Grads vs multiple exposure stacking

[NicholasR]

Hi All,

Just looking for some input.  Several people's work that I admire continue to use graduated ND filters when shooting digital.  Personally, once I moved away from film I began shooting multiple exposures, then doing grads in PP with layers, masks, etc.  Note I am not speaking of HDR merges.

While I understand getting as much as possible right 'in camera', assuming that one is shooting landscape on a tripod, what is the reason to continue using grads?  There seem to be less post processing options, less control, and it's at least (and probably more) fiddly in the field as exposure stacking is.

So, if you use grads, why is that?

[Guillermo Luijk]

The advantages of grad filters are derived from shooting only once:
- Save time, memory and CPU resources
- No alignment or ghosting issues
- Possibility of shooting HDR scenes in JPEG

If none of these advantages means anything to you, you can save the grad filter.

[MarkL]

For things that move like waves on the sea grads can be useful but honestly I have not used my lee grads with a digital camera. Multiple exposures means you can create any mask you want rather than a grad line with predetermined transition area, I've seen plenty of film landscapes with obvious grad lines especially with trees and rocks in the foreground.

I don't even exposure blend much anymore since most of my landscapes are 9-12 frames stitched I just exposure each frame properly and the blending in the stitching process deals with it since the software is so good now.

[Graystar]

Several people's work that I admire continue to use graduated ND filters when shooting digital.  Personally, once I moved away from film I began shooting multiple exposures, then doing grads in PP with layers, masks, etc.
...
While I understand getting as much as possible right 'in camera', assuming that one is shooting landscape on a tripod, what is the reason to continue using grads?  There seem to be less post processing options, less control, and it's at least (and probably more) fiddly in the field as exposure stacking is.

So, if you use grads, why is that?

For me, my hobby is photography...not computer art.  The pleasure is in making things happen in the lens.

[Chris_Brown]

So, if you use grads, why is that?

I used graduated filters when I shot film and chromes. They really helped balance the exposure between ground and sky. I found them limiting when shooting exterior architecture (i.e., tall buildings breaking up the horizon), and even then I resorted to layering multiple scans of the same scene (and dealing with the inherent lack of image registration).

(The thing that discouraged me the most was the price of the Cokin filter set up. The filters + the filter holder was priced too high for the build quality.)

I'm not a purist, I'm in it to satisfy my needs or that of a client. Layering different digital exposures in PS provides far more control, and was happy to sell off my Cokin gear years ago. The only filter I use to enhance a sky is a polarizer.

[NicholasR]

I'm not a purist, I'm in it to satisfy my needs or that of a client. Layering different digital exposures in PS provides far more control, and was happy to sell off my Cokin gear years ago. The only filter I use to enhance a sky is a polarizer.

Which was where I ended up as well.  It's about results for my clients and buyers, not a notion about purity (though there is nothing wrong with that for other peoples needs).  In practice I find it rare that ghosting is a problem, and anything that needs a grad is going to need significant computer PP time regardless in my world.   Realistically just about everything other than basic headshots gets PP time.  I guess I am a computer 'artist' as well then.

Thanks for the replies, I was just wondering if I was missing something.  I dont think I am.

[Guillermo Luijk]

Also think that the usefulness of a grad filter or doing a couple of shots can be substituted by a single shot with a high dynamic range sensor and no filters at all.

For instance my Canon 350D has 8 stops of usable dynamic range from sensor saturation to acceptably noiseless shadows, while newer cameras such as the Pentax K5 have 11 stops of usable dynamic range. If we are to shoot a sunset the following 3 options are equivalent in their ability to capture the scene's dynamic range since all of them can capture 11 stops:
- Shoot once on the 350D using a 3 stops grad filter
- Shoot twice on the 350D 3 stops apart without any filter
- Shoot once on the Pentax K5 without any filter

Grad filters have become unnecessary thanks to digital multiexposure. In the same way higher dynamic range sensors will make multiexposure unnecesary.

