Graduated Neutral Density Filters

[Jon_Are]

I'm considering getting one of these, but one aspect of it doesn't make sense to me.

Looking at the filters, the gradient ends virtually halfway down the filter. This means, in order to use it most effectively, I would have to shoot my landscape so the horizon is dead center. If I adjust the composition in either direction, the ND effect would be off.

What, do these filter makers follow the rule of halves?

What am I missing?

Thanks,

Jon

[Ads]

[smadab]

You'll want to get a Cokin Filter holder (I use the P-series) and adapter for your lens. The holder allows you to slide and/or rotate the GND filter to align the gradient with your subject.

Cokin Filter System

[NateS]

Your Grad ND should fade out....i.e. get lighter the farther down it goes. If this is the type of Grad ND filter you have then you should be able to shoot normal (say 1/3 land and 2/3 sky) and the Grad ND will fade out throughout the sky portion. So basically it will darken the top the most and gradually dissapear so that there's no prevalent line.

No there are Grad ND's that basically have a hard line where the transition is, but I wouldn't want to use one of those anyway for the reason's you are stating.

[Big Mike]

If you are looking at circular grad filters...they don't make sense to me either. I wouldn't bother with them.

If you want to use grad filters, get the rectangular ones (Cokin for example), that way you can place the split where you want it in your composition. There are filter holders that attach to the lens but I've also seen some pretty good photographers who just hold them in front of the lens for the shot.

I'm considering getting one of these, but one aspect of it doesn't make sense to me.

Looking at the filters, the gradient ends virtually halfway down the filter. This means, in order to use it most effectively, I would have to shoot my landscape so the horizon is dead center. If I adjust the composition in either direction, the ND effect would be off.

What, do these filter makers follow the rule of halves?

What am I missing?

Thanks,

Jon

[Canosonic]

Here is the same scene with and without a powerfull ND8 Filter. I bought one by accident , but still find it useful:
Without ND:



With ND:

[icassell]

On this topic ... I just got LR2 and I know there is a Grad ND function in it. Does this really work? Can I get rid of my Grad ND filters?

[Moglex]

On this topic ... I just got LR2 and I know there is a Grad ND function in it. Does this really work? Can I get rid of my Grad ND filters?

Not a good idea as that would, when used, eat into (or exceed) the exposure latitude you would otherwise have available.

[icassell]

On this topic ... I just got LR2 and I know there is a Grad ND function in it. Does this really work? Can I get rid of my Grad ND filters?

Not a good idea as that would, when used, eat into (or exceed) the exposure latitude you would otherwise have available.

Would you explain this further? Thanks.

[Moglex]

On this topic ... I just got LR2 and I know there is a Grad ND function in it. Does this really work? Can I get rid of my Grad ND filters?

Not a good idea as that would, when used, eat into (or exceed) the exposure latitude you would otherwise have available.

Would you explain this further? Thanks.

(I should have specified this only applies to RAW - obviously ND emulation is a bad idea if your shots are saved as JPEG's)

When you take a picture and store it in raw there is a 'lattitude' that you can access when you 'develop' the RAW image to TIFF or JPG. A sort of after the event exposure compensation.

If you are relying on using post processing to give you the effect of a graduated ND filter you will no longer be able to use some or all that lattitude because it will be used to perform the ND filter emulation.

So, if you expose perfectly (and the degree of ND is within the RAW lattitude) everything should be OK but if you are out what would have been recoverable may no longer be completely recoverable because the ND emulation is using some or all of the available lattitude.

[icassell]

Not a good idea as that would, when used, eat into (or exceed) the exposure latitude you would otherwise have available.

Would you explain this further? Thanks.

(I should have specified this only applies to RAW - obviously ND emulation is a bad idea if your shots are saved as JPEG's)

When you take a picture and store it in raw there is a 'lattitude' that you can access when you 'develop' the RAW image to TIFF or JPG. A sort of after the event exposure compensation.

If you are relying on using post processing to give you the effect of a graduated ND filter you will no longer be able to use some or all that lattitude because it will be used to perform the ND filter emulation.

So, if you expose perfectly (and the degree of ND is within the RAW lattitude) everything should be OK but if you are out what would have been recoverable may no longer be completely recoverable because the ND emulation is using some or all of the available lattitude.

Gotcha! That makes perfect sense (and I only shoot in RAW). I'll keep my grad ND

[Garbz]

On that note it works exceptionally well in cases where the highlights have only just been rebuilt. As lightroom works directly with the linear RAW data, and invokes highlight recovery algorithms every time you take the exposure slider into negative, it's amazing how much detail is actually still available in a photo, which is visually lost when it is adapted via a gamma curve to be displayed on a screen or processed in a camera.

Bear in mind though this works only on RAWs, and in Nikon's case works far better with RAW compression disabled. But then it also only works up to a point. If your scene is simply way too bright and the odd highlight is clipped in only 2 channels it'll work well. If you have actually completely clipped the sky it will do nothing for you.