Exposing for snow

[Chuck_P]

Last Winter I shot this with Velvia 100 on the F5 using AF settings.

No grey snow.

Copake, surely you are not suggesting that the OP take HP5 out to photograph in the snow and not take into consideration how the reflectance of the snow can be tricking the meter? Color renditions of snow from reflected readings may be a different thing, honestly, I am not versed in color.

Also, I assume you were in Nikon's matrix metering mode (similar with evaluative in Canon) that will probably do a better job at arriving at a more accurate exposure in an all automatic system setting. In my Canon, the evaluative metering evaluates many different segments of the viewfinder within the scene totally separate from an adjacent segment, then all segments are averaged, allowing for a more accurate exposure. But I am assuming the F5 was used in that manner, IDK.

Don't you thing reflected readings in general should be met with suspicion, especially in black and white?

Chuck

 

[Kilgallb]

Please explain by what mechanism the light source changes the albedo of snow.

Lee

From Wikipedia

Snow albedos can be as high as 90%. This is for the ideal example, however: fresh deep snow over a featureless landscape. Over Antarctica they average a little more than 80%.

If a marginally snow-covered area warms, snow tends to melt, lowering the albedo, and hence leading to more snowmelt (the ice-albedo positive feedback). This is the basis for predictions of enhanced warming in the polar and seasonally snow covered regions as a result of global warming.

 

[johnnywalker]

Rob Champagne's suggestion regarding the yellow filter is a good one. Even though I open up two stops for snow, there is a fine line between straight white and some texture in the snow. The yellow filter really helps with the texture, and a bit of bracketing on an important shot is a good idea as well.

 

[copake_ham]

Copake, surely you are not suggesting that the OP take HP5 out to photograph in the snow and not take into consideration how the reflectance of the snow can be tricking the meter? Color renditions of snow from reflected readings may be a different thing, honestly, I am not versed in color.

Also, I assume you were in Nikon's matrix metering mode (similar with evaluative in Canon) that will probably do a better job at arriving at a more accurate exposure in an all automatic system setting. In my Canon, the evaluative metering evaluates many different segments of the viewfinder within the scene totally separate from an adjacent segment, then all segments are averaged, allowing for a more accurate exposure. But I am assuming the F5 was used in that manner, IDK.

Don't you thing reflected readings in general should be met with suspicion, especially in black and white?

Chuck

If you re-read the OP:

I went out with a fellow photographer yesterday and was shooting snow. He was shooting digital and I was shooting a combination of HP5+ (whole plate) and Velvia 100 (35 mm). Now, my question is, should I have over or underexposed the snow to get it reasonably white? The question is for both b/w and color film.

He played with exposure on his digital and found that underexposure gave reasonable results. My limited experience of digital told me that it behaves similarly to slide film.

Diane indicates that she had shot both HP5 whole-plate and Velvia 100F in 35mm and was directing her query to both formats and film types.

I was addressing the 35mm, color portion of the query.

 

[dxphoto]

Add two stops with your handheld meter. (basic metering)

 

[Chuck_P]

If you re-read the OP:



Diane indicates that she had shot both HP5 whole-plate and Velvia 100F in 35mm and was directing her query to both formats and film types.

I was addressing the 35mm, color portion of the query.

Copake,

I must apologize as that did not register with me when I read it. Speed reading can be good or bad.

Sorry.

Chuck

 

[Mark_S]

Agree with the other posters about taking the reflected reading, and opening 2 stops. I shoot a fair amount of HP5 in the snow, and I love the texture that you can get on the snow. Much of the light on the snow is blue, so I will often use a yellow filter to increase the texture in the snow.

 

[Lee L]

From Wikipedia

Snow albedos can be as high as 90%. This is for the ideal example, however: fresh deep snow over a featureless landscape. Over Antarctica they average a little more than 80%.

