How do I remove vocals from a mix?

This is the most frequently asked question by far. Other variations are:

Can I remove the guitar from a mix?
How do I isolate one voice in a recorded conversation?

What follows is my attempt to provide the most complete and definitive answer to these questions so that we can all refer to it in the future.

The most direct answer to all of these questions is no. You can't really remove a sound from a recording after it's been mixed. There are a few techniques for manipulating these files that may allow you to reduce a sound, but not without unintended side effects. More on that later...

If the sound you want to remove isn't musical, but noise (buzz, hum, clicking, etc.), there may be some hope for you with audio restoration techniques. For the purpose of this FAQ, though, I'm assuming you've got a vocal line or some other sort of musical sound you want to remove.

Think of the mixed audio file as a cake. Just as it is not possible to extract the eggs or sugar from a cake after it has been baked, neither is it possible to extract a vocal, guitar, or any other musical part from music after it has been mixed.

The most helpful suggestion I can give you is this... If you are trying to remove the vocal line from any commercially-available CD to create a karaoke or backing track, you'll save yourself a lot of time if you just go buy the karaoke version of that song. Most of them are available to purchase on a per-song basis for less than a dollar from the iTunes Music Store, Yahoo Music, and other pay-per-download sites. Searching for and purchasing the song will take you five minutes, and the quality will be much, much better than anything you could do yourself with even the best tools available.

Although it is not possible to remove any part of a recorded mix, there are a number of techniques, gadgets, and software applications that attempt to do exactly that, with varying degrees of success. They all use some variation of a technique called center channel cancellation. Here's how it works...

The idea behind center channel cancellation is pretty simple. If you have a piece of stereo audio, you can invert the polarity of either the left or right channel, and combine them. This takes any material common to the left and right channels and cancels it out. Since it is common to place the lead vocal in the center of a stereo mix, this technique can effectively remove the vocals.

But there are some big problems with center channel cancellation. First, although it is common to put lead vocals in the center of a stereo mix, it is also very common to put effects (reverb, delay, etc) on the lead vocal that cover the entire stereo field. So although you might remove the (dry) lead vocal, what you'll leave behind is the reverb tails and the strange remnants of a vocal that's no longer there.

The second big problem is that the lead vocal isn't the only thing placed in the center of a mix. Center channel cancellation removes EVERYTHING in the center. Kick drum, snare, lead guitar, bass. Whatever's there gets cancelled. And even those instruments that aren't placed squarely in the center can suffer from center channel cancellation. Usually what you end up with is a hollow shell that used to sound like music.

This is the simplest form of center channel cancellation. You can do this yourself (if you really want to, that is) using just the phase invert button on your mixer (or in software) or a cable with the polarity reversed. There are also a number of products on the market which use this basic technique:

The Mackie DFX series of mixers...

The Tascam Vocal Trainer CD player:

Center channel cancellation can be improved somewhat by selecting which frequencies you are effecting and combining the phase inversion with EQ filtering. If you know, for instance, that you are working with a female vocal with no meaningful content below 200 Hz, you can allow everything 200 Hz and below (kick drum, for instance) to pass on through and only cancel frequencies 200 Hz and up. There are a couple of other products which use this modified technique:

The Alesis Vocal Zapper:


Perhaps the best known (and most expensive) device of this type is the Thompson Vocal Eliminator:



Some software applications (such as Adobe Audition) also employ this modified cancellation technique.

The success of any of these devices will depend greatly on the music you are working with. Generally speaking, the older the mix, the better this will work. In the early days of stereo recording, producers generally placed the vocals in the center and panned everything else pretty widely. Higher quality files work best. Original CD's and .WAV's work better than MP3's. And since all of this relies on stereo imaging, mono files won't work at all.

I'll repeat the warning that I started with, though... Most people who have attempted this have not gotten satisfactory results, regardless of which tools they use. You'll be much happier with a karaoke track if you can find one. Apart from that, the only real way to do a "no-vocal remix" is to get the original multitrack session and repeat the mixdown sans-vocal.

Mac vs. PC: Which is Better?

"I'm getting a Mac, because they work better with Pro Tools."
"Macs are only for graphic designers...right?"
"I'm not going with Windows because it's just not stable," or "Macs never crash."
"Windows is easier to use."

Ah, myths. Ya gotta love 'em. Chances are, if you've been around people who are involved in computer-based music, you've heard one of these phrases, or some other variation of them. Years ago, in Ye Olden Days of Windows 2000 and OS 9, these may have had some truth to them. Today, however, they just don't stand up as valid arguments.

