KnowBrainer Speech Recognition
Decrease font size
Increase font size
Topic Title: Is there any reason to use a different profile for transcribing recordings of my own voice?
Topic Summary:
Created On: 05/08/2021 01:05 PM
Status: Post and Reply
Linear : Threading : Single : Branch
 05/08/2021 01:05 PM
User is offline View Users Profile Print this message

Author Icon
wristofdoom
Top-Tier Member

Posts: 326
Joined: 09/03/2020

I have been making voice recordings on my phone. The voice recordings have lower quality audio then the dictation that I do with my good microphone at my desk. However, the transcriptions are acceptable quality.

Is there any reason why I might have better results with using a separate user profile for transcription? Will my primary user profile be somehow degraded if I use it for transcribing my low-quality recordings?

I would rather use my primary profile for both my at-desk dictation, as well as for transcribing my low-quality phone voice recording files because of shared vocabulary etc.

But if I expect better results from using another profile, then it might be worth it for me to maintain a separate profile specifically for transcription.

Also, here is a ride along question: I was reading in older posts about the varying quality of different sound file formats. MP3, for example, was not recommended for voice transcription. My current voice recorder creates sound files in M4a format. It's running on my android mobile phone. Are there any android mobile apps that are recommended for recording voice for transcription?

Getting another dictation gadget at this point is not an option. I'm specifically looking for android mobile apps.

Thanks



-------------------------

Dragon Professional Individual v15.6. Windows 10. Knowbrainer 2017.

 05/08/2021 02:30 PM
User is offline View Users Profile Print this message

Author Icon
R. Wilke
Top-Tier Member

Posts: 7881
Joined: 03/04/2007

You are obviously getting lost in information. An enormous load of information, all of which isn't correct.

Start following your intuition, and just start doing it. You can't go wrong for the most part.

If you really want to find out what is right or wrong, as regards the performance of a user profile, particulary with regards to how it develops over time depending on the route you follow, I have all the tooling enabling you to get real data points, as opposed to all the wild guessing and the BS posted about it in this place. But you aren't there for them yet, to be honest.


-------------------------



No need to buy if all you want to do is try ...

DragonCapture KB Download (Latest)
DragonCapture Homepage

 05/08/2021 03:12 PM
User is offline View Users Profile Print this message

Author Icon
Lunis Orcutt
Top-Tier Member

Posts: 39971
Joined: 10/01/2006

When creating a new profile, Dragon offers you the option of putting a check mark in This profile is to transcribe recordings of someone else speaking, which is supposed to be better. In our opinion, you may or may not experience better results. It didn't make any difference when we tested but it it is probably better for you to test yourself. It only takes a few minutes to make a new user profile.

 

As far as transcribing your Android phone recordings are concerned, we doubt it will work but please post back with your results.



-------------------------

Change "No" to "Know" w/KnowBrainer 2020
Trial Downloads
Dragon/Sales@KnowBrainer.com 
(615) 884-4558 ex 1

 05/08/2021 03:23 PM
User is offline View Users Profile Print this message

Author Icon
wristofdoom
Top-Tier Member

Posts: 326
Joined: 09/03/2020

I did create an alternate profile and tried it out, and I couldn't tell any particular difference. I'm just trying to understand what is considered best practice among the experts.

So I take it that you gentlemen do not personally have a separate profile for transcription?

Transcribing my Android phone recordings has been going fine--better than Lunis would expect, apparently. The accuracy is not terrible. In my most recent transcription, there were less than 10 mistakes in a 714 word transcription. I'm using the "Voice Recorder" app on a Samsung Galaxy S21 ultra with Sennheiser M2 Bluetooth headphones.




-------------------------

Dragon Professional Individual v15.6. Windows 10. Knowbrainer 2017.

 05/09/2021 01:47 AM
User is offline View Users Profile Print this message

Author Icon
kkkwj
Top-Tier Member

Posts: 947
Joined: 11/05/2015

I transcribe from MP3 files all the time and have no problems with the technology. Any problems I have are because I'm half asleep, speaking in broken phrases, or slurring my words and thoughts (or all four simultaneously :-)). If I take the time and care to speak clearly, the transcriptions are just fine. The biggest difference I notice is that when I'm dictating into a recorder, I don't have access to the full complement of corrective phrases and actions that Dragon and Word can apply in real-time as I dictate into Word. I've dictated into my little digital recorder using a $10 mic on the corner of a 4-lane highway with tractor-trailers thundering by without any consequential reduction in accuracy. IMHO, dictating in a recorder is just never going to be as "accurate" because of the lack of real-time corrections and vocabulary adjustments that you get with Dragon/Word in real time.

-------------------------

Win10/11/x64, AMD Ryzen 7 3700X/3950X, 64/128GB RAM, Dragon 15.3, SP 7 Standard, SpeechStart, Office 365, KB 2017, Dragon Capture, Samson Meteor USB Desk Mic, Amazon YUWAKAYI headset, Klim and JUKSTG earbuds with microphones, 3 BenQ 2560x1440 monitors, Microsoft Sculpt Keyboard and fat mouse

 05/09/2021 12:08 PM
User is offline View Users Profile Print this message

Author Icon
wristofdoom
Top-Tier Member

Posts: 326
Joined: 09/03/2020

>> dictating in a recorder is just never going to be as "accurate" because of the lack of real-time corrections and vocabulary adjustments that you get with Dragon/Word in real time.


Seems 100% true, although this is also the reason why I am trying to incorporate recordings into my workflow. I spend so much time making edits to my dictation as I go that I end up not writing enough. By edits, I mean refinements to what I have written, not necessarily fixing mistakes. I'm trying to get into the habit of just letting it all out and polishing later, for the sake of getting more words out there. Dictate drunk, edit sober.

-------------------------

Dragon Professional Individual v15.6. Windows 10. Knowbrainer 2017.

 05/10/2021 02:39 AM
User is offline View Users Profile Print this message

Author Icon
Mav
Top-Tier Member

Posts: 523
Joined: 10/02/2008

You don't have any need for a separate user profile, because it's the exact same person speaking with all of your individual pronunciation and custom words.

The only thing that differs is the format and quality of the audio input, that's why you have different audio profiles for different microphones.

For transcription of recordings, Dragon sets up a separate dictation source to account for different quality of your recordings compared to your live dictation.

 

hth,

mav

 05/10/2021 10:31 AM
User is offline View Users Profile Print this message

Author Icon
wristofdoom
Top-Tier Member

Posts: 326
Joined: 09/03/2020

Thank you Mav. This was the answer I was looking for.



-------------------------

Dragon Professional Individual v15.6. Windows 10. Knowbrainer 2017.

Statistics
32325 users are registered to the KnowBrainer Speech Recognition forum.
There are currently 1 users logged in.
The most users ever online was 12124 on 09/09/2020 at 04:59 AM.
There are currently 373 guests browsing this forum, which makes a total of 374 users using this forum.

FuseTalk Standard Edition v4.0 - © 1999-2022 FuseTalk™ Inc. All rights reserved.