This is too hilarious not to post. I got this a while ago when I was messing around with Suno.
I'd put my transcription of the song from the piano scene in Manamon 2 into Suno and generated tons of covers. This particular one came out when I asked it to make a rock cover.
It sounds pretty good, but instead of ending when the song finishes, it suddenly goes off into a completely random chromatic walk-down on the piano. For some reason, it just keeps descending, even though this is nothing remotely close to the rest of the song.
If that isn't weird enough, you'd think it would stop at the lowest note of an actual piano. Nope! Apparently, Suno thinks pianos have an enormous supply of extra bass strings below their normal range. It keeps trying to produce lower and lower notes, but the tonal component gradually gets swallowed up by increasingly strange noise. Then, just when you think it's finally done, it loops right back around to where it started!
I have absolutely no idea what possessed it to do this, but the result is hilarious.
@daygar and @FreakyFwoof, I thought you might find this interesting and funny.
ondrosik reshared this.
reshared this
Derek Lane reshared this.
like this
Peter Vágner reshared this.
You've never really heard accordion music until you've heard Ukrainian accordion virtuoso Oleksandr Khrustevych playing.
Here he is performing Vivaldi's The Four Seasons: Summer 🎶 🤩
None of this is thanks to private equity, venture capital, the Rockefellers, the Kochs, Peter Thiel, Silicon Valley or AI, just pure human heart, artistry and dedication.
reshared this
Peter Vágner likes this.
Peter Vágner reshared this.
Peter Vágner reshared this.
reshared this
Here's a torrent from archive.org: archive.org/download/500-sound…
ondrosik likes this.
reshared this
reFX Cloud is built with JUCE 8. It exposes its controls through UI Automation, but none of them accept keyboard focus, so Tab and the arrow keys do nothing on their own. This add-on moves real focus between the app's own controls and activates them, and it fixes labels that would otherwise be read as numbers or as bare words.
Download: github.com/tunmi13productions/…
ondrosik reshared this.
GitHub - ironcross32/Surge-XT-Patches: These are patches for the free and open source Surge XT synthesizer.
These are patches for the free and open source Surge XT synthesizer. - ironcross32/Surge-XT-PatchesGitHub
reshared this
So I built a bridge: a Midas MR18 and a Behringer X-TOUCH, with NVDA or short sounds telling you what the desk just did. Mic fader start, callers on CUE, cough button, talkback, ducking, silence alarm.
Free, open source, Polish UI for now.
github.com/michaldziwisz/moste…
GitHub - michaldziwisz/mostek-mr18: Mostek radiowy dla Midas MR18 i Behringer X-TOUCH, dostępny z NVDA / Radio broadcast bridge for Midas MR18 and X-TOUCH, accessible with NVDA
Mostek radiowy dla Midas MR18 i Behringer X-TOUCH, dostępny z NVDA / Radio broadcast bridge for Midas MR18 and X-TOUCH, accessible with NVDA - michaldziwisz/mostek-mr18GitHub
reshared this
ondrosik likes this.
reshared this
Release BlindFB 0.6.9 — Better video and notifications · serrebidev/BlindFB
Windows videos now use automatically installed mpv and bundled codecs, including HD AV1 playback. Native NVDA controls provide Play/Pause, Stop, seek, and volume without a mediaCtrl Tab stop. Load ...GitHub
reshared this
And Alt H is still help for me. You will not change my mind on this. I die on that hill. lol
Dennis reshared this.
Peter Vágner likes this.
Peter Vágner reshared this.
Added bunch of new tracks to my music catalog. Podcast, game, something else what needs music background? Download flac or mp3, individual files or entire library. No registration, attribution is optional. ondrosik.sk/music
I am enough of a nerd that I found this very funny.
Turning Bird Calls into (Metal) Breakdowns
ondrosik reshared this.
Johann reshared this.
