ND filter comparison pics

[stubbsy]

Just a quick unscientific demo of the impact of a Neutral Density filter. The two shots below have had no PP. The lens used was the Nikkor 12-24 DX. They were both shot raw 15 seconds apart at 12:28 pm yesterday. The first shot is without the ND4 and the second with the filter. While the second shot is noticeably darker it has much less blown highlights around the sun and so has more recoverable detail when PP is done.

[spada]

Hi Peter
This is very good used of ND , it allow longer exposure to revael more detail of the normally dark scene. I would like to ask which number is the darkest one ( Ie ND4..., NDXX), thinking of getting one and Cir PL for use with the wide angle lens, does it need to be a thin one ?

Thank
Spada

[Alpha_7]

Thanks for Posting Peter, it was visible even on the D70's LCD what a difference it made, but you can really appreciate it look at it on the PC. Looking forward to seeing what this filter and your eye can present out in the field

[wendellt]

I like the moody yellowish tone that the ND filter produces, same can be achieved with using a really high shutterspeed with f11

Has it got any other applikcations other than stoping down brightness

Investigating getting a NDX32 filter so i can take long exposures during daylight and get trails from moving objects.

[stubbsy]

Spada

The bigger the number the darker. I've even played with an ND400 (very dark - took 45 seconds in daylight to get the image)

ND2 reduces by 1 stop, ND4 by two stops, ND8 by 3 stops etc (and ND400 reduces by 9 stops to less than 1/500th of the original light intensity)

I have a "standard" rather than a thin CP filter - the standard ones can cause excessive vignetting, but I've had no real issues on my 12-24 DX. Thin ones cost way more $$$. The ND (which I've now purchased) is also not a thin one.

[stubbsy]

I like the moody yellowish tone that the ND filter produces, same can be achieved with using a really high shutterspeed with f11

Has it got any other applikcations other than stoping down brightness

Investigating getting a NDX32 filter so i can take long exposures during daylight and get trails from moving objects.

Wendell

Photographing solar eclipses and ultra-bright light sources can be extremely dangerous. This filter (the ND400) reduces light values by 9 stops to less than 1/500th of its original intensity and allows safe photography.

It can also be used to achieve super slow shutter speeds in daylight to render moving subjects invisible, and will enable you to use slow shutter speeds... to record movement in subjects such as waterfalls, clouds, vehicles, etc.

It will also decrease depth of field by allowing wider apertures to be used, which helps separate subjects from their background. And this filter allows you to decrease ... effective ISO ... so it can be used outdoors in bright situations.

I think ND filters would suit your style of urban photography. Note an ND32 reduces by 5 stops.

[wendellt]

I am getting one, it sounds like it has lots of potential for experimentation.
you think you can stack ND filters to get variable range of stops
like buying 3 ndx32 filters and using them in tandem or triplet to achieve different results?

[spada]

Thank Peter for those info, I've never use filter before, I am thinking of getting ND4 or ND8 and one CIR PL Multicoats 77mm , so I can use with my wide angle and 28-70 std zoom lens .

regards
spada

[birddog114]

The most sought and use are: ND filter are: ND2, 4, 8.

The more higher number is rarely used or just used it in some special application in photography and required sturdy tripod, head and timer remote.

Any ND filter over 8x is just packed in your bag for fun, found no life with them so far in our daily hobby of photography.

[Heath Bennett]

I have wanted to buy these for a while now... also the graduated ND's. Do you stock them Birdie?

Where did I put those damned lust pills...

[xorl]

I've been pretty happy with my ND8. I can always adjust the ISO a little to balance DoF & shutter if a ND4 would have been more appropriate.

[Heath Bennett]

From my research Nikon brand polarisers are much more expensive than the others, but don't really allow much less light in. This is a hidden positive of the Hoya polariser that I have - it works like a weak ND filter. For this reason I would only really want the ND4.

[MattC]

ND400 = 9 Stops = the filter in my welding mask!!!

Cheers

[Onyx]

Errr... it's just overall darker. What's the histogram comparison look like?

My understanding was, if exposed correctly there should be absolutely no difference between ND filtered and non-filtered images (camera's TTL exposure metering should compensate).

Thus was so when I first got a CPL for my kit lens. It loses 1.5 stops, versus the ND4 I had (2 stops), and stacked together - compared to no filters. All shots centre-weighted and shot in Aperture Priority, the only effect it had was prolonging shutter speed - and the numbers stuck to the theory very well. There should be no differences image brightness.

