Blue waterfall - Velvia shift?

[jdavis]

Howdy all,

I have a shot I'm very happy with of one of Oregon's many waterfalls. The problem is that the water has a very blue color tone that may or may not be accurate. The composition is of your typical dark forest, with some mossy green, dark woods, and lots of water - but the blue tone can be a bit dominating. The exposure was a very long one (30-45 seconds) with Velvia 100.

Is a blue shift normal for Velvia in these kinds of conditions? Or is the blue "true" ?

[Robert Oliver]

if you were shooting in shadows, which is typical of the waterfalls in the gorge anyway, your color will shift to blue. For accurate colors you will need to add a warming filter. Start with an 81A.

[roteague]

I would say that Robert Oliver's post is correct. You really need to warm it up a bit, an 81A may be a bit light, however. An 81B or even an 81C might be a better choice.

[Bruce Watson]

I have a shot I'm very happy with of one of Oregon's many waterfalls. The problem is that the water has a very blue color tone that may or may not be accurate. The composition is of your typical dark forest, with some mossy green, dark woods, and lots of water - but the blue tone can be a bit dominating. The exposure was a very long one (30-45 seconds) with Velvia 100.

Is a blue shift normal for Velvia in these kinds of conditions? Or is the blue "true" ?

The idea of "accuracy" may be misleading in a case like this. The problem is films are optimized to work with a given quality of light. Tungsten balanced films work best under the fairly yellow light you get from 3200K photofloods for example. Daylight balanced films for whatever the manufacturer thinks "daylight" is (D50? D65?).

All this if fine, and then you take the film outside. The quality of the light is almost certainly different from what the film was balanced for. All kinds of things effect the color of the light the film sees -- being filtered through green leaves, being lit by open blue sky, being lit by open gray sky, being lit by light reflected off a red cliff, etc. The film just records what it sees. So in that sense, your blue waterfall is "accurate." But it doesn't match what you saw at the scene because our vision systems can compensate for constantly varying light quality while the films can not.

This is just a long way of saying what they others have said. Use a warming filter. You'd be surprised how high a color temperature you really get, even in the darkest forest on a cloudy day.

[Alan Davenport]

I stopped using Velvia for waterfalls because of the weird color shifts. Velvia just doesn't have the reciprocity characteristics needed for long exposures of white water, IMO. After spending a lot of time trying to get waterfall shots to appear natural, I tried using Ektachrome. Wow. The Ektachrome renders the white water like it should be, and it never gives weird color shifts I sometimes got with Velvia.

I'm presently working on a freezer of E100SW, which gives the water the best color (i.e., NOT weird blues/purples/magentas) of anything I've tried. Exposures up to almost 30 seconds still have nice, neutral water.

The most comparable current offering is E100GX.

[roteague]

I stopped using Velvia for waterfalls because of the weird color shifts. Velvia just doesn't have the reciprocity characteristics needed for long exposures of white water, IMO.

Are you talking Velvia 50 or 100? Same name, two different films. I've had no problem using any of the 100 speed Velvia films for doing slow, water scenes.

[Alan Davenport]

Are you talking Velvia 50 or 100? Same name, two different films. I've had no problem using any of the 100 speed Velvia films for doing slow, water scenes.
Admittedly, my experience was with Velvia 50.

[roteague]

Admittedly, my experience was with Velvia 50.

Velvia 50 is quite well known for its color shifts. The new Velvia 100 has much better reciprocity characteristics minus the color shifts. However, I've never done any exposures longer than 3 or 4 seconds with it. Most of my exposures are in the 1/2 to 1 second range.