That would be incredibly easy to track I would think. Each request sent to the service will likely have a unique identifier that links the to the user. Hell, a web service company would probably even track it for you and send a nice itemized bill.
The $4 per million characters will get you a lot of ATC chatter. I think youāre right in that the āup-chargeā from the dev will be the same or even more, just because the Amazon charge is so small for this application.
That raises some questions:
Does that mean that time is only deducted when you talk, but not when ATC talks?
Is listening to ATIS deductable?
Does this imply that there will be no deduction for ATC comms happening between ATC and AI traffic at airports.
What about the comms that happens between ATC and injected en-route AI traffic like MSFS ATC currently does, or is BATC not even going to have that capability anyway?
If billing is per-character are they going to use reduced American or full English spelling, e.g. color vs colour to calculate cost??
Questions, questions 
1 Like
I was wondering that too honestly, but I think the expensive part is the voice recognition of the user. They sort of alluded to this in the interview with Overkill. Since the rest of the chatter is basically reading words from a database, the cost to produce it should be less than the cost of interpretation of the users voice. But yeah, it completely up in the air at this point how they will weigh that. It would stink to accidentally leave an atis running and burn up your time!
Which is why I think it will be deducted āmenu to menuā. Sometimes I get busy and AFK for periods of time and go way past my destination, so my theory will cost me for nothing.
I think āmenu to menuā is just a cleaner option for them.
Something else I thought was a little strange with the āOverkillā interview, was that the developer used a call sign, or alias. Unless I missed it, they never revealed a person or the name of the developing company.
Theyād probably use a Custom Language Model (trained on their ATC phrases etc) so typical costs would be $0.00378 per minute of speak to text conversion (about 22 cents per hour of listening, cheaper if in a bulk tier of use that a product would probably be).
TTS and STT services costs have fallen a lot. The Azure stuff is a bit more expensive with transcribe but cheaper for TTS output with neural voices. Thereās also a lot of suppliers doing this, which helps drive down prices.
I do agree with that! Does seem a bit oddā¦ Maybe he is a well known xplane developer 
It was mentioned by Overkill that the developer specifically requested that an alias be used as a condition of the interview. That was in this preview video if I remember correctly. Beyond ATC Exclusive Early Access | FULL FLIGHT DEMO! - YouTube
1 Like
Because he works at Asobo or WT 
.
Time to wildly speculateā¦
1 Like
Doesnāt want people to know he is Elvis
4 Likes
Or did. Seems a little odd heād use an alias, is he on the run? Iām sure he has to state his name for business reasonsā¦ at least thatās how it would be in UK to my knowledge. Itād be like me setting up a business, searching companies house (UK) and calling myself Johnny5Tanks12.
Can anyone remember when Alexa first started? Well, I had a dream last night.
āG-IM, Scottish Centre, good morning, pass message.ā
āG-IM, hi, type is a PA28, abeam Whitehaven, 3,000ft on 1002 hPa, heading up to Oban. Requesting basic service.ā
āOK, looking up omelette recipesā¦ā
4 Likes
The interview was with an AI bot, no human involved 
1 Like
Noticed this interaction on their Discord, which may shed some light on the aspect of when cost is involved:
And the reply from Captain:
āAny time any voice is generated from BATC it would count against your time. Not your voice to it, itās voice to you.ā
So additional chatter looks to count against your bought time. I image this is computed time on the cloud service, so on an average 1-2 hour flight weād be using, what 5 minutes of total generated voice time from BATC? Depending on the airports, of course, and how long you wish to stay on the ramp listening.
3 Likes
Hmm. Not liking what Iām reading here although of course much is speculation. I was all-out for this when first announced but now there are too many questions which wonāt be answered until the āproductā hits the market.
And that is of course also speculation 
. We donāt know whether they will drip drip more info on the go.
1 Like
Correct, a poor selection of choices.
Poor choices, because there cannot be an even somewhat good choice for offline ATC as long as default MSFS ATC is poor. There are way too many problems and missing features in default ATC for any offline ATC add-ons to reasonably suffice. All of the ATC add-ons are reliant on default ATC; this is the way it is. Add-on developers can try all they want, but there will always be problems for their customers, because where they succeed in one area they will fail in several other areas. Always a give and takeā¦hence the āpoor choicesā realization/reality.
If reasonable realism is not that important, and all the poor choices suffice, then this same cycle of inadequacy will continue. But this current cycle of sufficiency will keep the poor choices flowing at $40-$60 each, to even possibly subscriptions with promises that cannot be kept.
All the while MSFS default ATC will remain a total mess. And because the community is relying on add-on developers to be the cure, rather than calling for MSFS default ATC to be fixed first, this will never end.
Those interested in a more reasonable offline ATC experience will have to keep on waiting.
For there to be great add-on choices for offline ATC, MSFS default has to be fixed to a reasonable standard as outlined in the MSFS IFR feature presentation:
Feature Discovery Series Episode 8: IFR - YouTube
ā¦more promises from about 3 years ago that are still not kept.
I would even say that add-ons may not even be necessary, if default MSFS met the standards in the IFR feature presentation.
I wonder if there are going to be feature presentations for MSFS 2024.
1 Like
I donāt think that even if Asobo would decide to go all in and improve/rewrite their ATC, they could do a better job than what 3rd party tools like Pilot2Atc could do. No simulator in the history of flight simulaters ever had an offline ATC that came close to what external tools provide.
1 Like
Radar Contact was a stand-alone, offline ATC program which was a vast improvement on FS2004 and FSX default ATC, with features and functions which were not present in the default ATC.
1 Like