On october 9, my Zoom F3 field recorder started talking. Since then it learned a lot more things to talk about, and the patcher is finally available for everyone.
#accessibility #a11y #blind #FieldRecording #ZoomF3… (1/6)
reshared this
Everything is open source (MIT, the patcher page as a whole GPL because of eSpeak NG): the voice engine, the tools, and docs on how it works and how the firmware was decompiled. For those of you who answered the poll wanting this on other recorders, there’s a step-by-step write-up of the whole method, from reading the firmware to testing in an emulator: github.com/xogium/talking-f3
The reason why I do not wish to take on this work myself for other recorders is that simply put, I don't own the hardware. I couldn’t distribute a firmware patcher for hardware I do not own, with the hope that things would not break, permanently or otherwise.
I'm putting up this work in the hope other people might find it useful. (5/6)
GitHub - xogium/talking-f3
Contribute to xogium/talking-f3 development by creating an account on GitHub.GitHub
Thanks to Jookia for mapping every menu with me back in 2022, to tkna for reviewing the Japanese, to the friends who asked for faster speech, to @FreakyFwoof for helping to improve the patcher and for being the second actual user of a modded firmware, to @lexa_desj for being the third. And of course to Claude, who did most of the code reading with me. Voices by eSpeak NG, compressed with Opus.
Not made or supported by Zoom, just a blind user who wanted its recorder to talk. (6/6)
When starting a recording with this feature enabled, which is the only tweak of the actual code from zoom that has been done, the sliding switch no longer acts as a stop button. You can no longer acidentally push it a little too far to the recording/stop position when disabling the hold function. Only the actual stop button will work.
The number of times I accidentally did this and thought I was still recording but wasn’t.
I wonder if sighted folks hate this too, come to think of it.
WE have a new Getting Started with Clipman guide, contributed by the very kind @laurent.
Thank you for such a great resource. I hope it benefits everybody wishing to get into Clipman but who may have found it a little bit daunting.
Getting Started with Clipman
onjlouis.github.io/Clipman/get…
#Clipman #A11y #Tutorial #Blind #ScreenReader
CC @bitsacb @doubletap
ondrosik likes this.
reshared this
ssi263-speech 0.8.5 is out, and it's a great one!
A new voice: GW Micro's Sounding Board, running its own DOS drivers.
The Blazie emulator now runs the Braille 'n Speak 640, with the first Polish Braille 'n Speak firmware we've found, plus the double-speed upgrade, one-handed mode and Restart the unit. Android gets its own Blazie Engineering unit screen, and Export to WAV.
The big one for everyday use: Slovak, Hungarian, Polish, Czech, Russian, Korean and the other Braille Lite languages that never signal they're done no longer pause a full second after everything they say. Lists and dialogs feel snappy now, in NVDA, TalkBack and VoiceOver alike. In NVDA, Slovak is now simply a language of the Braille Lite 2000.
The Accent's volume wobble is (mostly) gone too, and the chip's pitch glides follow a law measured on a real SSI-263.
Thank you to everyone who helped: Leo, for two pull requests bringing the Blazie emulator to more BT Speak and BT Braille units, and for the idea of one-handed mode; spacepup, for the SSInger glide reports; grnowak, for his Braille 'n Speak 640's firmware and the board photos behind it; Jake, for the click report; Astra, for reviewing every change; and every tester and listener who sent reports and lent their ears.
NVDA add-on: github.com/tgeczy/ssi263-speec…
SAPI installer: github.com/tgeczy/ssi263-speec…
Android APK: github.com/tgeczy/ssi263-speec…
TestFlight (iPhone, iPad, Mac): testflight.apple.com/join/cJYh…
Linux, the emulators and everything else: github.com/tgeczy/ssi263-speec…
Release ssi263-speech 0.8.5: the Sounding Board, the Braille 'n Speak 640, double speed, and a unit screen on Android · tgeczy/ssi263-speech
ssi263-speech 0.8.5 brings GW Micro's Sounding Board as a new voice, the Braille 'n Speak 640 in the emulator (with the first Polish Braille 'n Speak firmware we know of), the double-speed upgrade,...GitHub
ondrosik likes this.
reshared this
My Zoom F3 talks now.
