8
\$\begingroup\$

Comments like this one got me wondering...

Assuming two light sources of equal size and relative intensity (let's say as an example monolights and the sun) and assuming that you're shooting raw to color balance, what's the difference between them, as far as quality of light? Is it only spectrum? Why would one produce a significantly different quality light than the other?

(I'm not interesting in the fact that one is easier to use or more flexible or always on or such - I'm more concerned with the quality.)

\$\endgroup\$
2
  • 1
    \$\begingroup\$ I think a lot of it has to do with even, diffuse light coming from everywhere. I bet if you could rig up a 100'x100' LED softbox shown through a 10'x10' window placed like 15' from the source it'd look similar ;) \$\endgroup\$
    – Shizam
    Jul 20, 2012 at 2:36
  • 1
    \$\begingroup\$ You might get a different perspective if you asked this on the physics site. \$\endgroup\$
    – ChrisF
    Jul 20, 2012 at 11:22

5 Answers 5

6
\$\begingroup\$

"Only spectrum" is a very major factor.

The following only simplistically 'scratches the surface' of an immensely complex subject area:

"Colour temperature" is a measure of the "warmness" of a white light source - this is a subject that rapidly descends in black (or white) magic and needs not be discussed here except as a means of comparing while light components. The colour temperature is the temperature that a black body radiator would need to be heated to to produce white light of the same equivalent "warmth".

Solar light is relatively continuously distributed in light frequencies.

Light sources such as a tungsten or halogen bulb which used a heated metal to produce light have a relatively continuous spectrum across a limited range of frequencies. The tungsten peak is centred around longer wavelengths / lower frequencies than the daylight distribution and is yellower and of a lower effective colour temperature.

Artificial sources which excite phosphors with one wavelength of light to cause them to emit light at other wavelengths, produce light in a number of relatively sharp frequency peaks with gaps between with less or no light. These peaks of wavelength are arranged such that the eye / brain system combines them to produce "white" light. While the eye may see white, the discontinuous spectrum produces photographic effects which are different than continuous spectrum natural light.

This method applies to fluorescent lights, CFL (compact fluroescent), & Phosphor LEDs. Similar results occur when a gas is exceited electrically or thermally so it emits light with sharply defined frequenices or when multiple mono-coloured LEDs are used. The resultant "White" is a phantasm of the brain. Source - CCA/SA. The curved solid line is the "Plankian locus" and is the colour that a heated black body would follow as temperature rose. Numbers 1500-10000 are the temperatures in Kelvin causing the associated colour. THe eye and brain see colours on this line as versions of "white" . The numbers around the outside of the coloured area are the wavelengths in nanometres of monochrome light at that point. Take any two points on the boundary, mix light using these two colours and alter the relative amplitudes and the effective colour will move along a line between the two. (It is not, alas, just a straight line drawn on this chart). Do this with 3 border co0lours and you can make colours which lie ~ inside the triangle formed by the 3 colours. BUT while you MAY be abale to make the eye/brain think you have a light of one colour, or a wide range of colours, a sensor system of film or filters or ... may react differently.

enter image description here

Modern white 'phosphor LEDs'typically use a short wavelength blue LED and a yellow phosphor. Some of the blue light is converted to yellow by "exciting" the phosphor so that it re-emits the energy as syellow light. The relative mix of blue and yellow and the exact emitted frequency ranges are varies to produc elight ranging from "warm white" (around 2500 - 3500 Kelvin effective colour temperature) up through daylight like whites in the 4000K - 7000k range and then to distnctly blue whites up to about 10,000 K equivalent. At or above about 10,000K the "white light" appears very blue. The yellow/blue mix is adjusted so that the vector sum lies on a spectrum line which true black body radiator colours track along so that the light "looks" white, within limits.

eg When you have continuous spectrum light you can apply filtering at any wavelength to remove or alter part of the light to change the overall mix. When you have a few finite peaks you may not have any light at the frequency range of the filter that worked OK with natural light. The results MAY be very substantially different.

