I will assume you aren't being silly and tell you I already have problems with this setup.

Originally Posted By: newton

at first I want to learn You very simple fact ( camera obscura)
I will back to below example later ...

A] bulb ( 4 watts) -----5 meters ------ camera

B] bulb ( 1 watt ) -----5 meters ------ camera

camera in situation A registered brightness of picture X
picture time = 1 sec. ( camera open/close)

camera in situation B registered brightness of picture X
picture time = 4 sec. ( camera open/close)

Looking at your proposed experiment I already have massive issues and concerns that I will be able to answer anything.

Electrical power is measured in watts not light energy ... no 4 watt globes you buy will actually put out the same amount of light and they have horrible efficiency conversions like 2% most of your 40 watts in a incandescent light bulb is going out as heat it is something like 95% to heat. Modern fluorescent lights are better but still bad.

So lets get to specifics have you actually measured the light with a light power meter or do we have even a basic manufacturer specification of the amount of lumens of light we get per watt of your bulbs.

The 1 watt globe is going to be totally different to the larger wattage.

Here you can see the problem on normal incandescent globes


40 W tungsten incandescent 1.9% efficiency 12.6 lumens per watt

60 W tungsten incandescent 2.1% efficiency 14.5 lumens per watt

100 W tungsten incandescent 2.6% efficiency 17.5 lumens per watt

Generally with light bulbs the bigger the globe the higher the efficiency and the more lumens per watt it will put out.

In general if you did you test above a 4 watt bulb on for a quarter the time for a smaller bulb you will always expect the bigger bulb to show more brightness if measured over time just based on incandescent globe physics it puts out more light energy.

We haven't gone into spread angles which ideally you would control by putting the bulbs in a reflector and then bring the light out thru a long metal pipe so all light energy is forced out down the tube in the same way. Essentially you are making the bulb setup act like a laser which is the easier way to do this test.

So done properly I would expect the bigger bulb to show more brightness over time (more light energy) but a lot depends on the bulb type here you really need to calibrate the light energy properly.

So that's my first advice measure the globes properly otherwise the test is really meaningless smile

Next lets talk about the camera as a detector of brightness it is really really terrible .. to show how bad.

Take a 1 second exposure on the 4 watt globe now open the exposure to 4 seconds. Now do the same two exposure for the 1 watt globe and look at all the images.

How do you quantify 4 times the light energy on a camera once it goes white as fully exposed it goes white and no additional arriving energy does anything. You can't meaningfully calibrate any of these 4 exposures.

So the camera is useless for any time or energy based measurement it is worse than non linear it suffers saturation when it goes full exposed and can't go any whiter.

So all in all the experiment is totally pointless and will tell you absolutely nothing meaningful you might as well go to a psychic and ask what it all means because a scientist sure as hell can't tell you anything with that terrible setup smile

If you were a scientist or really were an engineer you would know the drill assume nothing measure and control everything in an experiment Maciej Marosz, that setup is disgusting at all levels.

Edited by Orac (11/13/13 11:40 AM)
I believe in "Evil, Bad, Ungodly fantasy science and maths", so I am undoubtedly wrong to you.