[Sheetshooter]

Goodness me, the piffle is flourishing here. On a camera capable of 1/8000th shutter speed and f/22 in tandem with even the limitation of only 200 ISO the need for an ND filter for a shot like this is myth and illusion rather than reality and fact. A straight ND filter is simply a device to reduce the level of image-forming light entering the optical system. With any sort of truly NEUTRAL ND there should be no colour shift and the cast of which Wendell makes mention is more likely to be an artifact of the camera's malfunctioning rather than an inherent quality of given value.

There are specialist uses for ND filters in observing solar events and other sources of excessively high radiance or illumination - the illustration does not touch on one of them. A principal application of the ND filter is to allow shooting at wider apertures and/or longer shutter speeds in situations of high luminosity. With an infinity subject such as the clouds covering the sun with a wide-angle lens there seems very little point and, if all is functioning as expected, there should be no discernible effect. The fact that a difference is discernible would indicate that there were gross errors (possibly of negligence/possibly electro-mechanical) in the actual capture.

This prompts the important question as to HOW the photographs were made: Auto/Manual exposure, Auto/Set White Balance, etc.

With extended sessions of photographing something intensely bright such as the sun at the inifinity range the ONLY sense I could see in the application of an ND filter would be for the sake of the eyesight peering into the looken-peeper.

As one comes to expect, Heath is on the right track in waxing lyrical about the GRADUATED ND filter - a lighting kit in a pouch. But that is an entirely different beast with an entirely different set of purposes.

Cheers,

[MattC]

It seems that the most common use for ND filters is capturing water in motion.

A couple more uses:
When shooting outdoors in bright and contrasty conditions where I need to use flash, ND filters can be used to bring the shutter speed back into flash sync range or to allow the use of larger apertures. As an example, I sometimes find myself in the 1/500s and F8-F11 (or higher) range without a filter. A 2 stop ND allows 1/500s F4-F5.6 or F2.8-F4 with a 3 stop ND.
Another example of usage would be where it is desirable to shoot at a given aperture and shutter speed (no flash). For example, a panning shot of a downhill racer where it is desirable to show motion in the spokes. Required exposure may be 1/60s and F8, but ambient conditions may require a shutter speed 3 or more stops higher. This is a similar sort of usage as the running water - but less special effects.
These are just examples, the numbers do not matter, the important part is the concept.

If like me, you happen to use lowlight lenses (such as the 50/1.4 with its minimum aperture of F16) outdoors in bright conditions, it can be beneficial to have a couple of ND filters at hand to keep the lens in its optimum aperture range.

ND filters are just another tool to add to the list for exposure adjustment (the others being aperture, shutter speed and ISO) allowing for greater flexibility in the range of shutter speeds and apertures.
As already pointed out, if metering is working correctly in any program mode the image should be identical (brightness identical with no colour casts) with or without the filter. The only thing that will be different is the aperture and/or shutterspeed. Drop the camera into manual mode and not adjust the exposure between the two test shots, what would be seen is something like the two shots above and the meter will show x stops less.

Cheers

[birddog114]

MattC,
I agreed all your points and suggestion, I have and use the same techniques as yours in using the ND filters.
Nothing harm! everyone can play and experiment with it and find your own solution, theories sometimes are not working or becoming garbages as the way you want.

[MattC]

Nothing harm! everyone can play and experiment with it and find your own solution, theories sometimes are not working or becoming garbages as the way you want.

I know what you mean.
There did seem to be a misconception about what to expect from an ND filter. Of course that could be just the way I read the thread. I started to write something last night but decided to leave it until this morning - too tired. Sheetshooter and Onyx beat me to it, so I thought that I would throw in a couple of usage tips.

Cheers

[stubbsy]

Goodness me, the piffle is flourishing here. On a camera capable of 1/8000th shutter speed and f/22 in tandem with even the limitation of only 200 ISO the need for an ND filter for a shot like this is myth and illusion rather than reality and fact. A straight ND filter is simply a device to reduce the level of image-forming light entering the optical system. With any sort of truly NEUTRAL ND there should be no colour shift and the cast of which Wendell makes mention is more likely to be an artifact of the camera's malfunctioning rather than an inherent quality of given value.

There are specialist uses for ND filters in observing solar events and other sources of excessively high radiance or illumination - the illustration does not touch on one of them. A principal application of the ND filter is to allow shooting at wider apertures and/or longer shutter speeds in situations of high luminosity. With an infinity subject such as the clouds covering the sun with a wide-angle lens there seems very little point and, if all is functioning as expected, there should be no discernible effect. The fact that a difference is discernible would indicate that there were gross errors (possibly of negligence/possibly electro-mechanical) in the actual capture.

