*** WARNING: Originally Speechy was NOT meant to be anything other than simple voice to voice translation with amazing recognition but after tons of requests for other features, we kept the same UI but added-on as this is what our USERS wanted and we go according to what they want, even if this wasn't what we originally desired. ***
What is Speechy...?
A translator you can really talk to, literally. Say what you need and Speechy will translate by writing in the language you desire but also immediately speak out loud what you said for either others to know what you really want to say in a foreign language or as some are using it as a great tool if you're just trying to learn the language to prepare for your next trip.
Next time when you think you need to rent a Translator, Speechy is here!
Just remember, when you're at an Airport and no one speaks your language and you don't know the native tongue, you can say "How do I get to Terminal B" and know someone who speaks one of the pre-bundled languages can help you by listening to a voice or read the translation on your phone.
Did you know... you can now AIM your phone at a sign, magazine, or any form of a document with words and Speechy Camera Tool (SCT using a form of OCR) can actually scan the words, translate what it understood to the language you desire and even speak it back for you.
How is this DIFFERENT?
1). Mic/Recording Input
Simple! Our voice input function that grabs the audio you speak either close or whichever distance you may be and takes the input from the phone Mic. When using your "Mic", we've designed our own Equalizer that is built-in that actually has various functions and decreases/increases pitch, balance, volume and other aspects to obtaining the audio you speak in which it also breaks it up in to a bunch of "chunks" and sending it via Internet (Wifi Connection or 3G Required) in multiple parts to then have it run through a Speech Server where it is analyzed in each segment of your Mono input (Never Stereo).
2). Cloud Speech Server
When your multiple chunks are each processed one-by-one for better understanding, they are first run under another synthesizer to turn everything flat (Bass, Treble, Pitch/Frequency, etc...) and another technique that let's us better understand what you say, even if you have an accent to the best we can and as each few moments are separated, it allows us to promptly guess what each word is being said and if quickly it fails a built-in/server-run spelling database (Query's originally from Bing!) and if it is still valid it will continue on to check to see if it can find the next best thing using the 2nd Dictionary aka "Phrases" which uses a form of AutoComplete for quicker search which is why you may notice sometimes a little bit of lag time/timely response as it is doing it's very best to try and be 100% accurate.
3). Translated Voice Download
Once everything is assumed to be as accurate as possible it will then pass the string through the Bing Translator API and in this case grab the audio from the Bing Translator service and pass it through the App as a "Stream" to download it temporarily (removes every translation or application closing/back button pressed) but only to prepare for output in Mono or Stereo (where available).
4). Translation Output
Once the file is available it immediately streams back out at a specific pitch, frequency, mono playback but made a synthesizer that barely modifies the output a little more to make it louder and easier to understand/sound better on almost any device but more important if Stereo headset/playback is available where most App's just set Mono to Stereo for output but we playback in mono and modify it just a tad and again prepare it if Stereo should occur (Balancing the L/R Channels) but most may not even hear the difference in a loud room but the very least more clear.
Hope this gives you a better insight why a Speechy Pro exists for $0.99 where the free edition is available (1.2) but doesn't have all the features and also has Popup Ad's (60-second delay; Ad may not appear depending on the hourly usage from all the users... This means if 1,000 users used it in a 3-hour time frame or 333/Hr, we disable the Ad for 3-hours as a big THANK YOU but don't think we get paid big bucks, we get about 1-2 pennies per view as we use Smaato for Ads) that aren't really intrusive as you can still record/view/hear with it popping up. Once we hit a specific amount of downloads in the pro edition, we will only continue releasing pro but freely but at this time unsure when that will be but it will just pay for the lighting/power, A/C, Internet, and misc items that we needed to build this.
Thanks!
Pro Edition Download Here - 1.2 Available
Free Edition Download Here - 1.2 Available
*SPECIAL THANKS TO (No specific order):
nok_fusion, scoobysnacks, NeilM, diboze, boz, buffalosolja42, mmian, rajivshahi, fuxinyou, wpxbox, and FlashTek95.
Thank you for helping test Speechy as you've been a lifesaver to have testers all over the globe!
What is Speechy...?
