Soundscape Masterclass: Part 1

We show how you can use well-known Soundscape features in unusual manners, plus more in-depth explanation of some specific Soundscape functions, including details of how to use VST and Direct X plug-ins to process Soundscape audio.

When Mackie purchased Sydec in April 2001, many Soundscape users expressed concern, because the buyout marked the end of Soundscape Digital Technology, which distributed the Soundscape product range designed by Sydec. No one knew if Mackie would invest the time and effort necessary to keep the Soundscape product range going. As I was putting the finishing touches to this article, I received information from reliable sources suggesting that Mackie will be discontinuing Soundscape, but also that the original Sydec team will continue development and support for the product it created. Pending further developments, Mackie's influence on the Soundscape product range over its two years of ownership appears to have been largely positive: a generous trade-in offer has been extended to users of the original SSHDR1 wishing to upgrade to the new Mackie-badged hardware; the Editor and other related programs have received several free updates; version 3.6.1 of the Editor came with a new manual; and new plug-ins have been released.

This workshop is intended to provide information which is not included in the new manual, because either it is beyond the scope of 'normal' use of the software or because it pertains to third-party plug-ins and how to use them. There are also some ideas for 'lateral thinking', showing how you can use well-known features in unusual manners. In addition, I'll be providing more in-depth explanation of some specific Soundscape functions, including details of how to use VST and Direct X plug-ins to process Soundscape audio.

De-essing With Drawmer's SDX100

A mixer setup for de-essing using the Drawmer SDX100 plug-in.

The recently released Drawmer SDX100 dynamics plug-in can be used very effectively to tame excessive sibilance. However, its documentation does not provide any practical example of that application, so here's how to de-ess a mono vocal Part (the complete setup is shown on this page):

• Create a mixer channel with a stereo input and a mono output, and insert a stereo-in/mono-out (Mono + Ext Trig) SDX100 Element. This technique can also be used on stereo signals, in which case you would need a four-in/stereo-out mixer channel and a four-in/stereo-out (Stereo + Ext Trig) SDX100 Element.
• Set the input of this channel to use two of the internal busses. In this example, busses one and two are used.
• The audio Part that needs to be processed must reach the left channel of the SDX100 via buss one. In the example shown, the first mixer channel contains a track insert which will provide the main signal (a lead vocal Part).
• The key signal must reach the side-chain input via buss two. In our example, a send Element inserted in channel one provides a duplicate of the main signal, which is sent to the second mixer channel using buss three. That signal is equalised so that the sibilance is increased, and the EQ setting will be largely dependent on the material to be processed. In this case a boost at 10358Hz with a Q factor of 5.2 was effective. The gain applied to the selected frequency can be obscenely high, as this signal is not meant to be heard by the listener. Although low-frequency and high-frequency key filters are available in the plug-in itself, I prefer using the excellent Multi EQ for pinpoint accuracy.
• The Compressor section of the SDX100 can be switched off. The Ducker is used to lower the level of the sibilant portions of the signal. In the Key section of the SDX100's main window (bottom left), set Key Source to Ext. Set the Noise Gate section (bottom right) to Duck. With the audio playing in a loop, adjust the threshold so that only the 'S' sounds reach above the threshold (the plug-in provides excellent visual feedback in this regard, with the portion of the signal below the threshold represented in red in the meter, and the portion of the signal above the threshold represented in green. Ideally there should be green only when sibilance is present. Adjust the other Noise Gate parameters according to the characteristics of the material being processed.

  Web Resources     www.soundscape-digital.com The original address for the Soundscape web site, which is currently linked to the Mackie Broadcast Professional web site. The Sydec team hope this address can be retained for the future. www.mackie.com/mbp The Mackie Broadcast Professional web site, which, at the time of going to press, is still the official Soundscape web site. The forums can be read by anyone, but, unlike other Mackie forums, registration still involves entering a product serial number, and only registered users can post. Due to the current Mackie/Soundscape situation, it is not known for how long this address will remain valid. http://members.optushome.com.au/utopiaaudio/Soundscape/a.htm Besides the official web site, one of the most useful Soundscape resources on the web has to be the Soundscape page of user Christo Curtis. Christo is a noted high-profile user of Soundscape, which he relies on for live and orchestral recordings, including film scores and major live events. He maintains a Soundscape page which is full of useful tips, whether you use a Mixtreme soundcard or a full-blown multiple-unit DAW setup. www.spinaudio.com On Spinaudio's web site you can find ASIO FX Processor and VST Skinrack. There is a cut-down, freeware ASIO FX ProcessorLE available, and ASIO FX ProcessorSE, as used here, costs $40. It supports up to four VST effects per instance of the plug-in, and also hosts VST instruments. ASIO FX ProcessorME is similarly priced and, although it doesn't support VST instruments, up to eight VST effects can be loaded per instance of the plug-in. VST Skinrack costs $25. www.vb-audio.com At the VB Audio web site you can download the freeware FFX4. www.parkenet.com/apl/BatchFileRename.htm You can download Batch File Rename v2.4.12 from here. www.goldwave.com This is where you can download the shareware Goldwave utility. Registration costs 55 Canadian dollars — the shareware version requires that you close it after every 150 commands, but it will otherwise remain fully functional.

