File me.ronan.tt5-release.apk · 36.2 MB · click to show
Fixed: 1. Administrative options were shown to people without the permission. 2. Added external microphones choices in the voice activation long press menu. 3. Fixed TeamTalk transmitting sound through the "FAST" profile used for games. Now sound shouldn't go tinny through speakers. 4. Added aditional permission checks for administrative options besides those available.
P.S. If microphone sounds tinny or bad, turn off voice activation, talk through the TX button, then switch microphone modes to standard mic and it'll come back. I'm working on seeing where this is from and fix it.
So, we need to talk.
I know that some of you may wonder: Since TeamTalk pro was a commercial, paid product, are you ever going to charge for your TeamTalk build?
Let's make something clear from the beginning!
TeamTalk is a free platform, both on mobile, PC, Linux and even MacOS. None of it's features and functionality are paid or hidden behind a paywall, share TeamTalk wall, Join our server to unlock wall. TeamTalk has and always will be free, according to what Bjoern did over the years with it.
On that note, my client will forever be free of charge and not linked to any other service, any other RoApps identity or account or associated with RoApps itself.
Once I bring in more features and collect feedback from you, I'll pull request and mainstream them all into TeamTalk's official GitHub project, so that Bjoern can implement them.
Please share my client with the people, help us accumulate opinions and features, then contribute to TeamTalk together with my client. No, TeamTalk will never become a paid client.
File TeamTalk-5.23.0-arm64-signed.apk · 41.5 MB · click to show
To recap: Fixed: Distorting microphone if no microphone boost was active. Fixed: When using TeamTalk for the first time, you no longer have a microphone boost predifined. Fixed: If you enforce usernames by default, but the people on your server come from a blank, guest, admin or default account, their nicknames will be shown. This is to prevent the situation where you want usernames, nobody has usernames and your TeamTalk goes: "Has logged in" "Has joinned the root channel". Added accessibility actions to everything that can be long pressed through out the app. This includes servers in the server picker, users, channels, online users, maybe files inthe files tab, not too sure if I linked those options too. Removed the MSG button from near a user in the main tab. Refactored the pretty useless "Talking" status from near the RX/TX status. You will now see in realtime who is talking, who is not talking, who is sharing media, who is transmitting video, who is sharing a remote desktop sessions. Translated the app in over 80 languages, thanks Google Gemini for the quick and good batch work on translations! Todos: Upgrade the stream media option to include direct links and YouTube URLs, recording of conversations, streamming of videos and viewing of videos.
File TeamTalk-5.23.0-arm64-signed.apk · 41.5 MB · click to show
Fixed a bug where backing out of the settings menu caused transmission to abruptly stop, special thanks to Deborah for this bug! Added: In any text box, enter submits the writen content, like this: In single line text boxes, Enter simply submits and ends the popup E.G: Server editing, etc. In multi-line text boxes, Enter when pressed from the alpha-numeric keyboard submits the text, while Enter from the symbol keyboard inserts carriage return/new line. Thanks Ramblio for this design decision. Removed confirmation popups from actions clicked via accessibility node actions E.G: Kicking, banning, editing/deleting. Pressing these actions via the context menu clicks will still show the popups.
File TeamTalk-5.23.0-arm64-signed.apk · 41.5 MB · click to show
Fixed a bug where pressing Enter on keyboards dismissed the whole parent activity, I guess I hurried up too much. Test on editing/adding a server, creating a UAC, etc...
Direct message to Valeria Team!
You fucking kangers, stop posting stuff on ya channel without giving credit! I imediately ask you remove my app unless you credit my channels and my community! What you kangers are doing is totally against FOS and I'ma not going to let you do that! Credit me before posting my God damn stuff on your channel!
The objective of this project is to make it as good as to later be included in upstream TeamTalk.
For this reasons we want bugs fixed, issues reaching me, and people actually saying what they don't like. For you, Valeria Team, that have been here enough to see what we're doing here, then send no users to my community while still sharing the app on your communities is disrespectful towards our objective. It leaves problems unanswered, bugs unfixed, and your users stairing into nothingness, and I don't want that. I want the users of the app to actually reach here, speak their minds off, and let me fix everything I can fix. It's totally immoral to do this. As a lesson for everyone else:
Respect the flow of a project! It's not about entitlement to community, it's about not letting users scream when there are bugs and nobody to hear them. Do post the app on ya channel, but credit the community, mention that I, the developer, am open to feedback from users, and let's make the app as good as it can get together!
