Stereo Imager
Stereo Imager
Stereo Imager
Table of Contents
Register Now! ................... see WaveSystem Manual Chapter 1 ..................... About the S1 Stereo Imager .... 3
Before you start ............ 3 Features and applications ............ 3 S1 component plug-ins ........... 4
Chapter 7 ...............................More about shuffling .. 19 Chapter 8 ........................ History of stereo controls .. 20 Chapter 9 . What kind of stereo will the S1 work on? .. 21 Chapter 10 ...................................... Key commands .. 22 Chapter 11 .................................. Stereo monitoring .. 23
S1 Plug-in Manual
of the central, left and right stereo components of a mix, it is possible to pan the left/right and center components independently. In addition, the relative balance of the left/right and center signal may be also changed to either narrow or widen the perceived stereo image. S1s unique and informative stereo vector display allows the effect on level balance and positioning to be precisely visually monitored while making adjustments. Stereo level metering is provided in both left/right and M/S modes so that the effect on stereoism can be fully monitored. As with some other Waves plug-ins, correction is provided for the two most common stereo imaging errors: channel polarity (phase) errors and stereo channel reversal. This stereo remix capability makes the S1 suitable not only for sweetening and enhancing adequately mixed stereo masters, but also for more drastic corrective and creative remix work. Because S1 doesnt rely on gimmicky processes to achieve its results, it may be used to make subtle improvements even to refined purist stereo recordings. The S1 does not add any unwanted sound effects not already present in the original recording. Rather it is designed to enhance and rebalance what is already there.
S1 Component plug-ins
On most host applications and platforms you can select just the audio processing you need, and use only the power necessary to do the job. The S1 has several plug-in components in the menu. They are all stereo-outputonly plug-ins and can not be used for mono-to-mono processing. On some native platforms, 88.2/96kHz support is now available. Here are the current components as of this printing; new ones may be added in future updates, and will be noted in the plug-in Read Me files. Some examples for each component are listed: S1 Shuffler This is the full plug-in, with all controls, including Width, Rotation, Asymmetry, Shuffling, plus all channel controls (phase, gain, etc.), and does not count clips; (for mastering, to fix off-center or unbalanced mixes, precision image adjustment, change wideband width, expand low frequency image, Rotate true stereo sources without collapsing image, creation of strong antiphase content) S1 Shuffler (48) The full mastering plug-in (TDM only!), same as the S1-Shuffler, but 48-bit double precision resolution; (applications same as S1 Shuffler but when increased resolution is needed) S1 Imager Omits the Shuffler and Bass trim controls. Ideal for adjust Width, Rotation, and Asymmetry only; (saves processing power if Shuffling is not needed, excellent for stereo mixing to Rotate stereo-mic or submixes into place) S1 MS Simply an MS converter (a matrix) that converts left-right input into MS output, (or vice versa). It has no controls or metering, and is designed so that if the input does not clip, neither will the output. Due to the nature of the matrix, the gain is reduced by 3dB in this process to avoid clipping. Therefore, if you use two MS modules in a chain (say for equalization of just the M or S channel), then the total chain will have a nominal signal drop of 6dB (3dB per MS module, also assuming the EQ is flat with unity gain).
S1 Plug-in Manual
The stereo vector display does not show anything about the input signal. It shows you: 1 - the output positions of the original left, center, and right sounds; 2 - an accurate indication of their changed levels in dB so you can adjust level-balance between original left, center, & right sounds. The stereo vector display does not show the effect of shuffling, but only gain, width, asymmetry, & rotation. Shuffling is a process applied after these processes.
