Edit: Program has been available for some time here: https://github.com/Goodlookinguy/FFMediaCacheGrabber/releases
I’m gauging interest in obtaining Google Translate’s generated speech files. What I mean is, if you type in “hello” then you could obtain an mp3 that says “hello.”
I already have a fully functioning prototype. The problem I had was directly downloading from a given word or sentence. I would have setup my software to do that if I could, but I had issues and was being returned ReCaptcha and not sound files. So I went with the next best option, Mozilla Firefox. I used Mozilla Firefox’s caching system to look around and detect media files being added to the cache. This software is basically an assist tool and not really a full on tool to directly get sound files.
This is demonstration of the software. I only realized after I finished recording this that I forgot to record the sound. Trust me though, there was sound.
Update: I can now grab the input text thanks to the metadata Google inserted in the generated mp3 files. This is a major improvement that I didn’t think I’d get. Right now I’m working on making the Download button execute another piece of software for striping the metadata Google stuffed into the mp3 files to not only remove their ownership crap (so you can redistribute the files), but to further shrink the file size. GoldWave is high on my list of compatible software that can accomplish this task. I’ll also be looking into Audacity if it has command line options.
So if you watched that all the way through, here’s the steps to successfully get a file…
1) Open up Firefox and go to Google Translate (do not have any other media-type websites open like Youtube, Soundcloud, and Pandora)
2) Start up software (currently dubbed FFMediaCacheGrabber)
3) Go back to Google Translate and put in your text.
4) Click on the “Listen” button. This will generate the mp3 file which will be cached immediately.
5) The top file to be added to the list in the software is the sound file you just generated in Google Translate. Click on the empty space in the newest row in the column for “New File Name”
6) Enter a new file name for this mp3, typically whatever word or sentence you wrote
7) Click download. The file will be put into the directory “media” in the folder that FFMediaCacheGrabber is run in. It will automatically have the file type appended. In the case of these files, “xyz.mp3”
8) Repeat steps 3-8 as many times as needed.
Now, knowing that you have to do all of those steps to just get a single sound speech file, how many people are still interested in this? I will refine it to attempt to make it easier for regular users if there is enough interest.