1
00:00:00,000 --> 00:00:02,000
Thanks for making the time.

2
00:00:02,680 --> 00:00:05,680
I want to start with the part you said was
messy.

3
00:00:06,320 --> 00:00:11,520
Right, so the messy part is that the file
never arrives the way you expect it.

4
00:00:12,400 --> 00:00:14,153
Someone records on a phone,

5
00:00:14,233 --> 00:00:18,240
someone else on a laptop, and the levels
are nothing alike.

6
00:00:19,200 --> 00:00:21,600
We stopped sending the audio anywhere.

7
00:00:22,400 --> 00:00:25,501
Once the whole thing ran on the machine
that

8
00:00:25,581 --> 00:00:28,826
recorded it, half the problems were just
gone.

9
00:00:28,906 --> 00:00:32,224
What surprised me was how much of the time
went

10
00:00:32,304 --> 00:00:35,187
into the download, not the transcription.

11
00:00:35,267 --> 00:00:40,400
The model is about 200 megabytes, and
after the first run it is cached.

12
00:00:41,280 --> 00:00:44,800
On the laptop an hour of audio takes about
three minutes.

13
00:00:44,880 --> 00:00:49,911
On the desktop with a proper graphics card
well under one

14
00:00:49,991 --> 00:00:55,022
that was the moment it stopped feeling
like a demo let me

15
00:00:55,102 --> 00:01:00,402
add a pause here and then continue with
the second paragraph

16
00:01:00,482 --> 00:01:05,513
about subtitle formats SRT works on every
video platform.

17
00:01:05,593 --> 00:01:08,712
VTT is what browsers use.

18
00:01:08,792 --> 00:01:10,393
Play text is for notes.