How We Test
Our editorial standards are the rules. This page is the method: what we actually do before we publish a verdict.
Two kinds of tools, two kinds of test
Most of what we cover is a browser-based service — AI music generators, stem separators, distributors. These need nothing but a computer and an internet connection, so the hardware is not the test. What matters is whether the thing works on real material, what the output is actually like, and what the fine print lets you keep.
So we use them on real material rather than on the vendor’s demo file. Where the test is about what you get to keep, we use our own music and have something at stake: our DistroKid review follows a three-track EP we released ourselves, through nine stores and nine countries, with the royalty reports to go with it.
The rest is software you install — plugins, instruments, DAW tools. Those get tested in the studio.
Where we test
A Windows bedroom studio, which is the room most of our readers are working in too:
- Interface: Focusrite Scarlett 2i2
- Monitors: M-Audio BX5
- DAWs: Reaper, Pro Tools and Ableton Live, whichever suits the tool, with DaVinci Resolve for anything cut to picture
- Instruments: Fender guitars and basses, and Akai and Novation MIDI controllers, so a tool gets a real performance rather than a loop
Dan is an Avid Certified User in Pro Tools and has composed in Ableton for years, which is why reviews here tend to say how a tool fits a session rather than just what its buttons do.
What we listen for
- The output, on your kind of material. Dense mixes, layered harmonies and busy low end are where audio tools fall apart. That is where we point them.
- How it behaves when it matters. Stability, speed and whether it survives a real session.
- What the terms actually allow. Licensing is where music tools hide the catch, and it gets read properly: what you keep, what happens when you stop paying, and what “royalty-free” means on that particular site.
- The price over time, not the sticker. Renewals, per-release fees and extras, counted over the years you would really be paying them.
What we don’t do
- We don’t take the vendor’s word for it. We’ll read the manual and follow the recommended setup, because that is the tool at its best, but nothing gets called good because the press release said so.
- We don’t publish brand-supplied copy. A sponsored piece is written by us, from the client’s material, and labeled at the top.
When we haven’t tested something, we say so
Some pieces carry a “From the archive” notice. Those were published before these rules existed and have not been re-tested, so they keep the original byline and do not claim a current test. They stay up because the information is still useful, and the notice tells you exactly what you are reading.
When a piece is sponsored, it says so at the top. The verdict is ours either way: see our editorial standards.