The Controls
Gain - simply controls overall gain level without altering stereoism. Gain is displayed in dB. Width - This alters the width of the input stereo signal. For width values less than 1, the sound stage is narrowed, becoming mono for width zero. For widths greater than 1, the sound stage is widened, with sounds at the two edges of the stereo stage being moved beyond the loudspeakers. The stereo is unchanged when Width equals 1. Start around 1.2 for moderate enhancement, plus a little Shuffling. Asymmetry - This acts after initial width adjustment. This unique control does not affect central mono in-phase sounds in any way, but adjusts the relative level of left and right sounds. It differs from conventional balance control in that it keeps center sounds in the center. This control is particularly useful for altering the relative levels at the two sides of the stereo stage without moving a mix off-center, or biased in position to one side or the other. Rotation control - This control allows adjustment of the centering of the entire stereo image without any effect on relative sound levels in different parts of the stereo stage. It is particularly useful for centering a stereo mix biased to one side of the stereo stage without altering the mix level balance, and for mixing stereo-miked sources, serving as a true stereo 'pan' control. Useful tip: No separate stereo balance control is provided, but if you drag both the Asymmetry and Rotation controls together, then their combined effect is that of a balance control. For most platforms, they may be dragged together by dragging in the space between the two controls, or by selecting both controls first then dragging on one.
6
S1 Plug-in Manual
Shuffling - This increases stereo width at bass frequencies to help compensate for the fact that the ears hear stereo effects as being more narrow in the bass than in the treble. Subjectively, it has the effect of making stereo images more spacious. Unlike previous commercially available stereo shufflers, it is fully phase compensated, i.e. does not introduce unwanted phase errors between the stereo channels. This gives improved stereo sound quality with low listening fatigue. It is also carefully designed to minimize tonal alterations of the mix in the bass. It comprises two controls: shuffling and frequency. Shuffling controls the magnitude of bass width increase between 1 (no shuffling) and 3 (maximum bass width increase). The subjectively best value usually lies between 1.6 and 2.5. Frequency - controls the frequency below which the Shuffling effect is increased. This frequency may be adjusted between 350 and 1400 Hz. For normal stereo monitoring, a frequency between 600 and 700 Hz usually sounds best, although a higher frequency may be better if used with the very close loudspeakers often used in multimedia, portable, or TV one-piece reproducer systems where the stereo speakers are in the same box. A lower frequency is useful when adjusting stereo originating in stereo microphone techniques where the microphones are spaced apart by perhaps 20 or 30 centimeters.
LR/MS input button - Changes input mode to accept either standard Left-Right inputs or Mid/Side (MS) inputs. See the chapter About MS.
Bass Trim - Adjusts the level of bass when the Shuffler is in use. When set to 0, bass level stays more or less the same. On some material, higher Shuffler settings may need bass trim of 1 or 2 dB boost to compensate for any subjective shift in bass level. Reset - This button restores Width, Asymmetry, Rotation and Shuffling to settings at which they have no effect on input stereo. Hint: By repeatedly using the Undo button after using Reset, you may do A/B comparisons between unprocessed and processed stereo effects, without altering gain or channel polarity and swap settings.
Graphical control
A unique and flexible feature of S1 is the ability to control the processing from the graph. To control the S1 from the graph: (For ease of explanation, the following instructions assume that you have clicked the Reset.) Click and drag on the center line (pointing straight up). This line represents the center channel. Drag left/right to change Rotation. Drag up/down to change Gain. Option-drag left/right to change Width. (PC uses Alt key). Option-drag up/down to change Asymmetry.(PC uses Alt key). Hint for Mac: you can press and release the option key without re-clicking; this lets you change any of the four controls listed above in one step.
S1 Plug-in Manual
Mastering
Quick tip: Use the "one-click" adjustment of an input stereo image by using direct graphic control, described on the first page of this chapter. You can hold and release the option(Alt) key while dragging without having to let up on the mouse to easily adjust all four controls; Gain/Rotation, or Width/Asymmetry. To adjust stereo positioning without any effect on the internal level-balance of a mix, use the Gain and Rotation controls only (and the channel swap button if reversal of the i
Adjustment of Width and Asymmetry to modify level balance. Left, Center at 0 dB gain, Right at - 4 dB gain.