Regards

[Chris_Brown]

One example where a graduated filter would come in handy is when the landscape scene is being buffeted by wind, and everything that can move (e.g., trees, flags, tall grasses) is moving. This is when a single exposure provides a better solution than a layered multi-exposure image.

[ARP]

I have another reason to use ND grads with digital, but granted it is in specific circumstances.

My exposures are usually in the ~4 minute range because I tend to shoot when the light is ~EV 10 with a 10 stop ND.  I shoot two shot stitches with the 24pc-e for seascape panoramas, and when the light is increasing/fading fast I don't have the luxury of spending ~12 minutes bracketing exposures.  ND grads allow me to complete the entire shot with minimal difference in EV between the two shots.  Not only that, but I also want cloud movement to match that of the water movement (4 minutes of water movement vs 30 seconds of cloud movement, if bracketing 3 stops, just doesn't work for me).

[Guillermo Luijk]

Not only that, but I also want cloud movement to match that of the water movement (4 minutes of water movement vs 30 seconds of cloud movement, if bracketing 3 stops, just doesn't work for me).

What you say is interesting. I think you don't need to renounce to having the same movement for both water and clouds even if not using the grad filter. Imagine you use the same aperture and ISO in the following 3 methods:

METHOD 1 (grad filter)
1 shot 4 min with 3 stops ND grad filter (well exposed water and clouds)

METHOD 2 (blend method)
1 shot 4 min (4 min water movement and blown clouds)
+
1 shot 30 secs (underexposed water and 30 secs clouds movement)

METHOD 3 (grad filter simulation)
8 consecutive shots 30 seconds each (water will be 3 stops underexposed with respect to method 1 and none of them will have blown clouds).


Method 3 provides even more flexibility than Method 1. We average the 8 shots in 8 layers in PS (this can even be done linearly using a gamma=1 profile to perfectly mimic camera photon count) and push exposure for the water with a tone preserving curve and custom layer mask. If the 30 secs shots were taken right one after another you will not notice any time gap (times between shots will be negligible vs 30 secs), so the averaging will look 100% the same as a single 4 min shot both for the sky and clouds.

Regarding IQ, Method 3 will collect the same amount of photons for the water in the 8 shots of 30 secs as in the single 4 min shot with the grad filter in Method 1, so SNR in the water will be the similar (we could do more precise calculations here). And SNR will even be better for the clouds because you will collect the photons rejected by the grad filter in this area.

If the camera provides Liveview, the mirror can stay locked up all the time so the loss of sharpness because of misalignement will be minimised (only the shutter vibration remains).

Regards

[torger]

I'm still thinking about starting using grads, but will probably continue do stacking instead, blend in post. There's a certain purist romance in taking only one exposure, and adds some fun to the photography craft, that's the main reason for using grads as I see it. In some complicated scenes with moving stuff grads could be useful too, but that does not happen too often to me.

Dynamic range in digital cameras today are also so good, so often I end up using only one exposure anyway, with some grad applied in post to make it more printable. Postprocess grads have the advantage that they can be shaped in any form to perfectly match the scene.

[RFPhotography]

For me, my hobby is photography...not computer art.  The pleasure is in making things happen in the lens.

So where does that put the film photographer who blends multiple negs on the enlarger?  Is there really any difference?

[PierreVandevenne]

capture the scene's dynamic range since all of them can capture 11 stops:
- Shoot once on the 350D using a 3 stops grad filter

The key being of course that the 350D will map those 11 stops to 8 and that the Pentax with a grad filter would map 14 stops to 11.

[Josh-H]

For me, my hobby is photography...not computer art.  The pleasure is in making things happen in the lens.

Same Here. I get great enjoyment out of getting the shot in a single capture (and I use grad filters extensively). Sure, I know how to do it with multiple exposures in my computer; but it takes me less time (and provides more enjoyment) to slide a grad in and shoot it in the field. Each to their own.