I did this last with YouTube.First convert an audio file to an MP4 with a photo as footage. Then upload and wait until Youtube has analyzed the file. Then you can download and use the text of the subtitles.
Experience: Youtube already filters out a lot of UHS and Ahm’s and when it’s clearly spoken it can come along nicely, but I really had to go through the conversation to take out strange mistakes.
On one PC I had the text and on the other Youtube.Using foot pedals and some AutoHotKey codes, you can then control Youtube. This way you do not have to switch between applications and you can continue to work in your text editor (you can also use one PC if you are using AutoHotKey ControlSend). YouTube also lets you play back the file slower so you hear exactly what was said and you can rewind with the arrows for a few seconds which is convenient.
EDIT: I put the videos I used on “private” so the conversations are not for everyone to listen to.
I have a company that makes a lot of video as learning material.Subtitles are required, so subtitles must be offered with each video. Sometimes in several languages and then I have to send the Dutch subtitles in order to have them translated. Creating a transcription as a technology is called speech-to-text. Useful for when you’re looking for.
To take an advance on the conclusion.If you want to make spoken text really good to text, this is people’s work.Point. Even English spoken text in a studio with good recording equipment cannot yet be automated by artificial intelligence (deep learning, machine learning) to be converted to correct text. Today on 12 May 2019 in any case.
To immediately break a lance for companies that make paid transcriptions like for example Https://www.uitgetypt.nl/ the costs are not high.The result is often neatly broken down by speakers and in my view and experience, manually adjusting automated tools in terms of hours is more expensive than having to do well at once. But well, if someone with a lot of time can lie this relationship differently.
Here are some links of comparisons and services.The way to get free text from audio is always another service to try. Most services have a freemium model that can be used for ten minutes free and as Robert Ilbrink writes You can also extract text by first Uploading to YouTube and then with https://downsub.com/ you can get them for free and if Subtitle file (SRT).
For example https://www.happyscribe.co/ you get one hour transcription for free and that also works with Dutch.
Then there are also cloud service like Google that have a freemium model, but you have to do some effort for that:
There you get one hour per year free, and you pay $0.0004 per second.An hour transcribing for not yet a cup of coffee at cost.
But again, in my experience, both the free and paid tools cannot replace people’s work even though the claims.In Some situations you have luck and it works fine, or maybe it doesn’t have to be perfect.
Google “Free speech to text English” to find more services.
The 1st 30 minutes are free.If necessary, split your file into multiple shorter files. Audio to Text-Online transcription software-Happy Scribe