Real hi-res or upsampled CD: how to tell when you listen on a streaming service

Updated on 6 September 2026 · by the Hi Res Quest team

Short answer. There are two ways to check that a “24-bit” track really is one. If you have the file, a spectrum analyser tells you in ten seconds. On a streaming service you don’t have the file: the only check left is the edition — where the recording comes from, what it could contain, and what other platforms serve for the same track. That second check is what Hi Res Quest does, track by track.

1. Two checks, not one

2. What the spectrum sees — and what it doesn’t

• It sees the real sample rate of the source: a cut-off at 22.05 kHz signs a 44.1 kHz source; a wall at 16–20 kHz signs a pass through a lossy codec.
• It doesn’t see the quality of the master. A remaster crushed by the loudness war passes the spectrum test with flying colours and still sounds worse than the original edition. Compare at matched loudness, never raw.
• It is sometimes wrong: an old analogue recording, or a take with little treble, can have very little energy above 20 kHz and still be a genuine 24/96 transfer. Absence of high content is not proof of cheating; a sharp cut-off at 22.05 kHz is.
• It needs the file. On a streaming service it is unusable — which is the whole point of this page.

3. What the edition tells you

Without the file, you go back to the source. Five cross-checks, all documentary:

1. The ISRC code identifies the recording, not the song. Two ISRCs = two recordings (or two masters). The same ISRC served as CD on one platform and as “24-bit” on another raises a question, not an answer.
2. The recording date and chain. An album recorded in 16-bit digital — Brothers In Arms (1985), The Nightfly (1982, ≈ 50 kHz) — cannot contain more information than it did at the source. A “24/192” edition of those albums adds nothing: Hi Res Quest marks it “possible upsample” and proposes the version closest to the master.
3. The documented remaster. A genuine hi-res transfer from analogue tape is dated, signed, often announced (studio, engineer, year). When no trace exists, the pill says “unconfirmed label” and the score is shown without bonus — nothing is proposed instead.
4. The compilation. A compilation gathers masters of different origins; it is never proposed as the best version of a track.
5. Consistency across platforms. Same ISRC, same edition, different resolutions: we keep the highest one confirmed by the edition, never the highest one displayed.

4. Hi Res Quest’s pills

Each card says what we found, never a judgement on a track:

• possible upsample — the source is 16-bit digital; the 24-bit edition cannot contain more. We propose the version closest to the master.
• unconfirmed label — origin of the hi-res transfer undocumented. Score without bonus, no proposal.
• no remaster found — the proposed edition predates 1985 and carries no “remastered” credit: nothing proves a new transfer from the master. It is proposed, with that caveat.
• CD only on TIDAL — TIDAL only serves this track in CD quality while a hi-res edition exists elsewhere: exclusivity… or an upsample. Proposed with that caveat.
• to be checked — a signal without a documented reason; we say it rather than hide it.

The HRQ score (10 to 98) puts everyone on the same scale: 30 = CD 16-bit / 44.1 kHz, 49 = 24-bit / 96 kHz, 89 = 24-bit / 192 kHz. Details on the Understanding resolution page.

5. Do it on your playlist

First look at a real audit of 20 tracks (Kind of Blue, Dark Side of the Moon, Rumours, Thriller, I.G.Y., Porgy and Bess…): each card shows your version, the best authentic version found and the pill when there is one. Then Join the waitlist for early access — first playlist free (20 tracks), no credit card.

Frequently asked questions

Does the “Hi-Res Audio” logo guarantee real hi-res?

No. The Japan Audio Society label certifies that a device or a file can carry 24-bit / 96 kHz or more; it says nothing about where the master comes from. A CD upsampled to 24/96 carries the logo.

Is 24-bit / 44.1 kHz hi-res?

It is lossless above CD in bit depth, not in sample rate. The JAS threshold starts at 24/96. Spotify “Lossless” tops out at 24/44.1: CD quality, not hi-res.

Does Hi Res Quest listen to the files?

No, and it couldn’t: on a streaming service the file is not accessible. Everything is cross-checked from sources — ISRC codes, recording families, pressings, platform catalogues — and every statement is dated.

What if I still want to analyse a file?

Spek (free, Windows/macOS/Linux) or Spectro (macOS) do that job very well. Hi Res Quest doesn’t replace them; it answers the question those tools cannot handle: a track you don’t have on your disk.