Jump to content
UltravioletPhotography

The Electromagnetic Spectrum


Stefano

Recommended Posts

Hi,

I didn't know were to post this, hope this is the most appropriate section. Otherwise a relocation would be a good idea.

Anyway, nothing "serious" here, just some images made by me with Paint. It took me 3 days to make them.

 

If this image doesn't make you realise how LITTLE we see, I don't know what else will do.

post-284-0-20782100-1579552029.jpg

 

A version with some information

post-284-0-01586700-1579552169.jpg

 

LED "regions"

post-284-0-32581700-1579552234.jpg

red region = 210-265 nm

yellow region = 265-365 nm and 1050-1650 nm

green region = 365-1050 nm

 

NOTES:

  1. Some borders are not well-defined. For example, X-Rays and Gamma rays overlap by some definitions.
  2. Everything has been calculated. All the lines have been calculated using log base 10. Nothing has been done "by hand".
  3. In the visible spectrum, the violet (a bit more than 400 nm), cyan (500 nm) and orange (600 nm) lines colors have been "found" by observing a ~405 nm, 503 nm and 600 nm LED. Then I adjusted the brightness (eyeballing it this time) to make a "gaussian" brightness rainbow, to simulate the human eye sensitivity curve.
  4. I had to design every letter, made with circle quarters and straight lines. I didn't draw them "by hand", again.
  5. While I included very long-wave radio waves, I didn't include Very-high-energy and Ultra-high-energy gamma rays.
  6. Wikipedia's definition of SWIR is basically from red (700 nm) to 1400 nm. I instead defined the IR region as follows:
    NIR 700-1100 nm
    SWIR 1100-3000 nm
    MWIR 3000-8000 nm
    LWIR 8000-15000 nm
    FIR 15000 nm-1 mm
  7. For the LED range image, I used Thorlabs LEDs and I didn't include the quasi-CW (quasi-constant-wattage) ones. LEDs can sometimes have efficiencies above 40%, and I think that the most efficient ones are about 60% efficient. Still almost half of the input energy is wasted as heat.

Do you want to use this image for other reasons? Do it if you want.

Link to comment

This is fun. As you have the scale in log, you are crunching it a little, making it harder to see and a little artificial.

Maybe better to keep in base ten, but have linear seperation between the 100s.

As in you have artificially made 100nm to 290nm look huge but 500nm to 700 look tinny. In just one unit that maybe ok for energies, but when you jump to the next 10, you then make lower wider and upper equally squashed.

Using the same gap you have for the first 2 ticks, for each tick step at each energy jump would, make it easier to read.

Yes not valid from energy point of view, but easier to read and would be clear for each band.

 

Ahh read my comment 3 times and I can see how it can be confusing. I think better to have equal spacing for the tick marks, and I know its not energy equal, but would make it easier to read.

Link to comment

It is also interesting to compare the width of light to the width covered by electronics.

In those bands a frequency scale below the wavelengths would add useful understanding.

Link to comment

It is also interesting to compare the width of light to the width covered by electronics.

In those bands a frequency scale below the wavelengths would add useful understanding.

I may try to make another version with the frequencies.
Link to comment
Yeah, I feel like what I really want is an image that gives some idea of what technologies are available to image things in which bands. The exact size of the bands is of secondary interest to me. For example, there is a MWIR camera on eBay right now for $2000, which looked tempting until I saw it requires liquid nitrogen for cooling and the resolution was 64x64 pixels!
Link to comment

Ok not relavent, but I think I figured out how to explain my thought.

Your major ticks are equal spacing, not log spacing. So you have wide range.

Then your minor ticks are log spacing, artificially crunching up the range.

To be correct, all should be log spacing, or all linear spacing. So in energy the gamma is huge and radio is tinny. But you have radio waves appearing as more energy than gamma rays.

So best to use linear spacing for minor ticks as well.

 

 

Link to comment
Uh, I don’t think you can do this in linear. That’s 21 orders of magnitude. Radio would be like the width of a hair and gamma would be most of the spectrum.
Link to comment

Your major ticks are equal spacing, not log spacing. So you have wide range.

Then your minor ticks are log spacing, artificially crunching up the range.

To be correct, all should be log spacing, or all linear spacing.

 

To me the scale looks correct, a base10 log scale all the way, showing different wavelengths.

What am I missing here?

 

A wavelength scale is IMHO most appropriate in this forum as that is the parameter we use for our photography.

