Fullscreen HTML5Med Res (??MB)HTML4

Home Recording Studio By MICHAEL GOEBEL Learn how to produce audio CDs for musicians at a fraction of the cost of a recording studio Want to make good-quality audio recordings in your own home? All you need is a PC, a hifi system, a good-quality microphone and some software. D UE TO THE PRICES charged by professional recording studios (some in the region of hundreds of dollars per hour), many talented soloist and group musicians cannot afford the cost of producing their first CD. What’s more, attempts at recording an entire performance using a single microphone invariably lead to very 10  Silicon Chip poor results – results that are so bad that reviewers cringe when hearing even the most promising artists. But there is a way out. The latest PC recording software makes it possible to obtain surprisingly good results using relatively simple equipment in the comfort of your own home. Before describing how you go about making such recording, let’s take a look at the equipment you need. The list is as follows: (1) A standard desktop computer (the faster, the better); (2) A CD writer (writing speed not critical) and CD writing software; (3) A good quality sound card (at least 16-bit) or on-board sound system; (4) A good quality desktop microphone; (5) Adobe Audition 1.0 (Win98 users can use CoolEdit and/or CoolEditPro) or Diamond Cut; (6) A home stereo system capable of good performance; and siliconchip.com.au (7) A hifi “sampler CD” of known high quality and containing material familiar to you There are also a few important things to consider before you begin: • This project is intended for those who are fairly proficient in the use of computers, Adobe Audition 1.0 and have some experience with audio. • It is extremely unlikely that you will achieve adequate results using a bargain-basement laptop and/or a $99 combination CD/radio/cassette home “hifi” system. As with all systems, the chain is only as strong as its weakest link. • This project is much more easily carried out with the help of an experienced musician, preferably someone who can play an instrument (eg, guitar) and sing. • Screen resolutions of less than 1280x1024 will make it difficult to control and organise the program windows used in this project. • An absolute minimum of background noise is essential to the re­ cording process. If your computer has a noisy fan, then you should attend to this before embarking on this project. Also, remember to take the phone off the hook and switch off your mobile phone prior to any recording. • The objective is here is NOT to strive for bass levels that will rattle the windows or searing treble that causes bleeding from the ears. Instead, the aim is to achieve a well-balanced, pleasant and realistic level of sound. • Many stereo systems possess lots of “effects” options and it is vital that they be bypassed (disabled). The same goes for any “effects” on the computer (eg, “3D Depth”, “Stereo Enhancement”, “Surround Sound”, etc). • For best results, all file operations for this project should be done in WAV format. Avoid the use of MP3 or any other type of sound file extension. • If you are having trouble getting sufficient gain through the microphone, go to the volume control panel (Fig.1), select “Options”, “Advanced Controls”, then in the microphone section, click on “Advanced” and select “Mic Boost”. • In Adobe Audition 1.0, enable (tick the box marked) “Live update during recording” under “Options”, “Settings”. This will greatly facilitate monitoring the recording process. • While this project only uses a siliconchip.com.au Fig.1: this is the Windows volume control mixer panel. Set both the “CD” and “wave” level controls to about 80% and the “master” or “volume control” to zero. Fig.2: the Windows volume control mixer panel properties window. Be sure to select the Microphone option for recording. 2-track performance as an example, the principles and methods outlined can be extended to include up to 128 performers and/or instruments, by switching to the “multitrack” mode of Adobe Audition. Marrying a PC to your hifi The purpose of this step is to ensure that the backbone of the system (ie, the computer and stereo system combo) is functioning properly. First, connect the PC’s sound card output to the “aux” input of your stereo system. Avoid using a cable longer than necessary or one of poor quality. That done, boot the computer and once you are in Windows, power up the stereo system, select the “aux” input and set the system’s volume control to roughly 10 o’clock. Now locate the computer’s volume control panel (Fig.1), set both the “CD” and “wave” level controls to about 80% and the “master” or “volume control” to zero. Set all other level controls to zero. Now play a CD through the computer’s CD player, preferably one that has been well recorded and with which you are very familiar. Gradually increase the computer’s main volume control and adjust the volume control on the stereo until you obtain maximum volume from the system (before the onset of noticeable distortion) when the computer’s main volume is at 100%. Listen to the track in its entirety, November 2007  11 Fig.3: opening two instances of the Windows volume control mixer panel – one for playback and one for recording. preferably from a comfortable position in the room’s “sweet spot” and assess the quality of the sound. Important: avoid setting the “wave” level control to greater than about 80% – this is likely to cause overloading (clipping) of the computer’s audio system. If your stereo system has a CD player, it is a good idea to compare the performance of the two (ie, with the CD ROM drive in the PC). If you cannot detect any audible differences between them, then this is a good sign that the computer’s audio quality is adequate (it may also be a good sign that you have no ear for music, however!). Now switch back to the computer and try different tracks and different CDs. If