How is music remastered?

Factual Questions
1

Just curious as to how music can be remastered long after the original recordings were made to add more “depth” to the music or “re-equalize” it to give particular instruments or the vocals more presence and depth than the original releases. We’re starting to see a lot of this with Dolby 5.1 recordings and the SACD and DVD-A formats.

So, how do they do this? Do the recording studios maintain separate tracks of each musician that they can go back to and re-mix? If so, how can they create today’s digital audio fidelity using material that was recorded on older analog equipment? For example, the Led Zeppelin “remasters” album contains some material that is almost 30 years old, yet sounds great on a CD player (to my ears, anyway).

How do they re-issue any older material in a digital format without capturing “flaws” such as hiss that must exisit on the original recordings? How are 5.1 soundtracks, SACD and DVD-A formats created using such high sampling rates if the original recordings were limited by yesterday’s technology.

Just curious - I did not believe that there was any way to create a better fidelity recording off of a lesser fidelity recording…

2

Yep.

Studio quality analog equipment is much, much better than cheesy 1/4-inch cassette tapes that you’re used to. The sound is recorded on wider tape at faster speeds, to provide the maximum surface area to record on. A good analog studio master sounds nearly as good as a top quality CD does. Of course, up until a few years ago, most CDs came from analog masters, not digital studio recordings.

Digital interpolation can be used to great effect to “fill in” missing bits of sound and remove background noise, though in order for it to work the source recordings must be in pretty good shape to begin with.

3

While they can do a lot, there is onl;y so much that can be done. Read the fine print of most re-mastered material, it has a nice little explanation of this.

If the source analogue track has a hiss, that can’t usually be removed. It can be diminished, or you can play tricks on peoples ears’ by bringing up some of the other sounds.

What gets me is when I see a “digitally remastered” version of something that’s only been out in the last 10 years or so. WTF?

4

First off,

When artists recorded their material, they used what are called original master recordings.

The quality of these recordings is awesome, far far better than anything laid down on vinyl.You would be amazed at just how good studio recordings from the 1950’s can be.

Next those original master tapes are mixed onto another tape, the vinyl master tape, these too are very high quality, but, part of the mixing process is to address the limititations of the vinyl medium.

The faster any medium passes a replay head, the higher the quality of sound is that can be reproduced, original master tapes pass very quickly, and they are also very wide, meaning that there is a large area of media available to contain music information.
The same is true of vinyl, the faster the rotation speed, the better recording is possible.

Vinyl speed slowed over the years, from 78 down to 33, and some folk might even have 16rpm discs.This was made possible by improvements in recording technology, and disc manufacture, but the result was that dynamic range suffered.
To get around this a set of compressions standards (RIAA compression) became the industry standard, and it basically squeezed music into a narrower dynamic band that could be laid down on vinyl, and all turntables were built with amplifiers in them that could expand this signal range.

Even so, vinyl still could not hope to match studio quality. Vinyl master tapes are mixed down with all the RIAA equalisation characteristics on them, and also account for other limitations of vinyl.

When you master for a CD things are very differant, you don’t need all the jiggery pokery to mix the signal, a CD can cope with original master recording quality.

Artists and engineers will try to go back to those original master tapes if they can, and being inveterate meddlers, they will try to improve on this, knowing a CD can handle it.
Original Master recordings have a very neutral sound, they don’t emphasis things, nor take anything away, and just like the conductor in an orchestra, artists and engineers will want to bring certain instruments or voices to the fore at certain times, or they may spot things they wish they had corrected in the original recordings such as the odd bum note.
Lots of this can be edited by converting the original master tape to a digital format, and use software editing.

Sometimes the origianl master recording is not available, or damaged, and they will go to the vinyl masters, which then means trying to recreate those original tapes by reversing the effects of the vinyl mastering process.
This is not an easy task and sometimes there is quite a lot of artistic interpretation of what was intended, the result is then a CD master.

5

That covered the op very nicely, however, I feel it’s important to dispell common misconceptions some people not familiar with pro recording have.

CD quality isn’t that great. For regular, everyday, boom-box or whatever use it’s probably as good as anyone’s going to need. Master recordings, however, analog or digital, are of far higher audio quality than a CD. This means that even when mastering a CD, you need to make adjustments to account for the limitations of the medium. Sure, it’s more straightforward than mastering an LP, which is truly a daunting task, but it’s not something any dummy can do either.