  DSP Disparities

  The Mixpander adds a staggering amount of extra power to the Soundscape mixer. Each Soundscape rack unit, from the SSHDR1+ to the new Soundscape 32, contains two Motorola DSPs. One of these is allocated to the recorder, the other one is allocated to the mixer. The single mixer DSP allows a few plug-in effects to be inserted into the mixer for a 12-track SSHDR1 Plus. With up to 32 audio tracks per unit for the R*Ed and Soundscape 32 DAWs, many more mixer channels are likely to be used (up to the maximum of 128). In these conditions, if the aim is to mix completely within the Soundscape system then the mixer DSP in each rack unit will soon be maxed out.

  With a Mixpander, on the other hand, the limits of a Soundscape system are pushed back well beyond the needs of just about any user. Mixpander/5 and Mixpander/9 add five and nine DSPs respectively to the mixer. Note that, in either case, one Mixpander DSP is used for communication with the rack unit's mixer DSP. Therefore, with Mixpander/5, four DSPs are used for mixing, and, with Mixpander/9, eight DSPs are used for mixing.

  When building a mixer, you can use Mixpander auto routing. If you do, you can just go about your business, and the software will judiciously take care of DSP resources allocation. However, if you prefer to decide which DSP each channel is allocated to, and therefore turn auto-routing off (from the Preferences submenu of the Settings menu), then it is useful to know that these DSPs don't respond the same. While they are all the same model, DSP1 (Mixpander/5 and Mixpander/9) and DSP5 (Mixpander/9 only) come with extra memory: 1.5MB of fast SRAM, instead of 384K for the other DSPs. So these DSPs can be used for more memory-hungry plug-ins such as the TC Works TC Reverb. There appears to be a slight trade-off between available memory and available processing power, so it may be better to use the other DSPs for plug-ins which use lots of processing power.

  Soundscape Reviews In SOS     Soundscape Hard Disk Recording System: SOS September 1993 Soundscape SSHDR1 Hard Disk Recording System: SOS April 1996 Soundscape SSHDR1+ Hard Disk Recording System: SOS November 1997 Soundscape Mixtreme PCI Digital Audio Card & SS8IO3 Analogue Interface: SOS April 1999 Soundscape R*Ed PC-based Computer Recording System: SOS October 2000

  You are probably already familiar with the 'P' and 'M' values displayed in the mixer window's title bar, which represent the percentage of processing and memory resources used by a mixer channel. With a Mixpander, depending on the specific DSP block used, these percentages can vary significantly. For example, a stereo mixer channel containing: a peak meter, a track insert, an Aphex Big Bottom Pro plug-in, a four-band Multi EQ plug-in, an Aphex Aural Exciter plug-in, and TC Works TC Dynamizer and TC Reverb plug-ins uses 105.4 percent of the DSP1 processing and 50.7 percent of its memory, resulting in a warning message in the title bar. However, the same setup uses 99.6 percent processing and 60.1 percent memory on DSP2, which is fine. The Arboretum Hyperprism Tube-Tape Saturator plug-in uses nearly 10 percent less processing power on DSP2 than on DSP1, but a stereo TC Works TC Reverb plug-in uses around four times more memory resources on DSP2 than on DSP1. By way of contrast, the Aphex plug-ins show the same resource demands regardless of DSP allocation...

  As demonstrated by these examples, it is difficult to predict how much processing and memory resources a given plug-in will demand according to which DSP it runs on, other than by experiment, so I've tested a variety of plug-ins to find out — the figures can be seen in the big table on this page. This information should be useful when creating a mixer, and will help answer questions like 'if I get a Mixpander, how many compressors and reverbs can I run at the same time?'

