Why doesn't sentence-mode TTS count towards listening time?

I spend most of my time on LingQ reading ebooks. For several reasons, adding audiobook audio to ebooks is cumbersome and often unworkable. So what I have done is create simple keyboard macro that sends:

Shift+Right-Arrow (to turn the page)
Pause 250mS (to give page time to load)
a (to play sentence TTS)

This little script has increased my enjoyment of reading on LingQ tremendously.
However the time spent listening does not appear to be added to my total listening time.
Why is this?
I can understand how the TTS API might not report speaking times, but an approximation of how much time a sentence containing of X number of letters or words takes to be spoken could be made.

Couple of thoughts here: why not listen along to the audiobook on a separate device, and then just add your listening hrs manually? This can be done very easily and you’d get way more out of a professional read than a TTS voice.

I have a feeling the TTS count you’re suggesting would probably be too much hassle to implement for too little benefit, since I doubt many people use the TTS for long form reading.

But ultimately, I feel the hrs of listening statistic here on LinQ is probably the least important to keep track of because as soon as you’re intermediate, most of your listening should / probably will be outside of LingQ, listening to podcasts, books, TV shows etc… You could hypothetically sit there with a stopwatch every time you listen, but I really feel that as long as your listening time keeps up with the amount of words you’ve read, you’re doing fine and you don’t need to keep track of it.

How do you calculate words read to listening time, you ask? Here is the formula I use:

1 LingQ lesson = @2200 words = @7 pages of a printed book = @15 minutes of audiobook narration

So each lesson of text is about 15 minutes of listening – I have experienced this with three languages now and the formula holds fairly steady across languages.

BUT, if you listen to a podcast, the pace is way faster, so you’ll end up hearing a lot more than 2200 words in 15 minutes. This is one of the reason’s why I think hrs of listening is a somewhat pointless statistic. “Words read” is quantifiable. “Words known” is quantifiable. So just pay attention to those and listen as much as possible and you’ll be fine.

1 Like

The value of particular metrics is debatable, and certainly varies among users. And yes, I could do it manually. But isn’t the whole point of LingQ to automate the tedium of language learning – to be a tool to make things easier?
As far as listening on a separate device: Sometimes this is not possible, as the audiobook version is not available. Also, I find that managing audio on a separate device makes stopping to create LingQs an inelegant task at best. I very much prefer the integrated, synchronized audio listening mode available in LingQ’s sentence mode.
Where available, I would much prefer audiobook audio. However, since LingQ’s text length limit is much shorter than many book chapters, adding synchronized audio is not a reasonably workable solution. So instead I use TTS. If I could simply upload a chapter of text and a chapter of audio, that would address my gripes. But I can’t. Which I think is a shame because LingQ’s auto-timestamp feature for audio uploads works so very well.

I agree that we all have to find the method that works for us. In my case, I find that the “car mode” feature on the Audible app on my phone makes it very easy to hit pause and play as the phone rests near my left hand, while my right hand uses the keyboard shortcuts on LingQ to mark words. With a little practice this method can become very effective and provide benefits that outweigh any time stamping of TTS features. Depending on the amount of known words on a page, one can feasibly keep up with with the pace of narration for long stretches of text with this method.

I’m just bringing this up in case it helps because I simply feel that any method with a real audiobook is worth the extra hassle over an easier TTS method, since you will learn better from human pronunciation than from TTS – both directly and indirectly. Directly because the pronunciation is better, and indirectly because your engagement with a human performance will be way different on a psychological level.

1 Like

I appreciate your feedback, and agree with your assertion that natural speech is much better. I always listen to the audiobook – when available – in addition to going through the book in LingQ.