HITHER GATE MUSIC

Starting Point (varied inputs)

The starting point can be any soundfile. It doesn't matter if the end result is poor because the point is to explore what happens with different types of input soundfile. The processing sequence is a template, an exploratory tool. It is used not only to produce good end results, but also to learn what kind of inputs work well with the given sequence.

The input should be mono, nicely dovetailed and reasonably short, such as between 1 and 4 seconds, because the RE-ENTRY process lengthens the sound considerably. The sequence can be realised one step at a time with a graphic interface or all at once with a batch file. A batch file is an MS-DOS mechanism which runs commands. It is a text file with a .bat extension. It contains a series of commands and explanatory comments introduced with the word rem. In the Sound Loom GUI, the processing sequence can be constructed as an Instrument.

I prefer to use a naming convention which indicates which processing function has been used. Thus, as a new sound grows, you can read the processing steps in the name itself.

Step 1: Copy

Make a copy of the input soundfile, giving it a single letter name.

It is presumed that the input is mono and begins and ends smoothly, i.e., without a click or being suddenly cut off. We only have to name the outfile. The default values for the copy procedure are used. because no change to the format is being made, i.e., sampletype = 1, header = 0 and formatout = 0. The outfile name is specified. The .wav extension is added automatically.

Dialogue box:

Default parameters

New name is 'a.wav'

Command:

rem function infile outfile
copysfx bashdt.wav a.wav

Step 2: Reverse

Reverse the soundfile so that it plays backwards

This is a very simple operation!

Dialogue box:

Only outfile name needed

Command:

rem function mode infile outfile
modify radical 1 a ar

PLAY ar.mp3 (reversed soundfile)
New name: 'a reverse'

Step 3: Splice

Splice together the forward and backwards versions

There will sometimes be a dip in volume at the join. This could be overcome by overlapping the two sounds in a mix operation, but we are accepting a possible dip as part of this processing sequence. The command asks that the beginning and end be spliced to help smooth the result.

Dialogue box:

Select the files to splice

Command:

rem function infile infile [...] outfile splice_beginning splice_end
sfedit join a r arfb -b -e

PLAY arfb.mp3 (spliced soundfiles)
New name: 'a reverse forward+backward'

Step 4: Dovetail

Dovetail the ends of the spliced file

We find that mixing operations later on don't sound very good if the sounds cut off too suddenly. So here we put a fairly long slope on the end of the sound so that it ends smoothly. We use a linear (rather than exponential) slope to give a steady (rather than increasingly rapid) reduction in volume, for maximum smoothness.

Dialogue box:

Smooth end of file

Command:

rem function mode infile outfile start_slope end_slope start_type end_type
envel dovetail 1 arfb arfbdt 0.05 1.0 0 0

Play arfbdt.mp3 (dovetailed soundfile)
New name: 'a reverse forward+backward dovetail'

Step 5: Analyse

Prepare spliced & dovetailed soundfile for spectral processing

The analysis operation uses the Fast Fourier Transform ('FFT') to create an 'analysis file', converting time amplitude data to frequency amplitude data. The Blur and Time-Stretch functions we will use next require this type of data input.

Dialogue box:

Convert to analysis data

Command:

rem function mode infile outfile
pvoc anal 1 arfbdt arfbdt.ana

We don't need to play this result because it sounds the same as the input – it's just that the soundfile data is in a different format.

Step 6: Blur

Blur across analysis windows & time-stretch 4 times; Re-synthesise

The Blur operation averages data values across a specified number of analysis windows. A window is an envelope-shaped group of analysis data frames. We won't resynthesise until after the Time-stretch operation – but the interim analysis files can be played in Soundshaper by using F10. which calls pvplay.exe.

Dialogue box:

Blur analysis data

Command:

rem function mode infile outfile number-of-windows
blur blur arfbdt.ana arfbdtbl50.ana 50


Option to PLAY arfbdtbl50.ana with F10 before resynthesising (blurred and time-stretched soundfile)
New name: 'a reverse forward+backward dovetail blur'

Step 7a & b: Time-stretch (4 times in 2 steps)

Stretch the blurred analysis file to make it 2 times longer.

In this operation, we can imagine an image of the soundfile drawn on a rubber band. The rubber band is then stretched sideways until, in this case, it is four times longer. Just as in the visual image, each audio feature is extended, unfolding gradually.

Dialogue box:

Stretch file in time

We now take this ouput and time-stretch it again by a factor of 2. Time-stretching by only a factor of 2 each time maintains quality in the output. Large time-stretches can be a quick way to see what will happen, but may produce unwanted artefacts.

