Digital Audio Basics

Silverdaw is built so you can make music without knowing how audio works under the hood. But a handful of terms turn up when you import a file, set up a project, or export a mix — sample rate, bit depth, bitrate, and file formats like WAV and MP3.

This page explains those terms in plain language and tells you what to choose in Silverdaw. You do not need to read it to make a mix. It is here for when you want to know what a setting actually means.

What sound is

Sound is vibration travelling through air. Something moves — a string, a speaker cone — and pushes the air around it in rapid waves. Those waves reach your ears, and your brain hears them as sound.

Every sound wave has two basic properties:

• Frequency — how many times the wave cycles each second, measured in hertz (Hz). This is what you hear as pitch: high frequencies sound high, low frequencies sound low. The A above middle C on a piano vibrates 440 times a second (440 Hz); the A an octave above it is double that, 880 Hz.
• Amplitude — how big the wave is. This is what you hear as volume.

Those two ideas sit behind a lot of Silverdaw. The Tone and Filter controls in Mixing & Effects work on frequency, turning ranges of the sound up or down. Every fader, gain number, and meter works on amplitude.

How a computer stores sound

A sound wave is smooth and continuous, and a computer cannot store something with no end of detail. Instead it takes very rapid measurements of the wave and saves each one as a number. Play those numbers back in order, fast enough, and you hear the original sound again.

Two settings describe how that is done: how often the measurements are taken (sample rate), and how precisely each one is recorded (bit depth).

Sample rate

The sample rate is how many measurements are taken every second, written in kilohertz (kHz) — thousands of samples per second.

The rule is that to capture a frequency accurately, you need to sample more than twice as fast as that frequency. Human hearing runs out somewhere around 20,000 Hz, so 44,100 samples per second is enough to cover everything you can hear. That is where the familiar 44.1 kHz figure comes from — it was chosen for the CD, and it is still the standard for music.

You will meet two rates in Silverdaw:

• 44.1 kHz — the standard for music, streaming, and most sample packs.
• 48 kHz — the standard for video, film, and broadcast, so audio taken from video is usually at this rate.

Neither is "better" for a remix. What matters is that everything in a project agrees.

• Each project has one Sample rate, set in File ▸ Project Properties…. See Projects.
• New projects use the Default project sample rate from Preferences ▸ Audio.
• If you import a file recorded at a different rate, Silverdaw asks whether to Convert to the project's rate or Switch Project to the file's rate. See Importing & the Library.

TIP

If most of your material comes from music files, 44.1 kHz is a sensible default. If you are working with audio pulled from video, 48 kHz saves conversions. Either way, pick one and let the rest of the project follow it.

Bit depth

Each of those measurements records how loud the wave was at that instant, as a number. Bit depth is how many different values that number is allowed to take — in other words, how finely the loudness is measured.

Think of it as the markings on a ruler. A ruler with ten marks can only measure roughly. A ruler with thousands of marks measures the same distance far more precisely. More bits means finer markings, which means quieter detail and smoother fades survive intact.

• 16-bit gives 65,536 possible levels. This is the CD standard and is fine for a finished track.
• 24-bit gives about 16.7 million levels, with far more room to work while editing.
• 32-bit float is a more flexible way of storing the number that gives an enormous amount of room, so levels are very hard to damage during processing.

The distance between the quietest detail a format can hold and the loudest sound it can hold is called its dynamic range, and more bits means more of it. Editing, mixing, and effects all nudge levels around, and that extra room means those small changes never run out of space.

In Silverdaw you do not have to think about this while you work. Audio is processed internally in 32-bit floating point from start to finish, and only converted to a final bit depth when you export a mixdown. Export defaults to 16-bit, with 24-bit and 32-bit float available depending on the format.

TIP

When you export at 16-bit, turn on the Dither option. It smooths out the tiny rounding that happens when the audio is reduced to fewer levels.

The Bit Crusher in Track FX uses the same ideas deliberately in reverse: its Bits control coarsens the loudness measurements and its Rate control coarsens how often they are taken, giving a crunchy, lo-fi sound.

