A program for decrypting audio data. How to translate speech audio file to text using Google translate. Online transcription software

Publication date: 30-09-2018 5722

Good afternoon, dear freelancers!

We have already gone through what transcription is and how to make money on it in the last article. Now we will touch on another aspect of this phenomenon: such as audio transcription program.

There are a number of programs for transcribing audio to text and many ways to translate text from audio and video formats into written text.

We will analyze the methods of effective transcription in the next article, but for now, let's look at the key programs for this process.

Transcribe audio to text free software

I think it's worth starting with the most well-known program in the field.

Express scribe

Alas, this program comes only with English, but it is so simple and straightforward that absolutely anyone can handle it, plus everything - it is a bright representative of free.

Key benefits:

It is very easy to enter text and there is no need to rush between the field itself and the player.
It is possible to change the speed of text playback to make it convenient to type text.
It is possible to fix "hot" keys for the convenience of working with the player.
Perfect for Word.
There are time codes.

Only English, although this does not particularly complicate the work.

In another article, we will analyze in detail all aspects of working with this program and how to use hotkeys correctly.

LossPlay

Convenient software for text transcription, plus all this transcription program can be downloaded for free.

Key benefits:

The presence of hot keys.
Work with audio and video formats.
The ability to change the recording speed.
Availability of time codes.
The ability to adjust the rollback after a pause.
Perfect for Word.

You will have to switch program windows from time to time.

This software is able to replace ordinary applications in the same Google service where you also have the option to enter text via microphone recording.

As you can see, there is a number useful programs for successful transcription. Among other things, there are also many and video transcription programs. Below you can write your opinion about this or that program or any other that you use in the comments. All the best. See you in the next article.

Despite the popularization of video content, a text document will always win as when composing online documents, and when filling sites. Despite everything, transcription, or in other words, audio-to-text translation, is a service in demand to this day.

Read this article for a description of what methods you can use to translate audio or video into text with the least effort.

Transcription is the translation of speech into text. You can translate both audio files (recording on a dictaphone, songs, etc.) and video (speech from films).

Transcription is a rather laborious process. At the forefront is your ability to quickly ... (at first I wanted to write "typing", but tell me, didn't we write at school under the teacher's dictation, isn't this translating audio into text?) To print and write.

This profession has been in demand for several years. A modern blogger is also not averse to using an already disclosed topic in some kind of video, by translating audio into text.

In this article, we will analyze 6 ways to transcribe into text and start with manual transcription programs.

Audio-to-text translation software

As we understood, transcription is the translation of speech into text. That is, we must write down everything that we are told.

For this purpose, there are both basic and professional programs.

Microsoft Word text editor

The most commonplace way to translate audio into text is to use:

Run audio or video file.
Open text editor.
Listen and type text.

Initially, in this section I planned to write about Word, but on reflection, I decided that any text editor would do: Notepad, Word, Google Docs, LibreOffice Writer and others. The main thing is to be able to type text from the keyboard.

The banality of this method is that you will have to listen to the same section of audio several times, unless of course you have a fantastic typing speed.

And yet, this method has a right to exist.

Express Scribe professional transcription software

Have you decided to start translating audio into text? The best it has to offer is Express Scribe.

This program allows you to type text directly working with an audio or video file in the adjacent window. It works like this:

Upload an audio or video file to the program itself.
With the help of hot keys, set up the functions of playing or stopping the video.
In the next window, type your text.

Developers provide paid and slightly stripped down free version... As for me, if you are not professionally involved in transcription, the free version will suffice.

Of the minuses: it is in English and there is no russifier. Although, the interface is so clear that the language barrier should not become an obstacle.

From the pros:

For typing, no need to switch between programs.
Depending on the speed of typing, you can adjust the speed of the recording playback.
A huge number of customizable hotkeys are provided to help you work with the program.
There is a possibility of placing time codes.

Player for decryptors LossPlay

Player LossPlay is a modernized player for audio and video files. The idea of the functionality was based on the famous Winamp player.

Transcription occurs according to the following principle:

Launch LossPlay.
Add an audio or video file.
Open any text editor in which we will type text.
Start playing the file.

What is the difference between this method and the method proposed by the very first one?

Basically, yes, the idea is the same. And yet there are some nuances. The fact is that when you start an audio or video file, it is possible to minimize the player and work only in a text editor. The player can be accessed by means of hot keys, which can be used to:

start playback;
stop playback;
slow down playback;
set timecode;
and much more.

According to the developer, LossPlay is ideal for translating audio into text. The product, in terms of functionality, is on an equal footing with professional developments, but has the advantage of lower costs system requirements computer.

Free online programs for translating audio to text

As mentioned in the article, the profession of transcription is in demand. Of course, people try to make their life easier and try to create online programs s - assistants who will be in automatic mode, convert audio to text.

Looking ahead, I will say that the profession would not be so much in demand if such programs worked 100%, but some developments already exist and are actively used.

