Power, Corruption & Lies Original vs. 2008 Remaster

New Order’s Power, Corruption & Lies has been reissued so many times that “which one should I listen to?” is a real question. Loudness is the part of the answer you can measure. We took the first CD issue and the 2008 Collector’s Edition remaster, measured all eight album tracks on each with the same analyzer, and then went further: waveforms, loudness over time, average spectra, stereo image. Here is what the instruments say.

1. The remaster is louder on every track — by 4.4 LU on average (from +3.3 to +5.3). The album goes from −17.0 to −12.6 LUFS.
2. The loudness came from squeezing peaks, not raising them. Average level rose 4.4 dB while true peaks rose only about 1.1 dB, so the peak-to-loudness ratio fell from 15.9 to 12.5 dB and DR from 18.8 to 15.3.
3. The big-picture dynamics held. Loudness range barely moved (4.9 → 4.5 LU): quiet passages are still quiet, verses are still verses.
4. It is the same recording, on the same edit. Track lengths agree to within 0.7 s and the stereo image to within 0.02 in correlation. Nothing was re-mixed or re-sequenced.
5. After level matching, the remaster is bassier and a little darker: roughly +2.8 dB around 40–63 Hz and −1.1 dB through the 3–6 kHz presence region.
6. Both are lossy copies (~336 kbps MP3), which matters most for true peak. We tested how much in Limitations: at this bitrate LUFS holds to about a hundredth of a unit, while true peak can wander by a tenth or two, occasionally three.

Power, Corruption & Lies is New Order’s second album, released by Factory Records in May 1983. It was produced by the band and engineered by Michael Johnson (per the credits on Discogs), and its sleeve, Peter Saville’s design around Henri Fantin-Latour’s painting of roses, is as well known as anything on it. Discogs lists no mastering engineer or studio for either edition below, so we don’t name one.

Two digital editions are compared here, track for track:

A note on the word “original”. The 1983 vinyl was cut from the analog master. The CD is a digital transfer made later, and it is the earliest digital edition, which is the one we call the original here. Everything below is CD-versus-remaster; it says nothing about the LP.

Every number labelled LUFS, true peak, LRA, DR or short-term max is the site’s own measurement, produced by the same analyzer as mastering.to (ITU-R BS.1770 / EBU R128; see Methodology). The waveform, timeline, spectrum and stereo figures come from a separate script that decodes the same MP3 files to float. Its loudness agrees with the database to within 0.05 LU on all sixteen tracks, which is how we know the timelines are trustworthy.

Integrated loudness is the number streaming platforms and broadcasters use to decide how loud a track is. On the original CD, the album’s eight tracks sit between −19.2 and −14.7 LUFS. On the 2008 edition they sit between −14.6 and −11.4 LUFS. Every dot below moved right.

What does 4.4 LU mean in practice? It is about 2.75× the average signal power, and by the common rule of thumb that 10 LU doubles perceived loudness, roughly 36% louder to the ear. Not a night-and-day change, but easily enough to win a casual A/B, because listeners reliably prefer the louder of two otherwise-identical versions. That bias is the reason the fair-comparison advice near the end of this report matters.

Streaming adds a wrinkle. Take a player that normalizes to −14 LUFS (a common target; see the platform table). It would turn the 2008 edition down by about 1.4 dB. The original would be left alone or, on services that also turn quiet tracks up, raised by about 3.0 dB. Either way the gap in playback level mostly disappears, and the remaster’s lost dynamics are still lost. That is the general lesson of the loudness era: loudness is a gain you can add back for free; dynamics are not.

Loudness alone tells you very little about how a master was made. The interesting quantity is the distance between the loudest moments and the average. Two views of that: peak-to-loudness ratio (true peak minus integrated LUFS) and the site’s DR figure (a crest-factor measure). Both fell by about 3.4 dB.

