The offline voices sound ok, not great. Probably good enough, and certainly better than what I'm using now (FSHud). 

Quote

Offline Voice Model - 150+ locally generated voices using our custom AI engine. Uses ZERO Transmission Time.

Locally generated - does that mean already within Windows? Similar to MSFS Azure voices?

3 minutes ago, Bobsk8 said:

PF3 has been working correctly for years

Missed a PF3 statement since several pages from you. Thank you very much!

 

Would you appreaciate it if P3D or X-Plane users will post on every MSFS thread that X-Plane or P3D do this ot that better??
(ok, purely hypothetic as X-Plane or P3D couldn't be better, but ...)

12 minutes ago, guenseli said:

Missed a PF3 statement since several pages from you. Thank you very much!

 

Would you appreaciate it if P3D or X-Plane users will post on every MSFS thread that X-Plane or P3D do this ot that better??
(ok, purely hypothetic as X-Plane or P3D couldn't be better, but ...)

Solme posted about the up and down ATC instructions , and I told him that PF3 ATC  doesn't do that. If you don't like my comment, then don't read it. 

2 hours ago, Cpt_Piett said:

Locally generated - does that mean already within Windows? Similar to MSFS Azure voices?

(I guess) Yes and no. Azure is online, so it can't be that. Rather the default windows voices ("Microsoft David" and the likes) that are installed with windows, locally.

30 minutes ago, Fiorentoni said:

(I guess) Yes and no. Azure is online, so it can't be that. Rather the default windows voices ("Microsoft David" and the likes) that are installed with windows, locally.

BATC doesn't use SAPI voices.

3 hours ago, Bobsk8 said:

PF3 has been working correctly for years. No roller coaster approaches. 

And last I checked required yet another bridge program if you want to interact with using voice rather than assigning hotkeys for most-used functions.

That's why I picked P2ATC over PF3 back when I bought my last ATC addon. P2ATC does voice input natively. It's not very good by modern standards but, then, it came out before Google Home and Amazon Alexa showed us better voice interfaces. It doesn't understand even mild deviations from its preprogrammed phraseology. Responding "we're looking" to a traffic advisory results in confusion, for instance.

I fly in VR, and I use a Fulcrum yoke. Fooling with the keyboard is sub-optimal, looking at ATC and click-responding on another monitor like I used to do doesn't work anymore, and I don't have 35 buttons on my controller to assign to ATC responses. It's time for a new paradigm in ATC, and I'm hopeful BATC will bring it. If it can, it could sound like a Speak&Spell for all I care. 😉

 

4 hours ago, eslader said:

And last I checked required yet another bridge program if you want to interact with using voice rather than assigning hotkeys for most-used functions.

That's why I picked P2ATC over PF3 back when I bought my last ATC addon. P2ATC does voice input natively. It's not very good by modern standards but, then, it came out before Google Home and Amazon Alexa showed us better voice interfaces. It doesn't understand even mild deviations from its preprogrammed phraseology. Responding "we're looking" to a traffic advisory results in confusion, for instance.

I fly in VR, and I use a Fulcrum yoke. Fooling with the keyboard is sub-optimal, looking at ATC and click-responding on another monitor like I used to do doesn't work anymore, and I don't have 35 buttons on my controller to assign to ATC responses. It's time for a new paradigm in ATC, and I'm hopeful BATC will bring it. If it can, it could sound like a Speak&Spell for all I care. 😉

 

Yeah, I never liked the voice thing, hate trying to memorize the exact script to have it do what I wanted . Just easy for me to set up a couple of buttons on my joystick to do 90% of the functions I need. 

That's what I did with P2ATC when I flew with a regular HOTAS joystick that had a bunch of buttons. The Fulcrum has 4 buttons, 2 trim switches and a hat. Not as much of an option because they're busy doing other things like autopilot disengage and VR stuff (engage and center)

 

6 hours ago, Fiorentoni said:

Rather the default windows voices ("Microsoft David" and the likes) that are installed with windows, locally.

Just wanted to point out for everyone that the Windows default voices (e.g. Microsoft David, Hazel, Zira, George...) are absolutely *not* the 'locally generated' voices that will be used.

Here's a chat from the BeyondATC Discord:

Quote

KevDog: Hey Cap so the offline voices which don't use transmission time have 150 different voice?