Conversely, to adjust the internal level-balance of a mix without any effect on the stereo positioning of central images, use the Width and Asymmetry controls only. This adjustment may be done by depressing the option (Alt) key and dragging the mouse from the vector display screen. If the effect of a level-balance adjustment using Width and Asymmetry control is that the stereo image is too narrow, you may find that increasing the amount of Shuffling will help to restore width without much effect on perceived level balance. Dragging Rotation and Asymmetry control together acts simply as a conventional balance control, albeit one acting after the Width control. (You can conceptually think of the Asymmetry control as being a balance control with a rotation to restore the center sound to center position.) It is worth noting that the Rotation control has NO EFFECT on the internal sound balances of stereo mixes, only on their positioning. Besides making the sound stage wider or narrower, Width does have an effect on level balance, making non-central sounds louder relative to central sounds as the width is made larger. This can sometimes be useful for changing the relative balance of central and edge sounds. But beware: this only works in stereo, and does not change the balance heard by a mono listener, e.g. via a mono radio or TV. Asymmetry changes relative balance of left & right both in stereo and in mono.
9
S1 Plug-in Manual
Mono Compatibility
When stereo sound is broadcast via radio or TV, a significant proportion of listeners will be listening to mono. In that application, it is important that stereo mixes also be mono-compatible, i.e. that no sound in the mix becomes colored, disappears, or becomes excessively quiet.. The S1-Stereo Imager is designed such that, with material mixed on standard stereo equipment, the processing does not cause coloration in mono listening. The mono compatibility of reprocessed stereo material can be judged using the stereo vector display. The closer the vector or lines corresponding to left, center and right inputs are to being vertical on the display, the more mono-compatible is the mix, whereas the closer they are to being horizontal, the quieter will be that sound position in the mono mix. In practice, mono compatibility will usually be good if all three lines are within 45 degrees of being vertical, and adequate if all three lines are within 60 degrees of being vertical. If mono compatibility is particularly important to you, you are advised to monitor the mix via a mono reduction of the mix as well as in stereo. Many mixing desks or monitoring systems have a mono monitoring button.
10
S1 Plug-in Manual
MS mode is an alternative way of handling stereo signals that has many advantages and unique possibilities for stereo processing. Although MS mode has been used since the very start of stereo in the 1930s, it has only gained widespread use over the last decade, often being used in stereo TV production. But it's also used for various stereo microphone techniques and in other audio-only applications. Briefly, M and S signals are derived from left and right by a matrix process. M stands for "mid" and S for "side" signals. The signal produced by an "MS" microphone is in this format and needs conversion from the MS to standard LR. However, the S1 offers unique processing of the signal before it is converted, and also allows any LR signal to be converted into MS for the same processing. In the Special MS uses chapter, examples of MS processing are explained. The mid signal M, also termed mono, is simply obtained by taking the average of the left and right channels. Mathematically: M = (1/2)(L+R) where L is the left waveform magnitude and R is the right waveform magnitude. The side signal S is obtained by subtracting the right signal from the left with the same overall gain as used for the M signal. Mathematically: S = (1/2)(L-R). The process used to derive the M and S signals from L and R is known as an MS matrix. The MS matrix process may be applied again to the M and S signals to recover the original left and right, although an additional gain of 2 (+6 dB) is needed to get the original gain. Mathematically L = M+S and R = M-S. i.e. Left is the sum of the M and S signals and Right is the difference of the M and S signals.
How is MS used?
MS signals have various properties that make them useful for stereo processing. The M signal is simply the mono signal heard when a stereo signal is listened to on mono equipment. The S signal is zero when the left and right channel sound is identical as it is for sounds panned exactly to the middle of the stereo stage. This is why S is called the "side" signal it only contains that part of the sound positioned to either side of the center position.
11
S1 Plug-in Manual
The standard S1 Shuffler or Imager can be used with input stereo signals from either standard left-right (LR) stereo or MS stereo. In both cases, the output is in standard left-right stereo mode. Click the button above the channel swap/polarity buttons to select which mode you need. In MS mode, what is normally the left input channel now accepts the M signal and what is normally the right input channel now accepts the S signal. M and S outputs from an MS stereo microphone may be fed into the S1 using MS input mode, and the output will be in conventional left-right mode with the gain, width, asymmetry, rotation and shuffling controls affecting the sound in the usual way. MS signals may also be available from other equipment, such as stereo tracks recorded in MS mode on a video tape recorder.
12
S1 Plug-in Manual
13
The peak and clipping indicators only give correct indications in L/R mode, where they show peak levels and any clips independently in the left and right channels. Do not rely on peak and clipping indications in the M/S mode. Whenever switching back to L/R mode, please reset the peak and clip indications to obtain an accurate reading.
