim Jones]

Pfsor, star trail photographs often use exposures of many hours. Sky light can then wash out the faintest stars. Either stopping down or using a filter compensates for this. As I recall, a 11 hour polar area star trail on 35mm Kodak Tech Pan film in the Midwest worked best between f/5.6 and f/8 without a filter. I wouldn't have wanted to stop down smaller than f/8 due to diffraction.

[Pfsor]

Pfsor, star trail photographs often use exposures of many hours. Sky light can then wash out the faintest stars. Either stopping down or using a filter compensates for this. As I recall, a 11 hour polar area star trail on 35mm Kodak Tech Pan film in the Midwest worked best between f/5.6 and f/8 without a filter. I wouldn't have wanted to stop down smaller than f/8 due to diffraction.

Interesting. You realise that reducing the background light with a ND filter reduces also the star light so their mutual ratio doesn't change, does it?

[xkaes]

Let's go far afield -- from the original simple question -- as we often do!!

[LabRat]

Interesting. You realise that reducing the background light with a ND filter reduces also the star light so their mutual ratio doesn't change, does it?

But stars are very bright point sources, it's that pesky atmosphere and light pollution that's the problem...

Where's contrast when 'ya need it!?!!!

Steve K

[xkaes]

But stars are very bright point sources, it's that pesky atmosphere and light pollution that's the problem...

Where's contrast when 'ya need it!?!!!

Steve K

Like I said, the original question had nothing to do with astro-photography.

[Drew Wiley]

It's already been hinted at, but not all are created equal and can potentially affect sharpness or create a bit of flare just like other cheap filters, esp if stacked. The also can mess with color film due to often having a bit of hue bias.

[Drew Wiley]

...I almost always encountered density labeling, which is logarithmic. Every .30 added up equals a full stop or a EV. Pretty darn simple, even for me. For example, .90 means I count three fingers, equating to either three f-stops I need to open up the lens, or three full steps of less speed on the shutter, unless long time exposures are needed.

[Jac@stafford.net]

You guys are just too much! I shot stars using a 70's Instamatic with a flash cube!
Yea sure, it burned my forehead, damn near blinded me, but with modern gear...

Some amateur said I had the camera turned backwards, but I saw stars!
I STILL see stars.

[Pfsor]

But stars are very bright point sources, it's that pesky atmosphere and light pollution that's the problem...
Where's contrast when 'ya need it!?!!!
Steve K

Really? So why are the stars drowned by the ambient light then? And how does the ND filter helps them if it doesn't change the mutual light ratio between the two??

[Huub]

For me there are two reason to use a ND filter. First is when i want to shoot (nearly) wide open and the fastest shutter speed of the lens i use is way to slow, for instance when using a 360mm tele-xenar for a portrait. That lens is in a synchro-compur 3 of which i don't trust the fastest speed, so i try to keep it over 1/60 sec. The second reason for using ND-filters is when taking pictures of slow moving rivers, lakes and ponds where i want to create a wash out effect in the water. This calls for speeds of 30 seconds and beyond, which will be hard to reach by stopping down in broad day light.

[tonyowen]

Thank you all for your views.

Therefore, my understanding is :-
1] Given the optimal (recommended/preferred) aperture for large format images seems to be f16 or f22
2] Given a lens with a large aperture but no diaphragm.
Then I could use ND filtration on the lens to give an effective f16/f22 setting and use the appropriate speed for that pseudo f16/f22 lens aperture.
[Again I'm ignoring the effects of the ND filtration on the image due to its manufacturing characteristics, and the difficulties of focusing through high ND filtration]
Right or wrong???????
Regards
Tony

[Pfsor]

Thank you all for your views.

Therefore, my understanding is :-

Then I could use ND filtration on the lens to give an effective f16/f22 setting and use the appropriate speed for that pseudo f16/f22 lens aperture.

Tony

You would not get an effective f16/f22 aperture setting. You would just reduce the light to the value of those aperture settings. Those are two different things.

[Doremus Scudder]

Thank you all for your views.

Therefore, my understanding is :-
1] Given the optimal (recommended/preferred) aperture for large format images seems to be f16 or f22
2] Given a lens with a large aperture but no diaphragm.
Then I could use ND filtration on the lens to give an effective f16/f22 setting and use the appropriate speed for that pseudo f16/f22 lens aperture.
[Again I'm ignoring the effects of the ND filtration on the image due to its manufacturing characteristics, and the difficulties of focusing through high ND filtration]
Right or wrong???????
Regards
Tony

Tony,

Stopping down the lens makes the hole the light is passing through on the way to the film smaller. This has two important optical effects besides reducing the amount of light getting to the film: 1. the smaller the hole, the greater the depth-of-field. 2. the smaller the hole, the more (image-degrading) diffraction there is. Further to this latter: there is a "sweet spot" in most lens designs where stopping down eliminates lens-design flaws and aberrations, but where the image-degrading effects of diffraction are not yet enough to really reduce image quality. Any larger aperture and there would be more degradation due to aberrations; any smaller and diffraction would degrade the image more. This point is different for different formats; for large format it is approximately f/22.

