(i'm sorry for the title, hope in a good edit from a mod - not english native speaker here)

Hello, I'm in the search for a software able to do the following:

  • Take a wav, mp3, other audio format, even video (keeping only the audio) as input
  • Process the audio while ignoring possibly some audio noise
  • Write down in a well organized format the text recognized with timestamp and different persons speaking with auto-tagging, example: In a speech 3-4 persons talk, the software specify Person1, Person2, Person3, Person4. I can rename Person1 as "Professor" and rename, keep or delete Person2,3,4 talkings before exporting the text
  • Richness in the text format output: txt, csv, maybe json, but mostly word and pdf with all the tweaks to best fit the content into page.

(if taking advantage of multi core is better)

In short i want something like what the Android Voice detect do (look at the picture), but Offline, with offline word I Mean "not in realtime", i need a software for PC (windows, but also mac os and linux if possible) that can take advantage of a fully fledged computer CPUs to process the audio, maybe in batch.

(a bunch of x86 64bit cores cable powered is capable of a lot more than some arm battery powered cores - more filtering, fast fourier transforms and so on..)

The definitive note taking software for the students.

Language recognition should be in the main european talked languages:

  • English
  • French
  • Spanish
  • Italian
  • German
  • Portuguese
  • Dutch
  • more is better

I know the request is very specific and not easy to fullfill but please suggest me even softwares that do only some part of the process. I can consider making a personal workflow based on use multiple softwares taking the best they can do.

A similar software for this unrelated with note taking but also working for this purposes could be the auto closed captioning softwares existant for the video industry, that ones take note of the text and the timings, aswell the fact that probably the most good and expensive ones are able to detect, correct and isolate multiple persons and languages..

PS: Software or softwares can be free or paid, just not something cloud based or with monthly subscriptions i hate that kind of stuff and does not fit an offline (this time i mean not connected to the internet) workstation to do the job.

PPS: I know my request is hard, i want to achieve a result like i did in past with my Software to quickly join many JPG images into one PDF without recompressing question with a crowd of sources leading to a definitive rich argumentation for the benefit of the community.

  • Very well done! Nevertheless, I'll removing the comments now (as they might be distracting). Good luck, I'll definitely come back later to check the outcome! // PS: Any limit for "paid", e.g. "should not exceed a one-time fee of X"?
    – Izzy
    Mar 17, 2016 at 10:37
  • 1
    @Izzy well if the softwares starts to cost over 1500$ things are going to get hard :D Mar 17, 2016 at 10:39
  • Historically, 2 windows ones that have been around for ages, but not sure if/ how they fit your needs today - you'll have to check 1. IBM VoiceType 2. DragonSoft Naturally Speaking
    – Alex S
    Aug 18, 2016 at 6:51


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.