  Plug-in DSP Resources Usage With Mixpander/9

  These plug-ins were inserted in mono-in/mono-out, mono-in/stereo-out, and stereo-in/stereo-out mixer channels as appropriate, and the overall 'P' and 'M' percentages, minus those for the empty channels, were written into the table. The testing was done at 44.1kHz, using a R*Ed and Mixpander/9. Not all available plug-ins are featured. For instance, the Spinaudio Spin Delay was left out because its resources usage depends on the externally set configuration parameters. The Dolby plug-ins are not included because, by definition, they are not mono-in/mono-out, mono-in/stereo-out or stereo-in/stereo-out. Most of the standard plug-in mixer Elements, which tend to have modest resources requirements, are not included.

  Please bear in mind that when several plug-ins are inserted into the same channel the percentages do not simply add up: masking can take place, which means that the total used resources percentage can be less than the sum of individual percentages. Also, the percentages displayed are the maximum that a given plug-in will use. For further information, the Virtual Mixer Channels & DSP Processing Power section in the Mixer chapter of the Soundscape manual is recommended reading.

  Manufacturer	Plug-in	DSP1 or DSP5 Mono-in/mono-out	Mono-in/ stereo-out	Stereo-in/ stereo-out	Other DSP blocks Mono-in/mono-out	Mono-in/ stereo-out	Stereo-in/ stereo-out
  		P M	P M	P M	P M	P M	P M
  Acuma Labs	Sat Fat	23.3 15.3			21.2 15.3
  Aphex	Aural Exciter	10.7 5.1		21.3 10.3	10.7 5.1		21.3 10.3
  Aphex	Big Bottom Pro	7.6 3.7		15.1 7.6	7.6 3.7		15.1 7.6
  Arboretum	Hyperprism Tube-Tape Saturation	27.9 8.8	28 9.2	53.1 16	23.3 8.8	23.4 4.9	43.4 16
  Arboretum	Hyperprism High Pass/Low Pass	3.7 2.9	3.8 3.3	5.4 4.5	3.7 2.9	3.8 3.3	5.4 4.5
  Arboretum	Hyperprism Band Pass/Reject	9.7 4.7	9.8 5.1	13. 4 8	9.7 4.7	9.8 5.1	13.4 8
  Arboretum	Hyperprism Flanger	8.7 4.3	9 4.7	11.6 5.2	8 4.3	8.2 4.7	10.6 5.2
  Arboretum	Hyperprism Phaser	11.6 5.3	11.7 5.8	14.9 7.4	11.6 5.3	11.7 5.8	14.9 7.4
  Arboretum	Hyperprism Quasi Stereo		6.6 3.9	7.1 4.1		6.4 5.3	6.7 5
  Arboretum	Hyperprism Ring Modulator	7.4 3.7	7.1 4.1	7.1 4.2	6.2 3.7	6.3 4.1	6.3 4.2
  Arboretum	Hyperprism Sonic Decimator	5.4 2.2	5.7 2.7	8.4 3.2	5.4 2.2	5.7 2.7	8.4 3.2
  Arboretum	Hyperprism Bass Maximizer	13.3 6	13.6 6.4	24.1 11.4	12.6 6	12.9 6.4	23 11.4
  Drawmer	SDX100 Expander	4.3 7.4		5.8 8.8	3.9 7.4		5.3 8.8
  Drawmer	SDX100 Gate	5.6 8.4		7.4 9.9	5.1 8.4		6.7 9.9
  Drawmer	SDX100 Side Chain	5.6 8.4		7.4 9.9	5.1 8.4		6.7 9.
  Timeworks	CompressorX	26.4 11.4		29.1 13.9	25.5 12.4		27.4 13.9
  Timeworks	CompressorXSide Chain			29.5 13.9			27.4 13.9
  TC Works	TC Dynamizer			31.8 18			28.9 18
  TC Works	TC Reverb		29.7 11.8	29.7 11.5		26.8 52.1	26.8 51.8
  Wave Mechanics	Reverb		33.6 25.2	33.6 25.2		30.6 52.1	30.6 51.8
  Soundscape	Audio Toolbox Dither	2.4 1.5		3.8 2.7	2.4 1.5		3.8 2.7
  Soundscape	Audio Toolbox Delay (Short)	1.6 1.2	2.2 1.4	3.6 3	1.5 8.6	2 8.4	3.3 17.9
  Soundscape	Audio Toolbox Delay(Medium)	1.6 3.5	2.2 3.3	3.6 7.7	1.5 18.4	2 18.2	3.3 37.5
  Soundscape	Audio Toolbox Delay (Long)	1.6 8.2	2.2 8	3.6 17	1.5 38	2 37.8	3.3 76.6
  Soundscape	Audio Toolbox Chorus/Flanger	3.1 1.6	5.9 3.3	5.9 3.4	2.9 2.2	5.5 5.3	5.5 5
  Soundscape	Audio Toolbox Dynamics	1.7 1		2.1 1.3	1.6 1		2 1.3
  Soundscape	One-band Multi EQ	0.9 1.4		1.8 1.6	0.9 1.4		1.8 1.6
  Soundscape	Two-band Multi EQ	1.6 1.4		3.2 2.4	1.6 1.4		3.2 2.4
  Soundscape	Three-band Multi EQ	2.4 1.7		4.5 3.3	2.4 1.7		4.5 3.3
  Soundscape	Four-band Multi EQ	3 2.1		5.9 4.1	3 2.1		5.9 4.1
  Soundscape	MS Decoder			2.3 1.9			2.3 1.9

