Last updated September 28, 2006
Warren W. Gay VE3WWG
This FAQ will largely focus on recording guitar/mic digitally to your PC for producing digitally mixed content. There may be other bits of info that may be useful for other recording situations, however.
For recording classical, consider getting yourself a couple of microphones connected to a small desktop mixing console. Connect it to your sound input device's line-in jack. If your acoustic guitar also has a pickup, run that into your mixer as well. Record the mic and pickup simultaneously to allow yourself some mixing flexibility after the recording is done.
Yamaha also makes a small mixing desk with a sound card built in and a usb port to connect to any PC with a usb port.
Before you set out to acquire hardware for recording, you should answer these questions. The answers to these will help you to determine what your needs are.
If you are recording yourself, your needs are going to be rather simple. However, if you are planning to record a live performance by two or more musicians, you will need the ability to record several tracks simultaneously. This will influence the choice of audio gear you acquire and ultimately your PC soundcard hardware.
This FAQ will mainly focus on 1.1 above, for the purpose of creating CD content.
This largely depends upon how you want to record and what you want to produce from the recordings.
Be sure to read :
This FAQ will focus on mainly on recording on a PC for the purpose of creating CD or DVD content. It will also be assumed that most people reading this FAQ are interested in doing multi-track recordings of their own performances.
If you are recording yourself, one track at a time, then you want to do multi-tracked recording. You might lay down a drum or rhythm guitar track first. Then you would have your PC play back that track, while you simultaneously record (add) a new track, like lead guitar or vocals. The next time you play it back, you can then hear both parts (tracks) played together. It would sound as if you recorded them at the same time.
Note that multi-tracked recording is not limited to singular tracks. It is possible to record 2 or more tracks in multi-tracked fashion. A singular track is often what the small home studio user will do, however.
While most modern PCs include some capability to record and play sound, you may wish to consider buying better equipment for recording purposes. This depends upon the quality you want and how much editing you plan on doing.
Standard PC hardware usually includes support for recording and playback of 16-bit sound samples. This is good enough for typical PC use, but is generally considered inadequate for recording purposes (it depends again on your needs and goals).
While your production target may be 16-bit samples (for CD quality), recording in 16-bits has its limitations. Furthermore, digitally mixing two or more tracks of 16-bit samples introduces more numerical error (introducing noise). This error can be minimized by using sound cards/devices capable of wider samples (like 24 or 32 bits).
Even if your PC posses a 32-bit sound card (or larger) you may still need to acquire new hardware. Your sound card must work in "full duplex" mode if you want to do multi-tracked recording. Some PC hardware cannot do this. Sometimes the limitation is in the software. In either case, you want to acquire PC hardware that can simultaneously record while playing sound to do multi-tracked recording.
Your sound card should be capable of at least 44,100 samples per second in stereo (most are). It is better however, if your sound card can sample faster at some multiple of this (88,200 for example). This allows you to record higher quality samples for digital mixing operations, before you mix it back down to CD quality levels (44,100 samples/second). Whether you will record at higher sample rates or not will depend upon how fast your computer is and how much disk space you have.
A guitar pickup can produce a signal ranging from 100mV RMS to 1 Volt RMS [1]. A Les Paul with moderately hot humbucking pickups will produce a signal of about 425 mV [2]. The typical level of a guitar pickup might be considered to be approximately 600 mV for passive pickups.
The output of a professional quality microphone (like a Shure SM57 or PG58) is very small. The output signal level is less than 1 mV (1/1000th of a volt). [3]
The line level inputs are designed to accept line level signals. In short, a signal of about 0.3162 volts RMS.
The output signal typically ranges up to 2 Volts RMS. Output impedance ranges from 10 ohms to a few hundred ohms (30-400 ohms is typical). This allows the output to be connected to consumer electronic equipment with line level inputs (normal HiFi amplifier with -10dB nominal level line input) or powered computer speakers. [8]
It is wired on a stereo 3.5 mm jack as follows:
| Connection | Description |
| Tip | Left audio channel |
| Ring | Right audio channel |
| Sleeve | Ground (cable shield) |
Sometimes you might find RCA jacks instead, where they will be labelled left and right (or just white and red respectively).
The Mac input port accepts an audio signal in the 100 mV RMS range. [4]
Most PC sound cards mic inputs accept a signal of approximately 10 mV (10 millivolts). [3] Older 8-bit audio cards often required 100 mV of signal.
