Classifying light by photo statistics - poissonian, sub, super poissonian?

I've got three types of light:
- star light
- light from a discharge lamp
- light from a laser emitted high above threshold
and I need to decide on which type of counting statistics are expected for the detection of each source: poissonian, super poissonian or sub-poissonian. However, I'm unsure about how to do this.

The definitions of the counting statistics are:
$$\Delta n > \sqrt{\bar{n}} - superpoissonian$$$$\Delta n = \sqrt{\bar{n}} - poissonian$$$$\Delta n < \sqrt{\bar{n}} - sub-poissonian$$where the variance is on the left hand side, and the root of the average count is on the right hand side.

So, with star light, we've got light of more than one wavelength, and fluctuating intensity. I'm guessing that sub-possonian can immediately be ruled out. That leaves poissonian and super poissonian, but without actual numbers I can't do calculations using the above equations, so I'm a bit lost as to classify this. I'm leaning towards superpoissonian because it just seems like a very random source of light, but I'm really not certain.

The same things apply to the discharge lamp - I've read in a book that it's apparently super-poissonian, but again I'm not sure where to draw the line between poissonian and superpoissonian.

As for the laser light, we've got light of a single wavelength, emitted as a coherent beam. However, I'm thinking that sub-poissonian light has to be emitted at a constant intensity down to very short timescales, so wouldn't this depend on the type of laser used?
ie. am I right in thinking that a gas laser will have emission at fairly random times compared to, say, a semiconductor laser diode?

How do I classify these things?

PS. exam tomorrow, so I'm grateful for fast replies.