While the D3100 is an entry level camera its low light high ISO performance is commendable While not quite in the same league as the "big guns" it comes close enough to be useful.
DxO Labs publish their DxO Mark high ISO sensor performance comparison chart which states the highest ISO that a camera can be set to while still achieving certain benchmark performances. These are based on several factors which they feel are necessary to produce an acceptable quality photo. Their publsied figures are "normalised" to what they would be if the camera had a 12 megapixel sensor. this is done by multiplying their measure ISO result by square_root (megapixels/12).
You don't have to understand this - you just read the table.
I personally have some reservations about how well the table reflects reality in some cases, but it is good enough as a guide.
The table below is derived from their chart. I have added the D3100 in red. You will see that it performs slightly less well than anything in the table, but not vastly so.
The D3100's 919 ISO is so close to the 1182 of the D5100 that you would not notice in practice im most situations.
A better comparison is the Canon 5D MkII with an ISO score of 1815. This is about twice the D3100's 919. This 2:1 ratio corresponds to one aperture stop or a doubling in shutter speed. It says that if you scaled both pictures to 12 mp then if a D3100 could take a photo at say ISO 1000, f/5.6, 1/100 second and achieve a certain quality, then a 5D MkII would allow you to change any one setting to eg ISO 2000 OR 1/200 th second or f/8. The difference is useful but not so vast as tp be utterly stunning.
It is my strong personal impression that this assessment is a bit generous to lower ranked cameras and that there is liable to be somewhat more difference than it shows BUR even if there was another factor of 2 involved, the D3100 is close enough to the 5F MKII to be useful.
With a D3100, using a somewhat larger aperture than you otherwise would will help lift available low ISO performance. Affording such lenses is another matter :-).
More later maybe ...
The left hand column is the ISO score as DxO publish it. The column "RAw ISO" is my "denormalisation" by a factor of sqrt(12/mp) eg a 36 mp D800 gets its DxO score REDUCED by a factor of sqrt(12/36) = 1.73:1. This is the ACTUAL sensor performance. It is relevant if you look at a say 4000 x 3000 block of pixels from 2 sensors but not so relevant if you scale a whole image down and print it as say an A4 print from each camera. The top table is sorted on DxO's adjusted ISO score and the bottom one on the raw ISO score.
For comparison, the absolute king of high ISO "35mm" DSLR performance is the Nikon D3s with ~= 3200 ISO DxO benchmark performance (adjusted or raw).
The D3100 is "worse' by a factor of 3253/919 = 3.5:1. This is a tolerably small difference bwteen the best of the best and an entry level camera. In practice I feel that the DxO result rather understates the difference but, eve then, the camera is an acceptable if not stunning low light performer.