  Using VST & Direct X Plug-ins

  A feature of the Mixpander that has perhaps not received all the attention it deserves is audio streaming between the host PC and the Soundscape rack unit. Sixteen direct streaming channels are available in both directions via a TDM buss. The most obvious use of this feature is to have audio tracks from a MIDI + Audio sequencer sent to the Soundscape mixer, thereby increasing the total available track count. However, Mixpander also allows VST and Direct X effect plug-ins to be used in conjunction with Soundscape. This is an important benefit, because, while Soundscape-format plug-ins can be coded specifically to offer higher sonic performance than their PC native counterparts thanks to the dedicated DSPs, they are still thin on the ground and tend to be more costly than PC native plug-ins.

  Spinaudio's ASIO FX Processor lets you access VST plug-in processing from within Soundscape, while Vincent Burel Audio's FFX4 also allows Direct X plug-ins to be used — in the top screenshot you can see a Direct X version of Antares Auto-Tune within FFX4 and ASIO FX Processor. Although plug-ins without skins cannot be loaded in this way, a variety of different skins can be generated for skinless VST plug-ins using Spinaudio's VST Skinrack — two choices are shown in the other two screenshots.At present, the most convenient way to use VST plug-ins with Soundscape is via Spinaudio's very affordable ASIO FX Processor. This utility is a dedicated VST host application, of which multiple instances can be run at the same time. Multiple VST effects can be loaded per instance of the program. The Mixpander streaming channels appear as inputs and outputs in the Spinaudio software, with mono-to-mono, mono-to-stereo, and stereo-to-stereo configurations. It is important to choose the correct configuration for the type of send (output) and input (return) Element inserted in the corresponding Soundscape mixer channel. While the audio has to travel from the Soundscape mixer to the PC and back, the overall delay is reasonably low, thanks to the excellent Soundscape ASIO driver.

  On my test rig, a 1GHz Pentium 3 system, I was able to obtain a declared latency of 5ms by setting the buffer to 64 samples (input 1.6ms and output 3.4ms, as reported in ASIO FX Processor's settings window). A recorded test tone sent to the ASIO FX Processor software, returned to the original Soundscape mixer channel, and recorded via a track insert at the bottom of that channel was offset by 240 samples compared to the original (less than 5.5ms delay). Each plug-in loaded into ASIO FX Processor will obviously add its own processing delay, but this tends to be fairly small (I repeated the same test with two plug-ins loaded, and the offset was 294 samples in total.

  If you want to avoid the processing delay, it's extremely easy to record the processed audio and align it with the original, as the processing delay results in a flat line at the beginning of the waveform. Just cut the Part at the end of the flat line (ie. at the beginning of the wanted audio material), mute the resulting unwanted Part, and move the wanted Part to match the beginning of the audio in the original Part. (The quickest way to do this is to use the Move tool with the Snap option active and set to Prt+Mrk+L). The unwanted section of the newly recorded Part can then be converted into a Take (using the Mixdown tool) and saved with a suitable name, such as 'plug-in x offset'. This Take can subsequently be dragged into the arrangement next to further processed Parts to indicate where they should be cut.

  An important point to note when using VST plug-ins in this way is that ASIO FX Processor will not load skinless VST plug-ins. Fortunately, however, Spinaudio offer the VST Skinrack software as a solution. This keenly priced utility, which allows the user to create skins for VST plug-ins, is extremely quick and simple to use, and well worth buying in its own right. Also, there's no reason to be restricted to just VST plug-ins, because it is possible to load a Direct X VST wrapper into ASIO FX Processor, and therefore mix VST and DX plug-ins — the freeware FFX4 from Vincent Burel Audio can serve this purpose.

  Playing Back Soundscape Takes On Your PC

  Soundscape Takes use a proprietary file format, and when stored on a PC drive they cannot be played back by ordinary Windows audio applications. However, there are programs that will play a Soundscape Take if its file extension is changed to 'wav', and there are also programs which will batch-process such file-extension changes. For example, Batch File Rename (shown in the screenshot) can be used to change the file extensions and Goldwave allows you to listen to the 'falsely converted' files — see the 'Web Resources' box for details on where to get hold of these applications. Goldwave detects that the files are not genuine WAV files, and just asks how it should handle them (you are required to select a format, sample resolution and sampling frequency).

