OrynosMusicMixing & mastering guide
How to compare masters at matched loudness
Make a fair A/B decision, understand the meters and separate a louder master from one you actually prefer.
Match listening level before choosing a winner
A louder version can seem more detailed, fuller or more exciting even when processing has introduced a trade-off you would reject at the same playback level. Match listening volume before deciding whether the master helps the song. Then compare the final delivery level separately so both questions get a fair hearing.
Use the same passage and playback device for both versions. Start with a chorus containing the lead, kick, bass and cymbals, then check a quieter verse. Keep your listening volume comfortable and take short breaks. Avoid changing speakers, headphones and volume at the same time as changing the master.
Read the meters as different measurements
LUFS, peak level and dynamics describe different properties. Loudness standards define reproducible measurements and delivery practices; they do not assign a musical quality score. For example, EBU R 128 is a broadcast loudness recommendation. Its programme target should not be treated as a universal creative target for every music release.
Measure equivalent durations. A short excerpt and a complete song can produce different integrated loudness, loudness-range and peak readings. Silence, arrangement changes and the chosen measurement window matter. If two charts use different material, their numbers are not a controlled comparison.
| Measurement | What it tells you | What it does not prove |
|---|---|---|
| Integrated LUFS | Loudness over the measured programme, with the meter's weighting and gating. | That a louder master sounds better. |
| True peak, dBTP | An estimate of reconstructed peak level, including peaks between samples. | That the tone or balance is right. |
| RMS | Signal energy over a stated window. | A complete model of perceived loudness. |
| Crest factor, dB | The difference between peak and RMS for the defined measurement. | A genre-independent quality ranking. |
| Stereo Side energy | A Mid/Side energy relationship for that recording. | How pleasing, spacious or mono-compatible it will feel. |
Further reading: EBU: loudness measurement and R 128 (opens in a new tab)
Listen for specific trade-offs
Focus on one element per pass. Does the kick retain its attack? Can you follow the bass notes without the low end swelling? Are the vocal consonants clearer, or merely sharper? Check whether cymbals become brittle and whether reverb tails remain smooth. A waveform alone cannot answer those questions.
Next, listen to the arrangement. The chorus may need a stronger lift than the verse. Compression that makes a short chorus feel exciting can reduce that contrast across the song. A wider image may sound impressive in headphones while pulling an important part away from the centre. These are musical choices, not evidence that every measurement must increase.
Use the OrynosMusic comparison tools
On the homepage, play Orynos Test Song and leave Match listening volume enabled. Switch between the original and the available model renders while listening to the same passage. The example uses one recording; it demonstrates how processing differs on that source, rather than predicting how a model will handle every genre or recording.
Open Compare for the actual example audio, measurements, processing modes and test conditions. Check the units and the source before interpreting a chart. On your own mastering result, compare original and master at matched level, then use Translation Check to listen in Stereo, Mono, Small-speaker and Low-end modes. Those listening checks help expose different issues; they do not replace playback on real devices.
Compare models and modes without assuming a ranking
Choose a starting model, render a standard version and write down one or two observations. If you try a different model or a deeper analysis mode, keep the musical aim and other settings consistent. That makes the next comparison useful instead of changing several decisions at once.
Higher credit cost and longer processing time describe resource use, not a guarantee of preference. Power, Ultra and Max use deeper analysis where supported and produce one final mastering chain. The sensible choice is the version that translates well and keeps the character you want, with a processing time and credit cost that fit the session.
Check the delivered file as well as the preview
Download your preferred master and play the file from beginning to end. Check the opening, quiet transitions and final tail, not only the loudest section. If you need an MP3 delivery, audition that encoded file too. Lossy encoding can change peak behaviour, so a WAV measurement does not describe every property of its encoded copy.
Keep an untouched original and your final master together. For a revision, note the model, mode and settings as well as what you heard. A clear decision such as 'less sharp on the chorus vocal at matched volume' is more actionable than a larger number on a chart.