Some sound cards do however, allow the user to double, triple or even quadriple the sensitivity of their mic input (using some software control panel). This can add significant amounts of noise.
Given that a professional microphone provides 1 mV signals (see here), you'd have to shout your loudest just to register any signal at all in your PC (unless you are able to multiply the gain as described here).
Preamplification is normally the best option with a mic preamplifier.
See "Connecting the Microphone to a Sound Card" here.
Condenser microphones usually require a phantom supply voltage (PC sound cards do not provide this). Shure documents some work-arounds under "Adapting Condenser Microphones to Operate on Voltage from the Sound Card", at this link:
One option is to get a mixing console (mixer). Usually one or more microphone inputs is provided on a mixer.
When doing professional recording however, you may want to get a dedicated mic preamplifer instead. This will help to reduce noise and provide a cleaner signal. Some preamplifiers provide "warmth" to the recorded signal by applying tube circuits in place of solid state amplifiers.
Not really. Part of the answer lies in the guitar signal output and the sound cards expected input level.
The typical sound card interface for the PC, whether it be a separate sound card, or builtin to the motherboard, consists of the following jacks:
|
Colour |
Description |
|
Green |
audio stereo line output for speakers |
|
Blue |
audio stereo line input |
|
Pink |
audio mono microphone input |
The jacks are usually 3.5 mm (1/8 inch) stereo jacks (except for the mic input, which is usually mono).
Mic inputs on PC sound cards are generally optimized for telephone type voice applications and are usually "pretty crappy". [7] Hence, this input is generally designed for use with a "computer microphone" and may not be suitable for serious recording.
Some mic inputs are stereo, like that provided by Blaster soundcards (SB16, AWE32, SB32, AWE64, Live) from Creative Labs. The signal is still monophonic, but the connections are as follows:
| Connection | Description |
| Tip | Microphone signal input |
| Ring | +5 Volts (3 - 5 Volts) |
| Sleeve | Ground (cable shield) |
The +5V on the ring is current limited by a 2.2 kohm resistor within the soundcard. It may vary between 3 to 5 volts with no mic connected. This is designed for use with electret microphones (PC99 standard), but will work with other microphones using a 2-conductor (ring and sleeve only) plug. In this case, the ring (+5V) and sleeve are shorted together without causing harm.
According to the PC99 standard, the input AC impedance is a minimum of 4 kohm, but 10 kohm is recommended. The standard further specifies that 10.100 mV should deliver full-scale digital input.
Fancier sound interfaces may sport additional Mic or line inputs, surround sound outputs and earphone jacks. Some may also have digital interfaces like S/PDIF for digital receivers/surround sound (YELLOW on Sound Blaster Live cards). [5]
Others may have an orange and a black jack (surround sound) [9] : The PC99 standard also adds a brown jack.
|
Colour |
Description |
|
Black |
Analog line level output for rear speakers |
|
Orange |
S/PDIF digital output (sometimes used as an analog line output for a center speaker instead) |
| Brown |
Analog line level audio output for 'Right-to-left speaker'. |
A 3.5 mm stereo jack consists of a tip, ring and a sleeve (TRS). The input stereo audio connections are as follows [6]:
| Connection | Description |
| Tip | Left audio channel |
| Ring | Right audio channel |
| Sleeve | Ground (cable shield) |
The following link has some good information on the subject.
While most sound cards support MIDI output, some will not support MIDI input. C-media is notoriously bad at this. Apparently the hardware is capable (Linux can support it), but the Windows software drivers provided will not support it.
The following link has some PC soundcard information.
"Depending on the application, line levels are stated in units of either decibel volts (dBV) or decibel volts unloaded (dBu). Consumer audio equipment line levels are rated in dBV, and the most commonly used reference level for such equipment is -10 dBV, which corresponds to a signal of about 0.3162 volts RMS (this is 0.8944 volts peak to peak).
Professional audio equipment line levels are rated in dBu, and the most commonly used reference level for such equipment is +4 dBu, which corresponds to a signal of about 1.228 volts RMS." [10]
There is a handy dB Conversion Tool here.
For the purposes of this FAQ you need:
This varies rather widely depending upon your recording and editing requirements. It will also vary with the operating system being used on your PC/Mac.
These are general guidelines, and should not be construed as absolute limits. This part of the FAQ is likely to be the most quickly out of date, due to the pace at which PC features develop. Finally, this FAQ is Windows centric, even though Linux and MacIntosh may represent other perfectly acceptable (even superior) choices.