As a blind field recordist, I've modded the F3's firmware so it speaks: every menu item as you scroll, which option is selected, popups like "SD Card Full!", confirmation dialogs, and even the date and time setup. No sighted help needed.
Would you want a step-by-step guide, with no technical knowledge needed, to put this on your own F3?
#Accessibility #Blind #ZoomF3 #FieldRecording
- Yes, I own an F3 and want this (0%, 0 votes)
- Yes, I’d consider getting an F3 for it (14%, 3 votes)
- Just curious, following along (76%, 16 votes)
- I’d want this for other recorders (9%, 2 votes)
reshared this
The firmware itself is the one gating those from the bluetooth side.
Even a custom app for the mac for example wouldn’t help you access the features that are missing from the app, because they're missing from what the bluetooth side exposes.
If you want access to this talking firmware stuff, I can share you the stuff, before it is officially released.
But also, for the love of all that is holy, do not attempt this unless your zoom f3 is already on firmware 2.20 from stock; it might work or might not without it, and I haven't tested.
I'm sure you know the procedure for updates too, but just in case, never power off the device too fast after starting an update. Ideally wait until any AI description stuff confirms that the update is complete, better even if you get a sighted person to help.
Would you want it?
New: Stereo input. Put it on a stereo track or a whole mix bus, and each side plays into the room from its own place. A guitar panned left is heard from the left of the room, and the room still answers in both ears.
Mono recordings stay one source in the centre, and 1.0 sessions open sounding exactly as they did.
The Mac build is coming tonight.
tgstudios.app/spaces/
#TGSpaces #AudioProduction #Reaper #Accessibility #Blind
TG Spaces — TG Studios
TG Spaces is an accessible convolution reverb for Windows and macOS with 112 rooms, built from acoustic models rather than recordings, and designed to be used entirely by keyboard with a screen reader. VST3 for Reaper, Audio Unit for Logic Pro.tgstudios.app
ondrosik likes this.
reshared this
An Android bug can kill touch exploration for every screen reader until you reboot. I traced it to a stuck injected pointer in the input dispatcher and sent Google Disability Support a detailed report. Their answer: turn TalkBack off and on, clear Accessibility Suite's cache, and reboot. Only then would they consider passing it to an engineer. Their form wouldn't even accept my logs, only screenshots.
fireborn.mataroa.blog/blog/hav…
#Accessibility #A11y #Android #TalkBack #ScreenReader #Blind #Google
reshared this
GitHub - dgl1984/spacetrace: SpaceTrace: accessible binaural HRTF spatializer for Windows VST3 and CLAP
SpaceTrace: accessible binaural HRTF spatializer for Windows VST3 and CLAP - dgl1984/spacetraceGitHub
reshared this
I see there are three failing tests but we can fix those once we make sure that my Mac-related changes do not break anything on your end. I obviously hope this is the case and it looks to be so. :)
Add macOS port: Audio Unit, VST3 and CLAP with bundled heads by vick08 · Pull Request #1 · dgl1984/spacetrace
Summary Adds a macOS build of SpaceTrace: Audio Unit, VST3 and CLAP, universal (arm64 + x86_64), macOS 12 or later. All three formats come from the same SpaceTrace JUCE target, as on Windows. Head...GitHub
Release Latest Build · masonasons/FastRoute
Automated build from commit 5956773 Version: 0.6 Commit: 5956773 Built: 2026-10-09T18:30:14-05:00 Vendor the ASIO SDK and host one ASIO device Enumeration reads the registry, so it is safe to call ...GitHub
reshared this
reshared this
First, I released ElevenLabs Music and Sound effects.
