Whenever I finish a track in Famitracker or Furnace, I export it as WAV, open it in Audacity, normalize the volume, and that’s it. That’s been my “mastering” so far.

But I’m wondering if there’s more I could do to make my songs sound better before releasing them. Do you use EQ, compression, limiters, or other tricks? What does “mastering” mean to you?

I’d love to hear your workflows and tips, and see what else I could do to polish my releases.

I master ominously

This really depends on what you want your final song to sound like, mastering is as much an artform as it is a method.

If you find that just a raw render and normalization is what works for you, then just keep doing that, there's no reason to overcomplicate this aspect. This is preferred for BotB anyway and is what I do for this site.

If you want to colour your sound slightly more, then you can add some EQ as well and boost frequencies that you like and cut out what you don't like.

If you want to make it sound like those YouTube uploads with stereo and reverb and whatnot, or like those old 80s Konami cassette OSTs, you can also add a bit of reverb and chorus.

As for what I do: It might not necessarily be what you want, and my methods do vary per song - something like lakeside had me cut out the specific N163 hiss frequency, but doing the same for a 2A03 only song would sound weird.

My chain is mostly tiny amount of reverb (1.5-5% wet with pretty short settings) for something that you can feel but not notice, some compression to even out very quiet and very loud sections, and even this is at a pretty low mix level, EQ to boost/cut parts I like or not like, and then limiting to ensure it's loud enough.

How you achieve this, and what tools you use as well, are a great way to stir up an entire conversations about tools and whatnot that I am just not interested in doing personally. I use what works, and you should use what works for you.

For things that are not chiptune, I typically only use a limiter. The best "master" is having a good mix going into your master bus that you only need to use a limiter on for a desired loudness level.

for strict chiptune; no master!

I'll usually max the song's level to -0.3db and if there's like a lone peak or two that's way higher than the rest I'll let them get chopped. when making the tune I'll want to keep the kick and the bass out of each others' way if there's like full on bass. for lone dpcmless 2a03 this is easy enough bc the one triangle can do kick and bass but only one at a time. for sgen I'll often duck the bass out the way at the instrument stage by having a more gradual attack on the bass adsr for when the kicks land

for anything full daw I'll export the mixed song and master in FL Studio. soundgoodizer works for me and it's fast and easy. but first I EQ; I roll back any of the really low lows below 25hz. a next step I started doing, but in really small amounts, is the spreader. I leave everything 300hz and below as-is; for me this area's already more or less in mono. anything higher I open up just a little; iirc last time it was like 10%? just until the spreadered version sounds natural and the raw mix sounds like it's obviously missing something. my final step is two really light runs thru soundgoodizer; first at B, final at A. the first B pass gives me mids. like how all the 80s Slayer albums have the midrange that make the mix come alive? that's what I listen for! usually around 10% to 17% does it for me. maybe not even that much. the final pass thru A completes the mix. it doesn't take much. sometimes only 6%. the hi hats come out, the bass comes alive, and the mix just sets right. everything's just bigger and more present

mixing makes the internals fit the song. mastering makes the song fit the universe! in chiptune the module is the song. so I do no change at all on the master except raw volume boosts and cuts. those can be really important for a full album. some of the sine only or one channel tunes have no real drum peaks so the song has to be cut like 15db to fit what song came before and what song comes after

and yes the stages blend together some. some mix issues can clean up upstream at the arrangement stage. some mastering issues can be worked out in the mix

My mastering ends with wav export and conversion to mp3))

Normalize, light compression, limiter, normalize again, boom.

But also what MelonadeM said. You can go deep into the eqing, multiband compressor etc. But I feel like my biggest mistake is normally getting lost in details. What makes a track sound legit is arrangement > mix > master.

This singular Maximus preset has been on every song I've produced since 2021. That being said I do not write chip so this is primarily advice for mixists.

Aside from this on the master, I have a premaster bus I route everything through first, and I tend to hard clip it with Fruity Limiter by turning all the knobs down, and sometimes I do some slight EQ too if there's too much in the 300 or 3000hz areas. How much I clip (and whether or not I do) depends on the song and the intended atmosphere.