This prompts the important question as to HOW the photographs were made: Auto/Manual exposure, Auto/Set White Balance, etc.

With extended sessions of photographing something intensely bright such as the sun at the inifinity range the ONLY sense I could see in the application of an ND filter would be for the sake of the eyesight peering into the looken-peeper.

As one comes to expect, Heath is on the right track in waxing lyrical about the GRADUATED ND filter - a lighting kit in a pouch. But that is an entirely different beast with an entirely different set of purposes.

Cheers,

Walter

Thanks for your words of wisdom.

I shot these in Aperture priority mode so the TTL metering system should, if I understand you correctly, adjusted the shutter speed and I'd have two identical images. Metering was cntre weighted. This leaves me with the conundrum of why they are different.

From what I've read elsehwere ND filters also reduce the incidence of blown highlights (which is what I touched on at the top) in very bright scenes where you want a particular aperture or shutter speed setting. This appeared to confirm that reading so I am now officailly confused Or have I misunderstood you PS I'm at work so don't know the EXIF. Will look this up tonight

[Mj]

As already mentioned... an ND simply reduces the amount of light some amount... nothing more nothing less... if it is truely neutral in colour you should see no colour change, but many (and I think Peter's sample is one) will create some shift in colour. I suspect that the change Peter has shown is most likely due to changed exposure conditions keeping in mind that the cloud formation has clearly moved between shots and he has used centre weighted metering.

[Sheetshooter]

This raises an interesting point. I use AUTO nothing if I can help it. If nothing else, this allows me to be better informed as to what is going on. If I meter the sky and get a result that pleases and then add a two-stop ND filter I have sufficient faith in NAME-BRAND filters to simply open up the lens or the shutter by two stops.

Particularly with something as variable as digital data processing there is no way of predicting what slight biases may arise. Then there is the issue of shifts in illumination or luminosity by somethiung as variable as small clouds fleeting across the sun.

Meter inconsistencies are another oft-ovewlooked factor. If there are rules in photography then Rule Number 1 is NEVER ASSUME. To this end I have a set of very accurate Wratten ND gels covering 10 stops (none to a 3.0) and I regularly place these over the meter cell to check linearity of the meter response. I use two meters on a regular basis: a Minolta Flashmeter V and a Pentax 1º spot (modified by by Zone VI in the US of A). These are the only two makes of meter I have tersted that consitently give a linear response. In other words, the introduction of a 5-stop ND will result in exactly 5.0 stops difference in the reading. Not 4.7 and not 5.9, for example. Now this may all seem quite anal but let me assure you that there is one brand of reputable meter that I tested for linearity new out of the box (several samples over several years) and the only consistent thing I have found with these is that in 10-stops less light they underexpose by more than 3 stops - and the drift starts from the introductoion of a 1-stop filter.

With digital photography and the use of the histogram these factors can be readily seen in many ways. It always pays to check.

Another source of false readings is the meter cell's reaction to increased I.R. or U.V. light.

The other factor which can drastically influence results like this is the fact that a digital image and it's dynamic range is a LINEAR response whereas film is LOGRITHMIC and so if high values are placed too high the system may be introducing all manner of effects in order to attempt salvaging what it might consider a lost situation.

Let me add that with the D70s in particular I am happy to let the camera have oits own mind but I have only found that Natrix metering gives the consiostency or predictability of response that I seek. But for any commercial work where it matters I use hand metering of both the colour temperature and the illumination (and sometimes luminance).

Cheers,

[lejazzcat]

Using the cokin P system with grad filters, you can set your horizon point to balance sky/ground exposure.

[marcotrov]

Once again Sheetshooter you boggle the mind, well certainly mine! with the wealth of technical and practical information you possess a lovely marriage of theory and practice. You are the treasure trove of info we all value.

Walter you gracefully entwine the science and the art of photography. Keep up the detailed comments even if, occassionally you lose me with your explanations, not your fault but my neanderthal brain trying to come to grips with some of the technical concepts you spend time explicating for us all.
Either way thoroughly and intellectually absorbing
cheers
marco

[stubbsy]

OK

Here's the EXIF. Both shot at F6.3, Auto WB, Multi-Pattern metering

Pic without ND: 1/5000 sec

Pic with ND 1/3200 sec

[Sheetshooter]

Peter,

That info supports my lack of faith in auto metering. You said you used and ND 4 filter with a two-stop density rating. (This is a classification I am not familiar with - most are rated logrithmically and so I am more accustomed to a two-stop ND as being a 0.6). Anyway, the difference between 1/5000th of a second and 1/3200th of a second is just TWO-THIRDS OF A STOP - not 2 stops. It seems the metering system has seen fit to underexpose by one and a third stops!

