I've taken a closer look at the so-called "third step increments" in shutter speeds

[Compaq]

From 1/4th of a second to 1/8000th of a second, 33% more/less light only occurs four times out of a total of 33 stops. The average change in amount of light is
25.9 %. In fact, when going from 1/15th to 1/13th, we only get 15 % more light, which is half of what we think we're getting (unless this is common knowledge).

For those interested in knowing how I did this exceedingly simple math, here's an example:

How much bigger is 1/13 than 1/15? We just find the relationship between the two, with focus on 1/13's relation to 1/15 (to get a number > 1). The light difference will be the decimals in the number.
Light difference from 1/15th to 1/13th:

(1/13) / (1/15) = 1.153 = 15.3 % more light


This may not have much practical use, perhaps it's more of a "fun fact". Here's what we do, we SUE all the companies for calling it "1/3 increments stops" in their menus, and CLAIM free products until our days are over!

 

[Derrel]

There is a song for you. It was performed by the band Styx, and it was called Too Much Time On My Hands.

‪Styx - Too Much Time On My Hands Live 1996‬‏ - YouTube

 

[Trever1t]

bawaaaa!

 

[Compaq]

What, I was bored

 

[bentcountershaft]

So which one is lying? Is the shutter duration actually the same as what is indicated or have they rounded the fractions for display purposes only?

 

[Overread36]

Its not uncommon that many of the simple explanations we understand and use with regard to photography are, when placed under closer inspection, mostly simplifications of more complex relationships which, whilst more accurate, are not often needed by the average photographer in day to day shooting.

 

[ghache]

There is a song for you. It was performed by the band Styx, and it was called Too Much Time On My Hands.

‪Styx - Too Much Time On My Hands Live 1996‬‏ - YouTube

You love that song dont you?

 

[Compaq]

I don't know if they've rounded the numbers. Well, of course they have rounded the numbers, but whether we actually are getting 33 % more light by stopping up "one third" or not, I don't know.

Before these third stop increments came about, the shutter speeds went from 1/1000th of a second to one second, so this is what people are "measuring" from (in lack of a better word). I noticed something when looking at this, that is that if we go "one third" down from 1/8000th of a second to 1/6400th of a second, we can half this shutter speed several times, just as with 1/8000th of a second. Each shutter speed you have pre-set in the camera (except the highest ones) can be doubled and halved (or apprx doubled or halved). This contradicts the fact that each turn of the wheel will give you 33 % less / more light. Do you see what I mean.

Of course, this isn't really that important, just something I found interesting.

 

[Derrel]

There is a song for you. It was performed by the band Styx, and it was called Too Much Time On My Hands.

‪Styx - Too Much Time On My Hands Live 1996‬‏ - YouTube

You love that song dont you?

Look everybody--it's my own personal internet stalker!!! He loves following me around TPF, leaving little comments. Aren't stalking French-Canadians cute!?! lol...nice try Gayhomme!

 

[analog.universe]

So... I think the math that explains this goes this way:

If a whole stop is a doubling, then 1/3 stop would actually be the cube root of the double. 2 ^ (1/3) gets you something like 1.26, or a 26% increase, as opposed to the 33%

so, an applied example: if you start at 1/400... 400*1.26=504 (~500) 504*1.26=635.04 (~640) 635.04*1.26=800.15 (~800) which are the numbers we see when we actually click the dial

 

[Compaq]

Aren't you talking about apertures now? Surely if 1/500th lets in X amounts of light, a third of that will let in X/3rds amount of light, or 33% less?

 

[analog.universe]

But 1/3 of a stop isn't 1/3 less light. It's 1/3 the way down to 1/2 as much light... the next stop. When you're looking at the linear numbers of shutter time, a translation is being made to an exponential scale so that the times can be adjusted in stops. To maintain accuracy on the exponential scale, the divisions must occur such that each click multiplies the previous value by the same amount. 1.33*1.33*1.33 is not 2, so 3 clicks would be well over a single stop. 1.26*1.26*1.26 is a single stop (ignoring round off error).

 

[Compaq]

Ahh, of course. Logical shortcoming from my side Thanks for clearing that up!

 

[analog.universe]

:thumbup: