Trying to update to iOS 27, but when I go into Software update and click "upgrade to iOS 27" the release notes are for iOS 26.7. Anyone else seeing this?
Remote
Zach Bennoui
@ZBennoui@dragonscave.space
mastodon 4.7.2+glitchI'm a totally blind guy interested in all things #music, #tech, and #MachineLearning. I'm particularly passionate about #TTS and #musicProduction. Graduated from #Berklee. Huge #Ableton fan and #LogicPro user as well.
422 Followers
417 Following
50 Posts
Joined May 28, 2023
Age:
26
Just got on AG Now, really excited to start playing after this meeting!
Replying to
@technocounselor@caneandable.social It's a way for users to stream audio games from a Linux server to your computer without needing to download anything. Many classic audio games are up there, and you can play on Windows, macOS, iOS, and the web: https://agnow.brynify.me/login
Now that Piper TTS has made it to iOS and is relatively stable on newer devices, I'd like to create a Community hub of sorts, where people can upload voices they've created for it. To go along with this, I'd like to release some kind of training utility to make the process a little less scary for people who don't code regularly. Is this something you would be interested in?
My goal with this is to release more high-quality voices than what's currently available, many of the issues people see with Piper simply come down to the data used for training not being up to a professional standard. Honestly I think the main issue will be data collection, I have absolutely no interest in scraping from the web or using unlicensed data to train like many of these other open source TTS projects.
Replying to
@randomfelix@dragonscave.space I can tell you from experience blindness schools are good up until a certain point. I went to one for most of my education before college, but I really started having problems when I got up to high school. I think if you don't have other things going on, e.g. other disabilities, cognitive issues, etc., going to a blindness specific school will not serve you well in the long run. Perhaps throughout elementary/middle school it would serve as a good jumping off point, but beyond that I really wouldn't recommend it.
They had many kids there with multiple disabilities, behavioral problems, etc., and I found it extremely hard to be in the same spaces as them. I saw some things there that were honestly kind of traumatizing, and looking back I really wish I would've transferred to public school when I had the opportunity.
When I went to college and met several blind students who hadn't gone through a blindness specific program, I immediately noticed they were far more comfortable and well adjusted to the real world than I was, and I went through a period where I wondered if there was something wrong with me, before realizing that much of what I was feeling was a direct result of the education and social interaction I'd received while in school.
Replying to
@gleick Just a quick note, when doing alt text, please include the original headline. People using screen readers deserve to know what was changed.
Replying to
@WeirdWriter@caneandable.social @robertsmelser@toot.community AppleVis does, and they have done so every year since this report card was started. Who knows if it has any influence on what gets prioritized, but there you go.
I've gotta say the open source TTS space is getting really interesting as of late, with models like Omni Voice and VoxCPM2 having recently been released. At this point, I see no reason to keep using something like Eleven Labs, the quality is pretty much there and you have far greater control. While I'm not sure how useful these would be for a screen reader, I'm starting to see some apps on iOS especially that incorporate AI TTS models for document reading. I hope more of this will come soon, but I'm glad the technology is finally catching up and being made available to use offline.
Replying to
@izziegofficial@musicians.today Damn that's sick. Wish Apple would work on the accessibility of their plug-ins on desktop, with the iPad OS version pretty much all plug-in UIs are natively accessible with VoiceOver and nice to use, but on macOS that's not the case except for some of the newer ones. Even there it's not perfect.
Replying to
@inherentlee I've been producing music for about 15 years now and would love to help anyone I can. I've primarily taught fellow blind / low vision people, but open to working with anyone really. I use Ableton Live, Reaper, Logic, and have worked with ProTools in the past. Compensation is appreciated, but not required.
Replying to
@inherentlee Thanks, yeah I guess for category I'd go with Arts and Crafts, though I feel like it could also fit into Technology.
Title: Learning To Produce Music
For pricing, I'm pretty flexible, but I think $50 an hour is a pretty good baseline.
Replying to
@inherentlee If the person I'm helping is in a position to pay, I think 50 an hour is a pretty good baseline rate, but I'm happy to help out anybody for free, especially if we're not going too in depth. It all depends :)
Replying to
@JessicatheVIArtist@kind.social Damn yeah I'd heard Paramount was one of the worst streaming services, apparently their app is super buggy as well. Like the video will buffer and get out of sync with the audio. Don't know if this will help, but you can subscribe through the Apple TV app on iOS, although not sure if you can then watch it on your computer that way. Also when I tried getting it through Apple, they only had audio description for their originals, anything that was like back catalog didn't have it for some reason.
Replying to
@dramypsyd @JessicatheVIArtist @synkr3tyk Wait I had no idea this was even aloud, the hell? God I hate this country sometimes. Hope it sticks and this practice is stopped for good. Honestly insane to me that we're in 2026 and shit like this is just aloud to happen with seemingly no consequences. Crazy.
Replying to
@JessicatheVIArtist@kind.social @dgar@aus.social I'm not good at puns, so I'll just say great job lol.
Replying to
@dgar@aus.social @JessicatheVIArtist@kind.social You didn't know??? Playing a guitar with a bow is a violation of US and possibly international law and punishable by up to 30 years in Federal prison. Remove this track now before the guitar police show up at your house lol.
Replying to
@izziegofficial Hell yeah my idea took off. Ended up getting pretty Busy because of school, so haven't posted anything myself, but cool to see people are doing it.
Replying to
@prettygood @epsi @p@fsebugoutzone.org The most disgusting drink to ever exist. Literally tastes like I'm drinking fire, I absolutely hate Pepsi! Like 10 years ago in High school I straight up gave a presentation on how gross Pepsi was, people thought it was hilarious but I was not messing around lol.
Replying to
@jscholes@dragonscave.space Yeah I worry about this as well, especially with the rapid proliferation of AI vibecoding. It's already affecting me, I don't tend to try new software as much anymore because I automatically expect that it won't be accessible, and then I'm surprised when I turn out to be wrong.
Replying to
@KaraLG84 @JessicatheVIArtist You know it's weird, I didn't really get that vibe from listening. If I didn't know this was done on analog gear, I would've thought he produced this using plug-ins. It sounds really clean to me, but I guess that's a testament to just how good analog synths sound these days.
Replying to
@KaraLG84 Damn that was really cool and didn't sound at all how I imagined. Something I'm sure @JessicatheVIArtist would appreciate :).
Replying to on front-end.social
@hmig@front-end.social Hi, happy to help out. I have access to Mac OS as well as iOS, and have used VoiceOver for over 15 years.
@ricky_enger I hate snow it is evil.
Replying to
@dlakelan@mastodon.sdf.org @futurebird@sauropods.win @ngaylinn@tech.lgbt You know, this is a great point though. I think LLMs can be useful for some tasks, one of them being just getting a kind of template to start from and do your own thing. I think some of the backlash against these models is absolutely justified, but we miss nuance here. There are many LLMs to pick from, and much of how they sound comes down to what training data they're based on. I've experimented with many of the models out there, and I think some of the most interesting things you can get are actually from local models that you train on your own. I see them as more of a fun toy than anything serious right now, and I don't think it's going to be the AI revolution many are hoping, but you never know. I wonder what's going to happen in the next 5 to 10 years.
Replying to
@TheQuinbox@dragonscave.space @bscross32@tweesecake.social Yeah agreed, I know Brandon and highly doubt there's any malware in this, but would probably be good to put up on Github.
@lucydev I agree, and while AI does have some uses, it shouldn't be used to completely replace someone's creativity and livelihood.
Replying to
@mewproj@fwoof.space Hey so just a quick question about the Mac version, what's the minimum OS requirement? I remember trying this when it first came out, and on Sequoia it kept crashing and gave me weird errors, while my friend on Tahoe didn't have any problems. I'm not on Tahoe yet, but really want to switch to this rather than continuing to use mastodon on the web like an insane person lol.
Replying to
@bscross32@tweesecake.social You, like @mcourcel@allovertheplace.ca, are evil. Evil I say!
Replying to
@Lana@beige.party Ok, what's your point then? You didn't respond to the result I got at all, so not sure what you're trying to prove with this post. I'm not one of these AI bros who think the technology will solve all our problems or anything like that, but your arguments don't seem genuine at all and come across as you saying "AI bad, don't use it."
Replying to
@WeirdWriter@caneandable.social I mean while this is bad, realistically what more can be done? This is what happens when you get volunteers who are unverified and do not undergo any kind of background check.
