Composers Desktop Project Forum

In Windows CMD, you would write a batch script, the core of which is a sequence of CDP statements, something like this:

modify brassage 6 %1.wav %1(brassage).wav 0.7 5.0 100 0 1 0.5 5 5
reverb  %1.wav %1(reverb).wav 0.5 0.5 1 0 0 5
pvoc anal 1 %1.wav %1.ana -c1024 -o3
blur blur %1.ana %1(blur).ana 20
pvoc synth %1(blur).ana %1(blur).wav

%1 is a parameter you're passing to the script, in this case the name of your source file (e.g. "myfile.wav"), which is the common infile for each command line. You save the script to a text file and then change its extension to .bat (e.g. "mybatch.bat"). You then call the bat within CMD by typing:

mybatch myfile

...where myfile substitutes for %1 in the script. You don't need to call it as mybatch.bat, though you could. You mustn't put myfile.wav because that text would substitute for %1 and would become (in the first line):
 modify brassage 6 myfile.wav(brassage).wav [etc] which wouldn't work.

For blur, the infile is first converted to .ana using pvoc anal and afterwards back to sound using pvoc synth. Of course we could run a number of spectral commands with %1.ana. We can also delete the .ana files when we're done with them; after the pvoc synth line add:
 erase %1.ana
 erase %1(blur).ana

We also need to tell CMD where to find CDP programs and our files.  Unless you copy CMD.exe to your project folder, you need to make that folder the current one:
  cd \SNDS\MYPROJ
(or whatever the folder's called) - if it's in the same drive as CMD; if a different drive, put the drive name first, e.g.:
  E:
  cd \SNDS\MYPROJ
 
(for the folder E:\SNDS\MYPROJ)

These statements can be typed at the console before running the script.

To point the script to CDP programs, you need a PATH statement, e.g.:
 PATH C:\CDPR7\cdprogs
This can also be typed at the console, or put in the script before any CDP command lines.

The script itself is probably best put in the project folder, beside its infile. If not, the script needs the full path to the infile, e.g. call it as mybatch E:\SNDS\MYPROJ\myfile so that %1 picks up the whole path\filename. (My own preference is to use a single processing folder and copy all sources to that using copysfx, but that's another story.)

Finally, note that CMD takes only parameters %0 to %9, so there's very limited scope for passing values to the script. This is a pain, because CDP typically uses many parameters. The answer to this and other scripting issues would be to run CDP from something like PYTHON, but that solution will need some careful planning first.

Please see the thread "How to perform with spectral processes", which raises the same question. The answer is that a single-click is needed when you select the second input file from the grid. One of the replies in that thread describes it in great detail.

I will, however, change the coding in the next version of Soundshaper (or earlier) to interpret a double-click as a single one when the Add Input panel is visible.

Hi Phoenix cat,
When you select the first file, it will load into cell A_0 by default. Now click on cell B_0 and select the second file.  Click back on A_0. Select the process e.g. COMBINE CROSS (Menu: Spectral> Morph/Formants> Cross), and the red "Second Infile" panel appears for you to select the 2nd infile. Click on cell B_0 to do this and click the panel's OK button to proceed to the parameter page. CROSS has only one parameter-- REPLACE (degree of replacement - I usually set this to 1). Click OK or hit RETURN to run the process and return the result to Cell A_1.   

You are correct that CDP requires a conversion from wav to ana, but Soundshaper does this automatically. So the infiles can be soundfiles, as the program will auto-convert then into anas behind the scenes for a spectral process. (You can also convert them first if you prefer.) The output is also converted back to .wav if Option 5 (Auto-FFT) is checked. The output cell A_1 is coloured aqua for a spectral file, but the player will play the .wav conversion rather than invoke PVPLAY to play the .ana file directly. If you don't like the result, double-click the output cell (A_1) to return to the parameter page or hit CTRL+Z or the delete key to erase the output.

For the initial selection, the reason you have to click the cell you want to load into (actually you can just click its row) is so that you could load a file into any row. If Soundshaper just selected the next row down, this would not be as flexible. You don't need to use the file pool, but I find it helpful to stock this with commonly used sources -- it depends on your project. Also, you can load a file by drag+drop from any folder open in Windows Explorer. Again it will load into the currently selected row.  Soundshaper can also process stereo or multi-channel files for spectral processes. If you've chosen, say, one mono and one stereo source for COMBINE CROSS, it will combine channel 1 of the stereo source with the mono. (This generally works better than mixing to mono, but again you could do this yourself first.)   Hope this helps, Robert

Oops! I'll get that link changed right now. 
The correct link is: http://www.ensemble-software.net/CDPDocs/downloads/CDPDocs_PDF.zip
This is also accessible from the updates page (http://www.ensemble-software.net/CDPDocs/html/updates.htm).
Sorry about that.

The CDP reference documentation (excluding tutorials) has now been converted into printable PDF format and can be accessed here:
http://www.ensemble-software.net/cdpdocs.html

This page has links to download the PDF docs as well as the HTML docs as zip files, and also to view individual PDFs online. There is a link to the online HTML docs, which is also duplicated in the menu (CDP > Online Documentation > CDP Docs Home Page).

(Various charts and a couple of papers were already online as PDFs and are included within the HTML zip.)

Thanks for your replies.  The count sound should be OK. The warning message "<filename> not found" is just Soundshaper's message that the CDP program appears not to have worked. For most processes, a CDP error message will also normally be displayed, provided there is one.  Roomverb is working on my Win10 machine, both with mono and stereo inputs, and it does produce error messages (e.g. for room size= -1).

Have you checked the paths in Settings, as I mentioned above?  -- especially the temp folder and more importantly that the CDP programs folder is correct? Also that you actually have rmverb.exe in there? (note the spelling, not roomverb). [I know these checks sound obvious but they're the cause of most problems that have arisen in the past.]

It is possible to set up a terminal (command window) in Windows to run either single command lines or a batch file, including the one Soundshaper assembles (e.g. ~A_1.bat in your rmverb example; this should be in your TEMP folder.)  For the command line, the Soundshaper batch files currently name the in/outfiles in a non-friendly way (e.g. rmverb  %2_%3.wav  %n%_%o%.wav 1 0.5 0.5 0 0 0 5),  but you can amend this to something more obvious like count.wav (in) and countvb.wav (out). (You can't run these batch files "as is", unfortunately, without knowing the arguments that are being passed to the batch.)

There is more than one way of setting up a CDP terminal, but mine is to copy CMD.exe from the Windows folder to the TEMP file and then write and run a simple batch file (CDP.bat) which I also place in the TEMP folder:

REM cdp.bat for Release 7.0
echo off
PATH E:\CDP\cdpr7\_cdp\_cdprogs
SET CDP_SOUND_EXT=wav
cls
echo ***********************   COMPOSERS   DESKTOP   PROJECT   **********************

The PATH statement should lead to your CDP programs folder; mine is on drive E:, but yours may be somewhere else. In fact this folder can be anywhere unless you're running Soundloom, which expects the directory structure \_cdp\_cdprogs.

To run this, launch the terminal from the taskbar (right-click on the CMD.exe icon and select Pin to Taskbar) and type "CDP" (for "CDP.bat").

Back to your specific issues, I'd still like to know if VIBRATO runs OK or not! Also, I feel you may be better pursuing this directly with me by e-mail (rob@ensemble-software.net). I'll be happy to post the solution here when we find it: there is one!

As far as I know, modify loudness 4 is the correct command. For example, -l0.707946 will give -3dB output, whether the input is lower or higher than this (just tested both).