Tuesday, November 29, 2005

Stroboscopic snowflakes?

When I first came to Canada in 2000 I was living in Ottawa, so I got to see a lot of snow. Being from Australia I was fascinated by the snow. Less so about the ice and the minus 35 degrees, but the snow was almost magical.

Growing up in Geelong it never snowed, although I do remember once when there was a hail storm that dumped about 2cm of hail on the ground that I could pretend it was snow. But I was 19 before I experienced my first clump of snow on the side of the road driving through a National park in Victoria (AUS).

Now that I'm living in Vancouver (BC), snow happens much less frequently than in Ontario. But I'm still intrigued and fascinated by it.

Last night about 5cm of snow fell. I stood out on the balcony of our apartment and watched it fall. Watched it twist and move in the wind. It wasn't too long before I set up my camera to take some photos. It seemed like the thing to do.

There is a street lamp a few metres from the balcony and I wanted to use a long time exposure, in order to capture the dance of the snowflakes as they fell. When I went through the photos tonight, I noticed something that intrigued me. In the images where the snow was falling fast, so the flakes were captured as long streaks on the image, there was a regular variation in the brightness along the length of the streaks.

Canon 300D, 70-200mm f4L lens @ f5.6 + 0.5 seconds + ISO 400

But in the images where the flakes were not moving in such a fast and uniform motion, there wasn't any hint of the banding on the streaks.

Canon 300D, 70-200mm f4L lens @ f5.6 + 0.5 seconds + ISO 400

And this is another example where some of the streaks show banding and others don't appear to:


Canon 300D, 70-200mm f4L lens @ f5.6 + 0.5 seconds + ISO 400

I captured a lot of images last night and many of the images show this effect, it is remarkably consistent. One thing I do know is that it isn't caused by the snow flakes tumbling and varying in brightness across the frame. In fact it is always the opposite effect. The banding only occurs in those images where the snow was moving rapidly across the image.

I can only summise that it is caused by some sort of strobe effect from the sodium street lamp. Although I haven't been able to find any reference to this on-line. Although I may have used the wrong search terms.

Here are a couple of other images showing the banding effect - the street light illuminating the flakes is just above the upper left corner of these shots:

Canon 300D, 70-200mm f4L lens @ f5.6 + 0.5 seconds + ISO 400

Canon 300D, 70-200mm f4L lens @ f5.6 + 0.5 seconds + ISO 400

Can anyone confirm the mechanism that causes this intriguing effect?

5 Comments:

Anonymous Anonymous said...

It's very simple.

The street lamps do not emit a continuous light, but rather turn off and on in microseconds. The human eye perceives it as continuous, but since the photographic exposures are long, they manage to capture the sequence of on-and-off many times as they illuminate the snowflakes.

2005-12-02, 6:08 AM  
Blogger Daz said...

Yes, that sounds right. The ballast in the sodium lamps will operate at the line frequency of 60 Hz. But, I'm still a little perplexed as to why the effect doesn't seem to be a uniform effect - only occuring in some of the images and not all of the streaks...

2005-12-02, 8:28 AM  
Anonymous Anonymous said...

I guess it's because of the speed of the flakes. If a flake moves slowly, its' position doesn't change much between "exposures". A faster flake, on the other hand, moves so much that there is a clear gap between its' positions in consecutive "exposures". Makes sense?

2005-12-31, 10:58 AM  
Blogger Daz said...

Yes, makes sense to me... Happy New Year!

2005-12-31, 11:05 AM  
Anonymous jude said...

Gorgeous series of images, Darryl...so glad to have found your blog. Thanks for posting its url on PP.

I shall return.

2006-12-04, 10:06 AM  

Post a Comment

<< Home




Click Here