A translator you can really talk to, literally. Say what you need and Speechy will translate by writing in the language you desire but also immediately speak out loud what you said for either others to know what you really want to say in a foreign language or as some are using it as a great tool if you're just trying to learn the language to prepare for your next trip.
Next time when you think you need to rent a Translator, Speechy is here!
Just remember, when you're at an Airport and no one speaks your language and you don't know the native tongue, you can say "How do I get to Terminal B" and know someone who speaks one of the pre-bundled languages can help you by listening to a voice or read the translation on your phone.
Did you know... you can now AIM your phone at a sign, magazine, or any form of a document with words and Speechy Camera Tool (SCT using a form of OCR) can actually scan the words, translate what it understood to the language you desire and even speak it back for you.
How is this DIFFERENT?
1). Mic/Recording Input
Simple! Our voice input function that grabs the audio you speak either close or whichever distance you may be and takes the input from the phone Mic. When using your "Mic", we've designed our own Equalizer that is built-in that actually has various functions and decreases/increases pitch, balance, volume and other aspects to obtaining the audio you speak in which it also breaks it up in to a bunch of "chunks" and sending it via Internet (Wifi Connection or 3G Required) in multiple parts to then have it run through a Speech Server where it is analyzed in each segment of your Mono input (Never Stereo).
2). Cloud Speech Server
When your multiple chunks are each processed one-by-one for better understanding, they are first run under another synthesizer to turn everything flat (Bass, Treble, Pitch/Frequency, etc...) and another technique that let's us better understand what you say, even if you have an accent to the best we can and as each few moments are separated, it allows us to promptly guess what each word is being said and if quickly it fails a built-in/server-run spelling database (Query's originally from Bing!) and if it is still valid it will continue on to check to see if it can find the next best thing using the 2nd Dictionary aka "Phrases" which uses a form of AutoComplete for quicker search which is why you may notice sometimes a little bit of lag time/timely response as it is doing it's very best to try and be 100% accurate.
- e.g. "This is a test" may normally be recognized by other Speech Services as "This is a pest". In this case the word "pest" could be valid as it is a US-English Dictionary word but we have one last pass-through that has most common American English phrases that currently only has around 2,000+ but as people report more what they've tried to say and hasn't worked we can add it to the 2nd dictionary for accuracy.
3). Translated Voice Download
Once everything is assumed to be as accurate as possible it will then pass the string through the Bing Translator API and in this case grab the audio from the Bing Translator service and pass it through the App as a "Stream" to download it temporarily (removes every translation or application closing/back button pressed) but only to prepare for output in Mono or Stereo (where available).
4). Translation Output
Once the file is available it immediately streams back out at a specific pitch, frequency, mono playback but made a synthesizer that barely modifies the output a little more to make it louder and easier to understand/sound better on almost any device but more important if Stereo headset/playback is available where most App's just set Mono to Stereo for output but we playback in mono and modify it just a tad and again prepare it if Stereo should occur (Balancing the L/R Channels) but most may not even hear the difference in a loud room but the very least more clear.
Hope this gives you a better insight why a Speechy Pro exists for $0.99 where the free edition is available (1.2) but doesn't have all the features and also has Popup Ad's (60-second delay; Ad may not appear depending on the hourly usage from all the users... This means if 1,000 users used it in a 3-hour time frame or 333/Hr, we disable the Ad for 3-hours as a big THANK YOU but don't think we get paid big bucks, we get about 1-2 pennies per view as we use Smaato for Ads) that aren't really intrusive as you can still record/view/hear with it popping up. Once we hit a specific amount of downloads in the pro edition, we will only continue releasing pro but freely but at this time unsure when that will be but it will just pay for the lighting/power, A/C, Internet, and misc items that we needed to build this.
Thanks!
Pro Edition Download Here - 1.2 Available
Free Edition Download Here - 1.2 Available
*SPECIAL THANKS TO (No specific order):
nok_fusion, scoobysnacks, NeilM, diboze, boz, buffalosolja42, mmian, rajivshahi, fuxinyou, wpxbox, and FlashTek95.
Thank you for helping test Speechy as you've been a lifesaver to have testers all over the globe!
Attachments
Last edited:
