Make sure you're into your own mix. Mastering can do wonders, but there are limitations. If you are happy with the mix, you will be happier with the master.
Bypass all dynamic reducing elements in the stereo mix channel (master fader), like limiters or other plugins, enhancing just loudness.

They are mostly applied for a quick listen somewhere else, or to get mix approvals, or to hear how the mix reacts to a mastering treatment. If you like the way your limiter sounds, please send BOTH your untreated and treated version, so we can compare it. Limiting should be the final stage of mastering and all the mastering treatments applied (eg. EQ and compressor) should be before the limiting process, that is one of the reasons for sending a mix without it. The other is: using select state of the art and vintage analog mastering equipment needs a decent amount of dynamic headroom. For other plugins on your master fader: if you are not fully sure it is necessary (compare switching it on and off), take it off for the master source version, too.
The master source mix file you send, should not peak higher than -3dBFS in our experience. Peaks at 0dBFS make mastering much more difficult. For an in-the-box mix, run a print/render/bounce and check the peak meter on the master channel after it went through. If you are getting close to 0dBFS, try trimming all channel volumes if they go to the stereo out, or if you use groups, just take the groups down. Another way is using DCA groups, or subgrouping the whole mix and reducing the group fader, so the volume that hits the master channel settles at -3dBFS. (That of course only goes for the master source mix without the limiter or plug ins, not the compare mix with all the treatment you used initially)

For analog mixes, if you’re not mixing in-the-box or routing to analog and capturing a stereo mix back to digital, just avoid to clip the input on the way back to digital, and don’t apply any digital plug ins in the DAW. Export the recorded stereo mix at the native bit-depth and sample rate of the session and upload it.
Set the print/render/bounce settings to 24-bit floating point or above (eg. 24-bit, 32-bit or 64-bit, depending on what the DAW allows), regardless of the session’s bit-depth setting. Most DAW mixers and plugins are internally above 24-bits even though the individual source recording track files may be max. 24-bit.
Print/render/bounce your mix at the same sample rate as the mix session. We can apply any necessary sample rate conversions with a mastering-grade sample rate converter. DO NOT set the DAW to change the sample rate when you print/render/bounce your final mix files for mastering.
Listen to the whole once more only for noises and other anomalies you might not like in the mix, focus on quiet parts like intros, middle eights and endings, mute or fade down unused channels.

Best to use headphones.

Do another run to listen out for front/back depth placement and L/R stereo field and adjust the mix accordingly.

Solo Vocal Track

Listen to the vocal tracks in solo to check for any clicks, ticks, thumps, p-t-s consonants, headphone bleed, audible crossfades and other sounds that might be masked by the rest of the track but not intended to be in the mix. The vocal channels seem to be the source of most unwanted noises in a track from our experience. All these anomalies will be brought out even more in the mastering, we can remove most of them but we prefer not to have to alter the full stereo mix just for a small blip, that can be removed prior to mastering.


Visually scan your arrange page for crossfades and solo the relevant sections and check if they are audible. Set your channel and bus meters on peak hold and visually look at individual channel clips (red LED) after the mix went through. There, check the channel volume through the plug in chain and reduce ins and outs of each clipping plug in, and matching the thresholds. Mostly just reducing the output volume of the last channel plug in and adjusting the fade volume the opposite way does the trick. Reducing the audio file level and applying a brickwall limiter might also work in difficult cases.
Trimming the heads and tails during mastering takes no time and can be better controlled in the mastering process so please leave the natural noise floor in beginning and end and leave enough room for the last note or reverb to ring out and a bit more.

If there are channels or tracks that have a constant noise going through the song like a tape hiss or a guitar buzz and hum, you can’t remove or it is essential to the sound, leave a few seconds before or after the song including the base level of the hiss, buzz, hum or room sound and NOT MUTING OR FADING IT MUCH so we can generate a noise profile for reduction. If you already have faded in and clipped the mix, go back to it and just render a short snipped of the track noise (the quiet parts when the instrument/vocal is not playing) and upload it with the mix. You should not be afraid of slight noise reduction, contradicting old rumors, and from our own experience, current noise reduction algorithms, applied only in the necessary parts of the song like the very beginning or very end of a song sustain, actually make the master clearer rather than duller. If you want the mix to fade out a very certain way, send us a faded and unfaded version, so we can regenerate it with applied master devices.
Many times we have received a request to master an instrumental of the same song months or years after mastering the vocal mix song. That usually when artists sign to a publisher who then requests the instrumentals. So we recommend making them right away.

It’s not hard to master instrumental mix at the same time as the vocal mix master, but going back to do instrumental masters at a later time to achieve the exact same sound as the vocal mix can be tricky and costly. We surely make note of any analog equipment settings, but the exact reproduction is sometimes compromised when our studio is upgrading or replacing equipment, especially after a long time passed. So, doing the main versions and instrumental versions at the same time ensures much better compatibility. AND While you are making instrumentals, you could also do an instrumental and a vocal stereo stem for even better mastering control on our side. It is slightly more time consuming and costly in the mastering process, but 90% of the time it achieves much better results.
You and we have to know what version we are working on. Don’t use the word FINAL. Use the full song name. Numbering mixes sequentially while mixing is very useful and KEEP the original numbering of the mixes in sequence in the name to send us – don’t change it for sending the master. Example: “Rock And Roll Chicken mix 17 no limiter for mastering” and “Rock And Roll Chicken mix 17 WITH limiter for comparison” gives you and us all information needed. “Chicken FINAL FINAL mix new 01” does not tell you, which mix version it was and it does not tell us the song title and if it is limited. Using dates is not necessary, as the computer generates timestamp metadata, but useful for demo mp3s who tend to lose timestamp sometimes when they are imported to an audio player like iTunes. IF using dates, use it WITH the mix version number AND in following format YEAR-MT-DY eg: 2018-12-24 (to avoid national differences).
Listen to the mix file you upload BEFORE sending. Watch out for eventual glitches in plug-ins or sample clock, especially when OFF-LINE line rendering. It rarely happens that plug ins don’t do what they are supposed to do but it does. Especially in large sessions with lots of plug ins, an OFF line render is a better CPU use for some DAWs, but LISTEN to the rendered file afterwards, as you can not listen while rendering. Plug in failure can not be rectified in mastering.