Setting up your computer for transcription

Before starting automatic decryption to text, you need to properly configure your computer.

There are enough articles on the Internet in which it is proposed to use third-party developments for setting up an audio card. Perhaps this is necessary for professionals, but from my own experience, it turned out to be enough to configure the built-in audio card properly.

We have two options:

Include an audio file and substitute a microphone for the speakers (wildness). Speech from the speakers will go into the online program and automatically audio will be translated into text.
Set up the audio card and connect the Stereo mixer so that the sound from the system gets into the online program.

The first option is bad, because additional distracting noises may come across, the program will read them, and in the end you will get rubbish.

In the second version, speech will be sent from an audio file directly to an online program and perceived as it is. By the way, in this case, not everything is smooth.

Open the "Sound" window in the "Control Panel".

On the "Playback" tab, you need to set the default device to which the main speakers are connected.

Please note that I have both headphones (default communication device) and speakers (default device) connected on the screen. This is done so that I can transcribe while listening to the file through headphones.

I don't know, maybe it's just me, but if I turn off the main speakers, then the sound stops coming to the Stereo Mixer and the programs cannot detect the sound stream. If anyone has figured it out in more detail, write in the comments.

In the "Record" tab, we are interested in the "Stereo Mixer", which should be selected by default.

Note!! If you do not see "Stereo Mixer", click on an empty space in the window right click mouse and select Show Disabled Devices.

If you, like me, for example, do not need additional connections for audio output, then this is the end of the setup. In my case, I wanted the sound to continue to come from both the speakers and the headphones.

Select the properties of the "Stereo mixer" and go to the "Listen" tab. Here, in the item "Play from this device», Select the headphones or the device to which you want to output the sound.

Having done this procedure, we third-party programs and drivers, set up audio reception for translation into text, through system devices computer.

Google docs

I will put google documents first. This is a regular text editor that is analogous to regular Microsoft Word.

Choose "Tools" → "Voice Input" (or "Ctrl" + "Shift" + "S"). The microphone is activated and a message about the permission of the program for accessing the computer microphone (allow access without changing the parameters (should be set by default)).
Run the video of interest in any player or on YouTube.
Click on the microphone in Google Docs. If everything is configured correctly and Google Docs sees the audio stream, audio-to-text translation will begin. Otherwise, a message about missing sound will appear.

Personally, my opinion, it may be biased or subjective, but more is not required to transcribe audio into text.

Most clearly translates audio into text, without gaps.
Great functionality for editing text.

Unable to sync audio file with decryption start.

Speechpad Voice Notebook

Spechpad notebook, located at https://speechpad.ru/. Works according to the principle described above. An audio recording is turned on, the notebook begins to read the speech and translate it into text.

The advantage of this notebook is that in the settings there is a function "Run synchronously with recording". How it works:

Go to the notebook website (by the way, in order not to read the footcloth of the text, register on their website).
Click "Transcription".

Panel "Transcription", select the media type of interest. In my case, "Youtube video".
In the "Id file to play" field, add the code of the YouTube video (for example, for the address https://www.youtube.com/watch?v=Xihpth5vmmM, code Xihpth5vmmM).
Check the box "Run synchronously with recording."
Click "Enable recording".

The video will start from YouTube, and Speechpad will immediately start transcribing, translating the audio into text.

Synchronization of an audio file with text decryption.
Playback speed controller.
A huge number of settings.

When transferring the translated part of the text to the main block, some words are skipped.

The Dictation.io audio decoder, can be found at https://dictation.io/speech.

To start transcription, just start the audio file and press the "Start" button. At the end of the video, copy the result into a text editor to work with it.

Synchronization of the audio file with the beginning of the decryption of the text.
Copying the result is done by pressing one key.
There are a huge number of settings.

Not seen yet.

Online audio-to-text translation results

In conclusion to the section, I would like to show the work of the presented online programs - audio-to-text translators. For this, an identical segment was taken from the fairy tale about three pigs and this is what happened.

Google docs

In ancient times, 3 pigs lived in the world, they lived with their mother in a small house, one day it was time for them to leave the moment when the mother pig called them to her and said. My dear children, it's time for you to go out into the big world and children and start live your own life, but do not forget, whatever you do, do it conscientiously here The best way survive 3 little pigs, affectionately said goodbye to mom and hit the road. after a while they came across a log suitable for building houses. the youngest of the pigs decided to build a house from straw

In ancient times, 3 pigs lived in the world, they lived with their mother in a small house, one day it was time for them to leave their mother, then the mother called the pig to her and said my dear children, it was time for you to go out into the big world and children and start living your own life but do not forget whatever you do do it conscientiously here is the best way to survive 3 affectionate little pigs said goodbye to their mother and set off after a while they came across a log suitable for building houses the youngest of the pigs decided to build a house from straw

Online decryptor Dictation.io

3 little pigs times lived in the world 3 little pigs, they lived with their mother in a small house, one day it was time for them to leave their mother, then the mother called the pig to her and said. my dear children, it's time for you to go out into the big world and children and start living your own life but do not forget whatever you do do it on your conscience here is the best way to survive on a field but suitable for building houses, the youngest of the pigs decided to build a house out of straw

The main problem of all three programs is the lack of punctuation marks, but this is solved by editing the resulting footcloths of the text.