Neutral-density filters reduce the amount of light getting to the film by absorbing some of it on the way through the filter. This is independent of the effects of stopping down. Using an ND filter to get the same exposure at f/5.6 as you would get by stopping down to f/22 retains all the optical characteristics of an f/5.6 aperture, i.e., shallower depth-of-field and possible degradation due to lens aberrations (which would be corrected at f/22).

So, you can use an ND filter to get the right exposure with an aperture-less f/5.6 lens, but you'll have to live with the optical effects of the larger f/5.6 aperture.

BTW, you don't have to focus with the ND filter in place...

Hope this helps,

Doremus

[Pere Casals]

Well, also rather than using a ND filter we always can use a lower ISO film, even ISO 25, so we may get a technically better result from the lower ISO.

:) What I can say it is that my DIY dry plates won't usually require NDs

[tonyowen]

[QUOTE=Doremus Scudder;1429863]So, you can use an ND filter to get the right exposure with an aperture-less f/5.6 lens, but you'll have to live with the optical effects of the larger f/5.6 aperture./QUOTE]
Thank you, that answers my query without issue
regards
Tony

[ic-racer]

Thank you all for your views.

Therefore, my understanding is :-
1] Given the optimal (recommended/preferred) aperture for large format images seems to be f16 or f22
2] Given a lens with a large aperture but no diaphragm.
Then I could use ND filtration on the lens to give an effective f16/f22 setting and use the appropriate speed for that pseudo f16/f22 lens aperture.
[Again I'm ignoring the effects of the ND filtration on the image due to its manufacturing characteristics, and the difficulties of focusing through high ND filtration]
Right or wrong???????
Regards
Tony

Best to forget about the lens. It's aperture or effective aperture or shutter efficiency do not change at all with addition or subtraction of ND to the light path!
I teach that ND affects the ISO or Exposure Index of the film. It is as if you have a dense ISO trimmer coating on the emulsion. Nothing to do with the lens or shutter.

[andrewch59]

http://www.bwvision.com/complete-guide-long-exposure-photography-2016-edition/

Try this link, was very useful when I started dabbling in long exposures using nd filters, and some great examples.

[consummate_fritterer]

STOPPING DOWN...
1. Increases depth-of-field
2. Allows for a slower shutter speed
3. Decreases lens defects (chromatic aberrations)
4. Increases diffraction effects

ND FILTER...
1. Can decrease Depth-Of-Field (wider aperture)
2. Can allow for a slower shutter speed
3. Can be too dense demanding too wide an aperture (lens defects more pronounced)
4. Can allow for a wider aperture (less lens diffraction)
*** This is essentially similar to using a slower speed film!! Given the choice, I'd rather just use slower film but this isn't always practicable especially when the speed decrease desired is very substantial.

COMBINING BOTH (along with film speed choice)...
1. Allows finding the best combination for the lens, film and final image look

[William Whitaker]

Another consideration regarding ND filters germane to this forum and the netizens who frequent here is the type of lens you're using.

Some Portrait/Pictorial lenses, e.g., the Wollensak Verito (as a prime example), change the character of their image as they're stopped down due to the restriction of marginal rays which show a great deal of spherical aberration. So if you're working in a situation where the lighting may not be within your direct control, such as outside with a model, an ND filter will allow you to shoot with the lens aperture set to your desired level of diffusion, yet providing a shutter speed within the range of whatever shutter you're using.
In fact, speaking from experience, a robust ND filter large enough to cover the lens aperture is a necessary accessory for a Verito. Of course, there are any number of other portrait lenses which are diffuse wide open, then sharpen significantly as they're stopped down. It's just that most of my experience is with the Verito.

[consummate_fritterer]

Good point, Will, and this is why most (or many) Imagon lenses were supplied with a ND filter.

[Jody_S]

But stars are very bright point sources, it's that pesky atmosphere and light pollution that's the problem...

Where's contrast when 'ya need it!?!!!

Steve K

So use a film with bad reciprocity failure characteristics. That will significantly darken the landscape and light pollution.

[Jody_S]

I find I'm now using ND and grad filters for maybe 75% of my outdoor shots, sometimes more if I'm using barrel lenses with a Packard or Galli shutter. I also find the designations confusing, not to mention a 0.9ND might not be exactly 3 stops. I've solved this rather easily by spot metering through the filter.

[Jody_S]

One other factor to consider is color cast. Not all ND filters are truly neutral. The editor of On Landscape, an online landscape magazine published in Britain, looked at a handful and found significant color cast in some of them. As I recall, the resin filters were the worst, and one of the best was a glass ND filter from China. This may not be a big issue for B&W photographers, but for color film users is concerning.

Are the results of his test available online?

[BrianShaw]

I find I'm now using ND and grad filters for maybe 75% of my outdoor shots, sometimes more if I'm using barrel lenses with a Packard or Galli shutter. I also find the designations confusing, not to mention a 0.9ND might not be exactly 3 stops. I've solved this rather easily by spot metering through the filter.

I’ve read a thousand times that metering through a filter is wrong and inaccurate but I’ve always found it to be close enough. My SLRs do that successfully and so do my external meters... or at least it’s not so far off as to cause noticeable errors.

