Lossless vs high-resolution audio: what each means

Lossless and lossy describe encoding

Lossless encoding allows the input digital audio samples to be recovered on decoding. It can still reduce file size: FLAC is compressed and lossless. “Compressed” is therefore not a synonym for “lossy.”

Lossy encoding discards information while reducing the data. MP3 and AAC are examples described in Sony's digital-audio guide. That technical distinction is not a listening-test result. It does not establish whether a particular listener will hear a difference with a particular recording and playback system.

Lossless also has a starting point. If someone converts a previously lossy file into FLAC, the new encoding preserves the samples it receives; it does not restore information removed earlier. Nor does “lossless” mean that every byte of the file, including its metadata, is identical to another file.

Resolution describes a different set of numbers

For PCM audio, sample rate is the number of samples per second and bit depth is the number of bits used to represent each sample. A CD uses 16-bit, 44.1 kHz PCM as its audio reference. The definitions and higher-resolution source examples appear in Apple's Digital Masters brief, pages 2–3.

High-resolution audio is commonly used for specifications beyond that reference, but the exact threshold depends on the label or system. Sony's guide acknowledges that there is no single universal standard. When comparing files, record their actual numbers rather than relying on a badge.

A sample rate is not the pitch of the music. Bit depth is not the same unit as a streaming bitrate measured in kilobits per second. These PCM descriptions also should not be applied unchanged to a different representation such as DSD.

The two properties can combine in several ways. A 16-bit, 44.1 kHz FLAC can be lossless without exceeding CD resolution. A 24-bit, 96 kHz FLAC can be both lossless and higher resolution. Neither description tells you whether the file came from a master recorded at those specifications.

Keep the master and playback route separate

Increasing the sample rate or padding a file to a larger bit depth does not recover source information that was absent. Apple's mastering guidance specifically distinguishes genuine source information from such conversion. A larger set of numbers on the resulting file is not proof of a more detailed original recording.

Two editions of an album may also differ in mastering as well as encoding. Unless that provenance is established, a comparison between them cannot isolate the effect of file format alone. “Same album title” is insufficient evidence that all other variables match.

Playback adds another question. A player may decode a lossless file and then apply volume adjustment, equalization or sample-rate conversion. Those operations can change the digital samples without turning the source file into a lossy encoding. They can also be intentional parts of the system. Roon's signal-path explanation distinguishes processing stages and notes that software cannot see every later hardware operation.

Use a four-part comparison card

For each version you are considering, record these fields independently:

• Edition or master: exact release and any supported provenance. Write “unknown” if all you have is an album name.
• Encoding: for example, FLAC or AAC. Establish whether it is lossless encoding, without treating the filename extension as proof of earlier history.
• PCM resolution: bit depth and sample rate, when applicable. Keep these separate from encoded bitrate.
• Playback route: actual playing device, connection and any reported processing or output conversion. Record what remains unobservable.

Imagine two entries: both have an unknown master and an unverified playback route, while one is FLAC at 16-bit/44.1 kHz and the other FLAC at 24-bit/96 kHz. The labels establish different PCM resolutions and the same lossless encoding family. They do not establish different sound quality or which version reaches the converter unchanged.

Use the lossless playback checklist to investigate the route. If a device cannot accept a required format, the DAC guide helps distinguish a missing capability from an assumed upgrade.