We are normally not thinking about frequencies, wave-numbers or energy values that also can be assigned to electromagnetic waves.

(Not really my area. I might use some incorrect phrasing here)

Link to comment

 

 

To me the scale looks correct, a base10 log scale all the way, showing different wavelengths.

What am I missing here?

 

A wavelength scale is IMHO most appropriate in this forum as that is the parameter we use for our photography.

We are normally not thinking about frequencies, wave-numbers or energy values that also can be assigned to electromagnetic waves.

(Not really my area. I might use some incorrect phrasing here)

 

No problem. I think I used incorrect phases throughout.

 

I just would like to see equal spacing of the minor ticks, just like the major marked one.

 

Link to comment

I just would like to see equal spacing of the minor ticks, just like the major marked one.

The major ticks are NOT linearly spaced. They grow (or decrease, depending on the direction you read the graph with) exponentially.

 

I did something like

1, 10, 100, 1000...

 

Are you referring to something like

1, 2, 3, 4...?

 

Do you know that with 21 orders of magnitude, using 1 mm for the smallest interval, you would have a graph 1021 mm long, or 105 light-years? That's why you have no choice other than using something in which the order of magnitude matters, and not the magnitude itself, so a log scale.

 

Although I understand what you mean about making the small ticks linear.

Link to comment
I also want everyone to notice that I was limited by the 1200 pixel maximum image width. If I could do something like 10,000 pixels, I could have had a lot more detail of course.
Link to comment

This is the version with frequencies. I wrote them very small, they may be hard to read, but if you magnify the image you should be able to read them.

post-284-0-48398400-1579615670.jpg

Link to comment

Do you know that with 21 orders of magnitude, using 1 mm for the smallest interval, you would have a graph 1021 mm long, or 105 light-years? That's why you have no choice other than using something in which the order of magnitude matters, and not the magnitude itself, so a log scale.

Yeah, that is what I was trying to get across. I agree, it has to be log. The small ticks ARE linear ticks, plotted on a log scale, which makes the apparent spacing unequal. In fact, equal spacing (like the VHF, HF, MF intervals) corresponds to powers of ten, so not linear. This is how every log graph works.

Link to comment

Why is there nothing west of gamma rays or east of radio waves?

Because nobody named those regions. They are just more radio waves and shorter gamma rays, I think?

Link to comment

Because nobody named those regions. They are just more radio waves and shorter gamma rays, I think?

Yes, they don't have borders on the right side and on the left side respectively.
Link to comment

What I notice is that the visible spectrum is really tiny, it spans only 0.24 orders of magnitude in base 10. And still we see a lot of colors, think about it.

A normal full spectrum camera can see, say, from 350 to 1100 nm (theoretically from 190 to 1100+ nm), and in the first case it means barely half an order of magnitude. A TriWave, which can see from approx. 300 to 1600 nm (and that's a very wide range) spans 0.727 orders of magnitude. There is a LOT to see.

 

The "hard to image" regions of the EM spectrum are:

Long radio waves (not actually hard to detect, but don't think to see humans with 100 meters-long waves).

THz waves (we already talked about this region, which is at the border between antennas and oscillators and actual sensors (such as the microbolometer sensor of a LWIR camera)).

EUV (below about 100 nm and all the way down to X-Rays, everything is opaque, unless you have micrometer or nanometer-thin slices of it).

Gamma rays (gamma cameras are very big, heavy and expensive. They are basically a matrix of holes in a block of lead to allow only parallel photons to pass through. Forget lenses and mirrors there. And, most importantly, gamma rays are dangerous and very penetrative. They turn things radioactive (something X-Rays will NOT do). If you have a radioisotope source (such as Cobalt-60) you can not turn it off, and you need centimeters or even more of lead to shield yourself).

Link to comment

 

Yeah, that is what I was trying to get across. I agree, it has to be log. The small ticks ARE linear ticks, plotted on a log scale, which makes the apparent spacing unequal. In fact, equal spacing (like the VHF, HF, MF intervals) corresponds to powers of ten, so not linear. This is how every log graph works.

Yes, in a log scale, things that grow exponentially become linear, and things that grow linearly become "squished" (logarithmic).
Link to comment
Bill De Jager

That's a really nice effort, Stefano. I appreciate your labor on this project. Thanks for sharing it here.

 

They turn things radioactive (something X-Rays will NOT do).

 

Gamma rays won't cause any significant radioactivity in nuclei. Neutron radiation is the culprit there.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...