The core difference between Mac and PC is the operating system, or OS. An OS in general incorporates several things, not limited to:
* A basic set of instructions that controls how software and input devices are handled.
* A formatting system and file structure for data storage.
* A certain list of compatible hardware.
* A way to monitor system resources.
* A unique graphical user interface (GUI).


Some noticeable differences between the two GUI's include:
* OS X has the Dock (program launch bar) on the bottom, which does not exist in XP.
* XP's Taskbar (the bar with the Start menu and clock) can be placed on any of the 4 edges of the screen. OS X's upper Menu Bar cannot be moved, but the Dock can be reassigned to a different side.
* Windows look different in each OS.
* XP incorporates program menus in the individual window, whereas OS X has all menus on the top menu bar.
* Files are arranged in a similar way, but the OS's provide different mean of viewing the file structure.

Aside from these mostly superficial differences, the main question that seem to arise from audio users is how well each one performs: which is faster, which crashes more, which is just "better." The answer is: both. Or, neither. Half-empty, half-full. I like to look at the issue in this way:

Given identical hardware (processor, RAM, front side bus speed, hard drives, etc.), and proper configuration, each OS can be expected to perform identically.

"Proper configuration" means that the OS is adjusted, or "tweaked," toward maximum audio performance. Macs generally come out of the box ready to go; no major adjustments are necessary, although it is generally advised to disable any power saving so your system doesn't shut off during a recording session. XP machines, on the other hand, tend to require a lot of prep work. This is largely due to XP's design as a universal solution. It is meant to be accessible to the widest audience possible, so its preset structure is not the most processor-efficient for some applications. This characteristic has spawned numerous "tweak" guides in various industries. We have incorporated the ones that are most meaningful for audio production into our PC Optimization Guide, available in the Audioforums.com Resource Center. Also, there are a few different definitions of performance that also need to be addressed. "Performance" can be described in terms of raw processing muscle, long-term stability, compatibility across a wide spectrum of simultaneously running applications, hardware options available, and (insert your own definition of "performance here).

Raw Speed

The "speed" of a computer (how fast it completes a certain task) is mostly determined by it's hardware. However, certain aspects of the OS can slow it down. Processor scheduling in XP is one example. With it, you can define what type of processing gets priority, and if you're doing the other type, your performance will suffer. Bear in mind, though, that any any type of processing in your OS will still only go as fast as the hardware allows. Therefore, for raw speed, the CPU, front side bus, and total RAM still play the most important part. Like stated earlier, given identical systems, the two OS's should perform identically. This is especially true of newer Intel-based Macs. Mac Pro's and iMacs utilize the Intel Core 2 Duo processors. The same processor in a PC would yield the same results. Of course, differences in the proprietary Apple logic board and third-party PC motherboards will affect performance, so identical here is not really meant to be 100% literal.

Long-term Stability

In tech support, I see dozens of cases per week involving computers crashing. Only about half are PC's. Macs are just as subject to hardware failure, user negligence, and incompatible drivers as PC's. Hard drives especially, being mechanical device with moving parts, fail after a certain amount of time. it doesn't matter what OS is being used. Long term stability for a computer can be predicted based on the quality of the hardware when the system is built, and the time the user devotes to maintaining and caring for the computer. Computers can be considered like cars in this regard; they need regular maintenance to perform at peak levels for the long term. Having an Apple or Microsoft OS inside really doesn't matter.

That being said, the OS also needs care. Temporary files, registry keys, fragmented data, and old programs can clog up and conflict with current applications. Regular file cleaning is also essential for both Mac and PC. The main difference here is the type of file and the location. While these differ, the underlying principle of the need for routine care applies to ANY computer. Computer maintenance, however, is an entirely different discussion.

Compatibility

Macs and PC's have both undergone a long history of version changes, updates, bug fixes, hotfix patches, and other such enhancements to the OS. As the OS marches on, certain applications get left behind. These need to be redesigned and updated, and sometimes gutted and completely reassembled by the developers. Incompatibilities afflict all OS's, and generally speaking, the more an OS is updated, the more incompatibilities there are. OS X, for example, has undergone 10 major updates since version 10.4 was released. Each update brought the need to update individual software applications for compatibility. XP, on the other hand, has only undergone 2 major revisions ("Service Packs"), but dozens of minor updates, in addition to 5 sub-editions. These editions are Home professional, professional x64, Media center, and Tablet edition, and each bring about their own compatibility requirements. The Service Packs brought several compatibility issues as well, but most of the smaller "hotfix" updates did not.

