![]() |
KnowBrainer Speech Recognition | ![]() |
Topic Title: why dragon is great Topic Summary: auto transcribe folder Created On: 04/22/2022 01:10 PM Status: Post and Reply |
|
![]() |
![]() |
- matthewls | - 04/22/2022 01:10 PM |
![]() |
![]() |
- dilligence | - 04/24/2022 03:27 PM |
![]() |
![]() |
- Lunis Orcutt | - 04/24/2022 07:45 PM |
![]() |
|
I guess that for most people simple speech-to-text functionality is all they need--or think they want. But after decades of DNS use, building macro libraries etc., I just today discovered another great feature that I wouldn't ever want to lose: the auto transcribe folder. I had to write a powerpoint lecture for net viewing, complete with audio and video. The ppt features were insufficient, so I recorded my lecture in snippets using some Linux utility I found and added them to the ppt slide show. I don't remember what I said, and listening to the show would be painful. Voila, the ATF. All lecture snippets in text. Yippee! |
|
|
|
![]() |
|
For those who want to create a voice command to launch the agent in DPI 15:
AppBringUp "C:\Program Files (x86)\Nuance\NaturallySpeaking15\Program\tagent.exe"
Auto Transcribe Folder Agent indeed can be useful. Transcribing more or less "naturally spoken" audio may result in a better recognition rate compared to direct dictation in a window, where one might be tempted to use shorter sentences (I do :-().
But, I hope the next DPI version supports higher bitrates. Currently only WAV 16-bit 44100 will work (with a maximum around 1536kbps). The same appears to be true with MP3, no 48 kHz support. At least on my system...
I do a lot of screen recordings with audio and transcribing sections would be useful, but my screen recorder (Bandicam) won't allow me to record lower than 24-bit Waves. In order to transcribe I first need to convert to Signed 16-bit with Audacity which is cumbersome... ------------------------- Turbocharge your Dragon productivity with 40+ Power Addons |
|
|
|
![]() |
|
The average man’s speaking voice typically has a fundamental frequency range between 85 Hz and 155 Hz. A woman’s speech range is about 165 Hz to 255 Hz. Note that overtones can range as high as 6000 or 7000 Hz but 48 bit sampling won't make any difference because Dragon filters out higher frequencies. DragonSystems (original manufacturer) deliberately set Dragon's frequency response lower so that substandard microphones could reproduce acceptology accuracy. Nuance doubled Dragons frequency response in Ver. 11 because it was dropping some of the higher overtones. Simply put, increasing Dragons frequency response will not help and is unlikely to change. ------------------------- Change "No" to "Know" w/KnowBrainer 2022 |
|
|
FuseTalk Standard Edition v4.0 - © 1999-2023 FuseTalk™ Inc. All rights reserved.