Good morning, Deborah here!
A few days ago, in an Italian group dedicated to blind Android users has been sent a message for acquiring information about how, on Jieshuo, to recognize an image with OCR recognition.
First of all, it is not necessary to install any extensions, everything is accessible from the Jieshuo installation alone.
I personally am using the updated version of Jieshuo Plus.
1. Open Gallery, WhatsApp, or the app where an image is saved;
2. With a quick swipe from top to bottom and from bottom to top, without lifting your finger from the screen, get to 'OCR recognition'.
3. do a short swipe down, a flick to be more precise
Wait a few moments and the text-to-speech will start reading everything on the image, provided you have it focused.
I hope this helps. Have a great day, and I'm here to help should you need more information!
Over the past couple of months, I noticed some of the more capable record labels in Romania shifting away from complete AI instrumental generation, and it made me happy. I almost gave up on newly released music completely for this matter, and with good reason.
One of my earlier affirmations was that the cord progression makes half or more of the song.
That is true, but a song also means lyrics, vocals, and instrumental and/or production.
An instrumental made by human hands means attention to detail. It means sequential creation, not parallel creation. Sequential creation allows you, as a producer, to focus your whole attention on the current thing you are trying to produce at the given time, at and during runtime.
Parallel production, on the other hand, cannot see production as a multitrack, separated process. This leads to the absence of earcandy, the absence of personality in instruments, rhythm patterns, or even cord choices.
An instrumental made by an AI means absolutely no character and decisions taken at runtime, once, with the production choices, more poor than clearly thought out, made before the idea becomes an audio file.
At this point in time, the most capable and intelligent AI would be the AI that allows you to have the stems of the project before or while the AI spits the whole song.
Most, if not all, the AI projects and platforms available to date give you a stem separator long after the AI has "thought" out the song, chose instruments, and generated the final audio.
The "GET STEMS" option on AI song generators doesn't mean open project, solo each stem, render it out, save it in a folder. Or Open project, render dialog, render to, master plus stems before even making the track available. To AI music generators, that is what stem generation should actually mean. Instead, all it does is run it through a vocal remover, no matter how good or bad it is, most of the time it's not the highest end technologies from Hugging Face, because people can't wait for
A reply from sadra darbi which I thought should be posted on the channel and elevated:
Yep. That's why these absolute idiots must stop generating native audio output. That is a cool, novel idea, although it's not considered novel anymore because it's become repetitive, but it's not a good thing. I think they need to learn how to use actual professional music software by generating MIDI or code and creating music in that form. Of course, not just by running random code; the AI must learn how music works, with clear labels for emotions, notes, and so on. If the model is a general world music model capable of writing lyrics itself, then it can also define the emotions. This has already been done by multiple still-experimental models, or even public models like MIDI experimental models trained on massive MIDI databases. But that still does not matter, as the model needs clear labeling of all kinds of music around the world, along with enough data and documentation. Until these big companies do that, human artists are still better in all kinds of ways: emotion definition of the song, the ability to make as much song as they want, think on it as much as they want, use whatever they wish in the production, and much more. This doesn't quite relate to the message above, but just thought I should give my thoughts on that.
https://saves.eldrum.com/95d6ae73-4126-4da6-9ed2-50cd8a1e3e9d
Red Tide save.
Features:
1. All stats set to 8. Strength will be regulated after you wake up from the dream.
Unlimited Burning City timeslots, so you can explore everything possible in the burning city to unlock all the possibilities later when the time comes.
And that's it!
Enjoy!
Note:
Both Android and IPhone users can use this.
Black dust save.
Features:
Attributes set to 10. Trade points (for arena factions) set to 50.
Gold: Set to 9850 so you can pay the cartel as soon as you get the lockpick.
To get the lockpick:
1. Help Alma. Don't take the money.
2. Refuse to pay collecters.
3. Get a bad cartel reputation.
4. In the moalhouse, ask Reona about Alma. 5. Meet her at the dead end.
Now you can increase the rep with them, though I would recommend you hold on where required and make a high Orox Tribe reputation and relationship.