  A word of caution, though, when using this workaround. The Soundscape file extensions include the identity of the source SDisk (unit number and letter), and this will not be remembered. It is not a problem if your Takes are stored in specific folders according to the SDisk they came from — in that case you can easily rename them according to their origin. Also, I can give no guarantee as to the integrity of the files when they are given back their Soundscape Take extension. That said, I imported them back into a R*Ed and the time stamp was preserved.

  Spin Doctor

  Traditionally, the settings of Soundscape mixer plug-ins are saved along with the current mixer. In this respect, Spinaudio's Spin Delay is different from other plug-ins. Any change to the configuration of the plug-in made in the separate Configuration Manager application will affect the way settings are restored when a mixer containing instances of the plug-in is loaded. Therefore, in order to be able to restore Spin Delay settings exactly as they were during a given session, it is advisable to save the Spin Delay configuration file along with the corresponding mixer, and to back it up as part of the project. In a standard installation where 'C' is your working partition, the file path is C:\Program Files\Spin Audio\SpinDelay For Soundscape. The file has a 'cfg' suffix, and its name is the one you have entered when exporting an edited configuration. Using the name of the corresponding mixer when exporting a configuration is a sure way to help file identification.

  Editing INI Files

  Each time the Editor is launched from a shortcut, it checks the INI initialisation file specified by that shortcut (which will normally be located in the main Soundscape directory on your working partition), and runs with the settings specified in that file. If you need to check the name of the INI file referenced by a particular Editor shortcut, right-click that shortcut, click Properties in the pop-up menu, and look at the end of the line in the Target box. The location of the file is specified in the Start In box. Different shortcuts can be linked to different INI files containing different settings, and it pays to become familiar with the contents of the INI file. A few tips are provided here.

  You can customise the track colours from the Soundscape INI file. The screenshot shows how a lighter colour palette makes the waveforms themselves more easily visible.

  With the INI file opened in Microsoft's Notepad, a useful tweak can be performed to rename Soundscape inputs, outputs and busses. Alias lines need to be added at the end of the MixWnd section, as in the following examples:

  Inp1 alias 00000K388931C=Neumann U87;U87

  Inp2 alias 00000K388931C=AKG C414;C414

  Out27 alias 00000K388931C=Master (L); Mst (L)

  Out28 alias 00000K388931C=Master (R); Mst (R)

  Bus7 alias 00000K388931C=Reverb (L);Rev (L)

  Bus8 alias 00000K388931C=Reverb (R);Rev (R)

  As can be seen above, the syntax for these alias lines is:

  <Inp1...28, Out1...32, Bus1...32> alias <Soundscape unit serial number>=<alias for mixer full view>;<optional alias for mixer small view>

  Note that inputs and outputs are identified by numbers only, with no reference to their type. The TDIF ones come first (1-24 for R*Ed); followed by AES-EBU and S/PDIF (inputs 25 and 26, outputs 25-28 for R*Ed); followed by the analogue I/O (inputs 27 and 28, outputs 29-32 for R*Ed). Very long names may not be entirely displayed in the mixer itself, but they will appear as specified in the input/output selection menu.

  Sections of the INI file can be saved in a text document, to be pasted later into different versions of the file, for example to use certain pre-defined window sizes and positions (the sections have easy-to-understand headings such as ArrangeWnd for the Arrange window, BigClocWnd for the Big Current Time window, and so forth), or to open the Editor with particular tool page configurations (see the ToolWnd section).

  Sets of audio and automation track colours can also be saved. I find the default track colours too dark when editing at waveform level, and I have created a set of 32 lighter colours, against which the waveforms stand out particularly well. It is indicated as follows in the INI file:

  TrkColorsUnit1=14737357 13223847 14471623 13548974 14600658 13809602 14208992 12892369 13355752 11447772 13164006 11521498 12969181 11524815 12772042 11524793 14410949 13162406 15130288 14339723 14734782 14009508 15257288 14662835 15453156 14725591 14798073 13082355 13490936 10533361 12511727 10149351

  This colour set is especially useful when working with a lot of stereo Parts, with each track pair using a very light colour for the odd-numbered track, and a slightly darker variant of the same colour for the even numbered track. Copy it into your own Soundscape INI file and try it out for yourself.

  Thanks to Christo Curtis and Wolfram Dettki for their help with this article.