Another consideration is the imminent release of entirely new version of the OS. Microsoft just released Windows Vista recently, which forced manufacturers to rewrite their entire applications to ensure compatibility. Many have not yet been able to do so, for various reasons. Consequently, not everything is compatible with this "latest and greatest" Windows OS. When Apple releases their next OS version, 10.5 "Leopard," this fall, the same process will repeat itself.

The bottom line is that it is up to the developers to make sure the software or driver is compatible with the operating system. Checking the manufacturer's website or spec sheet will usually tell you what specific OS requirements are need. You can check you OS version in OS X by clicking on the Apple in the top left-hand corner of your screen and choosing "About This Mac:"

You can see that this Mac is running OS X, version 10.4.5. You can also see the processor and RAM specs. On Windows, a similar page can be found in the System properties window, which you can access by right-clicking on the My Computer icon and choosing "Properties" (or pressing the Start and Break buttons simultaneously, or going to Control panel, then System):



This system has XP Professional with Service Pack 2. You can see that it also lists processor and RAM specifications.

Hardware

This is where our ideal comparison falls apart. Hardware is the one major difference between the platforms. Macs all use 2 things that make them Macs: the operating system, and the logic board. The logic board (the motherboard on a PC) is kind of like the vascular system in the computer. It controls the flow of data to and from the processor, hard drives, video card, RAM, and peripheral (USB, firewire, etc.) connections. The front-side bus (FSB) of the board is the major factor in performance of the board. It dictates how fast data transfers. Since the logic board is fixed on a Mac (in other words, you can't add a non-Apple board), it will have an impact on how fast data is able to be routed inside the computer.

Macs also have a very limited processor selection. This is due to the way the operating system and logic board interact. Because of the proprietary way data is handled in the Mac OS/logic board system, the processing of this data needs to be strictly tied in. Therefore, a very limited selection of processor models is available in Macs. Additionally, some models are not expandable (adding PCI cards, for instance), while others are, but to a limited degree. usually, RAM and extra hard drives are all that can be added to a Mac.

PC's, however, are not limited to one board manufacturer. There are dozens out there, each with various capabilities for processors, RAM slots, expansion slots, speeds, built-in or card-added video controllers, etc. This is good because it helps keep overall cost down by the fierce competition. It's bad because you don't always know if what you're getting is compatible with your peripheral hardware. With a Mac, the inside is the same. With a PC, there are billions (perhaps trillions) of different combinations of internal hardware, and is expandable as far the individual component manufacturers allow.

Because of the strictness of Mac hardware and the open-ended designs of PC hardware, and because of the inability of one's hardware to work on the other, there really isn't a way to guarantee that the same hardware will be exactly comparable. In general, though, a Mac and a PC that both have, for example, a Core 2 Duo at the same speed, 2 GB of 667 MHz DDR2 RAM in dual-channel mode, and identical hard drives will be expected to perform, speed-wise, about the same. Of course, with PC's you have to look at the FSB speed of the motherboard, and make sure that your bus speed and RAM speed match. Macs are designed with this internal compatibility automatically. PC's also tend to have a lot more general incompatibilities, mostly due to the inability of manufacturers to test all 7,456,294,912 (or so) hardware combinations, vs. the 10 or so current Mac models. This is not the fault of Windows in general. This is simply due a competitive market and ever-evolving technology.

OK...So Which IS Better?

Given the ideas that identical hardware means identical performance, and that Macs and PC's both crash, and that the main differences in the actual operating environment are mostly superficial, and that both require care and maintenance....which do you choose?

There are actually several schools of thought on this question. Some people believe that the Mac OS is easier to learn if you've never used a computer before. Others contend that Windows-based computers are less expensive, so users will be able to get a better system for the same cost. Still others (myself included) believe that it is best to decide what software you will be using, and then make your choice based off that. In other words, the tools should be the first priority. Some applications are Mac-only, while others are PC-only. Some examples of PC-only software include:

* Sonar
* Samplitude
* ACID (and the entire Sony line)

Mac-only software includes:

* Logic
* Digital Performer
* GarageBand

Others, such as Pro Tools, Cubase, Live, Reason, and others are cross-compatible. Concerns such as interface compatibility, plugin compatibility, and version compatibility will need to affect the choice you make. If everything you plan on using is cross-compatible, then other considerations need to be made. Will you be using the machine as an office computer as well, or be using other programs on it? What kind of video display do you want, and can one system handle that better than another? Is price a factor? Compatibility with other musicians' work? If you've used both in the past, which is more comfortable for you? Does one's GUI just look and feel better?

