Video summary

Perbandingan Spotify vs Apple Music vs Youtube Music di Indonesia

Main summary

Key takeaways

Technology

Tech comparison focus (Spotify vs Apple Music vs YouTube Music)

The speaker compares streaming platforms by recording playback and running digital comparison tests (a “zero/null test”) between:

  • the original purchased file, and
  • a recorded playback output while streaming.

Key concept: they use a digital-to-digital out chain (so the signal remains “data” and is not re-amplified), then align/compare timing using a tool—mentions DeltaWave—with the claim that timings will be the same.

How the tests are done (methodology)

  • Song used: Snarky Papi – “The Jambone / … The Shortest …” Purchased from Snarky Papi’s site, described as guaranteed original.

  • Equipment includes: Ever Solo, Solo/Antelope (clocking mentioned), MacBook, and DAC/players such as UAPP / RU9 / DAPs.

  • Audibility benchmark: correlated null thresholds
    • ~ -90 dB and below: considered inaudible (“nothing can be heard”).
    • If the correlated-null number is smaller (closer to zero): the difference is more audible.
  • Important limitation: the recording chain isn’t fully “lab grade,” so it may not always reach perfect bit-perfect results. Factors include equipment clock sensitivity and clock drift / jitter, which can introduce small detectable differences.

Main findings: streaming vs “lossless” modes

1) On Ever Solo (direct streaming)

Spotify

  • Tests include Lossless/High settings; the free tier is more compressed.
  • With Spotify High (described as closer to ~160 kbps), correlated-null is around ~40 dB—described as quite audible (and can be made more audible).
  • Above ~90 dB is treated as inaudible.
  • Spotify Connect is described as good, with small differences vs direct playback.

Apple Music

  • Apple Music lossless is described as producing noise that remains below audibility, even when turning up.
  • AAC 256 (non-lossless) shows a small but more noticeable difference, correlated-null around ~40 dB.
  • Overall: lossless Apple Music is reported as very close to the original (numbers appear “safe” per the -90 dB criterion), while AAC shows more audible artifacts.

YouTube Music

  • Less detailed in the Ever Solo section.
  • Later tested on Android; generally lands in the “okay” bucket for timing/fit quality.

2) On Android phones (via digital out using SPDIF-capable DAC)

The speaker tests multiple Android devices (examples mentioned):

  • Vivo X200 Pro (MediaTek Dimensity 9400)
  • Xiaomi 14 (Snapdragon)
  • A “Google Tensor” / Pixel-related device (No iPhone/Samsung representative samples are provided.)

Trend on Android:

  • Correlated null gets worse (more audible compression artifacts) than on the desktop/streamer chain.
  • Vivo is described as best among Android samples (around ~50 dB, still bit audible).
  • Xiaomi and others are described as worse (examples include numbers in the 20s), also described as “messy” and clearly different.

Why Android playback differs (core technical claim)

The speaker attributes poor null-test results on some Android devices to:

  • unstable clock drift / jitter
  • phone SoC/chip behavior and internal resampling
  • a more aggressive internal resampler to save CPU (leading to audible artifacts and mismatch vs original)
  • possible OS audio processing effects (though they claim loudness normalization/enhancements were turned off)

Streaming rate conversion issue

The speaker claims Android resamples to 48 kHz during streaming, then re-sends to the DAC—changing timing/clock characteristics. Workaround proposed:

  • ensure the track is set to 48 kHz already when possible, to minimize resampling.

Offline vs streaming (UAPP / exclusive mode)

They test offline/high-res playback using UAPP:

  • Offline playback with exclusive/bit-perfect-style routing can yield very strong results (sometimes correlated null reported as ~0 or “safe”).

  • Still, they can’t always achieve “100%/99% perfect” due to equipment sensitivity and residual timing mismatch.

Additional notes:

  • Tidal exclusive mode exists (including laptop/PC), but the speaker says it’s not officially relevant for Indonesia.
  • Sony Walkman Player and other apps may improve fidelity if they output in an exclusive/bit-perfect manner.

Loudness / normalization analysis (LUFS + limiter concepts)

They measure loudness normalization differences:

  • Spotify and YouTube Music: around -14 LUFS
  • Apple Music: around -16 LUFS (slightly quieter)

They explain (conceptually) that streaming apps use:

  • loudness normalization
  • limiter/maximizer DSP

These adjust perceived loudness across tracks, which can color audio slightly and differ by platform.

UI / Search / Casting / Ecosystem features (non-audio)

Search & discovery

  • Spotify: sometimes slower/more inconvenient for certain queries.
  • YouTube Music: likened to Google search—appears to play faster on search.
  • Apple Music: similar complaints as Spotify (less immediate).

Recommendations / stations

All three offer “radio/station”-style discovery:

  • Spotify: “Song Radio” / similar-artist/similar-album recommendations
  • Apple Music: stations
  • YouTube Music: “Start Mix” (starting mix)

Casting / “Connect” features

  • Spotify Connect: sends/plays on a device over Wi‑Fi (described as more like remote control than Bluetooth streaming).
  • Apple Music: described as using AirPlay, with limitations on bit/sample rate.
  • YouTube Music: uses Chromecast / TV routing.

They argue Bluetooth is less important if AirPlay/Chromecast are available.

Desktop clients

  • Spotify: flexible and smooth on PC/Mac.
  • Apple Music: works on Windows, but requires manual configuration for exclusive/high-res switching (mentions Lossless Switcher for Mac/Windows).
  • YouTube Music: no dedicated app in their view; mostly via web browser (Chrome).

Dolby Atmos discussion (Apple Music)

They discuss Apple Music’s Atmos ecosystem:

  • Apple Music offers stereo vs Dolby Atmos channels (formats mentioned like 7.1.4, 9.1.6).
  • Warning: turning on Dolby/spatial audio on the wrong device (e.g., phone/headphones virtualization) may be “pretend” compared to real Atmos playback.
  • They distinguish channel/object-based Atmos from bit depth / sample rate—stressing these are different concepts.

Pricing summary (approximate from the speaker)

  • Spotify: around 39,000 (student pricing mentioned)
  • Apple Music: around 35,000 (plan mentioned)

Other pricing/codec notes (as described):

  • Spotify free-tier is highly compressed (around ~128 kbps), while premium “light/high” tiers are higher (described as ~160 kbps for some tier).
  • YouTube Music is described as Opus ~256 (non-lossless, per their claim).

They emphasize: the “winner” depends more on needs than on purely sound quality.

Practical recommendations / conclusion

  • Choose a service based on your needs first (search UX, playlists, Atmos, casting, subscriptions), not only sound quality.
  • For best technical fidelity, they recommend:
    • Prefer setups that support exclusive/bit-perfect routing (often via dedicated apps like UAPP).
    • Be cautious with Android streaming, because resampling/clock behavior can degrade match quality.

Main speakers / sources

  • Speaker/host: Fernand Ginens (also referred to as the “Fernand Ginens YouTube channel” owner throughout subtitles)
  • Referenced entities/tools: DeltaWave, Ever Solo, Antelope/Solo, The Headphone Show (benchmark equipment source), Lossless Switcher (GitHub), UAPP, and streaming platforms (Spotify, Apple Music, YouTube Music)

Original video