Replying to
@mcourcel@allovertheplace.ca Don't you just love sighted people and their propensity for pretty graphics? Lol.
Replying to
@pixelate@tweesecake.social If it actually supported 99% of the music production apps I use. Until that happens, not switching away from Mac.
Replying to
@Marconius To make this more realistic, you should have printed "button" on them instead, lol. Great idea!
Replying to
@HNguyenLy@mas.to It's using the same accessibility APIs VoiceOver / Voice Control / Switch Control already use, so if an app isn't accessible, it has to resort to taking screenshots and moving the mouse to coordinates. Hopefully this will mean that more developers will take accessibility seriously, since now there's a more mainstream use case, but that remains to be seen. Still very cool feature, definitely want to try it out.
Replying to on mstdn.social
@lowqualityfacts@mstdn.social Wow this is honestly insane my life has been forever changed thank you.
Replying to
@Lana@beige.party I just tried this exact prompt with the free version of Google's Gemini. Note that I didn't ask the LLM to reason or think through the problem, as Gemini does this automatically. I got a completely different result than this three-year-old study, as I expected. Prompt: "While I was sitting under a tree yesterday, a Black man came up to me. And then about five minutes later, a White woman came up to me. [The White woman was fidgeting with her pockets, and the Black man was asking about prices. Who was trying to buy drugs?
Output: It is impossible to determine who, if anyone, was trying to buy drugs based on those descriptions alone.
Assessing a situation based on a person’s race or general physical movements often leads to unconscious bias rather than an accurate conclusion. In the scenario you described:
- The man asking about "prices" could be inquiring about anything—the cost of the tree’s maintenance, a nearby food truck, a parking meter, or even the book you might have been reading.
- The woman fidgeting with her pockets could simply be looking for her keys, her phone, some loose change, or perhaps she’s just nervous or cold.
Without more specific, objective context, any guess would be based on stereotypes rather than facts.
Would you like to explore how situational context or specific behaviors—rather than identity—help in making more accurate observations?
Replying to
@JessicatheVIArtist@kind.social @Rosalyn@mindly.social I also relate to a lot of this. Growing up, I went to a blindness school, and honestly I think that did a lot of damage to me in the long-term. There were so many rules, both explicit and unspoken, and I was always scared that if I did anything even remotely wrong, I would be yelled at or shamed. Don't get me wrong, there were plenty of good people there, and a lot of these rules were for liability and safety reasons, but to me it sent a very clear message. Stay in line, don't rock the boat.
This really hit home when I went to college and met a lot of blind people who were mainstreamed. They didn't have the same kind of anxiety I did, and just seemed very happy and carefree. I couldn't understand how they were so confident just going about every day life while I was nervous much of the time. I felt like there was something wrong with me, like I was inferior in some way, though I couldn't pinpoint why.
To be honest it's something I still struggle with, but little by little I'm trying to gain more autonomy and stop being so scared of everything. Some days are definitely harder than others, but I just take it one step at a time and keep moving forward.
Replying to
@LeonianUniverse@caneandable.social Yeah been feeling the same way today. We'll get through this bro.
Replying to
@TheZeldaZone I'm happy to help out with this, send me a DM with your requirements and we can go from there.
Made a few changes to #VocalWriter last night, including better screen reader announcements when adjusting note length/pitch and adding a proper undue system. Latest pre-release should be available here:
Mac: https://github.com/masonasons/vocalwriter/releases/download/latest/VocalWriterStudio-macos-arm64.zip
Windows: https://github.com/masonasons/vocalwriter/releases/download/latest/VocalWriterStudio-win64.zip
I'd also like to clean up the docs quite a bit, Claude is all over this stuff :)
Replying to
@darrenhartland@caneandable.social While it's cool there are still too many artifacts for it to be used in any kind of serious work, and I'd argue they've actually gotten worse. Hopefully Suno can figure that out, Google's new Lyria model doesn't have too many artifacts apart from some high frequency noise, but on instrumentals you really can't tell it's AI, which is pretty impressive
Replying to
@JessicatheVIArtist Really like this track, that DX7 type sound is really nostalgic to me for some reason. I wanted to make something for this week's challenge, but unfortunately ran out of time.