While listening to music made before the 2000s, you may remember experiencing some noise in the recording. The older the music, the more the distortion and crackle. Additionally, during the in-between .mp3 phase, after CDs entered the market and before streaming apps became common, a lot of the music just sounded awful in quality.
The first is because of discrepancies that occurred when the music was transferred from a cassette or a vinyl record to a CD. The second is because the formats of music on CDs, either Waveform Audio File Format or Audio Interchange File Format, are compressed when reproduced in the .mp3. This causes a significant drop in quality, since much of the data is removed during the conversion. It’s not what the music composer intended. This is where Sreejesh Nair comes in.
Nair is a re-recording mixer for such Hindi productions as Jodhaa Akbar, Kaminey and Gangs of Wasseypur II (for which he won a National Film Award). In his spare time, the Dubai resident cleans up and produces top-quality versions of some of the most popular Hindi, Tamil and Malayalam film tunes recorded between the 1980s and the early 2000s. Nair releases these versions on his YouTube channel The Mastering Project.
Nair launched the video channel in 2017, and has over 46,000 subscribers till date. “I began doing this simply for my love for these songs,” he told Scroll.in. “These were the songs that made me read the cassette’s back covers to see who had done what and got me interested in music and sound. The good thing about YouTube is that the rights to this music don’t belong to me, and YouTube matches the sound signature. So the entire money off my version goes to the original owner. I don’t make a penny.”
Why is The Mastering Project important? For that, you need to plug in your headphones.
One of the most popular uploads is Nair’s remastered version of Kadhalikum Pennin from AR Rahman’s soundtrack for Kadhalan (1994). The bass and the beats sound more pronounced and have more depth, and the opening vocals of SP Balasubramanyam and Udit Narayan sound as if they were recorded yesterday. Every instrument and every recorded track used in the final mix are crisp and distinct.
Out of 428 uploads on The Mastering Project channel till date, 169 are remastered versions of Rahman compositions mostly from the ’90s, on which the legendary H Sridhar worked as a sound engineer. These include Jiya Jale and Dil Se Re (Dil Se, 1998), Taal Se Taal Mila – Western (Taal, 1999), and Varaha Nadhikarai (Sangamam, 1999).
“You could call my channel a tribute to H Sridhar,” Nair said. “We were close, and he was like a mentor to me.” Sridhar died in 2008.
Among the most liked non-Rahman uploads are Yeh Kaali Kaali Aankhein from Baazigar (1993) and Mushkil Bada Yeh Pyaar Hai from Gupt (1997). What immediately stands out on the Baazigar track is the richness of the most minute electronic samples and beats, particularly the bass. “When I worked on this song, I kept thinking about how the hell Anu Malik came up with these sounds at that time,” Nair said.
For Muskhil Bada, one of Viju Shah’s best tracks, it’s astounding how fuller the haunting synths that kick in at the 1.04-min mark sound and add to the texture.
Nair has a simple criterion for his remastering choices: nostalgia, “that initial rush of emotion which I felt on hearing the Roja, Gentleman or Thiruda Thiruda soundtracks”.
Nair added, “Over time, we may have forgotten how exactly these songs sounded the first time, but we remember the emotion. I pick songs that you’d listen in the 1990s with your family sitting with you in front of the television for music programmes on a Sunday morning, like Chitrahar. There’s Vidyasagar, Lucky Ali, Junoon, Anu Malik, Jatin Lalit, all kinds of stuff.”
There are also 38 remastered versions of Ilaiyaraaja compositions. While speaking about why problems crop up in digitised songs first produced in analogue formats, Nair spoke of an Ilaiyaraaja song whose CD version turned out oddly following the transfer from the audio cassette.
“I don’t add or subtract any new thing to the songs or tweak the vision of the track,” Nair explained. “What I do is help realise the best version of a song according to what its maker wanted. Some people say I could’ve added more bass or more treble. That is not my intention. So with this Ilaiyaraaja song, in the CD mix, everything was tilted to the right channel, so you’d hear more stuff on the right unit of your earphone, which is wrong from a mixing perspective. That needed to be fixed.”
What is behind this persistent problem in most digitised mixes of analogue audio tape recordings? One reason is the Dolby Noise Reduction, or Dolby NR system. “You may remember that in the cassette decks back in the day, if you pressed the Dolby NR button, the noise was cancelled,” Nair said. “Often, while transferring these songs to a CD or reproducing a digital file, this noise reduction system wasn’t taken into account. At the start of the 2000s, mobile phone did not have much storage capacity, and people ripped songs to 128 kbps bitrate. Thus, an entire generation grew up listening to bad quality songs.”
Today, a music streaming app such as Spotify streams songs at the maximum bitrate of 320 kbps, while the rest stream at bitrates up to 256 kbps. Tidal, which is not available in India, streams music at a maximum lossless level of 1,411 kbps.
In the absence of top-quality audio mixing and mastering technology, there were other problems with the final versions. Nair listed them: “There was obviously distortion. Then, the consistency levels of different tracks inside a song varied. Some instruments would sound too loud, some would barely be heard.”
Another culprit is what Nair terms the “loudness war”, which peaked in the West in the early 2000s. At some point in the ’90s, the music industry believed that the louder a song, the faster it would zoom up the charts. Tunes would be mastered in ways that exhausted the highest audio levels that could be reached through a CD recording. This trend exploded in the 2000s, and reached a crisis point with the 2008 Metallica album Death Magnetic. Its songs were so loud that when they crossed the maximum decibel levels, the loudest parts got clipped, leading to tracks with distorted portions.
“I have been wary of this loudness war in India, which is why with The Mastering Project, I strictly adhere to YouTube’s loudness levels,” Nair said. “If the song is louder than YouTube’s standards, the song will be forcibly compressed and it will sound awful.”
Nair mainly uses Avid Pro Tools, along with bunch of other software. “Today, with machine learning and artificial intelligence, the time you’d need to find a certain flaw in a recording has been reduced from two hours to five minutes,” Nair said. “The company Izotope does that, for example. With the saved time, I have an additional one hour and 55 minutes to work on the error. Then, there’s ADX TRAX, which automatically finds out the vocal elements in a song and spots out what’s wrong where in them.”
The availability of high-end audio mixing and mastering software have led to adventurous but amateur individuals releasing warped versions of professionally mixed songs. One such example is the 8D audio recording, in which sections of the song oddly shift between the left and right channels.
Nair laughed off this trend: “What the 8D does is it changes the intent of the song. Furthermore, there are no eight dimensions in the audio in the first place. But I am not questioning or criticising people who do enjoy such stuff. Maybe they really like it.”