The unique phase compensation feature of the S1 assures low subjective phasiness of the stereo image. This uses phase manipulations to modify and improve stereo quality. However, even at the unity-gain default or reset settings of no shuffling or width alterations, these phase manipulations have a side-effect of altering the levels of peaks in the audio signal, so that on normalized or peak limited sound files, you may sometimes need to reduce the gain level by maybe 2 or 3 dB to avoid output clipping. Set the S1 gain so that no clipping registers on the output level meter in L and R mode. For this reason, to get maximum output level, it is very strongly recommended to do normalization of level or peak limiting using the Waves L1 peak limiter AFTER processing by the S1 rather than beforehand. Here is an example of how to use the S1 and L1 together for stereo image adjustment while maintaining unity gain. Insert the S1-Shuffler first, followed by the L1 limiter or +L1. (This example is for systems that support multiple real-time processes, whether DSP or Native-based). Assuming you have high-level files, or wish to be sure to avoid any clipping, first set the S1 gain to -6.0dB. Next, set the L1 Threshold to -6.3dB, and the Out Ceil to -0.3dB. By using the L1 merely as a limiter to control any new peaks created by the S1 process, unity gain is maintained while giving 'safe' adjustment headroom for the S1 processes. Of course, additional gain beyond unity can be achieved with limiting, by lowering the L1 Threshold even further.
14
S1 Plug-in Manual
MS Equalization
The small modules of the S1, along with those of the Q10, allow a kind of EQ that is different from conventional equalization of left and right stereo channels. It is known as MS equalization. This kind of EQ can achieve unique effects on stereo signals that cannot be obtained any other way, altering not only the overall tonal quality, but also the spatial effect of a stereo recording. MS equalization is obtained by the following cascade of processors
S1 Shuffler or S1 Imager or +S1 Shuffler
MS Matrix
Q(n) Equalizer
using an MS matrix followed by an "n-band" stereo EQ component from the Q10 (n= up to 10 bands of EQ), followed by an S1 module having the desired output facilities. If the output of the MS equalizer is to be fed to other processes in TDM or in 24 bits, the S1 Imager or S1 Shuffler modules may be used depending on whether shuffling is desired or not. One other option is to use another MS plug-in at the end instead of the Shuffler or Imager, but only if you don't want to use Width and other options in those components.
S1 Plug-in Manual
15
The effect of this processing chain is that what is normally the left channel of the Qn equalizer now equalizes the M signal and what is normally the right channel of the Qn equalizer now equalizes the S signal. By giving the M and S channel signals different gains at different frequencies, you can vary the width of the stereo in a frequency-dependent way. For instance, if you boost the S channel at 1kHz, then the output will have greater stereo width around 1kHz. If you boost the M channel at 1kHz, then the output will have a decreased width (and stronger 'phantom' channel) around 1kHz. A practical example would be to increase the midrange of Crown SASS microphone outputs, which many times have a 'hole-in-the-middle' effect on some sources; a reasonable boost around 400Hz of the M channel makes a significant difference without affecting the general stereo imaging in other frequencies. You are now ready to start using the processing. Here are a few of the possible applications.
Uses of MS equalization
Overall EQ Open the Q10 (or Renaissance EQ) interface. Ensure that the stereo channels are set to "Strap" or "Link". Then adjust the EQ in the conventional way to equalize both channels.
Frequency-dependent width This is a new range of adjustments. Open the EQ interface. The procedure to adjust the width as a function of frequency involves the following steps. [1] Start with identical EQ settings in both the M and S channels. Make sure the eq is in "Strap" mode. [2] Select in a given band the type of filter you want, its frequency and Q, and any overall gain adjustment in that band that you wish. Then "unstrap" the EQ and click the Right button so that you can change just that side, which is the S channel. [3] In that band now adjust the degree of boost and cut required to alter width in that band. Increase of gain will increase width and decrease of gain will decrease width, but only of that band's frequencies. [4] In general, this adjustment will have altered tonal quality. To restore tonal quality, select "strap" and then adjust the gain in that band up and down until the tonal quality is what you want again. By adjusting both bands while strapped together, the new width adjustment is maintained (as a difference in gain between the two channels). [5] If further adjustment is required, return to step 2. It is important that you make these adjustment with strap on or off as indicated in each step. Otherwise the adjustments in the stereo sum and difference channels will no longer track together. Please note it is preferable to do overall bass shuffling using the final S1 Shuffler rather than adjustments on the EQ, since the former is a design dedicated for the job.