If a marginally snow-covered area warms, snow tends to melt, lowering the albedo, and hence leading to more snowmelt (the ice-albedo positive feedback). This is the basis for predictions of enhanced warming in the polar and seasonally snow covered regions as a result of global warming.

We all know that over time the sun's energy melts snow, which in turn changes the reflectance characteristics of snow from closer to Lambertian to more specular. That's what drives the greater insolation and snow melt feedback mentioned in the article you quote. But that's not the answer to my question.

My question was: In what way does the difference in cloudy/diffuse vs sunny/collimated light striking snow change its percentage reflectance (aka albedo). You say that it does in your earlier post.

Lee

 

[copake_ham]

Copake,

I must apologize as that did not register with me when I read it. Speed reading can be good or bad.

Sorry.

Chuck

Chuck,

Thanks. No problem.

Just be warned, we users of sub-miniature 35mm cameras can be a feisty lot!

 

[photobackpacker]

Hi Diane:

I shot this scene several years ago:

(there was a url link here which no longer exists).

I metered the bark of the tree and placed it between zone III and IV. The snow in direct sun fell on Zone IX.

I wanted to retain the texture of the shadows so I accepted the Zone IX placement but gave the negative N-2 development. This caused the bark to retain detail with high Zone III placement and caused the brightest snow to fall on Zone VII.

In other words, I purposely accepted an 2-stop overexposure to retain shadow detail but "tamed" the highlights via underdevelopment.

 

[Snapshot]

I just finished shooting a roll of APX 100 in a snow filled park. I metered off of some pristine white snow and set my camera for a 2 stop overexposure . My pictures turned out great. The snow was crisp and white, without looking like they were blocking up. For slide film, however, you may reduce it to 1 2/3 stops overexposure.

 

[Steve Smith]

I'm not sure if anyone else has mentioned this but one other thing to remember is that the basic sunny 16 rule (1/ISO at f16) for a bright sunny day is one stop less on light sand or snow. Therefore an incident reading should be reduced by one stop due to extra reflections from the snow (or sand) surface. An incident reading will not take this into account as it is a simple measurement of the light source (sun).
This is usually shown in the simple exposure guide inside film boxes.


Steve.

 

[MattKing]

I'm not sure if anyone else has mentioned this but one other thing to remember is that the basic sunny 16 rule (1/ISO at f16) for a bright sunny day is one stop less on light sand or snow. Therefore an incident reading should be reduced by one stop due to extra reflections from the snow (or sand) surface. An incident reading will not take this into account as it is a simple measurement of the light source (sun).
This is usually shown in the simple exposure guide inside film boxes.


Steve.

Steve:

I'm not sure this is necessarily the case. I think it depends on how you take your incident reading.

IIRC, when the incident meter is positioned so as to be influenced by the light reflecting off the snow or sand, the additional light will be taken into account. If you shield the incident meter from that extra reflected light, then you do have to open up a stop or so.

I haven't tried this for a while, so comments and corrections are welcome.

Matt

 

[Lee L]

IIRC, when the incident meter is positioned so as to be influenced by the light reflecting off the snow or sand, the additional light will be taken into account. If you shield the incident meter from that extra reflected light, then you do have to open up a stop or so.

This take is consistent with everything I've read about using incident meters and with my experience when shooting with properly employed incident meters for the last 25 years. The incident meter will take into account the extra reflected light and fill from the snow or sand.

Lee

 

[Steve Smith]

The incident meter will take into account the extra reflected light and fill from the snow or sand.

I will have to try it if we get any snow this winter!

The Sunny 16 less one stop, i.e. becoming Sunny 22 on snow is still relevant though and in bright to average light, is probably accurate enough (especially for black and white).


Steve.

 

[Sirius Glass]

When I used Sunny 16 on the snow as f/22, I got all the contours of the snow and moguls but anything else tended to be to dark. That is when I learned to take a light reading off my hand.

Steve

 

[Lee L]

I will have to try it if we get any snow this winter!