Shoutouts Thingerthing for introducing the Fruity Limiter clip method to me btw, an actual gamechanger for my production (it's a super dope sound design tool too!), been doing it for a few years now.

My dude, I have also shamelessly used nothing but Clear master for the past 5 years. Then I go to the built-in channel EQ and turn up the highs and turn up the lows and maybe turn down the mids because it makes my ears feel good. Then if I want it to go a little harder I put all the non-melodic or percussive instruments on an insert routed to master and layer an automation clip to duck that insert's volume with the kick. Sidechain. And aside from that it's just putting reverb, delay, and stereo widening in places I think it would sound good.

I haven't really played with mastering chiptune stuff but I've shied away from it cause it sounds cursed. at least where I notice it. shovel knight ost mastering does not feel good on my ears

i kinda just beep the boops

melonadem has it right. these days my philosophy is that the mix is like 90% of the sound quality, and that's divided into maybe 70% arrangement and 20% balancing frequencies and stereo mix

personally i wasted about 10 years trying to find a magic compressor/eq combination to make my tracks pop when what i really should've been doing was tastefully panning my channels lol

i dont

Yep, I'm the same as pretty much everyone else, and I agree that the mix is 90% of the sound quality. Ideally your mix is good enough that you pretty much just need a limiter

I've always assumed that in formats where the submitted file is not an mp3, editing the render in any way beyond normalizing and fading in/out is against the spirit of its purpose (a faithful rendition of the song for those who can't or don't want to play it natively).

In formats where the program allows EQing and compression (like Bespoke and VCV) I do those while rendering (I almost never compress the master though). This means that if I don't like the EQ I have to render it again.

Yeah, I assumed this topic was about non-Botb renders, but it should definitely be restated in case anybody didn't know:
Botb renders should be accurate to the original file as much as possible, which means no mastering except for normalization.

i'm by no means an expert but for mastering i mainly focus on loudness as the primary control factor. this is most important if you're making an album or some collection of tracks and you want the volume to be consistent. "loudness" is typically measured in LUFS which i'm sure has some scientific basis.... i typically aim for -14 LUFS with -1dB peak but that decision is up to you, it's a little bit arbitrary. if you want to upload your music to some service they may have certain requirements for this.

there can be some difficulty hitting your loudness goals if they are too high, because of the correlation between dynamic range (the variation of amplitude throughout your track) and loudness. if you don't compress your drums heavily enough for instance, your audio can hit the limiter pretty hard which sounds irritating in my opinion. i tend to err on the side of limiting the loudness if i know my audio isn't compressed much to avoid this.

i use a plugin called LUveler ( https://luveler.blogspot.com/ ) with ardour which works reasonably well i think. first i run a loudness analysis on the audio to get the starting point, which allows me to estimate the max gain to set into LUveler. e.g. if the loudness is -23 LUFS and the target is -14, i'll set the max gain to about 9dB. i'll set the peak to -1dB. if you set the max gain too high, the quiet parts of your track will become too loud and it'll be too inconsistent. ardour has more loudness analysis and correction tools, so you don't even need a plugin i think, but this works for me for now. i do quite like getting a loudness chart and stats when i export audio so i know i'm in the right range.

create channelf music only: always perfectly mastered, no additional work needed

imo it's always been about taste as much as any other thing in your mixer rack. overall what you're attempting to achieve with the sounds you make should dictate the way to approach mastering it. shoutouts MelonadeM

i do think a collection of sounds, album or whatever, might sway you a little more to assure they all share a similar loudness profile, shoutouts agarak.

and i do think there are staple tools that are generally agreed upon to handle what anyone probably desires out of a master. i like to primarily control the dynamics (clipper/compressor/both) then use a limiter at the end chain for loudness, it's a very well-rounded setup for most cases. i'll sometimes put something before that i think will sound good for any individual track, be it OTT at 2%, reverb, EQ, name any effect, and usually just a subtle application. if it's an OHB on this website, you're lucky if i remember the limiter.

here's a workflow that i somewhat unintentionally do myself when i decide i want to go through with a "real" master, but i also do it when developing tracks:

- render and listen outside the DAW, or close your eyes, basically not looking at the project and pondering. my brain listens differently when the options to tweak things are in front of me, therefore, take away DAW. (ymmv, some people just bounce into a different project) also consider taking a break before your listening session if you've been at it a while.
- note what enhancement or adjustment you want out of a process that impacts the entire track (if anything), good time to compare with a reference if you have one
- apply in mastering quickly, i like to do it fast so i don't think too much, i've already noted what i want.
- render again, decide from there.

usually i just slam the fapfilter pro-l with one of the clipping presets

. if i'm mastering an album i might adjust EQ on individual tracks if they feel too off (typically few dB adjustments in mids/highs). don't feel there's much else to do, assuming the mix is done well.

for chiptunes i do nothing. unless it's on the snes, then i might do some phase distortion magic to make the echoes not disappear completely in a mono downmix (i made a reaper template for it only to use it for a single tiktok video)

for an album, i use reaper to get all the tracks to the same LUFS, and then adjust individual track volumes if i feel like one of the tracks is too loud or quiet, and throw a limiter if there's some big peak on one of the tracks.

if gameboy record off dmg because the color sounds groedy. If .xm load in FTII and muck around with all the silly audio options. If .mod load in various ST mod trackers until I find one that sounds more fuzzier and gooder for the song but also plays without goofy mistakes (possible/advantageous ~80% of the time) I'm happier all the time with the tone balance I manage to strike when tracking but when it comes to deliberate EQing and all that I'm an absolute caveman and I think I'm ok with that.

I export an .nsf on Dn-Famitracker and then i use nsfplayer to export the audio and on Audacity i adjust the volume, when i use the N163 expansión i export the N163 channels separetly and adjust the volume because they end up being really quiet or really loud

i wanna put my money where my mouth is!

so i've submitted this to spring tracks xv: https://battleofthebits.com/arena/Entry/Make+Our+Own+Way/84016/

the render is just 0CC-FT output normalized, right? exactly what it should be for botb

I've also given it a master for my own personal reasons, which you can listen to here: https://drive.google.com/file/d/1x9xyare86DIgdlX-fUeqaJN7mgXiqBeM/view?usp=sharing

you tell me which you prefer, but here's what i've done to it if you're curious:

- first thing i'm doing to this before anything else is applying valhalla vintageverb. it's on the default preset with the decay turned down to 0.49s, predelay turned up to 34ms, and the mix is set to 3.6%. i wanted to have a sense of "space" without making it obvious there's reverb applied, so i don't take away from all of the details within the 0cc itself.

- eq (I used Pro-Q 4 here), with it i cut out sub-20hz signals, boosted 70hz range by about 1db, and then boosted the highs (6.5k) by about 2.5db. nothing very extreme here. the goal here was to bring back the highs and emphasise some of the bass, NSF stuff sounds oddly dark and muted i feel

- compression (Pro-C 2 here), "mastering" style with a -10db threshold, 1.77:1 ratio, almost maxed out attack and almost minimal release, knee set to the "hard" setting, auto gain turned off. the goal here was to provide some glue, through balancing between loud and quiet parts without making things seem too unnatural. the gain reduction hovers around -1db, brief spikes reach -2db, so it's definitely there but isn't an extreme setting or anything. obviously, threshold will depend on the song and so will the rest of the settings, but this is what i felt worked for this track

- limiting (Pro-L 2). default pre-set, the only thing I have enabled is 4x oversampling, made sure true peak limiting is on, set the output to -1.0 dBTP, and then cranked the gain up to as far as needed to reach an average of -9 LUFS. I think I went for 7dB gain, and the integrated loudness ended up at i think -9.8. Although quite a bit of gain, it only really captures some of the peaks from transients.

i think i spent 10-15 minutes on getting the master for this song. working fast so that you end up on a good setting instead of trying to finetune every minute detail is a legit tip that might save some of y'all. and, again, keep it simple! you might think 10 plugins is a good thing on a master chain but only if you know what each of them does for you and how you want to use them in an instant. i don't, so...

i'm happy to answer any questions if y'all have any!

i just slap a limiter preset on master ✅

I do not master currently, instead focusing on good mixing. This could have the benefit of easier stem creation.
I don't really aim for a certain LUFS (mistake or not, you tell me), just a certain loudness based on other tracks I listen to (some MrKoolTrix stuff -3 dB are one of my loudness references).