Bitrate

Bitrate is how much audio data there is each second, so together with the length of the track it decides the file size. For uncompressed audio it is simply the other numbers multiplied together:

sample rate × bit depth × channels = bitrate

44,100 × 16 × 2 (stereo) = 1,411,200 bits per second ≈ 1,411 kbps

That is why a WAV file is large. Formats like MP3 shrink it by discarding detail the ear is least likely to notice, which is why you choose a Bitrate when exporting an MP3 from Silverdaw:

• 320 kbps — the highest quality on offer, and the largest file. Use this unless you have a reason not to.
• 192 kbps — a middle ground between size and quality.
• 128 kbps — the smallest files and the lowest quality of the three.

Mono and stereo

Channels are the separate streams of audio in a file. A mono file has one, so it sounds the same from both speakers. A stereo file has two — left and right — which is what gives a mix width and lets sounds sit in different places.

Silverdaw handles both, and a project's mixdown is always exported as a single stereo file.

• Pan, on each track header, places that track in the stereo field.
• Transform ▸ Split Stereo Channels… splits a stereo clip's Left and Right sides onto their own clips, which is useful when the two sides carry different material. See Arranging the Timeline.
• Clips can be drawn as a Single waveform or as Left and right lanes — set Waveform display in Preferences ▸ General. Mono clips always show a single waveform.

Loudness and levels

Audio levels are measured in decibels (dB). It is a relative scale, so what matters is the difference between two levels rather than the number on its own. In a DAW, 0 dB on a track means "leave it as it is": positive values make it louder, negative values quieter. Double-click a track's gain number to type an exact value.

When audio reaches your speakers, or is saved to a normal 16- or 24-bit file, it has a hard ceiling. Push the level past it and the tops of the waves are chopped off, which is called clipping and sounds harsh and distorted. Watching the master output meter and leaving a little space below the top is the simplest way to avoid it. Silverdaw also has a Safety Limiter, whose Protect final output option caps the final output at -1 dBFS. It is on by default — see Mixing & Effects.

Two more terms appear in the Export Mixdown dialog, where loudness normalisation can put your export at a consistent level:

• LUFS measures perceived loudness over the whole track — how loud it actually seems, rather than how high the peaks are. Streaming services use it to play everything at a similar volume. Target LUFS is the loudness you want the export to land at.
• dBTP (decibels true peak) measures the highest peak the audio reaches when it is played back, including peaks that fall between the stored measurements. Ceiling dBTP sets the highest peak allowed, keeping a little space below the ceiling so nothing clips.

File formats

Audio formats fall into two groups.

Lossless formats keep every bit of the original audio:

• WAV — uncompressed and universally supported. A safe choice when you may want to work on the audio again later.
• AIFF — the same idea as WAV, originally from Apple.
• FLAC — compressed, but with nothing thrown away. Smaller files than WAV with identical audio.

Lossy formats make files much smaller by discarding detail permanently:

• MP3 — the universal standard, and fine at higher bitrates.
• M4A — the file type used for AAC audio, common in downloads and Apple Music.

Silverdaw imports WAV, AIFF, FLAC, MP3, and M4A, and exports WAV, FLAC, AIFF, and MP3.

TIP

Work from the best-quality source you have, and save lossy formats for the final export you share. Detail thrown away by compression cannot be recovered, and compressing audio that was already compressed makes it worse.

Quick reference

Setting	Choose	When
Sample rate	44.1 kHz	Music, samples, and most remixing
Sample rate	48 kHz	Audio taken from video or broadcast
Export bit depth	16-bit (with Dither)	Sharing and general listening
Export bit depth	24-bit	Keeping maximum quality for later work
Export format	WAV, AIFF, or FLAC	Archiving, or taking the mix elsewhere
Export format	MP3 at 320 kbps	Sharing and uploading

Where next

• Importing & the Library — getting audio into a project.
• Mixing & Effects — shaping levels and tone.
• Exporting a Mixdown — turning the project into a file.