The Sunny 16 less one stop, i.e. becoming Sunny 22 on snow is still relevant though and in bright to average light, is probably accurate enough (especially for black and white).


Steve.

A single observation from last Monday, a sunny day with full snow cover within 24 hours of the snow falling, no melting yet, sun 18.5 degrees off the horizon in the late morning, incident metered for a subject fully front lit by the sun (photographer's back to the sun). Note that the sun doesn't get higher than 26 degrees off the horizon here this time of year, so I couldn't meter with the sun much higher than it was.

ISO 125 incident meter reading gave EV 15.67. That's f:16 +2/3 at 1/125th, or 2/3 of a stop less exposure than "sunny 16". The meter I used reads in 1/3 stop increments, so it could have been anywhere between 1/2 to 5/6 of a stop less exposure than "sunny 16".

BTW, the reflected reading off the front lit snow on the ground in front of me gave EV 17. I only aimed below horizontal far enough to fill the meter coverage with snow on the flat ground in front of me (avoiding my shadow), so the difference between incident and reflective reading of front lit snow only was 1.33 stops.

I also did this same check aiming toward the sun and wrote the readings on a small receipt in the car as I left for errands. I put it where it would be safe, and it's still safe, wherever it is... So don't hold me to the following data. There was obviously a somewhat more specular character to the reflected snow reading with the sun only 18.5 degrees above it, and IIRC the difference between incident and reflective readings was 4 1/3 stops. You also need to take into account that the standard "correct" way of reading this scene in incident mode would be to have the dome of the meter pointed away from the sun in nearly full shade with the sun at that altitude. You'd likely get better results by holding the meter dome pointed straight up or tilted 45 degrees from straight up (angled away from the sun) in this case.

Lee

 

[Joe VanCleave]

When exposing for yellow snow, use an 81A filter.

~Joe

 

[mikepry]

Might I throw the BTZS wrench into the thread? This is a case where BTZS really takes allot of the work out of tricky scenes such as this. If your familiar with that approach, I take care when I meter (w/incident meter) my low EV's as the reflections from the snow can taint that reading. I usually hold a grey card UNDER the meter as I read the shadow. I've seen as much as 1 full EV difference with and w/o the grey card underneath. I was out the other day and had a high reading of 14.9 (here in Wis.) so the low EV recorded wrong would be a huge problem. I have not noticed much of a difference with and w/o the grey card under the meter as I read the high. (meter dome facing camera)

A winter scene from a couple years ago using this approach (there was a url link here which no longer exists)
As usual the scan sucks but I was able to have detail in the dark shadows and up through the highlights.

 

[RobC]

As you can see from the replies on this question, the reflectance from snow can vary signifcantly so when using an incident meter, there is some guess work involved. That is one of the reasons why a spot meter is a good idea for oudoor work. It removes the guess work and gives a spot on (pun intended) accurate reading for the reflectance of snow. My previous post assumed a spot meter was being used.

 

[thuggins]

Shooting landscapes in Colorado required opening the aperture 1/2 - 1 stop to get proper exposure (i.e. white snow with realistic texture). The correct exposure ended up being what I would have used for the same scene without the snow. This was with VS, so minor changes in exposure (1/3 stop) are noticeable.

 

[John Koehrer]

When exposing for yellow snow, use an 81A filter.

~Joe

Don't you eat that yellow snow, That's where all the huskies go.
Frank

 

[Eric Rose]

For snow shots I either use a green filter or no filter at all. I find using a yellow filter increases contrast to an unacceptable level. Even in shady areas. The green filter also makes tree trunks look better as well. I have included a shot I took last week on an overcast day using a green filter, HP5+ and souped in PyroCat-HD. PyroCat is amazing in how it handles snow scenes. Check my gallery for some snow examples.

 

[colrehogan]

I will have to scan the results from my snow shooting expedition and post them in the gallery here. Thanks to all who replied.