Is it possible to control frequency of the light? Generating light is the process of energy conversion. I mean is it possible to control the frequency of light by controlling directly the input like heat, current... not by filters or medium arrangments.
I thought (thanks to a discussion in other media) for such a control technique we should control the movement of electrons between the energy levels. If I am right, is it possible to do that?
 A: The thing you're looking for is a tunable dye laser. It works just like a radio transmitter: a broadband amplifier amplifies its own output which it receives via a resonator. Instead of a transistor powered by a DC source, the amplifier is a dye cell "pumped" by another laser.
The tuning is indeed accomplished by adjusting the resonator, but that's just like a radio frequency "variable frequency oscillator".
A: Light is typically taken to mean the visible part of the electromagnetic spectrum. But you can extend it - infrared light, ultraviolet light. Or any electromagnetic wave.
At visible wavelengths, it is often created by atomic transitions. The energy levels are largely fixed. You get lines in the spectrum. However, electron spin and orbital angular momentum interact with magnetic fields. A magnetic field can change the energy of the electron and shift or split the line.
There are other ways to get visible light. As mentioned heating an object makes it emit light.
At lower frequencies, it is easier to control more directly. Radio waves are created by pushing charges back and forth in a wire at megahertz frequencies. Vary the frequency of the current and you vary the frequency of the radio waves.
A: It depends on the range you are looking for: a tunable frequency in the MHz-GHz range is easily achievable with diode lasers. In the simplest case, the applied current and environment temperature have a direct effect on the supported wavelength.
For the temperature dependence, there is also a useful visualization. The higher the temperature of your gain medium is, the more it expands, so it gets longer. This means, that wavelengths of light supported by this medium also increase, so the frequency decreases.
If you have a grating for further filtering the output light, this also gives you further tunable range by adjusting the distance between the two grated mirrors. These elements can also be used to stabilize the wavelength, which is important for a variety of experiments.
Larger spans are more difficult, because at some point you see that the laser hops onto a different mode, changing the mean frequency by tens to hundreds of GHz (this cannot easily be controlled).
A: A Free Electron Laser can be directly tuned by varying the speed of the electrons.
A: 
control the movement of electrons between the energy levels

One example is "magnetic field laser". Shift the levels by magnetic field and enjoy variable frequency.
http://www.cchem.berkeley.edu/rjsgrp/publications/papers/1980-1983/8_evenson.pdf
A: Black-body radiation is an emission of electromagnetic radiation which takes a characteristic spectrum dependent only on the temperature of an object. You can directly estimate the temperature of a hot object from the color of its glow, and you can change the color of an object's glow by changing its temperature. As an object heats to about 800K, it begins to emit a dull red glow, and as the object continues to get hotter, the glow becomes yellow and then bluish white. This is one example of how it's possible to control the frequency of light - by controlling the temperature of a black body.
A: Radiation is the field quanta (or force carrier) for the electromagnetic interaction.  Its absorption causes an electron to increase in energy level (causing in change in at least one of the quantum numbers associative with the absorbing electron).  The reverse process, emission, involves the production of a photon
$$\Delta E = \frac{h c}{\lambda} $$
where
$h$ is Planck's constant $c$ is the speed of light $\Delta E = E_2-E_1$ and $\lambda$ is the wavelength of the photon.  The most complicated component is determining $E_2$ followed by $E_1$ as these involve solving (or approximating a solution to) Schrödinger's Equation, which is determined by the position of the nuclei and the composition of the perspective material.  Given that the position determines these one could, in the correct material, compress it to alter the energy levels (which are unlikely to be uniformly modified) resulting in a different $\Delta E$.
