What’s Dat Noise?
Much music today is considered to be an esoteric collection of broken vocal chords and noise generators parading as musical instruments. The fault though does not always lie with the musicians. Technicians connected to the industry contribute heartily.
Case in point, I was listening to the soundtrack of the bollywood movie ‘Partner’. Music by Sajid Wajid and performed by a host of biggie singers, some of the tracks on the album are distinctly catchy. So when faced with an original CD replicated by Super Cassette Industries (aka T-Series), I was looking forward to blasting the music on my Creative Computer Speakers.
What shattered were however my ear-drums and hopes and I went – “Blast! The sound sucks”. I wondered why and set about investigating it.
I extracted the Digital Audio (Track #4, ‘Soni De Nakhre’) and opened it up in Sound Forge to analyze it. Immediately noticeable is the compression in the sound. To maximize the loudness factor of the music, the technicians amplify every nuance of the music and damp out the loud portions so that the sound does not clip.
Such music is quite loud even at low volumes and is characterized by complete lack of sharply defined instruments. While this kind of compression may be fine for dance music, the loss of Q-Factor is quite telling when cymbals no longer sound crystal and sharp, but turn out as FM noise.
When overdone, compression has a nasty side-effect where loud portions of the audio disintegrate. It sounds as if you have turned the volume of your speakers all the way up and your speakers can no longer handle the torture.
Opening the Spectrum Analysis Tool reveals another surprise! The reason for cymbals sounding like FM noise is also because the complete frequency spectrum required to accurately produce the sound of the cymbal is missing.
Cymbals are high frequency instruments and produce sound waves in frequency right up-to 20 KHz. The spectrum analyzer revealed that this audio track did not have noticeable audio above 16 KHz.
Missing audio over the 16 KHz spectrum is characteristic of highly compressed MP3 files. When audiophiles complain that MP3 files downloaded from the Internet just don’t have the quality of an audio-CD (even though the MP3 sounds just fine to most of us), they are complaining about missing audio.
So what was happening here? Did the music studio email 128 Kbps MP3 files to T-Series to decode and master into a CD? Or did some T-Series audio engineer had the brilliant idea of converting Studio-Output (Uncompressed WAV files) to MP3 so that he could make space for porn on his HDD?
Big questions are:
- Audiophiles! Don’t you dare to hold the Audio-CD format up on a pedestal anymore. Audio-CDs suck equally and carefully encoded MP3 files sound as good as Studio-Output. On a serious note, this is a big reason for audiophiles to get worried since they are paying through their noses and receiving a bogus product.
- Who is responsible for this fiasco? Is it the music director/s? Is it the studio? Is it the more than one company that is involved in the actual process of mastering and replication of the CD?
- Does the paying customer deserve a refund? The customer maybe expecting a 20 KHz cymbal but the Musicians+Studio+Mastering can simply claim artistic license and ‘deliberate EQ’ and get away with it.
- Does any one monitor the quality of the product at any point in the production chain? Does anyone have the power to Observe-Object-Rectify?
- Is it even worth it buying high-fidelity audio equipment in India considering that much audio-video content is so heavily compressed and lacks quality? Though serious recordings are probably pristine in production quality and compromises are limited to popular records, how long will it be before a mistake is made in the production of a audiophile desired recording? Case in point: I have been checking out various MP3 versions of the soundtrack from James Cameron’s movie ‘Avatar’ and I have noticed the exact same kind of issues.
Next I checked the quality of music from the soundtrack of the bollywood movie ‘Metro’ (Track #12, ‘In Dino Revisited’). The audio quality was distinctly better but nowhere near the clarity I expected. The spectrum analyzer acted as the snitch yet again. It revealed that the audio was highly compressed and the Q-Factor was muted (but not entirely gone). Frequencies over 16 KHz were present but only just about. This CD was replicated by Sony-BMG.
Simply listening to Rahman’s ‘Sunta Hai Mera Khuda’ (CD: The Rahman Experience, Track #2) was a delight. My tell-all friend revealed the music in all it’s glory. Hardly any compression. Excellent Q-Factor. Sounds spread through the entire audible-spectrum. No wonder Rahman’s music seems low in volume but very melodious. This CD was replicated by Sony Music.
On a hunch I also checked the music from ‘Dreamz’, an initial effort by Aman & Ayaan Ali Khan. The audio was pleasant but not anywhere near Rahman’s quality. My digital eyes revealed the audio as compressed with dynamics, low Q-factor and passable high-frequency presence.