Command:

rem function mode infile outfile number-of-times
stretch time 1 arfbdtbl50.ana arfbdtbl50x2.ana 2
stretch time 1 arfbdtbl50x2.ana arfbdtbl50x4.ana 2


Option to PLAY arfbdtbl50x4.ana with F10 before resynthesising (blurred and time-stretched soundfile)
New name: 'a reverse forward+backward dovetail blur stretchx4'

Step 8: Re-synthesise

Convert the time-stretched analysis file (.ana) to a soundfile (.mp3).

In the command, we don't have to add the .mp3 extension because this is done automatically by the CDP software.

Dialogue box:

Convert data to soundfile

Command:

rem function in_analysisfile out_soundfile pvoc synth arfbdtbl50x4.ana arfbdtbl50x4

Play arfbdtbl50x4.mp3 (blurred and time-stretched soundfile)
New name: 'a reverse forward+backward dovetail blur time-stretchx4'

Step 9a & b: Time-varying Transposition (two versions)

Steady glide downwards. We will create two versions, to be mixed together later.

We make the sound glide downwards by a time-domain transposition which goes steadily downwards 1½ octaves over the whole duration of the sound. A second version goes down 2 octaves. The first column in the breakpoint file is the time (adjust end value to actual length of the soundfile) and the second column is the transposition in semitones ('-' for lower). Thus the transposition breakpoint files atvpch1.brk and atvpch2.brk read:

 0	  0		 0	  0
19	-18		19	-24

Dialogue boxes:

Steady downward glide for 1½ octaves

Steady downward glide for 2 octaves

Command:

rem function mode infile outfile transposition_file
modify speed 2 arfbdtbl50x4 arfbdtbl50x4tv1 atvpch1.brk
modify speed 2 arfbdtbl50x4 arfbdtbl50x4tv2 atvpch2.brk

Play arfbdtbl50x4tv1.mp3 (glide down 1½ octaves)
Play arfbdtbl50x4tv2.mp3 (glide down 2 octaves)
New name: 'a reverse forward+backward dovetail blur time-stretchx4 time-varied-transposition'

Step 10a & b: Mix (two versions)

Mix together the two downwards gliding files

To make the final sound fuller and with more internal changes, we will now mix together the two descending glides. As the goes down further than the other, when put together, the sound spreads out in pitch range.

The first mix just uses the two transposed versions, both starting at time 0. The mixfile aftv1-2a.mix sets this out:

soundfile            starttime channels pan  level
arfbdtbl50x4tv1.mp3  0	   1        -1   -0.9
arfbdtbl50x4tv2.mp3  0	   1         1    0.9

Notice how the level is reduced a little and the files are panned far left and right: the first file will come out of the left speaker, and the second out of the right.

The second mix uses the same inputs but restarts them at 0.5 and 1.0 seconds respectively. The mixfile aftv1-2b.mix sets this out:

soundfile            starttime channels pan   level
arfbdtbl50x4tv1.mp3  0         1        -1    0.9
arfbdtbl50x4tv2.mp3  0         1         1    0.9
arfbdtbl50x4tv1.mp3  0.5       1        -0.34 0.8
arfbdtbl50x4tv2.mp3  1         1         0.34 0.8

Now the level is reduced a little more, and the pan is adjusted to fill the horizontal space evenly, the second set of files being placed closer to the centre (0).

Dialogue boxes:

Mix two soundfiles

Mix four soundfiles

Command:

rem function mixfile outfile
submix mix aftv1-2a.mix arfbdtbl50x4tv1-2mixa
submix mix aftv1-2b.mix arfbdtbl50x4tv1-2mixb

Play arfbdtbl50x4tv1-2mixa.mp3 (mix 'a')
Play arfbdtbl50x4tv1-2mixb.mp3 (mix 'b')
New name: 'a reverse forward+backward dovetail blur time-stretchx4 time-varied-transposition mix'

Step 11: Dovetail

Smooth end of second mix with a dovetail envelope edit

When we listen to arfbdtbl50x4tv1-2mixb.mp3, we find that it still ends rather abruptly, in spite of the dovetailing of arfb.mp3. So we will complete the processing sequence with a final dovetail.

Dialogue box:

Smooth end of file

Command:

rem function mode infile outfile start_slope end_slope start_type end_type
envel dovetail 1 arfbdtbl50x4tv1-2mixb arfbdtbl50x4tv1-2mixbdt 0.05 1.0 0 0

Play arfbdtbl50x4tv1-2mixbdt.mp3 (dovetail of mix)
New name: 'a reverse forward+backward dovetail blur time-stretchx4 time-varying-transposition mix dovetail'

Complete Sequence: Batch File reentry.bat

Use reentdel.bat to remove the outputs

rem day1.bat - 'RE-ENTRY' - batch file using first set of programs
rem Description: long slow modulating descent of a rich sonic complex
rem Last updated: 27 Mar 2003

rem The source file should be about 2-4 sec. long, mono.
rem Consider using a softsnd bat before or after this one.
rem Sources (with comments on how it sounds after processing):
rem   aklcdp.wav  (1.38 sec) - rhythmic element produces good changes
rem   bashdt.wav  (1.5 sec) - attack works well
rem   frogcdt.wav (1.8 sec) - surprisingly good, with high-pitched 
rem	  sounds at end: becomes 2'10" long
rem   hoggdt.wav  (3.7 sec) - lovely rich descent with extra high 
rem	  pitched component towards end
rem   tomcdt.wav  (1.3 sec) - not so good, not enough to work on
rem   touccdt.wav (3.0 sec) - rich texture
rem   whdtm.wav   (2.4 sec) - (whirling tube) only OK, not enough 
rem	  differentiation

echo on

rem Step 1. Copy to simple name
copysfx bash.wav a.wav

rem Step 2. Reverse
modify radical 1 a ar

rem Step 3. Splice forward & backward versions
sfedit join a ar arfb -b -e

rem Step 4. Dovetail end of the spliced file
envel dovetail 1 arfb arfbdt 0.05 1.0 0 0

rem Step 5. Analyse the spliced & dovetailed file
pvoc anal 1 arfbdt arfbdt.ana

rem Step 6. Blur the spliced file
blur blur arfbdt.ana arfbdtbl50.ana 50

rem Step 7a. Time-Stretch the blurred file 2 times
stretch time 1 arfbdtbl50.ana arfbdtbl50x2.ana 2

rem Step 7b. Time-Stretch the blurred file 2 more times
stretch time 1 arfbdtbl50x2.ana arfbdtbl50x4.ana 2

rem Step 8. Synthesise the 4x stretched file
pvoc synth arfbdtbl50x4.ana arfbdtbl50x4

rem Step 9a. Time-varied transposition (glide down 18 semitones)
modify speed 2 arfbdtbl50x4 arfbdtbl50x4tv1 atvpch1.brk

rem Step 9b. Time-varied transposition (glide down 24 semitones)
modify speed 2 arfbdtbl50x4 arfbdtbl50x4tv2 atvpch2.brk

rem Step 10a. Mix the two transposed versions (spread at bottom)
submix mix aftv1-2a.mix arfbdtbl50x4tv1-2mixa

rem Step 10b. Mix the two transposed versions, with 2nd & 3rd later
submix mix aftv1-2b.mix arfbdbl50x4ttv1-2mixb

rem Step 11. Dovetail end of the final file
envel dovetail 1 arfbdtbl50x4tv1-2mixb arfbdtbl50x4tv1-2mixbdt 0.05 1.0 0 0

echo off

Further fine-tuning of the result might typically include adding some reverb and directional movement (panning).

---

SCREENSHOT UPDATES

(by Robert Fraser, March 2016)

The Soundshaper screenshots used in this tutorial are out of date in two respects:

• Output files are no longer named on Parameter Pages: instead, the output is saved to a temporary file and appears on the Main Page patchgrid in a new cell named after the process.
  If you want to save the output permanently, highlight the cell (if necessary) and click File/Save or CTRL+S to obtain a standard SAVE dialog:
  
  
  
  
  There is normally no need to do this until the end of the sequence.


• FFT conversion (PVOC Analysis / Resynthesis) now uses current settings and the FFT page is only used to set these, not to perform a conversion, as shown. As the tutorial examples use default values, there is nothing to set!
  Furthermore, Soundshaper does these conversions in the background, so you do not even have to select ANALYSE or SYNTH – just select the next process.

To run the RE-ENTRY chain of CDP processes in Soundshaper:

• STEP 1 (COPYSFX): Open the tutorial folder. Drag and drop the source bashdt.wav onto Soundshaper's Main Page.
  Program runs COPYSFX and outputs to cell A_0.


• STEP 2 (REVERSE): Select Soundfiles/Extend/Reverse. No parameters – output appears in cell A_1.
• Copy cell A_0 to B_0: click on cell A_0; click the Copy Cell START button; click on cell B_0.
  The copy is needed so that the reversed sound can be the second input in STEP 3.


• STEP 3 (SPLICE): With B_0 highlighted, select Soundfiles/Edit /Join.
  'Add Input' dialog appears: click on cell A_1 to select this as the 2nd input. Click dialog's 'OK' button.
  Parameter page: set parameters as in tutorial – 15ms splice and tick both options (splice start and end).
  Click OK to run; output appears in cell B_1 ("Join").


• STEP 4 (DOVETAIL): Select Soundfiles/Edit/Fades (Dovetail).
  Parameter page: set parameters as in tutorial (Fade-in=0.05 Fade-out=1.0).
  Click OK to run; output appears in cell B_2 ("Fades").


• STEP 5 (PVOC ANAL): The settings used are the defaults, so there is nothing to set.
  Hit Ctrl+F9 or click Spectral/PVOC/Analyse. Output appears in cell B_2.
  (This step is actually unnecessary, as Soundshaper would run PVOC automatically before performing STEP 6.)


• STEP 6 (BLUR): Select Spectral/Time/Blur. Set parameters as in tutorial (Windows=50).
  Click OK to run. Output appears in cell B_4 ("Blur").


• STEP 7a/b (TIMESTRETCH): (a) Select Spectral/Time/Stretch. Stretch parameter: 2, as in tutorial. Click OK to run.
  (b) Hit F8 to repeat the process (or select it again). Parameter is already set at 2. Click OK to run.
  Output appears in cell B_6 ("Timestretch").


• STEP 8 (RE-SYNTHESISE): Hit F9 or click Spectral/PVOC/Synth. Output appears in cell B_7 ("Pvoc Synth").
• Copy cell B_7 to C_7: click on cell B_7; click the Copy Cell START button; click on cell C_7.
  The copy is needed because the next step involves two different sets of parameters.


• STEP 9a (TRANSPOSE): Click on Cell B_7. Select Soundfiles/Pitch/Speed.
  Parameter page: click the T-V checkbox for the TRANSPOSN parameter, as we'll be using a datafile.
  Also uncheck Auto-scale .brks at the top. (The datafiles are too long, so their times would otherwise be scaled.)
  Drag and drop the datafile atvpch1.brk onto the Parameter page (or click the Open button and select it).
  Click OK to run. Output appears in cell B_8 ("Speed").


• STEP 9b (TRANSPOSE): Click on Cell C_7. Hit F8 to repeat the process Speed (or select it again).
  The T-V checkbox is already checked. Uncheck "Auto-scale .brks" at the top, if necessary.
  Drag and drop the other datafile atvpch2.brk onto the Parameter page (or click the Open button and select it).
  Click OK to run. Output appears in cell C_8 ("Speed").


• STEP 10a (MIX): Click on Cell B_8 (Speed). Select Edit/Mix/Mix. 'Add Input' dialog: click on cell C_8 as 2nd input.
  Mix Parameter page: both inputs are listed in the left panel. Click on the first (~B_8.wav) and set its parameters as in the tutorial.
  Click on the second input (~C_8.wav) and set its parameters as in the tutorial.
  Click the MIX button to perform the mix. Output appears in cell B_9 ("Mix"). Play the result.
  (OPTIONAL: to preserve this result, before altering it in STEP 10B, select File / Save(As) or CTRL+S to obtain a standard SAVE dialog.)


• STEP 10b (MIX re-edit): Double-click on cell B_9 (Mix) to edit it.
  Mix Parameter page: previous inputs are already listed. Click the SELECT SOUNDFILE button and select ~B_8.wav (again).
  Click the SELECT SOUNDFILE button again and select ~C_8.wav. We need to edit only the newly added files:
  
  
  
  
  Click on ~B_8.wav (2nd listing) and set its parameters as in the tutorial (see screenshot); do the same for the 2nd listing of ~C_8.wav. Click the MIX button to perform the mix. New output appears in cell B_9 ("Mix"). Play the result.


• STEP 11 (DOVETAIL): Select Soundfiles/Edit/Fades (Dovetail).
  Parameter page: set parameters as in tutorial. Click OK to run; output appears in cell B_10 ("Fades"):



view full size



• SAVE FILE (OPTIONAL): Save the result to a permanent file. Hit CTRL+S or File / Save(As) to obtain a standard SAVE dialog.
  Enter a file name, e.g. re-entry.wav (or arfbdtbl50x4tv1-2mixbdt.wav if you prefer).


• SAVE PATCH (OPTIONAL): Uncheck "Include Sources", unless you want to save the copied source (~A_0.wav) with the patch.
  Click the patch SAVE button (central panel). Select a save name (e.g. ReEntry.ssp); click OK. Optionally enter a description; click OK.
  


• RUN PATCH (OPTIONAL): To run the patch with a different source (assuming the patchgrid has not been cleared), click on cell A_0, select a source (e.g. by drag+drop) and click the central RUN button, with the "Whole Patch" checkbox checked.


• RELOAD PATCH (OPTIONAL): If the patch is not on the grid, click the central LOAD button. Select the .ssp file (e.g. ReEntry.ssp). If you have not saved the source (~A_0.wav), the patch will still run with it, provided it still exists in the same location. Otherwise, you will be prompted to select a source. To run the patch with a different source, see RUN PATCH above.