Captain: Yeah, it's a homegrown AI model, so we'll add more all the time. It's the accents it has trouble with. It actually can handle about 40 accents but the problem is when you play them.....they sound extremely racist. Like a stereotype of a person from that area of the world not a real person from that area. So I've decided to exclude them until I get it better.

KevDog: Thanks It's  great to know I won't just be hearing the exact same voices in offline. Looking forward to it!

Captain: No it will be quite varied indeed, to the order of hundreds of voices. Just with limited accents.

 

And here's a reminder of what the different tier voice options will sound like:

 

 

7 hours ago, eslader said:

it could sound like a Speak&Spell for all I care.

Completely off topic, but if you listen really closely to one and press the 'z' key it actually pronounces it *static* followed by "Lee". Guess they couldn't synthesize the zee. Of course being Canadian, it should have been "zed" anyway. 

Carry on. 

20 hours ago, virtuali said:

The problem with AI

AI, as of today, is both incredibly interesting and powerful, but it's also too expensive, especially in the context of a flight simulator, if you want to do some kind of add-on that needs to be perceived to be fairly priced.

…..

So, in the end, we decided it was a nice learning experiment, which we'll surely want to revisit when the time is right, perhaps AI costs will drop in the future due to more powerful hardware, or we might see the appearance of consumer-level GPUs with dedicated AI chips and memory, so you can run your game and the AI together, locally, with no performance loss.

…..

If BeyondATC fails because customer’s aren’t willing to pay the costs of the AI, they will hardly be the first tech company to fail because they were too far ahead of their time.

The great thing about the technology market is that, more so than in other industries, prices generally tend to fall over time, sometimes dramatically so.  The problem for startups or businesses that have innovative new ideas ahead of everyone else is generating enough revenue to sustain themselves until the market is ready to accept them.  History is replete with companies that couldn’t cross the “valley of death” to reach prosperity.  Perhaps BeyondATC won’t survive, but eventually someone will come along and make AI work for this application.  I’m cheering for BeyondATC, and I will be buying their product.  I’m not at all concerned about the cost, as it will make a fantastic addition to my (rather expensive) hardware A320 cockpit.  Hopefully there will enough serious users like me to keep them going.  The costs of AI will drop, it’s just a question of how long it will take.

So let me get this straight - if I stick to offline voices for everything, then offline voices will be used for AI traffic and for my own voice communication (Push To Talk) with ATC, and I won't be charged anything extra and I won't even use up any transmission time at all?

I just want to buy the product for $60 USD and be able to talk to ATC without paying anything additional after my $60 purchase.  So I don't want to even use up the transmission time that's given to me for free.

The transmission time has me confused a little, because I'm not sure if they are charging for transmission time if I keep AI traffic on.

I am also reading that voice recognition (ie. if I use Push To Talk) with ATC will cost transmission time? So even if I use offline voices for everything, the moment I use Push To Talk, to talk to ATC, then my transmission time will be deducted from? (this is bad if Push To Talk costs transmission time because while I don't care if the voice of the ATC controller isn't that great, the fact that every time click on Push To Talk to communicate to ATC, my transmission time will be eventually depleted).

Edited February 22, 20242 yr by abrams_tank

As far as I understand, PTT needs a very small amount of transistion time. The included amount should last for years. 

But yes, otherwise there is no extra cost after purchase if you use just the 140 offline voices. 

 

 

2 minutes ago, guenseli said:

As far as I understand, PTT needs a very small amount of transistion time. The included amount should last for years. 

But yes, otherwise there is no extra cost after purchase if you use just the 140 offline voices. 

 

 

Oh, ok. If the PTT uses up very little transmission time and it can last for years, I think that is fair. I just didn't see this with their pricing scheme, it's a very confusing for me.

5 hours ago, kholt said:

If BeyondATC fails because customer’s aren’t willing to pay the costs of the AI, they will hardly be the first tech company to fail because they were too far ahead of their time.

I don't think it will fail, because as far as I understood from the reviews, AI is only used for Premium voices, but the actual ATC procedures are coded normally and works the same with the Offline voices so, it's just a matter if users will find $60 a fair price for the improved ATC functions.

Had the product relied on AI entirely (like our abandoned AI Copilot product or the upcoming AviMate), then yes, those products would either succeed or fail depending on users' willingness to pay the subscriptions costs, because with no AI, the product would be useless.

 

Edited February 22, 20242 yr by virtuali