eg A photo sensor may react in a certain way to natural light with a wide range of frequencies present. Artificial lighgt with the same apparent colour temperature to the eye will present the sensor

eg If you have an eg Sodium Lamp as are found on some highways with a vey orange light, you have a few closely spaced orange emission lines and nothing else. No amount of filtering will 'correct' this to look like natural light. Whilke that is obviously ext=reme it is just an extreme case of what is happening with the limited wavelength output sources sources mentioned above. Source CCA/SA

enter image description here

\$\endgroup\$
1
  • \$\begingroup\$ Great answer in regards to the spectrum - much more like what I was looking for. Any factors other than spectrum though? \$\endgroup\$
    – rfusca
    Jul 20, 2012 at 17:26
8
\$\begingroup\$

There are several kinds of artificial lights - fluorescent, tungsten, LED, halogen, xenon, explosives, electric arc etc. And there are also several different kinds of natural lights - sunlight, moonlight (sunlight reflecting from the Moon), light from other stars, fire, lightning, volcanoes, aurora borealis, glowworms etc. Obviously, both classes contain very different light sources and any differences between such wide classes can only be found when overgeneralizing the classes to a couple of most common examples of both (e.g. xenon speedlight vs sunlight).

Most natural light sources are remarkably further away than reach of artificial light sources, therefore the drop in intensity (falloff) of artificial light is faster, since the light source is so much closer. Therefore, the area you could lit with a single artificial light is much smaller. Try lighting up a landscape, or the sky with a monolight :)

The most common forms of natural light - sunlight and moonlight - are always on, while artificial light sources most commonly used in photography are synchronized to switch on during exposure. Thus natural light provides for easier modeling of the lighting, and your camera's max sync speed is irrelevant, and there won't be any blinking to flash.

Dispersion of sunlight over the sky implies that the shadows cast by sun are not pitch black, but filled with a bluish tint.

Since artificial lights can be moved easily, you can easily create lighting schemes that would be impossible with natural light only (you might have some luck with directing fire or glowworms, not so much with others).

Finally, a few words on "quality" in business context (superiority), as opposed to philosophical context (property or attribute).

Here, artificial light thrives on

  • availability (you can bring it any time, day or night);
  • repeatability (you can get same lighting by using same setup again; the sun and moon are moving, weather might change);
  • reliability (weather has much less effect on artificial light because there's much less of it between the light source and scene; with artificial light, uncharged batteries are your fault, not the lighting's).

Note that for artistic results, the unpredictability of natural light might be preferable.

Natural light easily beats artificial on

  • expected lifetime;
  • initial cost;
  • running costs.

There's no difference in signal-to-noise ratio if the illumination level of subject is the same. Sunlight (especially non-diffused) will provide stronger illumination than most artificial lights, and therefore better signal-to-noise ratio; other natural lights are weaker than a flash close to subject.

\$\endgroup\$
8
  • 2
    \$\begingroup\$ Just one minor clarification: diffused window/skylight lighting (that is, a window or skylight with a diffuser placed over it, as opposed to diffuse lighting coming through a window) will have the same falloff as, say, a softbox of the same size at the same distance, since the diffuser becomes the light source. \$\endgroup\$
    – user2719
    Jul 20, 2012 at 15:49
  • \$\begingroup\$ This really doesn't answer what I want...it talks about the differences in terms of ease of use and such, but the question is regarding the quality of the light. I edited the question to highlight that fact (although it was already there originaly). \$\endgroup\$
    – rfusca
    Jul 20, 2012 at 16:02
  • 1
    \$\begingroup\$ I'd have thought that falloff, lit area, or ability to use any shutter speed are quite important qualities for a photographer. If these were strictly quantity issues, you'd be able to compensate with a different ISO. So it might not be what you want, but it does answer what you asked. \$\endgroup\$
    – Imre
    Jul 20, 2012 at 16:12
  • \$\begingroup\$ Maybe there's confusion here, because your edit only seemingly made it less ontopic. I'm not talking about what advantages or disadvantages natural light has in general for photographers...I'm asking VERY SPECIFICALLY what influences the quality of the light in a picture. Being always on or being able to move the light around, has ZERO effect on the quality of the light. \$\endgroup\$
    – rfusca
    Jul 20, 2012 at 18:48
  • \$\begingroup\$ If it's none of those things... then what do you mean under word quality? \$\endgroup\$
    – Imre
    Jul 20, 2012 at 20:52
6
\$\begingroup\$

Strictly speaking, if you could really emulate everything about natural lighting with an artificial light, they'd be exactly the same. Since we lack an artificial light source of the same intensity of the sun, not to mention radio triggers with a 93 million mile range, the best we can do with artificial lighting is to simulate sunlight.

