Your housemate is wrong. Or, not thinking it through. Or, has ever narrated as a primary income earning job.
That's like saying "Those 20 lines of code you wrote are only like 90 words. You type at 40WPM, so it took you what 2, maybe three minutes to write?" You and I both know that's utterly uncorrect.
Your housemate might be thinking only of the "talking into the microphone part", and assumes that the narrator only makes mistakes on a line a few times in an hour. There are some excellent cold readers out there, but in my experience that's just not the case for the majority.
And for the vast majority of the time in the ACX world, the narrator is also doing your housemate's job.
As someone who's been narrating for about 10 years with hundreds of hours of finished audio under my belt, here's how it breaks down:
- Rehearsal / Prereading. A narrator needs to know what's going to happen so they can properly act a line. So you have to read the text in advance. No one should expect a narrator to read a work of fiction cold and get it perfect on the first try. YOu need to rehearse, or at least get familiar with the text. Let's say that's 30 minutes because you read faster than you narrate. If it's a work of fiction and you've got a number of different characters you need to create and voice consistently it can mean additional prep work. Characters don't just come out, first try, fully realized.
- Actual recording. By definition this cannot be less than one hour, but that assumes an absolutely perfect read for one hour. In my experience it's probably a minimum 1.5x multiplier for retakes, mistakes, line fixes if you do punch and roll. So 1.5 hours. Complex fiction with lots of characters could take a significantly longer. Of course there are amazing narrators out there who can do this, but that is the vast minority.
So, We're up to 2 hours already, minimum.
Now that it's recorded you need to do a QA pass -- you should listen to what's been recorded and make any adjustments to timing, breaths, volume riding, effects, compression, EQ, and to identify and correct errors. You shouldn't ship untested code, you shouldn't ship untested audio.
In many ACX cases, the narrator also performs this job.
If it's absolutely perfect on the first try it's another hour (unless you're QA'ing at a faster than 1.0x reading speed, then you're ignoring all the technical checks, and only validating narration checks). Again, in many ACX cases, the narrator does this job too.
There are almost always pickups (redoing mistakes you find) in an hour of text, for missed words, dropped lines, bad line interpretation, reversed words, etc... The audio should be letter perfect to the text, so even the slightest mistake should be fixed.
Any pickup will take time to get right. If you interpret a line wring it can influence the outcome for the character. You need to get the line reads right. Skilled narrators can punch in a correction perfectly with just a take or two, but some lines can require multiple tries and full minutes to get right. Especially if it's the next day and you need to warm up again so you can sound like you did before, (your voice changes as you narrate) so there is a multiplier. call it 1.5x
Now we're to 3.5 hours. Your friend may disagree, this is my experience after doing this a long time.
Finally you have to master it and get it delivered to ACX. This can be largely streamlined, but it takes time to ensure you meet ACX's technical standards. Let's say you get it perfect on the first try. Call it 15 minutes to render, label, upload and pass the ACX checks.
We're to 3 hours 45 minutes.
A narrator needs to warm up before a session so that they sound consistent from session to session. When you're warming up for a gig, you're not working on another gig, so the time goes to the current project. Maybe you can multitask and QA yesterday's work while your warming up, but it's hard to do critical listening while you're humming and stretching and reciting toungue twisters. Call the warm up 15 minutes. Maybe less maybe more. But Most narrators cannot narrate for 8 hours a day. There is a limit to preserve the voice, and breaks must be taken to preserve your voice not only for the session, but for your work. So warming up is part of the gig, you're working and you need to factor that time in to creating the finished audio.
So, We're at 4 hours. And that's if you're really skilled and great at it and working at high efficiency. You know your DAW inside and out, and you have the voices down perfectly every time.
For many narrators there is an efficiency factor to build in. Narrating is physically demanding as it requires a great deal of sustained concentration -- not to dissimilar to coding. After a long narrating session you might not have the mental energy required for the sustained attention to detail to perform QA. (Similar to how it's difficult to debug code right away, and you find errors the next day when you're fresh)
In my experience that adds 30 minutes work to get the finished audio right and out the door.
So, yeah, a finished hour of audio can easily take 4 to 5 hours for an experienced narrator to create. Less skilled narrators who make more mistakes, take longer to edit, take longer to punch in corrections, take longer to meet the technical standards can easily get to 8 to 10 hours PFH. Just like a junior dev makes more mistakes and takes longer than a senior dev.
I coach newer voice actors all the time. I know from real experience how long this takes.