6

Hmm. So, what you seem to be implying is that, in a few years, someone will invent a new medium for selling music and all music will have to be remastered again for that new medium.

7

Well, that’s certainly what many people in the record industry are hoping. First you bought the LP because that’s all there was. Then you bought the cassette because you wanted to play it in your car. Then you bought the CD because it sounded much cleaner. They tried to sell you the DAT, and the MD, and now, maybe the Super Audio CD. Of course, SACDs sound a lot better than plain old CDs, which means that they’ll need to go back to the masters to account for the improvement.

Of course, those new formats are mostly a ploy to get people to buy music they’ve already bought again. The same can definitely be said about many remasters.

However, my point was just that a) CD quality is nowhere near pro quality and b) digital audio mastering isn’t just a question of copying data to a new format.

8

“If the source analogue track has a hiss, that can’t usually be removed.”

The hiss can be removed, and pretty easily at that. It’s just that removing the hiss usually involves removing musical information as well, so many feel it’s better to leave the hiss in and maintain the dynamic range of the music.

The same goes for the surface noise on some masters from the pre-tape era. It can be removed digitially, but the price is loss of detail in the music.

For the pre-tape era you can often get better results mechanically restoring a 78 commercial copy than you can get applying a digital scrubbrush to the master. This is proven by the career of my own personal remastering god, John R.T. Davies, who has done some astonishing things:

http://www.jazzrescue.com/faq's.htm

9

Also, keep in mind that the first batch of cd reissues (from the 1st 10 years or so of cd’s) were ‘remastered’ very poorly. Actually, many weren’t really re-mastered at all. A good amount of reissues were recorded to digital by playing an un-opened, clean record on a high-end deck directly into a digital recorder.

Most of the current re-issues are remastered directly from the master tapes that were recorded by the musicians. This is an extrememly time consuming process. You need to digitize all the analog tracks, process each track (to remove noise, or other artifacts), and then completely remix the album. Some albums are remastered from a high-quality stereo mix-down track. This can sound good, but removes the ability to digitally edit each track. Also, to create a 5.1 or surround sound disc, you need to go back to the original multitrack.

10

I correct another small piece of misinformation, audio master are not substantially wider than home open-reel machines. The tape is wide (two inches is the largest I’ve ever seen, but there could be wider somewhere. But they fit many more tracks.

Home audiophiles gererally used what was called 1/4-track open reel recorders. On a 1/4 in tape, you would record a total of 4 tracks. two each direction - so you only had two record heads and two playback heads. There were other head configurations that would allow you to record 2 or 4 tracks over the entire width of the tape (in one pass) for higher fidelity and more mixing options. 8 track machines use 1/2 in tape.

The largest studio mastering machines I ever saw used 2-inch wide tape and recorded 24 tracks, which does in fact make the tracks slightly wider that the others. But the main difference was the tape speed. Tapes would typically be mastered at 30 ips (inches per second), and sometimes 60 ips. A typical home machine recorded/played at 3-3/4 ips, or maybe 7-1/2 if “-philest” for the audiophiles. Remember, doubling the speed also doubled the cost of the recording tape.

At 7-1/2 ips, a 10’ reel (a standard selling size) of 1/4 inch tape would record for about 60 minutes. A good professional quality tape of that size cost about $12*

At 60 ips, the same amount of tape only records about 8 minutes. 2" tape cost more than a hundred bucks per reel. Gets pricey real fast.

*These prices are about 20 years out of date.

11

Huh. I’ve noticed that some of my older CDs are “quieter” than a lot of more recent ones – meaning you have to turn the volume up much higher. Could this be the reason older CDs are much harder to hear?

12

I wouldn’t think the volume would be related to the quality. The engineer should have recorded the cd louder IMO. Recording music too quiet does limit the sound quality though… Picture a digital wave… With digital there is a fixed number of ‘steps’ vertically thru the wave (referred to by the ‘bit’ size - 16 for CD). If the wave only takes up 50% of it’s total verticle limit, then it is only using 1/2 of the verticle ‘bits’. The same sound, recorded the loudest possible (without clipping), would use all the verticle ‘bits’, providing more clarity. BTW, the sampling frequency, (44,100 for CD) refers to the horizontal ‘bit’ size (44,100 per second!).

13

This is more the result of a recent trend of recording engineers compressing the dynamic range of recordings and artificially boosting the signal level. This makes the music sound louder but can result in the clipping of peak levels and giving a flatter sound. Listen to Queens of the Stone Age’s Songs for the Deaf for an extreme example of this.