16
S1 Plug-in Manual
Overall (wideband) width is also preferably altered on the output S1 plug-in, using the Width control. Altering the gain of the M and S channels of the Qn separately will also alter width, with an increase in S gain increasing width and a decrease decreasing width. But to avoid confusion, it is preferable to use the Qn gains only for overall gain adjustment and to avoid clipping in the Qn by altering both channel gains together.
Similarly, dummy head recordings or spaced identical microphones pointing in the same direction (typical spacing 15 to 50 cm) can be "Blumlein shuffled" by the same technique but using a higher boost frequency, say + 3dB at 300 Hz on the S channel, along with say a -1 dB shelf on the M channel to compensate to the resultant exaggerated bass. The channels may then be relinked and the frequency slid up and down to adjust width or adjust for different microphone spacings.
One-sided Equalization
The small S1 modules with small Q10 modules can also be used for a unique mode of EQ termed one-sided EQ which allows the EQ on one side of the stereo stage to be adjusted without affecting the EQ of sounds in the middle of the stereo stage. For example, you can make sounds on the left have a brighter treble without affecting the sound of center-stage vocals. This can be very useful in remastering stereo mixes or submixes, as one has much finer control over the EQ effect in different parts of the stereo stage. The basic process uses three small modules in cacade:
S1 Imager Q(n) Equalizer S1 Imager
although the last stage can also be an S1 Shuffler or +S1 Shuffler if desired.
S1 Plug-in Manual
17
If one wishes to EQ the left side sound without EQ'ing the center, the first S1 Imager is loaded with the setup file "S1 LeftSideIn" and the second S1 Imager is loaded with the setup file "S1 LeftSideOut" If one wishes to EQ the right side sound without EQ'ing the center, the first S1 Imager is loaded with the setup file "S1 RightSideIn" and the second S1 Imager is loaded with the setup file "S1 RightSideOut" The input and output S1 imager files are designed such that if the Qn is set to flat, then apart from an overall gain change, there is no effect on the stereo. If the channels of the Qn are strapped and the EQ's varied together, this adjusts overall EQ in the usual way. What is new here is what happens if one unstraps the two channels of the Qn. When using the LeftSide setups, adjusting the EQ of the left channel of the Qn adjusts the EQ of the left-side sounds of the input stereo signal, but has no effect at all on sound at center stage, which remain unequalized and stay at the center position! When using the RightSide setups, adjusting the EQ of the right channel of the Qn adjusts the EQ of the right-side sounds of the input stereo signal, but has no effect at all on sound at center stage, which remain unequalized and stay at the center position! Excellent when needed.
18
S1 Plug-in Manual
19
20
S1 Plug-in Manual
S1 Plug-in Manual
21
For complete information on interface controls, read the WaveSystem Manual. (PCs use the Alt key instead of Option, for example).
22
S1 Plug-in Manual
The best stereo monitoring will be obtained if the loudspeakers are set up symmetrically towards the left and right, with the listener placed symmetrically so that he or she is equally distant from them. Normally, stereo monitoring is optimal if the angle subtended by the two loudspeakers at the listener is around 45 to 60 degrees, i.e. if the distance between the loudspeakers is between 0.7 and 1 times the distance of the listener from each of the loudspeakers. The subjective stereo effect heard via headphones will generally not be the same as via loudspeakers. If it is important to know the effect heard on both loudspeakers and on headphones, both should be used to check a stereo mix effect. One cannot with certainty predict the effect for headphone listeners by loudspeaker monitoring or vice-versa. In general, reverbs, and other uncorrelated materials will sound louder in headphones, sometimes by 6dB or so.
S1 Plug-in Manual
23