necessary, “fine-tune” the system with the stereo system’s bass and treble controls (or equaliser) but avoid using any more boost or cut than is absolutely necessary. Do not use any equalisation facilities (or “effects”) that may be available on the computer. Ensure that these are disabled. Testing the PC’s audio A number of computer audio per­ formance testing programs are available, with RightMark Audio Analyser (http://audio.rightmark.org/download. shtml) being the best tested by the author. This program will test the playback and recording capabilities of your PC’s sound system and report such specifications as frequency response, distortion, hum and noise, stereo separation and more. While considerable variations of performance will be found between various types of motherboards and sound cards, you should rely on your ears to make the final judgement. It is not much use having a system with superb specifications if it is not pleasing to listen to (it does happen). If you are intending to use a laptop computer for this project, then it is highly recommended that you evaluate its performance using a program such as this. Some laptop computers, even fairly expensive ones, have relatively poor audio performance specifications and hence, poor sound quality. Treating room acoustics To measure the overall characteristics of the entire system (or of any audio system, for that matter), just follow these simple steps: (1). Generate a 20Hz to 20kHz sine­ Fig.4: setting the format of a new file in Adobe Audition. Choose the mono, 16-bit & 44,100Hz sample rate options for recording. 12  Silicon Chip siliconchip.com.au wave sweep <at> -6dB of approximately 1-minute duration (you do this using the “Generate Tones” function in Adobe Audition). (2). While playing the file at medium/ high volume through the stereo system, listen carefully to the room (not the speakers). Record the results using a microphone placed roughly in the centre of the room. In most cases quite a few rattles, buzzes and assorted resonances will make themselves annoyingly apparent at certain bass frequencies. Loose wall panels, windows, hanging pictures and various standalone appliance panels are usually the worst offenders, followed by such items as glasses and other objects on shelves, and even pens on a table. Removing these resonances will greatly improve the sound quality (or “resolution”) of the system. As with any audio system, the subject of room resonances is not to be treated lightly if the very best results are to be obtained (in fact, the best place for recording is outdoors but this is rarely convenient). One big problem is the standing-wave bass resonances present in any room. The room’s dimensions determine these frequencies. To learn more, see http://www.silcom.com/~aludwig/ Room_acoustics.html In most cases, the effects of these resonances can be minimised by thoughtful equalisation. Higher frequency resonances are also a problem. These are due to sound reflections from nearby objects which cause destructive interference. The more soft furnishings present in the room, the more these resonances will be attenuated. Fig.5: an example of a track recorded at an optimum level. The waveform peaks should not exceed the -3dB levels shown on the righthand scale. Recording the instrument The purpose here is to successfully record the performer’s instrument and play it back with good fidelity. First, ensure that you have access to the microphone level control on your computer’s volume control panel. If you cannot see it, go to “Options”, then “Properties” and select “Microphone” (Fig.2). Do this for both the recording and playback modes. The mode is again set using “Options”, “Properties” and then selecting either “Playback” or “Recording” (Fig.2). If other options such as “surround”, “3D depth”, “subwoofer”, etc are present, disable them to keep the control mixer panel simple. siliconchip.com.au Fig.6: an example of a recorded instrument track complete with time markers. These timing reference points are used later on in order to synchronise the recorded tracks for mixing. Now place the microphone in its intended position, preferably not too close to the speakers. That done, open two instances of the computer’s volume control panel, one for playback mode and one for recording mode – see Fig.3. Next, ensure that the main playback level control is set to zero, then set the microphone recording level to around 50%. When these settings are correct, open the Adobe Audition program and open a new file (mono, 16-bit, 44,100Hz) for recording – see Fig.4. Now press the record button and have your assistant start playing approximately 30cm from the microphone. Watch the resulting waveform and adjust the microphone recording level so that the average level of the November 2007  13 waveform is as high as possible without the peaks exceeding -3dB. Fig.5 is a good example of a track that has been correctly recorded. As shown, the peaks do not exceed this value. Note: ensure that the level scale on the righthand side of the window is in decibels. This is done by right clicking the scale and selecting “decibels”. If the recorded level exceeds -3dB at any time, stop and start again with 14  Silicon Chip a slightly reduced recording level. Let it record for about 30 seconds and then stop. Now set the microphone playback level to around 80% and play back the recorded track. Set the main playback level control for a subjective loudness level similar to the original instrument’s volume. The objective here is to obtain a crisp, clear reproduction of the original with (ideally) no de- tectable distortion, background noise or hum. At this point, it is necessary to make it a habit to lead in any recording you do with a “1-2-3-4” (depending on the beat of the music). Use something that creates a sharp, short pulse in the recorded waveform as shown in Fig.6 (the author uses a pencil on the edge of a wooden table). This is very important, as