[PD-dev] Audio Regression and Unit Testing
abez at abez.ca
Fri Jun 4 19:37:00 CEST 2004
I apologize if this is the inappropriate forum for this discussion
Reading through the archives I noticed there were attempts at regression
testing Pd but I couldn't find any concrete results of these tests. I
suspect the scope of the problem might have been beyond the scope of Pd
itself thus it wasn't worth the effort. I do not as I am an outsider to
What I'd like to know is what functionality would one want to test and
provide regression tests for in Pd? What has been done in the past? What
difficulties have arisen in unit testing or regression testing. If one
wanted to unit test what type of assertions would one need, and if one
wanted to regression test what kind of functionality would be required
from the test framework?
When you're developing your applications and instruments do you ever
wish you didn't have to listen to it over and over again to verify
nothing serious has changed? Do you ever wish you could just test a
filter or generator to see if it was producing the right sound rather
than listening to it. Do you ever wish you could apply test first
methodologies to your filters, signals generators and instruments?
Such a tool would be great but how would you want to use it? Would you
want to integrate it into your testing framework or would you prefer it
provides it's own test framework. What would you want the test to cover.
Would you want it to cover IPC and latency issues? What common
functionality would you want from it. What kind of assertions would you
like to see.
There are difficulties in testing audio, you need complex functions to
figure out: how loud it is; whether or not certain frequencies are too
loud; is it producing a proper distribution of elements,.
What paradigms would like it to use. Would you prefer a procedural
interface or an OO interface. What languages would want the test suite
to directly support?
It's an interesting problem and if it sufficiently explored and
implemented it could definitely help the audio programming community.
If you are interesting in providing input regarding the requirements of
such a system please read the attached Invitation Letter as I am
executing a study on what the open source audio community would like to
see in the way of audio testing.
========= Invitation Letter =============
Dear Potential Participant:
Hello, my name is Abram Hindle; I am currently a masters student in the
faculty of Computer Science at the University Of Victoria. I am
investigating the requirements of a Audio Testing Suite. as part of a
directed studies course supervised by Dr. George Tzanetakis .
The purpose of this research is to extract the testing requirements of
audio programmers. By gathering the requirements from potential users of
an audio test suite we can build an effective audio test suite. This
will help enable audio programmers regression test their code as well as
employ test first methodology to audio development.
Participants should be experienced in computer instrument design or
audio programming. Participants will be asked to fill out a
questionnaire. Data from the questionnaire will be recorded for future
use but the results anonymized to protect the participants identity.
Participation in this study is completely voluntary and the participant
may withdraw at any time without any consequences. Data will be
collected only with your consent. Participants will participate
Your participation is very important as it can improve the state of
software engineering in the audio domain. You may contact the PI, Abram
Hindle by email abez at uvic.ca or by telephone 250 721 9066 for further
information about the research.
Thank you for your assistance.
You may participapate by visiting http://abez.ca/~abez/audiotest/
Graduate Student of the
Faculty of Engineering
Department of Computer Science
University of Victoria
http://www.abez.ca/ Abram Hindle (abez at abez.ca)
More information about the Pd-dev