News:

If you need instructions on how to get through the hotels, check out the enclosed instruction book.

Main Menu

Loquat pre-alpha released (Snow Leopard)

Farted by Loki Clock, July 30, 2010, 09:31:50 AM

Previous topic - Next topic

0 Members and 0 Guests are viewing this topic.

Loki Clock

Now an untested Leopard version is available.

As some of you may know, I've been working on a Text to Speech program, which I have last-minute-titled "Loquat" for a good number of reasons. This is the very first release of my very first released program, and I made it just for you Clock Crew! So I'd appreciate it if you Intel Mac users downloaded the thing. More platforms coming eventually.

http://www.lokiclock.com/rsrc/dl/Loquat_pre-alpha_10.6.dmg
http://www.lokiclock.com/rsrc/dl/Loquat_pre-alpha_10.5.dmg

Read the latest news post off my website for more info: http://lokiclock.com/

Related:
June 20th progress video
July 15th progress video
TTS Progress Journal
Media Package from Journal 5 - Zip; 7z
Output for a test string - Drǿmdi mig én drǿm í nǫ́tt, spoken in a cruder version of the voice from Norse Grammar Lesson 1. An updated Norse voice is coming soontimes.

More help:
Raw audio and Audacity
From raw files to voice

I1I1I1I1I1I1I11111I1I1I1IIIIIII1I1I1I1I11I

Could we see a demonstration before we download it please?
Quote from: PezDispenserclock;1948598Abba, I might not smoke weed, but I experiancing it being hit with a crowbar on a modded TTT server. Flashing colours, screen flipped, screen flying. Yup, I know how it\'s like.

Loki Clock

Quote from: Jamie's Elegant Moustache;1773043Could we see a demonstration before we download it please?

There's some demo material included in the package. You can also download the media package from Journal #5, or view these progress videos of the program's operation:

June 20th
July 15th

(OP edited)

screwdriver_clock

Wait I'm mildly confused. Isn't this the same as the "say" command in OS X already, since you can output sound files with it?

Loki Clock

Quote from: ScrewdriverClock;1773574Wait I'm mildly confused. Isn't this the same as the "say" command in OS X already, since you can output sound files with it?

No, because the Macintosh voices, like with most TTSes, are made by the engineers or technicians who created the product, and them alone. Though it can be argued that this is because their methods make producing the voices an intractable task for commoners, the real reason in my opinion is so that the voices are distinctive enough that they can take royalties off usage if they want, or at least stamp their brand all over your animation/song/whatever. I'm pretty sure the Clock Crew owes Speakonia a chamillion dollars.

Not only is my program free (I'll be setting up a donation box for tips soon enough), but the built-in editing tools for the voice and map make it able to adapt to any language - or even just personal characteristics.

The main purpose of my program is to make it possible for people to always be able to create a truly unique voice for their CC characters while maintaining the robotic character of a concatenative TTS voice, beyond screenshots of your favs Speakonia settings.

AstronautClock


Loki Clock

Quote from: AstronautClock;1773659I am using windows : (

Sowwy. If I knew how to cross-compile already I would probably start development right away. But I also have to deal with SDL later on. I have no idea how to get a cross-platform SDL program set up, actually, because SDL now runs through Cocoa on Mac, not directly in C++, which is incredibly obnoxious. Establishing the need to maintain cross-platform compilability early on will hinder later stages of the project, such as said SDL affair. If I learned how to do the SDL bit I would do it anyway, though.

AstronautClock

could you maybe upload some kind of mp3 somewhere? so that I atleast know how awesome it sounds

Loki Clock

Quote from: AstronautClock;1773663could you maybe upload some kind of mp3 somewhere? so that I atleast know how awesome it sounds

Oh yes! I forgot to post this.

http://lokiclock.com/audio/tts/Droemdi_mig_een_droem_in_navtt.mp3

Flounderpa

Quote from: Loki;1773667Oh yes! I forgot to post this.

http://lokiclock.com/audio/tts/Droemdi_mig_een_droem_in_navtt.mp3
sounds like an African tribal chant

also: It sounds a lot like this ( never mind the language)
[U2]xrhvjlGOmao[/U2]
ââ€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"¬à®Å"Û©ÛžÛ©à®Å"ââ€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"¬
SHUTãâ,¬â,¬Ã¯Â¼ÂµÃ¯Â¼Â°Ã£â,¬â,¬Ã¯Â¼Â¡Ã¯Â¼Â®Ã¯Â¼Â¤Ã£â,¬â,¬Ã¯Â¼Â¥Ã¯Â¼Â®Ã¯Â¼ÂªÃ¯Â¼Â¯Ã¯Â¼Â¹Ã£â,¬â,¬Ã¯Â¼Â´Ã¯Â¼Â¨Ã¯Â¼Â¥Ã£â,¬â,¬Ã¯Â¼Â­Ã¯Â¼ÂµÃ¯Â¼Â³Ã¯Â¼Â©Ã¯Â¼Â£Ã¯Â¼Â
ââ€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"¬à®Å"Û©ÛžÛ©à®Å"ââ€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"‰â€"¬

Slurpee

I fucked around with this for a while and then concluded that it's too advanced for me

I followed the instructions in Enneativedoc.txt and managed to output a file identical to /Loquat/_Voices/Enneative/Other Resources/song
and then I tried importing it Raw into audacity and it sounded like a squeaky robot fart
so I lowered the sample rate and got this
[flash]http://files.myfrogbag.com/8xu8snn0tggygnz/fartfartfart.swf height=40 width=40[/flash]
and I'm pretty sure that's not what was supposed to happen? I've never worked with raw audio before (in fact I'm quite sure I don't even know what it is), so I feel like I'm grossly mishandling it

so I was like okay maybe that's something different
so I skimmed over CaliAccent.txt and ReadMe.txt
and I think I see what you're doing, and it's brilliant, and I have a cali accent so I probably wouldn't have to fuck around with the phonetics too much, but I can't really do a test run because I have no idea how to output raw audio from audacity, and I wouldn't know how what to do with the datO output even if I did.

hahahaha sorry
I'm trying to be helpful but the entire thing is just like :confused:

Loki Clock

Quote from: Slurpee;1773752I followed the instructions in Enneativedoc.txt and managed to output a file identical to /Loquat/_Voices/Enneative/Other Resources/song
and then I tried importing it Raw into audacity and it sounded like a squeaky robot fart
so I lowered the sample rate and got this
[flash]http://files.myfrogbag.com/8xu8snn0tggygnz/fartfartfart.swf height=40 width=40[/flash]
and I'm pretty sure that's not what was supposed to happen? I've never worked with raw audio before (in fact I'm quite sure I don't even know what it is), so I feel like I'm grossly mishandling it

Oh snap. I forgot, see, to mention that one must also adjust for Audacity's miscalculations. The endianness (100 vs. 001 for a hundred) of the samples is big endian (100), and the byte offset should be 0. It would say that when you check the info on the voice if I'd set it to translate the encoding flags. I can read them, because they're my flags, but obviously no one else can.

Quotebut I can't really do a test run because I have no idea how to output raw audio from audacity, and I wouldn't know how what to do with the datO output even if I did.

To export raw audio from Audacity:
  • File > Export (Selection)...
  • WAV, AIFF, and other uncompressed types
  • Format: Other...
  • Header: RAW (header-less)

QuoteI'm trying to be helpful

You succeed. I'm adding this info to the ReadMe. Also, congratulations on being the first end-user with published media!