That is the fingerprint of dynamics processing: transients get pushed toward a ceiling, and then the whole signal is turned up until the average is where the target says it should be. It is also, to be fair, moderate processing. A peak-to-loudness ratio of 12.5 dB and a DR of 15.3 are still spacious by 2008 standards, and the loudness range didn’t shrink (4.9 → 4.5 LU), so the music still breathes between sections. The remaster is not a brick wall.

Numbers become intuition when you can see them. Below, each track’s waveform is drawn twice on the same scale: the original on top, the 2008 edition underneath. The pale outline is the peak envelope; the solid core is RMS, the body of the signal. Dashed lines mark full scale. On the remaster the core fattens toward the outline: less space between peaks and body.

The short-term loudness curve (a rolling three-second window) shows how each track builds and recedes. Because the two editions are the same take on the same edit, the curves have the same silhouette. The interesting thing is the gap between them: the shaded ribbon. It stays close to constant from quiet intros to full-band peaks, wandering by only ±0.4 LU (standard deviation) around its mean. That is what you would expect from one gain move plus a limiter that mostly catches peaks, rather than from re-balancing sections of the song.

Loudness is only half of what changes in a remaster. To compare tone fairly you have to remove the level difference first, so figure 6 pulls the 2008 edition down by 4.4 dB before overlaying the two average spectra. What remains is the difference in balance. Across all eight tracks it is remarkably consistent:

• Deep bass is up. Roughly +2.8 dB at 40–63 Hz, and +1.1 dB through the 80–160 Hz range.
• The middle is essentially untouched (−0.2 dB around 630 Hz–1.25 kHz).
• The upper mids ease off. About −1.1 dB from 3 to 6 kHz, and −1.1 dB at 8–10 kHz.

A caution on reading this. It is an average spectrum over whole tracks, of a signal that has also been dynamically processed, so it isn’t a clean picture of an EQ curve. A limiter working harder on the loudest, bass-heavy hits also shifts average balance. What we can say is that the remaster, at equal loudness, carries more weight below 100 Hz and a little less energy in the upper mids. Whether that is a tonal correction or a taste is for your ears.

It is just as informative to list what stayed put. If a remaster had been re-mixed, re-edited or re-sequenced, these would have moved.

Same lengths, same stereo image, same loudness range: this is the same recording on the same edit, re-transferred and reprocessed. That is the signature of a remaster rather than a re-mix, and it means the differences in the earlier figures can be attributed to the mastering chain rather than to the mix.

Is −12.3 LUFS loud? The database can say. Below is the median track loudness of every decade with at least 40 measured tracks in the database, with the two editions’ medians overlaid as vertical lines.

The original sits near the database’s 1980s median (−16.5 LUFS): a record mastered for the dynamics of the analog-to-CD era. The 2008 edition moves about 4.8 LU up, into the range of the database’s 2000s recordings (median −11.0 LUFS), though still 1.3 LU below that median. It’s a partial concession to the loudness war, not a full surrender. For a 2008 catalogue reissue that is arguably the point: loud enough to sit next to modern releases in a playlist, not so loud that the songs stop moving.

The Collector’s Edition adds a second disc of singles and B-sides: Blue Monday, The Beach, Confusion, Thieves Like Us, Lonesome Tonight, Murder and two instrumentals. It averages −11.7 LUFS, a touch louder than disc 1. The loudest is Thieves Like Us (Instrumental) at −9.2 LUFS; the quietest is The Beach at −13.2. A Discogs contributor also notes that some disc-2 tracks sound as if they were transferred from cleaned-up vinyl, one more reason not to assume every track on a reissue shares a source.

A comparison is only as good as its caveats. The important ones, most significant first.

Both editions are lossy copies, not lossless. The files measured are ~342 and ~330 kbps MP3s, not WAV. In our experience integrated LUFS is nearly identical between an MP3 and the WAV it came from, while true peak can differ by a tenth or two of a decibel, because MP3 coding adds and removes tiny amounts of energy and can move inter-sample peaks. Rather than assume that, we tested it: we took a seeded random sample of 30 lossless CD rips already in the database, encoded each to MP3 at several settings, and measured every copy with the same analyzer that produced the numbers above.

