One commonly held but false assumption, is that “...¹³C NMR takes a bucket of material...”.    Many texts on the subject use examples from the early days in NMR, when sample sizes were measured in grams.  A ten milligram sample would have been thought near impossible.   Now-a-days, it is normal if not generous.    

Here are a couple of salient aspects to this story.  First, we acknowledge the fact that carbon NMR will never compete with proton sensitivity and therefore we adjust our expectations. 

The 10 mg/ml spectrum shown above shows an average of about 20:1 signal to noise across the spectrum (not including the solvent).  And yet, it has enough information content to identify the compound.  Carbon has a 200ppm chemical shift range.  An individual carbon peak is 10⁴ times narrower.  Its spectrum often resembles a series of spikes on a baseline of noise.  Each peak belonging to just one unique carbon atom positioned in the molecule. 

It’s worth noting that the reason for this exquisite simplicity is the same reason for its low sensitivity.  If it were ~100% abundant like 1H, then its spectrum would resemble the complex pattern of splitting’s proton NMR is known for.