Settings info- Please read before use!

A. Gentle.

Use this to add that final bit of polish to your audio, or if you are generally happy with the sound but just want minor improvements and smoothing. Technically speaking, this is what will happen:
-LUFS to -24 to fit into the gain structure cleanly.
-Cut subsonics below 60hz @ 12db/Octave using linear filter.
-Remove clicks, crackles and pops.
-Level to -6 target dbfs.
-Some light de-essing
-Very light transparent limiting with final ceiling at -1dbfs.
(Not designed for music tracks)

B. Standard

Identical to "Gentle" setting, but with two additions. It will also detect and reduce background noise by 3db (not much). Also its approach to removing clicks, crackles, and pops is a little bit stronger.
(Not designed for music tracks)

C. Detailed

This setting has a focus on speech intelligibility. It cuts subsonic rumble, emphasizes the human speech range, firmly denoises, and runs restorative anti distortion tools at a high setting. It is also firm in its leveling out of volume changes.
(Not designed for music tracks)

D. Podcast Enhancer

Designed for podcast creators who have recorded an episode where their host is connected via a low quality remote link like a poor Skype or phone connection. It uses the same approach that traditional FM radio took to solving this problem, a process known as "high frequency emphasis/limiting". This is where the top end of the frequency spectrum is juiced up before being all squashed back down again with a limiter/de-esser combo that only works on the treble range. This idea of this is that if there is a mismatch between the amount of treble in the host, and the remote guest, this will iron it out. It is part of what makes fm radio listenable when someone calls in by phone. It is also part of what gives fm radio that characteristic "sizzle." If you have ever listened to a podcast and found yourself constantly riding the volume control depending on who was speaking, even though the remote guest seems loud enough yet remains oddly unclear, then your podcast creator should have used this approach.
(Not designed for music tracks)

E. Restorative tools only

Use this if you are already perfectly happy with the EQ, levels, and general sonic quality of your audio, and are just seeking technical improvements. It will mitigate clicks, pops, crackles etc and leave the rest as is. This setting uses no leveling, compression, or limiting, so your audio will come back to you at almost the exact same level that you put it in. Great for taking mouth clicks out of voice over, narrations, dialogue etc
(Not designed for music tracks)

F. Forensic

This setting will focus on attempting to help you understand what is being said. It attempts to draw out only human speech and destroy everything else. After that it adds an intense dose of human speech range EQ. Works great on old recordings, old speeches, home movies, people talking in a noisy car, videos where people are talking with a noisy tv in the background, and generally anytime you are struggling to hear what people are saying.
(Not designed for music tracks)

G. Music level, mix, and master

This setting will give your demo track a polish, round out the frequencies and get it up to a healthy level without overcooking it. If you want to show a demo to a friend or enjoy it yourself, but do not want to spend time/money on mix engineers or Landr, then this should help. Technically it doesn't seem much different to anything you could do yourself with some presets across your master bus, BUT the difference is gain structure control and range control. The gain structure control makes sure that your track is coming in at the right level for the subsequent leveling and compression processes to be effective without biting down too hard, while the range control prevents them from over compressing beyond their predetermined db threshold. So while it is still a generic process, it is a little bit more intelligent than just slapping a plugin on your master bus and hoping for the best. To get the best results out of this setting, make sure that you haven't already applied compression across the master bus.

H. Spotify Preview

According to its own documentation Spotify is moving towards a loudness matching approach where all of its content is forced to be about -14 LUFS. LUFS is a newish kind of algorithm that attempts to measure how loud something is more similarly to how a human being would. In the past, VU meters and other approaches weren't very good at copying the behaviour of the human ear when it comes to deciding how loud something seems, but actually the LUFS approach does a pretty good job. So using -14 LUFS as the target, if your master is too soft Spotify will turn it up and vice versa if its too loud. The other thing of note here is that a limiter *might* be applied to keep the "true peak" of the content below either -1 or -2. If you have a finished track you haven't uploaded to Spotify yet, and you went get an idea of how Spotify will react to it, use this preset.