I'm Getting a (insert computer here)

Hopefully, this helps you discern some of the difference between Macs and PC's, and dispel the myth that one is better than the other. In the end, "better" is a totally subjective word. What is "better" to one may not be "better" to another.

What is Word Clock?

This is a question I get a lot. Word clock is one of the most commonly misunderstood and enigmatic topics in the studio. Here's my explanation...

Every digital device, from a simple portable CD player to a Pro Tools HD rig, has a word clock. Not to be confused with sync clock (like MTC or SMPTE), word clock is what tells a A/D or D/A converter when to take the sample. It fires an impulse to the converter 44.1k (or 48k, or 96k, etc) times per second. The reliability of this clock, how evenly spaced those pulses are, determines the accuracy of the conversion process.

If you have digital connections in your studio (S/PDif, lightpipe, etc), you MUST have a single clock source. If multiple devices are each trying to be the master clock, you will probably experience pops, clicks, or sometimes chirps in your audio. The solution is usually to make sure that you only have one master clock, and that all slave units are getting a clean clocking signal.

But even without obvious problems like pops and clicks, you may still have more subtle word clock problems. Think of it like this...

Ever seen clips from the very early days of film? Back in the days of hand-cranked cameras, it was extremely difficult to get the shutter speed to be perfectly even. The result was a moving image that, when played back at a constant shutter speed, seemed "jittery." the motion was often jerky an unnatural. This was due to the imperfections of the original film speed.

The same thing happens in digital audio. The more "jitter" present in the clock source, the less accurate the resulting conversion will be. Jitter is manifested in audio as phase distortions and discepancies, particularly in the higher frequencies. They can lose some of their sparkle, transients can lose some attack, and stereo images aren't as vivid.

Improving the quality of your clock is an easy way to upgrade every A/D and D/A converter in your studio. In most cases, adding a high-quality master clock, such as Apogee's Big Ben or Antelope Audio's Isochrone OCX will result in a marked improvement in the overall quality and clarity of your finished product.

But what if you don't have a master clock? One of the hallmarks of Apogee and other high-quality converters is usually a low-jitter clock. If you have a great converter in your system, you can feed that clock source to everything else with an A/D or D/A converter on it (like an audio interface, for example). Replacing the built-in clock of your interface with the clock signal from the master clock will dramatically improve the quality of the conversion being performed by the interface. There are still other factors at play, so it's not identical to running the same signal through the high-quality converter, but it is usually a very noticeable improvement.

What is white, pink, blue "brown" noise?

What is noise?
noise is random frequencies at random sound pressure levels or random voltage spikes populating a given spectrum.
you are all familiar with the static on your tv on channel (pick a blank channel) it is like light, white noise is like white light it has all of the frequencies that we can hear (even the ones that we cant hear).
a calibrated white noise is random frequencies in a given frequency range that all have the same loudness. the graph would look flat across the board.

Your static on your tv is not a calibrated white noise, it is a random white noise and cannot be used for calibration because it is inherently colored (this color can be graphed)

so what is pink noise or red noise?
is there really such a thing as brown noise?
how about blue noise?

well yes, you can have any color of noise just like you can have any color of light.
by adding more "volume to higher frequencies, you can generate a blue colored noise this type of "hiss" would sound like it has more treble added to it.
some scientific research would use blue noise to calibrate instruments, also, speaker manufactures would use different colored noise to calibrate their loudspeaker cabinets.

here is a graph of blue noise

So what is all of the Hoopla about pink noise?

well pink noise is special
when you have a sound system you need to calibrate it to the human ear.
the human ear it turns out has a frequency response just like a microphone (see frequency response).

if you were to graph the frequency response of a healthy human ear, you would get something like this:

see how the frequency response dips in the higher frequency range and also note how there is a peak in the lower end of the spectrum?
in pink noise, also called "1/f noise" is a signal or process with a frequency spectrum such that the power spectral density is proportional to the reciprocal of the frequency. WHA???

basically, pink noise has more energy on the frequencies added to the low end to compensate for the humans frequency response to "flatten it out"

for more in depth information on noise, look here
http://en.wikipedia.org/wiki/Noise_%28physics%29