The lower the resolution (or dispersion), the fainter objects can still spectroscopied. Low Resolution R < 1000 is likely to determine the spectral class of stars. It maps the entire optical spectrum at a time on the CCD chip. This resolution range is also suitable for flat objects such as novae, comets, galaxies, etc. (for flat objects the slit spectrograph is best).

Medium resolution at 1000 < R < 10,000 already shows many details of the optical spectra. However, it will cover only a relatively narrow wavelength range around 200 to 500 angstroms at a defined grating position, so that the collection of overview spectra requires several recordings at different grating positions.

To study line profiles and their temporal variability may be even higher resolutions envisaged. 10,000 < R < 50,000 are appropriate here. The Lhires III finds its limit at about R = 20,000 (but with narrow slit width in the red I got up to 25,000).

High-resolution spectra with R > 100,000 will be required if one wants to recognize the hyperfine structure of lines (for example, the Zeeman splitting), by magnetic fields. At this high dispersion is needed very much light, so that for accessible amateur telescope apertures probably "only" remains the sun as an object of study.

The graph shows a recording of the Wolf-Rayet star WR140 (6.7 mag) with a dispersion of 1.47 angstrom / Pix . The red peaks represent the corresponding calibration light of the LHIRES III built-in Neonlamp. The Natriumdublett (D1 and D2) is resolved. The spectrum of this star shows broad issues and hardly any fine structure, so a higher resolution is not necessary.

Spectrograph, with or without a slit ?

Flat objects such as nebula, planets, galaxies can best spectroscopied with slit spectrographs. The use of a resolution defining slit (the slit width is smaller as the diameter of the seeing disc of the object in the focus of the telescope) in principle leads to loss of light.

Slitless spectrographs with a high aperture ratio and a short focal length of the telescope (eg 1000 mm focal length, f / 4) concentrate much light in the focussed star picture so even fainter point-like objects (stars) can be observed. Even with slitless spectrographs can be reached resolution R = 20,000 if a high dispersive grating is used (2400 lines / mm).

The picture shows the spectrum of the Orion Nebula, taken with a slitless spectrograph with 1.5 angstrom spectral resolution. The three horizontal stripes spectra belong to 3 stars. The nebula itself shines in different sharp emission lines, which is why it is also depicted several times in the light of these lines. From such images one can measure the different intensities of emissions from different atoms / ions in the different regions of the nebula. Recordings of this kind are only possible in strong emission lines (emission nebula). Flat objects with a smooth continuum (reflecting nebula, galaxies, comets) are not accessible by slitless spectrographs.

Personal preferences and restrictions of the private environment.

• Who does not want to stay up late at night, the sun should be selected as an observation object.
• Who can afford only irregular periods of observation time, should choose objects that are bright and if possible the whole year can be seen through (circumpolar objects). It is sometimes depressing, if in our region because of the frequent overcast sky one cannot measure for weeks. And just at the beginning of a fine weather phase, the object is no longer at the seasonal visible sky.
• The observation of temporal variations (spectroscopic binary stars, variable stars with emission lines, such as Be stars or pulsating stars) assume that you can measure often. The apparatus should be always installed and ready to measure anytime. Observing systems, which first must be established, orientated and collimated in this respect bring little fun.

The graph shows up as dots results of radial velocity measurements of the famous variable star Mizar A, plotted over days. Each pair of points (blue and red) stands for an observation. You can see the time gaps caused by cloudy skies, but also the distinct fine weather phases on days 30 to 50 (early 2008). The lines represent the calculated Doppler split.

Emission line stars
(explorable slitless and with slit)

These include the well-known Be stars but also massive stars, supergiants and very (O) hot stars showing emission lines.

The Be stars have a circumequatorial disk of gas with diameters several times the star's radius r (typically 10 to 30 r). It is about 10,000 K hot plasma consisting of H and He (and some metals), which is held in a ionized state by the UV light of the hot and very bright star. The recombination of protons and electrons produces excited H atoms (HI). They emit photons, for example the visible lines of the Balmer series. The gas in the disk comes from the star itself, but can also be transmitted from the outside from the filled Roche-lobe of an companion (eg bet Lyr).

Massive hot stars such as WR or O stars emit from regions not concentrated in a rotating disk (disk), but spreaded surround the star in more spherical regions (star winds, lobes).

The emission phenomena are often time-variable, if the emitting regions are subject to a development. This leads to quite complex line profiles and temporal variability (LPV = Linien Profile Variability). The study of time trends is therefore a valuable tool with which the theories of stellar evolution and the Be phenomenon can be improved. Here, the amateur can make significant and important contributions (see the website of the FG spectroscopy, topic publications).

n.	The graphic shows the H alpha line of gam Cas in emission. Gam Cas in a variable Be star, who also has two companions (triple star).
.	This is the H beta line of the same star, also in emission (not calibrated).
	The HeI 5876 line also in emission (not calibrated).

For the emission lines of gam Cas is seen that the profil type of the lines is equal. The line is shaped in the emission as a double peak, with approximately equal intensity of the two line components (V / R = 1). But it is also striking that the splitting caused by the Doppler effect increases in the order of H alpha < beta H < He5876 (all three pictured spectra have similar dispersion. The splits are therefore directly comparable). This means that the Kepler rotational speed of the disc in the regions where the lines are emitted, increases in that order. This is indeed the case, the H alpha line is emitted far out in the disk, where the lowest Kepler velocity exists in the disk. The H beta line is formed further inside and the HeI line is emitted inside, near the star, where the differential rotation speed of the disc, according to Kepler's law is greatest.

Orbital elements of spectroscopic binaries (SB2)
(explorable slitless and with slit)

SB2 type binary stars show periodic line splitting due to the Doppler effect of roughly the same bright stars, one of which approach us and the other goes away. In SB1 binary stars is the apparent brightness of stars different (> 3 mag), so we only see the lines of the brighter star, the lines of the fainter star are hidden. The lines move back and forth like in the case of SB2 also, however, can only grasp this shift in line with the slit spectrograph (absolute calibration with a calibration light required).

When using a slitless spectrograph to observe the splitting in SB2 binaries only a part of the orbital elements (orbit parameters) are accessible, but this with high accuracy (no calibration error does go in). For the theory and evaluation methods, see my lecture, Part II, from 22nd page. If you want to deal seriously and intensively with the issue, please contact me. I can give you evaluation methods and literature on this topic.