later on you will need siliconchip.com.au A lot depends on the quality of the microphone and the author uses the Logitech unit shown here. Although relatively inexpensive, it has a fairly flat frequency response and quite good sensitivity. success is when the performer likes what they hear. Gain structure Fig.7 left & above: these are the windows required for ease of control in duplex mode. This involves opening two instances of Adobe Audition & setting one for playback and the other for recording. The playback & recording volume control panels must also be opened. these timing reference points in your waveforms in order to synchronise them for mixing. Recording the voice The purpose of this stage is to successfully record and play back the singer’s voice, again with good fidelity. Repeat the steps described above in “Recording the instrument”. A word of warning – advise the vocalist not to sing directly into the microphone as this can create “woofing” and “whooshing” artefacts that sound terrible and are impossible to edit out later. With practice, you should be able to record the performer’s voice and play back a faithful replica at will. Again, review each recording attempt in its entirety and only continue to the next stage when you have a satisfactory result. One indicator of siliconchip.com.au The signal chain between the performer and the listener’s ear consists of a number of stages, each of which has an effect on the overall gain (or amplification) of the system. Aside from the numerous volume controls, you must deal with and the original loudness levels of both the instrument and the voice, the distance between the microphone and the performer is another factor affecting gain. It may take some practice to establish the correct level settings for each stage. The configuration of these settings is collectively referred to as the “gain structure” of the system. This is a subject of tremendous importance in sound engineering. There exist any number of combinations of settings that will result in the same apparent loudness but obtaining the best sound quality depends on finding the optimum settings for each control. If, for instance, the microphone level is set too high, then distortion will result from overloading of that stage. Reducing the volume somewhere else cannot compensate for this. Do not move on to the next stage until you feel confident about the preceding stages. To learn more about gain structure, see http://www.mmproductions.co.uk/ gain1.html Duplex operation We will now learn how to perform playback and recording simultaneously. This involves opening two instances of Adobe Audition and setting one up for playback and the other for recording. The playback and recording volume control panels must also be opened and the four windows arranged conveniently on the screen as shown in Fig.7. It is necessary to become proficient Extra Tips & Techniques (1). A refinement of the duplexing technique described in this project is to play back the instrument to the performer through headphones and not via the main system. This eliminates any presence of the instrument track in the vocal track. Once again, it’s best to let the performer decide which method they prefer. (2). If you are experiencing burn speed/audio quality related problems, try lowering the burn speed when writing to the CD. (3). Adobe Audition can also be used in 32-bit mode. November 2007  15 Fig.8: the “multi-track” mode of Adobe Audition 1.0. This gives access to no less than 128 tracks! in setting up both recording and playback gain structures, starting both programs (recording first) and monitoring the process until completion of the recording. Sometimes, during a crescendo (a significant increase in “vocal level”), the waveform will overload (or “clip”). If this happens, the recording must be halted and the levels reset. Remember that the distance between the performer and the microphone is an important factor. Only practice and patience will yield the optimum levels, placements and settings. To perform this stage, open a new file in the recording instance of Adobe Audition (Fig.4) and start recording. Then open the playback instance containing the instrumental track, having first verified optimum levels. Setting the marker In order to set a synchronisation marker in the voice track, instruct the 16  Silicon Chip singer to tap the microphone lightly with a pencil on the last lead-in marker. The timing precision of this operation will determine the final accuracy of time-alignment (synchronisation) when mixing the two tracks later on. The vocals should now commence and be recorded simultaneously with playback of the instrumental track. Note that the voice track will also contain a faint background signal of the instrumental track. While an equivalent, higher amplitude copy will later overlay it, it is suggested that the instrument playback level be as low as possible to minimise this effect. This is usually no problem if the microphone is not too close to the speakers and the vocals have been adjusted to give a good strong recording signal. At the conclusion, you will have two files, one of the instrument and one of the voice, both with accurate and clearly visible synchronisation markers. Both recorded tracks should be listened to and evaluated, particularly by the performer. Often the performer will recognise instances in either or both tracks where improvements could be made and ask to re-record them. Remember that it is difficult for any performer to accurately perceive how they sound while they are actually performing. Track signal processing It’s at this point that you may elect to process the tracks – eg, equalisation, noise reduction, dynamic range enhancement, reverb and echo, etc. Adobe Audition is a very powerful audio editing program, capable of generating studio-quality results if it is used competently. However, it is suggested that a minimum of “special effects” be applied. Only use those