By locating an artificial light source closer to our subject (by a few million miles), we can produce a similar intensity of illumination on the subject, but it's pretty tough to replicate the diffusion caused by all those miles of space dust and atmosphere between us and the sun, among other things. You touched on spectrum, too, which again, I think we can emulate, but is really difficult to duplicate.

One of the challenging aspects of duplicating natural light, of course, is that natural light is changing all the time. Given all the factors that can flavor natural sunlight, it's really an almost infinite number of different light sources, isn't it? If you shoot near sunrise or sunset, this is clearly apparent as your exposure changes from one shot to the next. I'd expect that despite the occasional happy surprise when natural light does something unexpected that we like, that variable nature is actually one area where artificial lighting improves on natural lighting.

\$\endgroup\$
2
  • \$\begingroup\$ For an exact emulation, you wouldn't need the radio triggers - the sun is on all the time, thus available at any shutter speed. \$\endgroup\$
    – Imre
    Jul 20, 2012 at 5:01
  • \$\begingroup\$ Right, but the artificial light you're using to emulate the sun wouldn't be. It's ok -- it was meant in a light-hearted fashion. \$\endgroup\$
    – D. Lambert
    Jul 20, 2012 at 13:01
5
\$\begingroup\$

From a practical perspective, what mainly characterises the sun during midday, other than its spectrum, is the fact that it is an omnidirectional light that is shining from above, very bright (luminosity of 3.84 × 1026 W), and has a small angular diameter, 0.53 degrees, resulting in rays that are close to parallel. An artificial source with the same illuminance and angular diameter would nearly reproduce many of the lighting effects you see with the sun provided it is far enough away from the scene, namely: sharp, very dark shadows, and fill lighting reflected off of nearby objects (which is typically diffuse, but may not be if they are mirror-like - it also tend to acquire those objects' colors).

For a light source at distance d to have the same angular diameter and illuminance as the sun, it must have an actual diameter of about d/108 and a luminosity of 17200d2 W. So if your light source is 1m away, it must have a diameter of 9mm and a luminosity of 17.2 kW. If it is 10m away, it must have a diameter of 9cm and 1.72 megawatts, while if it is 100m away, it must have a diameter of 93cm and a luminosity of 172 megawatts.

For comparison, a typical high-end strobe studio flash goes up to 1000 wattseconds, which at a typical max speed of 1/1500th second gives you 1.5 megawatts. By locating such a flash at a distance of about 9.3m above the scene, you can get a similar effect to the sun as long as its diameter does not exceed 8.6cm, which is plausible. However such a setup would require a substantial investment.

On-camera flashes on the other hand don't really stand a chance of reproducing sunlike effects - the Nikon SB800 puts out about 60 kW max, assuming no loss from reflectors and diffusers. So it must be located 1.9m away, and have a diameter of 1.8cm, which it doesn't have.

\$\endgroup\$
5
\$\begingroup\$

If you want a simple answer specifically regarding the difference between "artificial" and "natural" light:

The breadth and continuity of the spectrum involved.

Remember your physics classes. The color of objects we see is governed by how much light they absorb and how much they reflect, and the distribution of absorption and reflection across the visible spectrum. A blue object is blue because it absorbs less and reflects more blue light, an orange object is orange because it absorbs less and reflects more orange light, etc. If you illuminate a scene full of blue objects with artificial narrow-spectrum tungsten light, your blue objects are going to appear more dull and less colorful than if they were lit with natural broad-spectrum light.