Now, ElevenLabs speech. here's my companion app just in-time for you to enjoy the V4 model for free, for the next I think, 10 days. No cost for using V4 during this time period, even via the API. If I were you I'd take full advantage.
Go download and have fun.
github.com/OnjLouis/ElevenLabs…
Release ElevenLabs Speech Generator 1.0.0 · OnjLouis/ElevenLabsSpeechGenerator
ElevenLabs Speech Generator 1.0.0 Initial release of an accessible native speech and audio application for Windows and Mac, by Andre Louis. Included workflows Text to speech and multi-voice dialog...GitHub
ondrosik likes this.
reshared this
well it would appear I need to log in and give it a permission that is missing. oopsy! got this when I hit the test key button.
Voices: ElevenLabs returned HTTP 401: The API key you used is missing the permission voices_read to execute this operation.
MAX_VIDEO_SIZE=83886080
Kaveinthran reshared this.
And yes, it went through some serious testing. For example at one point the preview voice button would, without fail, dump you into the file name box, no matter where you were. Control shift space or pressing the button manually.
Stuff like that. I quite enjoy the testing process anyway. Some people will make a thing, put it out there and never test it, just excited to have people use it and that's not at all the best method.
youtu.be/KVjaNQuKTA8
ElevenLabs Speech Generator: Accessible software for Windows and Mac
First, I released ElevenLabs Music and Sound effects.Now, ElevenLabs speech. here's my companion app just in-time for you to enjoy the V4 model for free, for...Andre Louis (YouTube)
reshared this
For those of you that might want to hear this. here is chapter 1 of a harry potter fanfic called Ah Screw it read by me genorated by eleven labs. I have to give props to @FreakyFwoof for making the program that let me do this. If you would like to try out the very awesome ElevenLabsSpeechGenerator that he wrote you can get it from here.
github.com/OnjLouis/ElevenLabs…
I did also have to use sound forge to join all the pieces together.
Release ElevenLabs Speech Generator 1.0.0 · OnjLouis/ElevenLabsSpeechGenerator
ElevenLabs Speech Generator 1.0.0 Initial release of an accessible native speech and audio application for Windows and Mac, by Andre Louis. Included workflows Text to speech and multi-voice dialog...GitHub
reshared this
@Derek here you go.
Ah, Screw It!
Author:
mjimeyg
Summary:
Harry goes to sleep after the final battle... but he wakes up at his first Welcoming Feast under the Sorting Hat. Harry has been thrown back in time into his eleven-year-old body. If he's going to have to suffer through this again, he's going to do all he can to make sure he enjoys himself.
Category: Stargate: SG-1 + Harry Potter Crossover
Status: Completed
Updated: 09-15-16
Total chapters: 37
Calling all indie musicians and those who support them! (Boosts appreciated - let's get the word out)
We'd like to hear from you.
Actually, what we'd like to hear is your music. So drop a link to one of your tracks so that we can listen.
reshared this
Andrew "Ace" Arsenault likes this.
Our music goes in a few different directions at once, but our latest album is centered largely around an acoustic sound and a classic songwriting approach.
multivariateoutliers.bandcamp.…
(P.S.: I run both the band page and the Blue Grazer Records page - I'm looking forward to listening to all of your music!)
Middle Of Nowhere, by Multivariate Outliers
from the album All The Colors On DisplayMultivariate Outliers
@Oriol Gómez Sentís @ondrosik Frankly I see the fact it's a web based app as a strong selling point so there's nothing to install and the guest can just open it in their browser.
I think there are two things to implement:
- So called Wakelock API generaly available since 2025.
- Or for older devices playing looped video with no audio track in background positioned off screen.
Do you see this as something you would consider adding?
Peter Vágner reshared this.
Zvonimir Stanecic
in reply to ondrosik • • •ondrosik
in reply to Zvonimir Stanecic • •