necessary to correct obvious inadequacies or to add a bit siliconchip.com.au of “shine” to the performance (eg, a small amount of reverb for the singer’s voice or a slight brightness boost for the instrument). If you do choose to edit the tracks, be sure to first save copies of the original unedited tracks! While Adobe Audition allows you to “undo”, you can quickly get lost, especially if you neglect to save and correctly name each track variation (remember to always use WAV format). Synchronisation and mixing When you are satisfied with the instrument and voice tracks and know that they both contain accurate timing markers, you are then ready to synchronise and mix them. The method of combining the separate tracks into one involves copying one track and “mix-pasting” it into the other track, positioning the cursor so that the two are in sync. This is where the lead-in markers are used. Fig.6 shows a magnified portion of the start of the instrument track. First, position the cursor so that it aligns with the centre of the last lead-in marker, then select the track from there onwards to the end by holding down the <shift> key and then pressing the <end> key. Now copy the selection. Next, go to the other instance of Adobe Audition and zoom in to the beginning of the voice track, where the marker should be clearly visible. Position the cursor directly in line with the marker and then select “mixpaste” from the EDIT menu. After pasting, the resulting file will be a time-aligned mix of both instrument and voice. Again, it may need some practice to align the two files, however if the result is unsatisfactory, simply hit “UNDO” and try again. It is usually not necessary to align the tracks to any greater degree of accuracy than about one millisecond (1ms). Now listen to the resulting mix to determine if the relative volume levels of the two components are in balance. If, for example, the voice is too loud, then simply apply an amplitude reduction to the original file (typically 2dB at a time) and try another mix. Remember to save the original file and rename the amplitude-edited version. Writing to CD It is now only necessary to write the mixed track to a CD and this can be done using almost any CD-burner siliconchip.com.au Frequently Asked Questions Question: sometimes room resonances are easy to hear but difficult to locate and if the sweep time is made shorter, they disappear. What to do? Answer: select a narrow band of frequencies either side of the frequency where the offending room resonance is being triggered and set Adobe Audition to “Loop Play”. Question: what do you do if the instrument is not acoustic (ie, electric)? Answer: use the computer’s “line input” and rig the system so that the performer can hear the instrument properly while they are playing it. Question: when I do a sweep, the recorded result looks like a map of the Himalayas! Is the response of my system really that bad? Answer: do a number of trials with the microphone in various positions throughout the room, take the average and only attempt to correct overall response imbalances. Do not worry too much about individual narrow peaks if they change when the microphone is moved. These are caused by room reflections and are of little concern. Concentrate on the most severe frequency anomalies and treat them via suitable equalisation. Question: is it possible to analyse the finished product for frequency content (ie, do a frequency histogram of it)? Answer: yes - select the entire waveform and go to “Frequency Analysis” under “Analyse”. Higher FFT settings will increase the number of samples. (Select “advanced” in the analysis window to gain access to this). Question: why this duplex business? Why not just make two tracks and mix them? Answer: no performer, regardless of skill or ability, can sing a track and play it separately and maintain perfect timing (unless a metronome is used). The method adopted here (and found to be popular with musicians) is to record the instrument track first and then while playing it back, record the performer singing along to it, hence the need for the two instances of Adobe Audition. The performer will find it much easier to sing along to the instrumental track than singing without accompaniment. Question: I am getting feedback between the microphone and speakers, which I cannot eliminate. Answer: at no time should you be using the microphone in playback mode through the speakers. Whether recording or playing, the microphone playback level control should be set to zero. program. Ensure that you use a good quality CD blank, though. Although there’s some debate over whether any CD could be “optimised for audio” (considering that all CDs store data in digital format), just choose the type that gives you the best subjective results. Also make sure that the burner program is configured to convert the WAV file to a CDA file before writing to the CD, otherwise normal stand-alone CD players will not be capable of reading it. A nice touch is to fill in the title, artist and authoring information section. Even more swish is printing a label on the CD and making artwork for the CD cover. A final evaluation of the overall tonal balance should be made by playing the CD on other systems. Sometimes it may be found that the entire track could benefit from a touch of equalisation. Advanced functions Adobe Audition has many advanced functions, including “DeDynamic” (a DirectX-based algorithm) which dynamically expands the dynamic range, effectively eliminating the cross-modulation problems associated with earlier analog expansion methods. Applying this transform to virtually any music track will make it seem “more alive”. If you are interested in the more advanced multi-track capabilities of Adobe Audition 1.0, switch to multitrack mode (press F12) and you will have access to 128 tracks (Fig.8). Examine the demonstration multi-track “ses” file and refer also to the Help SC section. November 2007  17