The more continuous and broad your illuminant is, the greater the color fidelity of your scene will be.


Simple answer over, on to the specifics.

Artificial light sources do not necessarily emit a broad spectrum, and rarely emit a "full" spectrum, nor do they always emit a continuous spectrum for the range they cover. The quality or fidelity of color and detail that we see from a lit subject is very dependent upon the breadth and continuity of the light that illuminates it. Artificial lighting also generally tends to have an unnatural wavelength distribution, in that its spectral curve will usually peak either too warm or too cool relative to sunlight, producing the shifted white balance that requires correction in post. If your working with tungsten (halogen) lighting, your working with a very narrow band of mostly continuous, but very warm light. Some subjects will appear just fine under such lighting with white-balance correction in post, as they primarily respond to more redshifted wavelengths of light. Other subjects, however, may lack detail and color fidelity when lit by tungsten light because they primarily respond to more blueshifted wavelengths of light.

While some forms of artificial light offer a more broad spectrum, there are usually either limitations to their bandwidth, or there may be holes and gaps in the spectrum emitted. Lights based on black-body emission, or in other words light sources that emit light by heating some kind of element (usually meta), will usually provide more continuous spectrum lighting that has more limited bandwidth. Lights based on gas emission, or in other words light sources that emit light by passing an electric current through a gas of some kind, will often provide broad bandwidth but spotty continuity (lots of gaps). Neither form of lighting is perfect, although many specialized types of lights greatly mitigate the negatives while enhancing the positives...such as providing as broad a spectrum as possible with as few gaps as possible.


Natural light, on the other hand, is not just broad spectrum...its "full spectrum", including all wavelengths from radio, through the entirety of the visible spectrum, to EUV and X-Ray. Natural light includes everything in the visible spectrum, so it is broad bandwidth and fully continuous, with an ideal spectral curve that peaks right in the middle of the visible light spectrum (yellow-green green, a band around 555nm).

The benefit of having full spectrum illumination is that the full color fidelity and detail of your subject can be brought out. If you have spotty lighting with gaps and limited spectral bandwidth, and your subjects respond more to wavelengths of light not within the band of primary emission of your artificial lights, you'll have color-anemic results. Thats not to say you can't correct such a problem in post, but it generally won't look as good as when you use broad spectrum or full spectrum lighting. There are artificial lights that emit a broad spectrum, or emit a spectrum of light that is as broad as possible via artificial means, and that replicate the spectral curve of sunlight as closely as possible. With such source lighting, the quality light and shadow in your scene would then boil down to how you configure and arrange your lighting...but that you should have total control over.

\$\endgroup\$
3
  • \$\begingroup\$ Strictly speaking, sunlight isn't continuous-spectrum either (there are those pesky Frauenhofer lines), but that's a minor niggle. A somewhat more important possible deficiency in some artificial light sources is that not all of the colours originating at the subject arise from reflection; some arise as the result of absorption and re-emission at a different wavelength (eg fluorescence, which usually requires a UV-rich impinging light, ruling out most tungsten sources). Natural light ain't perfect, but it's what we're used to and adapted for. \$\endgroup\$
    – user2719
    Jul 21, 2012 at 6:26
  • 1
    \$\begingroup\$ @Stan: True, there is some fluorescence involved. I guess that might be of particular interest for anyone wanting to photograph prints printed on papers with optical brighteners, or (certainly) any subjects that are fluorescent. As for Fraunhofer lines, those are pretty thin, and a minimal degradation of the light we get from the sun. On the other hand, an RGB "white" LED produces light that has three distinct and narrow peaks in the red, blue, and green wavelengths with significant troughs or gaps between them. A tungsten bulb, or even a flash, have single distinct peaks. \$\endgroup\$
    – jrista
    Jul 21, 2012 at 15:42
  • \$\begingroup\$ It should be noted that absorption (Fraunhofer) lines do not actually mean a lack of light at those wavelengths. Absorption lines simply mean that there is a sharp, but relatively minor, dip in the spectral curve of the light emitted by a subject. \$\endgroup\$
    – jrista
    Jul 21, 2012 at 15:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.