Speech Recognition – How To Create Content FAST

Speech Recognition

There’s so much I want to write about, I decided I needed to reduce the time I spend on content creation. Use of speech recognition is one part of achieving that. This post was going to be about a very different topic – but it morphed along the way! I’ll write a second post today to make up for it.


So – this is the next update on my business blog which is recording what I do as I go about developing the business which you see on this site.

There’s certainly no shortage of material to write about. In fact each day I think of about three different topics I’d like to cover, but right now I don’t have the time to write content at such a fast rate.

But having said that, I’ve just changed my mind about the subject matter for this post. What I’m now going to write about is how to create content fast. More specifically, a technique for creating content much faster than normal. It’s what I’m using right now…

Speech Recognition

Talking to a number of colleagues recently, I was surprised to learn how few of them realised that speech recognition technology is now many times more efficient than it was just a few years ago.

What I’m doing right now is speaking into my smartphone and creating this post in an Evernote note using the built-in speech recognition facilities of Android. I just click on the microphone button on the keypad and the input mechanism switches from keypad to microphone.

The quality of the speech recognition is quite amazing. It works at its best if you speak at normal speed because as you complete a sentence the system looks at the whole structure and goes back and changes words which it might have misunderstood initially.

The only annoying factor is that the speech recognition doesn’t always recognise punctuation commands.

For example – if I say full stop new line it very often just insert those words – exactly as it has right now (!) instead of actually putting in a. And starting a

Ha! In that instance it actually put in the punctuation!

Because you can’t be sure whether it’s going to put in the punctuation or not, it ends up being quicker to always switch into keyboard mode to do it.

The fastest way to change words or phrases

For some types of change the best way is to switch to keyboard mode. But often it’s fastest to stay in speech input mode and do the following :

  • Position the cursor just after the incorrect word or phrase.
  • Click the back icon – which is to the right of the microphone (on an my phone at least). In speech input mode this deletes a complete word rather than just a character – which is exactly what you want.
  • You can then, of course, just re-speak that word or phrase – and the change is done.

Practice to improve your speech

To begin with you may find that using speech recognition is hardly any faster than simply typing in the text. But keep experimenting for a while and you’re likely to find you get a very significant speed improvement.

You will need to get good at articulating the words clearly, but don’t fall into the trap of thinking you need to speak slowly and pause between each word. That will actually give you worse results.

Also don’t be put off if you have a strong regional or foreign accent. You may be surprised at how well this technology handles strong accents.

Practice to improve verbal and semantic structures

( I’m almost certainly using the wrong terminology here but I’m going to assume that not many linguistic students (at most one, probably) are going to be reading this.)

There are two points I want to make here .

The first is that the way you think when you’re speaking is somewhat different to when you’re writing. And your sentence structures end up being somewhat different.

You may need to get used to seeing your spoken sentence structures preserved in text as opposed to disappearing into the ether as soon as you finish speaking.

But with a bit of care and practice you can create more than half decent sentence structures that you’re happy to preserve in print.

The second structure point I want to make is that I personally end up with a slightly different attitude to how I structure the complete article when using speech input.

If I’m speaking the article I’m more likely to make a start without a clear idea in my head about how the whole thing will be organized. I’m very unlikely to create an outline beforehand.

This can mean that the article ends up being a little bit meandering – perhaps like this one! Sometimes that matters and sometimes it doesn’t.

You just need to make a judgement and adjust accordingly.


So if you’ve not already started using speech recognition I do encourage you to give it a go – and let me know your thoughts and experiences in the comments below.




Don’t forget to subscribe to my newsletter to ensure you don’t miss any posts!

[ulp id=”ZRCymgXhv976FdCa”]

3 thoughts on “Speech Recognition – How To Create Content FAST”

  1. Hi Alex, Very nice article thank you.

    I note your first point, and I thought you might find this YouTube video of interest. https://youtu.be/UmvOgW6iV2s?t=3m46s It’s not necessarily as new a problem as you might think…

    I have used speech recognition software for years. I originally used the Microsoft one that came with Microsoft Word and then MS made it standalone. (“At the boat” speech type O example{Twice!}) At about that time I switched over to Android.

    When I’m on my PC I use an online speech processing app called SpeechPad. I’m using it now. I understand it uses Google speech Technology. For some reason I perceive it as being better than speech processing on Android, I’m not sure why! But really there’s not much to choose between either and speaking to your phone is usually more convenient. Here’s a link to it:- https://speechpad.pw/?autostart=1&vid=compact&pagelang=en-US

    SpeechPad is definitely much better than the speech processing software in Google Documents. I find the Google Documents speech recording facility particularly annoying! Give it a try, I think you will see what I mean…

    There are two main issues with using speech recognition:-
    First- When the speech processing software makes a “typo” (so to speak) they are different than the typos you are used to. You can easily miss them! You need to train yourself to spot them.

    2nd- You can get carried away speaking, you can say something really profound, in other words you are doing a good job, are unfettered in your speech. However when you look back through you discover that the speech software has misinterpreted what you said… And for the life of you you cannot remember what you actually said, although you know the words flowed perfectly and were exactly right! That can be extremely frustrating! However I have a simple solution to this:- “Online Voice Recording” it’s not perfect, you have to watch yourself because sometimes the recording doesn’t happen, or terminates early. The online voice recording program here:- https://online-voice-recorder.com/

    The other thing you need is a good quality USB headset.. Look to spend at least £30.00 — What you are looking for is a headset with a USB… The USB plug also contains the speech processing software. In other words you are not relying on your PC to process the speech, it is being processed by a dedicated chip inside the USB. The end of the cable where the USB plugs into the USB on your PC contain such a chip…

    Currently I am using:-
    “CHATMAX HS-720 USB Headset £29.99”

    It’s not the perfect headset. I noticed one person commented that they have broken (their hand Typo Example) theirs, and indeed I broke mine, even after seeing the warning! and have them (typed up Typo Example) taped up. They are of a very very weak construction.

    My favourite headphones I cannot get anymore they were perfect very strong, sturdy. I wore them out! And they were the very comfortable behind the neck type:- “CREATIVE HS-3?? Headset”

    Ones to watch out for in particular (“IV” – typo – Should be “are thee”) are the USB dongle type, without a cable. I haven’t had any real success with these although I must say I did buy some refurbished ones for about £30… The problem with (them worse they Typo) them was they had this constant annoying clicking sound and the voice recording level wasn’t constant.

    1. Thanks, Tony.

      A LOT of very good info here. We’ll have to get you guest-posting sometime!

      By the way, if anyone want to know ANYTHING about MS Access – Tony is your man!


  2. Re:- “We’ll have to get you guest-posting sometime!”

    Hi Alex I could do a post on moving from MS Excel to MS Access. It’s not as simple a process as it might at first appear. There are some “Major” Gotchas in the process. You could save yourself days, if not weeks of heartache if you are aware of them and plan your migration accordingly.

Leave a Comment

Scroll to Top