We'll consider three major categories of PCs:
This is not a complete list. The author of this FAQ has used the Audiophile 2496. Others have used other cards that may be listed here. If you have information about other good choices to be included here, please email the FAQ author.
One popular choice for small home studio use is the M-Audio Audiophile 2496 PCI card. Visit this link for more information.
SoundBlaster Live :
See [12] for a description of some of its problems.
[13] has a summary review
Some good information about the basics of DVD surround sound can be found here:
There are many digital audio file formats today. Visit this link to view a list of them and their file suffix types. When recording and when saving tracks for later editing and remixing, it is always best to save them in an uncompressed (or lossless compression) file format. The following section about MP3 Files discusses why. On Windows platforms this usually means that you'll be using a WAV file format.
The Wave file format is a subset of Microsoft's RIFF multimedia file format. There are some nitty gritties about the WAV file here. It is important to remember that WAV files can contain various sound sample formats including compressed audio (ADPCM for example). Uncompressed WAV files store the audio in PCM sample data.
Normally however, when saving data to a WAV file, the data is stored as uncompressed audio (PCM) unless you specify some compression algorithm to be used. But you should be careful to check this in case your software tries to be smart and default to using a compression technology.
As you would expect, uncompressed audio data requires more disk space. To calculate how long you can record, go here.
MP3 format is ok for MP3 players and other end-user applications as long as the listener is willing to accept the degradation in the audio signal. As a studio editing format, it should be avoided at all costs because the file's compression is obtained at a slight loss of information. This translates to signal degradation.
What this means is that every time you open an MP3 file for the purpose of editing it and resaving it in the MP3 file format, the signal will get progressively worse (more noisy/distorted).
For editing purposes, save your audio data in a lossless format such as a WAV file. This requires more disk space, but it is the only way to preserve its audio quality.
If an MP3 file is your desired target for production, then produce an MP3 from your final mix. This should be the last step in your production, and the result should never be re-introduced into your sound editor. In other words, always retain your master files in case you need to make adjustments to your production at a later date.
Some FAQ information about MP3 files can be had at these resources:
If you are tight on disk space, then you're going to be very concerned about how fast a recording session is going to consume it. Or you may have plenty of space, but you're concerned because you want to make a long recording. Sometimes software will tell you how much recording time you have left (by disk space). But if you need to estimate it for yourself, you need to know how to calculate it.
This section is going to assume that you are recording uncompressed (PCM) data to a WAV file. There is some overhead information to manage the blocks of data recorded, but this is insignificant compared to the recording itself. So we will simply ignore the overhead information.
To calculate the disk space required, you need to know:
If you are recording in unusually sized samples like 20 bits, then you need to round each sample to the nearest multiple of 8 (computers like to have a minimum size of a byte, which is 8 bits). With this information at hand, we'll use the following symbols:
The number of bytes B consumed by each sample is this:
B = ( S / 8 ) * C bytes / sample
So if the sample size was 24 bits, in stereo, there would be:
B = ( 24 / 8 ) * 2
= 6 bytes used for each sample
In one second then, a recording would consume this many bytes of information:
B = ( S / 8 ) * C * R bytes / second
So if we recorded 24 bit samples, in stereo, at 44,100 samples/second, we would consume this many bytes in one second:
B = ( 24 / 8 ) * 2 * 44,100
= 264,600 bytes / second
= 258 kb/second (approx.: Note 1 kb = 1024 bytes).
If you want to calculate this in terms of GB (G) then adjust the units accordingly as follows (1 GB = 230):
G = ( ( S / 8 ) * C * R ) / 1,073,741,824
Simplifying this algebraically, we arrive at:
G = S * C * R / 8,589,934,592
or close enough as:
G = S * C * R / 9,000,000,000 bytes/second
If we have 4.7 GB of space available on our PC, what is longest stereo recording we can do, for 24 bit samples, at 44,100 samples / second?
S = 24 bits (3 bytes per sample)
C = 2 (stereo is 2 channels)
R = 44,100 samples/second
Now let's compute the cost of each second recorded: in terms of GB:
G = S * C * R / 9,000,000,000
= 24 * 2 * 44,100 / 9,000,000,000
= 0.0002352 GB/second
Since we have available 4.7 GB (A) available, the maximum amount of time (T in seconds) before we run out of disk space is then:
T = A / G
= 4.7 / 0.0002352
= 19,982 seconds (approx)
= 5.5 hours (approx)
If you have time to kill, visit these sites: