A singer with a guitar is the simplest recording there is, which is exactly why so many of them sound worse than the performance deserved. Two sources, two mics, one small room, and every decision you make in the first ten minutes is baked into the files forever. Here is how to make those decisions so the mix has something to work with.
You need a two-input interface, two mics and two stands. Nothing more.
First decision: together or one after the other
Recording guitar and voice at the same time gives you the performance. The push and pull between strumming and singing is the whole point of the genre, and it never quite survives being split into two passes. The price is bleed: the guitar ends up in the vocal mic and the voice ends up in the guitar mic. That bleed cannot be removed later, and it means every mix move on one file drags a bit of the other along.
Overdubbing, guitar first and voice second, gives you clean files and total control. The price is that some players cannot play the same way without singing, and the timing goes stiff.
If you can play the guitar part convincingly on its own, overdub. If the song only works when you sing and play together, record live and manage the bleed with placement. Both ways produce two separate files, which is what matters for mixing. One mic in the middle of the room capturing both at once produces a single file that cannot be mixed at all, for the reasons in Can You Mix a Single Stereo File?
Mic placement that keeps the two apart
For the voice, a dynamic mic, close. Ten centimeters, a fist between lips and grille, pop filter on. A dynamic mic hears what is in front of it and little else, and at that distance the voice is so much louder than the guitar that the bleed drops to a level the mix can live with. A condenser on the voice in an untreated room hears the guitar, the room, and the fridge. How to Record Vocals in a Rehearsal Room applies word for word.
For the guitar, a small-diaphragm condenser if you own one, otherwise the SM57. Point it at the spot where the neck meets the body, around the twelfth to fourteenth fret, from about 20 to 30 centimeters away. Not at the sound hole, which gives you boom and nothing else. Angle the mic so its back faces the singer's mouth: most mics reject sound from behind, and that rejection is free bleed control.
Then apply the three-to-one rule. Whatever the distance from each mic to its own source, the distance from that mic to the other source should be at least three times as much. Vocal mic at 10 centimeters from the mouth means the guitar should be at least 30 centimeters from the vocal mic. In practice that happens naturally if the vocal mic sits high and the guitar mic sits low, both angled away from each other. The rule is not about volume, it is about phase: when both mics hear both sources at similar levels, the two files partially cancel when summed and the result sounds thin and hollow.
Gain: set each input so the loudest moment peaks around -12 dB. A strummed chorus is louder than you think.
A pickup is not a third option, usually
Many acoustic guitars have a pickup, and plugging it into the second input removes bleed entirely. It also removes the guitar. Under-saddle pickups produce the plasticky quack that every listener recognizes as "live from an open mic". If the pickup is all you have, use it, and the mix will tame the harshness, but a 100-franc dynamic mic on the body will beat it every time.
Export and label
Two files, WAV, both starting at zero, both the full length of the song. No reverb, no compression, no EQ printed into either one. Do not normalize them and do not try to balance them against each other; that is the mix's job.
In Bandmixr, label the files Lead Vocal and Guitar Acoustic, and pick Acoustic/Folk as the genre. That profile is built for exactly this: it keeps the dynamics of the performance instead of flattening them, uses gentle subtractive EQ rather than hyped brightness, and puts the voice in a natural room instead of a pop hall. The engine also checks the two files against each other for the polarity problems that bad placement causes, which is your safety net if the three-to-one rule slipped.
If you added a second guitar pass, a shaker or a harmony, each gets its own file and label. How Many Stems for Mixing? has the grouping rules.
What the mix will fix
It will place the voice on top of the guitar, tame the boom around 150 to 250 Hz where every acoustic guitar builds up, cut the boxy room ring in the vocal, control the difference between the quiet verse and the loud chorus without squashing it, and master to a streaming level that does not fight the dynamics you kept.
Upload the two files to bandmixr.com, choose Acoustic/Folk, and listen to the first mix against your rough bounce at the same volume. The first download is free.