[LabRat]

So use a film with bad reciprocity failure characteristics. That will significantly darken the landscape and light pollution.

I've used that, and it works, and would work with a ND to enhance LIRF effects...

Even better for very long B/W exposures is to use a #29 red, as there is a good amount of red in starlight that passes, but holds back bluish light well for a dark background (but light pollution from red/orange streetlights and moon still enters in, but somewhat reduced)...

Steve K

[LabRat]

I’ve read a thousand times that metering through a filter is wrong and inaccurate but I’ve always found it to be close enough. My SLRs do that successfully and so do my external meters... or at least it’s not so far off as to cause noticeable errors.

Yea, I tested ND's to work fine metering through, though there's a point where LIRF kicks in, even in daylight with very heavy ND's, so don't forget to use the filter factor when calculating exposure for LIRF (but reading the base exposure through the ND is OK)...

Steve K

[Oren Grad]

I've deleted a bunch of griping that descended into an ad hominem food fight.

Please make your substantive point, no matter how critical, without framing it as personal attack. If in the heat of the moment you're feeling too frustrated to do that, please take a break and simmer down until you can.

[Drew Wiley]

Meters are not color neutral, so metering colored filters through them can be misleading, not to mention the spectral unevenness of pan films. ND filters are supposed to be color-neutral, as their name implies, but do often have a minor bias. I've measured them using an instrument at least a hundred times more accurate than either a handheld or TTL light meter, and the quality ones are typically within .05 density - a sixth of a stop.

[Jody_S]

... the quality ones are typically within .05 density - a sixth of a stop.

My cheapo plastic Chinese ones are off by way more than that. In fact, reading through the filter is the only way to get acceptable exposures with them.

I don't know if they have a color cast because I've never used them with color film, but I doubt it's enough to affect metering (@ a dozen readings in different areas to calculate exposure) or color shots that are to be scanned for processing (beyond what can easily be corrected in Photoshop). I expect the processing of color film introduces more color variation than the filters, especially since there are no more labs batch processing for commercial photographers. I doubt a Jobo with small chem kits of dubious vintage (no matter if they were just purchased, at least some of the chems could be years old) is anywhere near as precise as the old pro labs were (I know, we all have horror stories of poorly processed E6 somewhere, once upon a time, which is why we also had our favorite labs that didn't screw up our film).

[Two23]

Funny this topic should come up. In the past I've rarely used ND filters. When I do, it's usually to slow the shutter speed way down (like to 1/2s) when using ancient lenses that have no shutter and I have to use my hand. However, the past several weeks I've been photo'ing a genre new to me--ice racing! It's blinding bright out there on the ice. To minimize DoF so I can make my subjects pop out from a very distracting background, I need to stop down wide open. However, even with f2.8 Nikon lens that's not nearly enough! And often I can't stop down that far because even with ISO 100 I need a shutter speed faster than 1/8000s and my Nikon camera just doesn't have it. So, I bought a Hoya 3-stop ND filter to tame things down. With that I can slow shutter down just enough (1/300s) to allow me to pan a little. So while not large format shooting here, I did explain why stopping down won't work as well as using an ND I think.

Below shot: ice racing on Brant Lake, SD.

Kent in SD

[Doremus Scudder]

Funny this topic should come up. In the past I've rarely used ND filters. When I do, it's usually to slow the shutter speed way down (like to 1/2s) when using ancient lenses that have no shutter and I have to use my hand. However, the past several weeks I've been photo'ing a genre new to me--ice racing! It's blinding bright out there on the ice. To minimize DoF so I can make my subjects pop out from a very distracting background, I need to stop down wide open. However, even with f2.8 Nikon lens that's not nearly enough! And often I can't stop down that far because even with ISO 100 I need a shutter speed faster than 1/8000s and my Nikon camera just doesn't have it. So, I bought a Hoya 3-stop ND filter to tame things down. With that I can slow shutter down just enough (1/300s) to allow me to pan a little. So while not large format shooting here, I did explain why stopping down won't work as well as using an ND I think.

Below shot: ice racing on Brant Lake, SD.

Kent in SD

You do mean "opening up" to maximum aperture don't you?

[Pere Casals]

I'd like to add something about color shifts from ND filters.

Taking a look to datasheets we realize that quality filters have a really flat spectrum, ND filters only changes colors in case of bad filters in their denser version, in special with some variable density filters based in cross polarization.

...but there is another effect, long exposures on their own may change colors, color film have different reciprocity failure for each layer, we all know the need of M filters for Velvia 50 long exposures... but color shifts may also happen with long exposures and any color film.

...this is well known, just pointing that (IMHO) there is a myth around saying that strong ND filters changes colors when it's about different reciprocity failure of color layers.

This can be seen in a single image: say we expose for 4 seconds (with ND filter ), while "Zone-V" is neutral it may happen that shadows (that would require 32s, for example) would have a color cast. This cast disapears if we can open up the diafragm (with ND filter still in place) by 3 stops.

For BW film it also may happen that spectral sensitivity changes with exposure, but I'm not sure about that...