You could achieve the same result without employing a costly filter just by under-exposing. In fact, it begs the question: are you absolutely sure that you actually put the filter on the camera? I'm not asking that to be a smartarse but it seems that such an error in the metering is pretty substantial. If you DID use the filter then it could indicate that there is something a bit on the bugle in the state of Denmark and the camera could do with a check-up.

Cheers,

[phillipb]

Walter,
I don't know about the camera, but I can attest that Peter used the filter, I was there when he did it at the mini meet on Saturday.

[Sheetshooter]

Stubbsy,

You don't mention the ISO setting. Assuming 200 ISO then an exposure based on the 'sunny 16' notion would indicate an exposure somewhat at odds with what you have posted - about 2-stops in the case of the unfiltered shot as it happens. Given the fact that the meter was reading from the light source I could imagine perhaps allowing a stop, but two-stops would seem exccessive. Add to that the fact that the filtered shot is a further stop and a third under. Little wonder there are colour shifts.

May I ask how you processed the shots? Did you capture in RAW? If so, how did you convert them ... and to what?

I shoot RAW and use ACR for the conversion. In its default state ACR applies a series of 'AUTO' settings to each file, attempting make them all look the same. I disengage these AUTO settings (Command-'U' on a Mac) and set the parameters myself. This negates any further errors entering the equation.

There's something going wrong somewhere my friend and I'd love to be able to get to the bottom of it for you.

[stubbsy]

Stubbsy,

You don't mention the ISO setting. Assuming 200 ISO then an exposure based on the 'sunny 16' notion would indicate an exposure somewhat at odds with what you have posted - about 2-stops in the case of the unfiltered shot as it happens. Given the fact that the meter was reading from the light source I could imagine perhaps allowing a stop, but two-stops would seem exccessive. Add to that the fact that the filtered shot is a further stop and a third under. Little wonder there are colour shifts.

May I ask how you processed the shots? Did you capture in RAW? If so, how did you convert them ... and to what?

I shoot RAW and use ACR for the conversion. In its default state ACR applies a series of 'AUTO' settings to each file, attempting make them all look the same. I disengage these AUTO settings (Command-'U' on a Mac) and set the parameters myself. This negates any further errors entering the equation.

There's something going wrong somewhere my friend and I'd love to be able to get to the bottom of it for you.

Walter

This is all a puzzle to me too.

Yes, ISO 200 with an EV of -0.7 when shot. Raw image, No PP, didn't alter EV - just opened in Nikon Capture, resized and saved as jpeg.

I'm certain the ND 4 was on. The images were handheld (hence the image movement) and taken 22 seconds apart. So far as 2 stops for the ND4 filter - that's taken straight from here(exerpt below). This is consistent with what you've said (ND 4 = 0.6 = 2 stops). BTW the filter is labelled ND4 not 0.6.

Hoya Neutral Density filters are available in most sizes and in three strengths:

.3 (2X) Reduces the light one f-stop.
.6 (4X) Reduces the light two f-stops.
.9 (8X) Reduces the light three f-stops.

[sheepie]

...Metering was cntre weighted...

Which. to my simple mind at least, suggests the reason for the apparent 'error' in metering. Put simply and briefly - perhaps the cloud structure moved slightly between pics, thereby meaning a different overall balance of light for the meter to read.

As these were never really intended to be scientific tests - rather a quick try to see the difference - I think they served their function well. Perhaps it would have been better to use (say) matrix metering but even then there would have been differences. Nature doesn't stand still waiting for us to change filters!
A pure scientific test of this would have required identical lighting conditions - something unachievable in the natural world, and totally achievable in the studio.

I appreciate the well-informed posts here talking about the usefulness of, and conditions where, ND filters work. There are some good learnings here - some of which I'll try to remember when next I look to use one.
But let's not read too much into what was essentially a show-and-tell of Peter's first (I think!) experience using these things

Peter - thanks for sharing these in a bigger version than we were able to see on the day (the camera screen).
I think actual real-world use of these filters will be a good test, and look forward to seeing some of your results

[Hlop]

Technically, to avoid any metering interferences to the tests results, manual metering must be used.
1st step: Proper exposure without filter
2nd step: Just put ND filter and image should look underexposed
3rd step: Correct exposure (2 stops in this case) Resulting image must be quite similar to step1 without introducing colour casts with the same brightness as mentioned above

From what I've found a while ago, only expencive ND filters work this way (B+W for example) cheaper ND filters are greyish and bringing colour casts