Point the microphone at the ceiling. Not at the speakers.

Ceiling, ear height, on a stand

Denon prints the same sentence in every model manual: point the tip of the microphone toward the ceiling, and adjust the height to match the height of the ears of a listener in a seated position. It reads oddly the first time because the natural thing is to aim a microphone at what you are measuring. This is an omnidirectional measurement mic whose calibration assumes it is vertical, so tilting it toward the front speakers skews the result. Use a tripod or a stand, get the tip to the height your ears sit at on your own sofa, and keep hands off it while it runs. A microphone held in a hand is also measuring the person holding it.

Take the subwoofer’s own controls out of the loop first

This is the single most skipped preparation step and it happens on the back of the subwoofer rather than in the receiver menu. Denon asks that a capable subwoofer be set to direct mode on, with its volume adjustment and crossover frequency setting disabled, before the measurement runs. The reason is that the receiver is going to set a crossover itself, and if the subwoofer is also applying one you get two filters in series and a hole in the response near the handover. Do it before you start rather than after, because changing subwoofer volume afterward is one of the things that invalidates the whole calibration.

The subwoofer distance will look wrong. Leave it alone

This one costs people the timing of their entire system. After the run, the receiver may report the subwoofer as further away than it physically is, and Denon explains why: subwoofers add electrical delay, and the measurement correctly reports the acoustic result rather than the tape measure result. The reported figure is what makes the sub arrive in time with everything else. Owners see a number that disagrees with the room, assume the calibration made a mistake, type in the physical distance, and undo exactly the correction they paid for. If the number looks too big, that is the system working.

Measure where people sit, and nowhere else

Most AVR-S models take up to six positions and AVR-X models take up to eight, and the temptation is to spread them around the room to cover everything. That averages the room rather than the seats. Keep every position inside the area where people actually listen, do not measure in corners or hard against a wall, and do not push positions far apart. The first position is the primary seat and carries the most weight, so put it exactly where the main listener’s head goes. If you have one sofa, all your positions live within that sofa.

Set the layout before you measure, not after

Amp Assign and Speaker Layout tell the receiver what system it is looking at, and they belong before the measurement. Channels you set as unused are skipped, which makes the run shorter. If you have Dolby Atmos enabled speakers, the setup asks for the distance from the speaker up to the ceiling, and that number should be measured rather than estimated, because it is what the height processing works from. Getting the layout right first also prevents the more expensive version of this mistake, which is adding a speaker afterward.

The calibration is invalidated by exactly the tinkering that follows it

Denon is direct about this: do not change the speaker connection or the subwoofer volume after Audyssey Setup, and if they change, run it again. Separately, do not enable new speakers in the speaker configuration menu after the setup has run. This is worth understanding rather than just obeying, because the calibration is a description of one specific arrangement of one specific room. Move a surround speaker six feet, add a pair of heights, swap the sub, or nudge the sub gain because a film felt light, and the stored correction is now describing a system that no longer exists. Re-running takes half an hour and is the honest fix.

Let it finish, and keep the room quiet

Analysis takes several minutes and gets longer with more speakers and more positions, and Denon states plainly that the power must not be turned off while results are being saved. Plan for the run to be uninterrupted. The measurement itself needs a quiet room, which in a real house means the dishwasher off, the HVAC off if you can, and everybody else out for twenty minutes. Ambient noise is measured as part of the room whether you meant it or not.

Read the results, because they catch wiring mistakes

Audyssey measures polarity along with level and distance, which means it will flag a speaker connected out of phase, and that is a wiring error the ear tends to miss on its own. It is the best free check you will get on the work that just happened behind the walls and under the carpet. Look at the detected distances and confirm they roughly match the room, allowing for the subwoofer exception above, and look at the crossover values. Anything wildly out is usually a cable at one end, not a measurement problem.