What audio file format should I use when I’m bouncing Mix Masters?
What bit depth should I use when I’m bouncing Mix Masters?
What sample rate should I use when I’m bouncing Mix Masters?
What should my master fader level (peak amplitude) be when I’m bouncing Mix Masters?
Why aren’t my mixes as loud as other CD’s? How can I make my mixes louder?
Should I bounce mix stems, or mix reductions?
__
What audio file format should I use when I’m bouncing Mix Masters?
The Producers and Engineers Wing of The Recording Academy recommends that digital mix masters be delivered as split or stereo-interleaved Broadcast Wave files (BWF). Other uncompressed audio file formats like AIFF may also be acceptable. Any audio file format that uses data compression (lossless or otherwise) is generally discouraged.
What bit depth should I use when I’m bouncing Mix Masters?
The Producers and Engineers Wing of The Recording Academy recommends that digital mix masters be delivered as 24-bit audio files. Among several benefits, 24-bit mix masters facilitate higher resolution processing during pre-mastering. 16-bit audio files should only be used for listening reference, like CD-DA.
What sample rate should I use when I’m bouncing Mix Masters?
Short answer: For music = 88.2kHz or 176.4kHz; for multi-media = 96kHz or 192kHz
Safe answer: the same sample rate as the source session.
Best Practices: Ideally, digital audio recordings (multi-track sessions and Mix Masters) should be made at a multiple of the target consumer sample rate. Sample rate conversion requires a common multiple between the source and target sample rates. If we choose a production sample rate that isn't a multiple of the target consumer sample rate, an intermediate conversion to a much higher, common multiple sample rate will be required (along with the necessary filtering for that process).
By choosing a production sample rate that is a whole number multiple of the target consumer sample rate, we can ensure that sample rate conversion process will only require a single stage of filtering. Regardless of the quality of the filtering being used, less filtering means less group delay and frequency response distortion.
Music audiences listen to material at 44.1kHz. DVD-Video and multi-media delivery tools usually use a sample rate of 48kHz. That difference initially accommodated the need to have a digital audio sample rate that was evenly divisible by standardized film and video frame rates.
Multiples of 44.1kHz and 48kHz are recommended throughout the production cycle to facilitate higher quality A/D and D/A conversion. At any sample rate, phase shift (high-frequency group delay) and aliasing in the A/D and D/A conversion processes lead to a progressive loss of audio quality as audio frequency increases toward its upper limit (which is established as half of the sample rate). By increasing sample rate to a multiple of the consumer sample rate, a significant proportion of this phase shift and aliasing is moved beyond the audible bandwidth of human hearing (less than 20kHz).
Here are some good guidelines to follow:
- If you’re starting a new music project, work at 88.2kHz or 176.4kHz
- If you’re starting a new film or multimedia project, work at 96kHz or 192kHz
- If you’re already outside of these guidelines, stay at the current sample rate. There’s no logical benefit to converting to a new sample rate until the project is being mastered.
- Always use the best A/D and D/A conversion you can afford. Well-designed, dependably manufactured A/D will perform better at 44.1kHz than bargain converts running at 88.2kHz. If it makes your music sound bad, is it really a bargain?
What should my master fader level (peak amplitude) be when I’m bouncing Mix Masters?
It is recommended that 24-bit mix masters be bounced with peak amplitudes between -6.0dBFS and -3.0dBFS. While there is no reason to leave any headroom for the mastering process itself, several circumstances suggest this range as a best practice:
- Metered peak amplitudes greater than -3.0dBFS can often include significantly higher inter-sample peaks, including material beyond -0.0dBFS. This should be avoided.
- Many mixing environments lack the metering and monitoring tools to detect these inter-sample peaks.
- A 24-bit digital audio file has about 20dBr more dynamic range than human hearing. There’s no need to achieve full scale.
Simply adjust the master fader until peak amplitude falls between -6.0dBFS and -3.0dBFS. That’s precisely what it’s there for. It is not productive to use peak limiting to set master fader level. The peak limiting process (even in small quantities) can have additional audible effects that are best managed during the mastering process.
Why aren’t my mixes as loud as other CD’s? How can I make my mixes louder?
Remember, there are Mix Masters and then there are Replication or Download Masters. Your project isn’t finished until it has been mastered, so the relative loudness of a mix does not represent the final level of the project. Comparing the loudness of a Mix Master with a finished commercial CD is not particularly useful.
That said there are a lot of aspects of mixes that directly contribute to the eventual loudness of a finished master. So what should you be listening for while you’re mixing? Here’s an example scenario:
My client has brought me a set of 5 multi-track recordings to mix. The client is very concerned that her project fit in with the latest CD from Artist X as much as possible, including being equally loud. Here are some things I’ll be sure to pay attention to while I'm mixing:
- What is the loudest instrument in Artist X’s mixes? The answer is probably pretty consistent across the whole CD. I’ll be sure to use a similar approach with my client’s project.
- What are the relative loudness relationships between kick, snare, bass, and lead vocal? These relationships go a long way in establishing the fundamental structure of mixes in a lot of genres. Give your reference a listen; you might be surprised.
- What is the brightest instrument in Artist X’s mixes? Transience is a big component in apparent loudness. Pay particular attention to tonal contrast between instruments to get the most impact out of any particular tone.
- Apparent dynamic range can have a lot to do with panning and depth. Do Artist X’s mixes have a lot of subtle panning, or are they 3-channel stereo? What is the contrast between the instruments that seem nearest to the listener and furthest away?
These sorts of musically relevant aspects of mix structure will help you create consistent, engaging mixes that fit into a genre in a lot of fundamental ways. The mastering process can then help prepare those mixes for their commercial audience, including addressing their market loudness.
Should I bounce mix stems, or mix reductions?
Instrumental and A Cappella mix reductions can be very useful in a lot of scenarios, including live performance, remixing, music for advertising, and creating ‘clean’ versions (a.k.a. radio edits). Be sure to bounce Full, Instrumental, and A Cappella mixes from the same timeline selection to ensure flawless synchronization. In general, it’s better to take the time to carefully bounce these sorts of additional 2-track resources at the end of a mixing session than to rely on mix recall to generate them later.