The takeaway: the 4.4 LU difference between these editions is hundreds of times larger than the 95th-percentile LUFS error we saw at 320 kbps, and still more than 17 times larger than the worst error at 192 kbps, so the headline finding is safe. Peak-to-loudness and DR are more sensitive than LUFS: at 320 kbps the DR reading is off by a median 0.17 dB (worst 0.77), and MP3s tend to read slightly hot. Against a fall of about 3.5 dB, that still holds. True peak is where to be careful: a track reading +0.1 dBTP on an MP3 could be a hair under 0 on the lossless master and vice versa, so counts of tracks over full scale (2 versus 1) are suggestive, not conclusive.

Other limitations worth stating plainly:

• We can’t prove which pressing was ripped. Discogs and MusicBrainz list several CD pressings of the original master (Factory, Qwest, London Records 90, others). Track timings match Factory FACD 75 and the 2008 Collector’s Edition listings, and MusicBrainz lists more than one Collector’s Edition release (including a “corrected 2nd edition”). Pressings cut from the same master would measure the same; a different master would not.
• “Original” means the first CD, not the LP. The vinyl was cut separately, and analog-to-digital transfers of the same tape differ.
• Spectra come from MP3s. Above roughly 16 kHz the encoders’ low-pass filters dominate, so the top bands are shown dimmed and excluded from every finding. Average spectra also blend EQ with the effects of dynamics processing.
• One album, eight tracks. This shows what happened to this record. It is not a claim about all remasters, or about all 2008 Collector’s Editions.
• Later remasters exist and weren’t measured. MusicBrainz lists a 2015 digital remaster and a 2020 “Definitive Edition” of the album. The 2008 edition is one waypoint, not the latest word.
• Louder is not better, and quieter is not truer. Loudness and dynamics are facts; preference is yours. We did not run listening tests.

If you want to decide by ear which edition you prefer, match the levels first, or the louder one will win regardless of merit. With the two editions in a player that lets you trim gain, pull the 2008 edition down by about 4.4 dB: individually, between 3.3 and 5.3 dB depending on the track (the table below lists each). Then switch back and forth blind if you can. Listen for what remains once loudness is equalized: the bass weight, the size of drum transients, the top-end air. Those are the real differences.

“Match gain” is the trim to apply to the 2008 edition to equal the original’s integrated loudness. The disc-2 bonus tracks are in the database too; the two editions’ pages list every measurement: original, 2008 disc 1, 2008 disc 2.

• Measurement. Each track was decoded to 32-bit float (so lossy decoder overshoots aren’t clipped) and measured with the mastering.to analyzer: integrated LUFS, true peak, LRA, DR (crest factor) and short-term maximum, per Methodology.
• Extra analysis. Envelopes, the short-term timeline, average 1/3-octave spectra (8192-point FFT, mid channel) and stereo statistics come from a separate script decoding the same files; its loudness matches the database to within 0.05 LU. Scripts live in the site repository under tools/articles/.
• MP3 error test. 30 random lossless rips from the database, encoded with LAME at 192 kbps CBR, 256 kbps CBR, VBR V0 (~245 kbps), 320 kbps CBR, decoded to float and measured by the same analyzer.
• Edition details from Discogs and MusicBrainz. Sleeve thumbnails from the MusicBrainz Cover Art Archive; artwork belongs to its owners.
• Reuse. The data is free to use, including professionally, with credit; see Our Data. Suggested credit: “The Loudness Database (lufs.to), Power, Corruption & Lies: Original vs. 2008 Remaster, September 26, 2026.”

• New Order
• Power Corruption & Lies remaster
• Power Corruption and Lies 2008 Collector's Edition
• remaster vs original
• loudness comparison
• LUFS
• dynamic range
• true peak
• loudness war
• album loudness