[L2Ork-dev] Expression of interest in GSoC 2020

[Jonathan Wilkes]

On Thu, Mar 5, 2020 at 8:45 PM Tsz Kiu Pang <osamupang at gmail.com> wrote:
>
> On Fri, 6 Mar 2020 at 12:27, Tsz Kiu Pang <osamupang at gmail.com> wrote:
>>
>> Hi all,
>>
>> I am writing to express my interest in participating in GSoC with Purr Data, in particular the project "Fake News Audio Library".
>>
>> I have been using Pure Data for almost four years now, and would like to further develop my skills by participating in GSoC this year.
>>
>> Something about me
>> ========
>> I am currently doing my last year of Master of Engineering (Electrical) at the University of Melbourne.
>> I have solid foundations in DSP, and am experienced in implementing filters for audio signals using C.
>> I have previously written some simple audio externals for pure-data after learning from https://github.com/pure-data/externals-howto
>> which I think would be the right skill for this project.
>> I also have experience working with the Pure Data code base, and have recently made a pull request: https://github.com/pure-data/pure-data/pull/853
>>
>> Demonstrating my skills
>> ========
>> To further demonstrate my skills, I am planning to write an prototype external to transform audio signal into a "warmer vinyl sound" in the coming week.
>> However right now I am having trouble building purr-data from source, in particular this error:
>> clang: warning: argument unused during compilation: '-fast' [-Wunused-command-line-argument]
>> pdlua.c:45:10: fatal error: 'lua.h' file not found
>> #include <lua.h>
>>          ^~~~~~~
>> The problem seems to be that gcc is not able to locate the include file.
>
> Sorry I forgot to mention that this occurs upon running the script ./tar_em_up.sh -X on mac10.14.6 after installing all the dependencies via brew.

Tsz Kiu,

It's great to see someone familiar with Pd getting interested in GSoC
for Purr Data!

For the error: it looks like you're somehow still missing the lua
headers. Did they install lua successfully with homebrew?

Check /usr/local/include and /usr/include to see if the header is there.

As far as GSoC conflict with your school-- I don't think you could
successfully complete a project given those scheduling constraints.

Beginning and end of the summer are crucial times. And 30 hours per
week is an average AFAICT.
You'll likely hit "crunch time" at some point in both the very
beginning and end of the project.
Not having adequate time to focus and ensure the code works during
those times would be
catastrophic.

Perhaps the other GSoC admins can chime in here?

As far as the projects: I think you're on the right track for the
"warm vinyl" idea. For the ambition of GSoC
that would probably be one of a small collection of "fake news"
objects. I'm happy to give some more feedback
on that and the other project ideas if you'd like.

Best,
Jonathan



>>
>> I have been attempting to solve this problem in the past hour, and would continue doing so tomorrow
>> However it would be great if you could please point me in the right direction.
>>
>> Question regarding the project
>> ========
>> In regards to the project "Fake News Audio Library", I would like to know more about cross-referencing the "fake news".
>> Since we are dealing with fake news, I am assuming that there would not be much reliable source (peer-reviewed journals, for example), and would have to rely on sources such as:
>> https://www.oregonlive.com/music/2014/11/does_vinyl_really_sound_better.html
>> where it says the vinyl sound is "very mid-range-y and very warm", and
>> https://hydrogenaud.io/index.php?topic=73636.0
>> where it says the vinyl sound is "a prominent mid-bass with a slow decline toward the high frequencies".
>> I am just wondering would this kind of referencing be sufficient enough as a starting point?
>>
>> Another question regarding the project is that how should the filters be implemented?
>> Specifically, should they be implemented as externals written in C? Or should they be implemented as abstractions patched in Pd?
>>
>> Other projects
>> ========
>> I am also interested in other projects such as "JIT-compiled Signal Graph for the Audio Engine" and "Data Over Audio Messaging". However, upon considerations, "Fake News Audio Library" seems to be the project which I am most confident about. However, if other projects are more significant to Purr Data or Pure Data, I am happy to look at those as well. Please let me know what you think.
>>
>> Time commitment
>> ========
>> Since I am from Australia, where the summer in the US is our winter, so I would only have approximately one and a half month in which I am able to contribute full-time.
>> I have also expressed this issue last year and decided not to proceed with the application due to the time commitment.
>> However this year I have a lighter study load as well as a better understanding of the Pure Data code base, therefore I would be able to make up the time overlapped with the semester.
>>
>> From the GSoC guideline, it seems like students have to spend at least 30 hours per week for 12 week, from 19/05 to 11/08, and the dates overlapped with my semester are from 19/05 to 31/05 (14 days) and from 27/07 to 11/08 (another 14 days).
>> I would be able to contribute to this project for around 10 hours per week during the semester time, and around 35 hours per week during my winter holiday.
>> And if necessary, I can also start researching once I got accepted into the project.
>> On top of these, I am confident with my knowledge in DSP and Pure Data, I would be able to complete this project even with the limit of my time commitment.
>>
>> Thank you very much for your time.
>> Please let me know if you have further questions.
>>
>> Kind regards,
>> Tsz-Kiu
>
> _______________________________________________
> L2Ork-dev mailing list
> L2Ork-dev at disis.music.vt.edu
> https://disis.music.vt.edu/listinfo/l2ork-dev