Six cards here call themselves a headset, and five of the six state a microphone. The word keeps its promise, but it doesn't work backwards: eighteen of the forty-eight called headphones have one too.
Comparisons
A comparison is only worth reading when both sides are measured the same way. These pages put two models next to each other on the figures both makers publish, under the conditions each maker states, and stop where the shared ground stops.
That last part does most of the work. Sensitivity quoted in dBV/Pa cannot be set beside one quoted in mV/Pa until it is converted; a noise figure with no weighting curve named cannot be ranked against an A-weighted one at all; peak power and continuous power are not the same quantity. Where a number exists for one model and not the other, the page says which, instead of quietly filling the gap.
Each comparison also states who the two models are actually for, because the more useful answer is usually not "this one wins" but "these are aimed at different jobs, and here is the job each was built for".
A microphone can be cardioid and condenser at once, and seventeen listings here are. The pattern row predicts nothing about the transducer, while the transducer row predicts whether the pattern can change at all.
A hundred dollars separates the two Beats headband lines at the dealer. Two rows change with the price and four move against it, while the driver diameter stays the same on both cards.
Of 82 microphone listings, twelve can do omnidirectional and ten of those also do cardioid. Three publish the numbers for each pattern, and all three agree on what the switch costs.
Cardioid is the default: 41 of 82 cards say it, 12 say supercardioid, and three say hypercardioid. Eleven publish a number for any of it.
Same $109 and the same −56.0 dBV/Pa. The differences that matter: 10 Hz at the bottom, rejection figures only one of them publishes, and two cartridge part numbers.
Twenty-seven decibels and $298 apart, and RODE sells four microphones called NT1. What each sheet publishes, and which room decides the answer.
$140 less, 36 dB of built-in gain and 4 dB more on XLR against the lowest-output dynamic here. Which of the two answers which desk.
$70 buys 4.5 dB and a supercardioid pattern, and costs the published rejection figures. Which of the two answers which stage.
Cheaper, 4.6 dB hotter, wider at both ends, and the owners still prefer the other one. What both sheets publish, and what one of them does not.
Ten dollars and 2.5 decibels apart, and the RE20 keeps its sensitivity in a PDF the shop never reads. What both sheets publish, and which desk each suits.
Four dollars apart, and the V7 wins almost every published row. What it leaves out, and why the pattern decides more than the 2 dB does.