File GoogleTTS_Player-v1.8.0-arm64-v8a-releaseSigned.apk · 95.1 MB · click to show
Added support for TalkBack's language selector feature. You can now apply voices to be used for a specific language. Maybe fixed AutoTTS, I'm not sure as I don't have it purchased, someone may please test and tell me. Added the ability to use faster speech rate. Working on artificial rate for Seanet and Lemonbalm voices. Bypassed sound touch DSP by default, tell me if you encounter any volume/rate issues, I didn't encounter any. Added an audio upscale feature. It doesn't work so well now, but I'm looking into more computational heavy sollutions. Fixed idle battery drain, fixed lag.
PLEASE REMEMBER!
This version bumps up the GoogleTTS revision!
Finally! It's been more than a year, maybe a year and a half since Google released a voice/language revision bump! Now one is here! All the voices have been bumped to higher revisions! Please clear your data after installing this update and download the new voices! Enjoy!
File GoogleTTS_Player-v2.0-releaseSigned.apk · 95.1 MB · click to show
Version, two, point, zero, is, here! Hello everyone! it's been a very long time since you've seen a large jump like this one, and this version has emotional value for me. I want to thank you, from the bottom of my heart, for being here close to me, testing this closely, and reporting every, single, bug! We had an amazing path together through this -- testing, fixing, all that sort. Before I come in with what's new, I want to thank you, once more and again, for being here, up close, and making this, through your testing and ussage, worth my while. This app started from a simple wish I had, to run neural Google voices free of MultiTTS. Even now I remember the first 1.0 and 1.5 UIs with MultiTTS layout, lag, slowness, and all the imperfections. But it was a path. A small path, but still, a path ahead. Ever since then, we've expanded in a GoogleTTS experience no other GoogleTTS implementation is able to provide, and this all because of and with the help of you. GoogleTTS version 2.0 fixes, once and for all, the TalkBack language features, but most importantly? Artificial speech rate for Seanet and Lemonbalm voices. Now you can finally use Seanet voices at high speech rates, without switching away into Berdock half the speed curve because the speech becomes incomprehenceable just like the official GoogleTTS does. But most important, Lemonbalm, the highest quality architecture within GoogleTTS, can now be used at high speeds. We've come from stupid ZIP files to a voice downloader, from all the voices shown without any order to an amazing, clear, and beautiful interface, from nothing aditional to our own improvements over GoogleTTS. Using the neural voices at any speed with any screen reader was, in my opinion, worthy of a version jump. Before I finally stay back and have a rest from this project for a while, I want to thank you from the bottom of my heart for being here testing with me, an app that came from an obsession! Now GoogleTTS Player offers a unique experience wi
File GoogleTTS_Player-v2.0-releaseSigned.apk · 95.1 MB · click to show
Version, two, point, zero, is, here! Hello everyone! it's been a very long time since you've seen a large jump like this one, and this version has emotional value for me. I want to thank you, from the bottom of my heart, for being here close to me, testing this closely, and reporting every, single, bug! We had an amazing path together through this -- testing, fixing, all that sort. Before I come in with what's new, I want to thank you, once more and again, for being here, up close, and making this, through your testing and ussage, worth my while. This app started from a simple wish I had, to run neural Google voices free of MultiTTS. Even now I remember the first 1.0 and 1.5 UIs with MultiTTS layout, lag, slowness, and all the imperfections. But it was a path. A small path, but still, a path ahead. Ever since then, we've expanded in a GoogleTTS experience no other GoogleTTS implementation is able to provide, and this all because of and with the help of you. GoogleTTS version 2.0 fixes, once and for all, the TalkBack language features, but most importantly? Artificial speech rate for Seanet and Lemonbalm voices. Now you can finally use Seanet voices at high speech rates, without switching away into Berdock half the speed curve because the speech becomes incomprehenceable just like the official GoogleTTS does. But most important, Lemonbalm, the highest quality architecture within GoogleTTS, can now be used at high speeds. We've come from stupid ZIP files to a voice downloader, from all the voices shown without any order to an amazing, clear, and beautiful interface, from nothing aditional to our own improvements over GoogleTTS. Using the neural voices at any speed with any screen reader was, in my opinion, worthy of a version jump. Before I finally stay back and have a rest from this project for a while, I want to thank you from the bottom of my heart for being here testing with me, an app that came from an obsession! Now GoogleTTS Player offers a unique experience wi