14

30 years? My god! How did they cope with that ol’ Stone Age equipment?

The moral is: newer is not always better. Do not underestimate analogue recording technology - decent studio tapes capture a very high quality sound.

15

Remastered is more a term of art than of description. Many audio recordings, including the first one ever made (IIRC, it was Edison reciting "Mary Had a Little Lamb) are digitally re-recorded, and remastering refers to the attention and effort paid to the process. If the original source is good enough, or conversely if the people footing the bill are cheapskates, “remastering” can involve little more than plugging a playback unit into a recording unit, balancing the two machine’s input/output levels against a reference tone, and hitting record and play.

OTOH, the only limits to how you can play with the source material are your own imagination and market constraints. You could reduce “Hotel California” to only those notes where Joe Walsh’s guitar plays middle C. Joe being on an isolated track in the original would make it fairly easy, but you could get a fairly close approximation of the same thing with some more signal processing from a CD stereo mix. You could also take an audio narrative originally recorded on a single channel (mono), and give it apparent location and movement by mixing it among the six channels available to most home theaters, maybe adding some ambience recorded elsewhere to define a space for the performer in the listener’s mind.

16

Edison’s original recording from 1877 of him reciting “Mary Had a Little Lamb” no longer exists. Purported recordings of it are actually latter-day recreations.

17

The medium has already been invented. Optical and magnetic media are obsolete.
It’s going to be solid state. Look at all the usb flash drives,portable mp3 players and flash memory cards for digital cameras. IMHO, it’s just a matter of the companies all sitting down together and agreeing on a physical size and format. And what sort of reader one would use. Well, there’s some more tweaking to be done also, but my call is read only memory cubes (or something) are in our future.

18

This is incorrect. The RIAA standard is for equalization, not compression. The RIAA recording curve reduces bass and increases treble. The bass reduction reduces the wave size for loud bass passages, making them easier to track. Increasing the treble helps to swamp out the hiss in the recording medium. The RIAA playback curve simply undoes the effect of the recording curve.

Before the RIAA curve was introduced in 1953 there were many equalization curves. Every record company had its own, and they tended to change over time. Equalization was introduced in 1925 with electrical recording, and the amount of equalization tended to increase over the years. Some older amplifiers had controls to select the equalization curve. Modern amplifiers all have the RIAA curve built into their phono amplification circuits.

By the way, 33 1/3 LPs and 45s generally have better dynamic range and frequency response than 78s. Although the speed of rotation is slower, 33s and 45s have much narrower grooves, allowing for improved frequency response (a smaller groove and stylus allows for shorter wavelengths to be recorded). 33s and 45s were made of vinyl (and sometimes polystyrene), while most 78s were made of shellac (although many post-war 78s were made from plastics such as vinylite). Vinyl is a quieter medium, which allows soft passages to be recorded and played back successfully. Also, improvements in recording and playback equipment allowed for improved frequency response and dynamic range.

19

Actually, it doesn’t affect dynamic range so much as remove signal. Dynamic range is the ratio between the loudest and softest signal. Overly agressive noise reduction tends to remove frequencies that are part of the music, making for a dead-sounding recording. That is, the top end will be missing and some detail will be gone.

Hiss reduction can work well as long as the remastering engineer isn’t too aggressive. It’s possible to improve the sound of a noisy recording using digital noise reduction - I do it all the time myself with my collection of 78s.

20

My a cappella group recorded recently, and probably since I’m a freshman voice major, most of the terminology in the equipment room went right over my head. <Our group leader was smart and brought in the guys two or three at a time, randomly, not making the artistic decisions autocratically> There was a mic for each section, two solo mics, and two room mics, making eight tracks. Apparently, last year the group tried isolation booth recording for each part, but we couldn’t make music that way. Singing along to recorded earlier takes of other parts in headphones didn’t cut it for tuning, toning, timing, dynamic matching, etc. Perhaps professional musicians can do that - maybe I can call myself one in a couple years.

The isolation booth thing is exactly that - you do a preliminary recording as well as you can, then put each part into a soundproof box to do many takes without any sound from the other parts leaking into the track. As it is, with twenty guys standing in a circle, each track has bleedover from the others. Directional mics help, but they’r not perfect. It’s worse for sections that have to be quieter, as their track can have half the sound coming from other sections <go baritones!>. With multiple independent tracks, with several takes, you can splice together the tracks with the fewest mistakes.

That’s all I know about recording, besides some book larnin’.