Ways to speed up the work on decryption

Having vast experience in typing texts, I made the following conclusions for myself how you can speed up the decoding of audio into text.

When the speed of video playback is reduced, the transcription is performed with the smallest errors.
When typing text manually, the scheme is ideal: listen → remember → write down. This scheme will allow you not to postpone editing until later, but to do everything at once.
Using commands to help add a piece of text. Imagine that in the text, the word "Transcription" often comes across, we add it to the command and, using the burning keys, display it in the right places.
It doesn't matter if you type manually or using online programs, all edits and corrections are at the very end.

If you didn’t spend hours transcribing audio to text, endlessly restarting the same recording, cursing the speaker’s unwillingness to pronounce the words completely without swallowing the endings, then, therefore, you don’t know what transcribing audio to text is. By the way, this way suitable for freelancers, since often on sites offering work you can find a task for transcribing text and it costs not so little. Therefore, in this article I decided to talk about various programs that help you quickly perform audio-to-text transcription.

Transcribing audio to text with Google Docs

Similar in functionality to the same Word, they have a built-in dictation tool called Voice Input. Voice decoding happens almost independently. Those. when you speak, this tool understands your speech and automatically translates it into text, which is immediately printed in Google Docs. This feature is suitable for those who cannot type quickly or those who prefer to dictate text.

In addition, you can use the tool to minimize the time it takes to transcribe an audio recording in about the same time as the recording itself. Basically, you will be engaged in listening to the recording (on the phone or computer) through headphones and. This program for transcribing audio to text will translate your spoken words into text automatically.

When I used the 1 hour interview transcription tool, it came up with sentences with decent accuracy. This tool spelled out proper names without mistakes, knew when to use pronouns, and when a verb, as well as when I finished pronouncing a sentence, automatically corrected the words in which I initially made a mistake. I finished translating the text in about 70 minutes.

Below I have given step by step instructions for transcribing text via Voice Input:

Find a quiet place

To transcribe text through voice input, listen to the recording through headphones, then speak the words aloud. This instrument does not have the ability to transcribe the recording through the player, it needs to hear the voice without extraneous noise. If there are extraneous noises in the recording, or a lot of people speak at the same time, then the transcription will turn out to be inaccurate.

Use a microphone and headphones

Use headphones to listen to the recording, repeating the words behind the speaker, so that Voice Input can only distinguish your voice, not the recording.

Opening Google Documents

Go to the home page of Google Docs and click on "create document".

This is how a blank sheet of document looks like, in principle, it is not much different from the Word document.

Opening the Voice Input tool

Checking for the presence of a button

After you have pressed "Voice Input", a record button should appear on the screen to the left of the document:

Make sure the microphone is turned on

I hope there is no need to explain that to record speech, you need to connect a microphone to your computer, or if you have a laptop, use the built-in microphone.

Voice input has the ability to distinguish between many languages, including different dialects and accents (for example, it distinguishes over 20 adverbs in Spanish). To select the language you want, use the drop-down list under the record button, select the one you want and start dictating the text:

Press the record button and start reading the text

To start transcribing, click on the Voice Input button. The microphone icon will turn red, it means the recording has started. Start listening to the audio recording through headphones and speaking the text into the microphone. The tool will immediately start converting it to block letters.

During transcription, do not switch to other tabs in the browser. If you go to another site, the tool will stop transcribing.

Try to pronounce all the letters in a word, speak in a regular tone and not too quickly. If you speak too quickly, the program may start to make mistakes.

While the program is running, you will see a red circle in the Google Document tab:

A red circle that appears in the browser tab indicates that the entry is being processed.

Make sure the recording is not interrupted

To ensure that the transcription has not stopped, constantly monitor for possible problems. If this tool stopped translating the recording into text, you should again press the button with the microphone so that it first turns off, and then press again to turn it on. While I was transcribing an hour-long recording, I had to turn the instrument off and on again about 3 times, as the recording was sometimes interrupted.

Express scribe

But, nevertheless, not everyone is ready to dictate the text into the microphone. Someone is more convenient. At the same time, if you use ordinary programs like Word and listen to the recording while typing, you can waste a lot of time switching between windows. Thus, you can lose potential money for the work performed.

For convenient transcription, the Express Scribe program has proven itself well, with intuitive functionality, and also free. The only drawback is the English interface.

But, in principle, you do not need to know all the functionality in it. For our purposes, it is enough to use just a few functions, so you can easily figure it out without translation into Russian.

So, what are the advantages of the program:

the ability to slow down the playback of the recording
the text is printed in the same window as the audio recording, so there is no need to switch between tabs
there is a rewind function if some recording area is not clear
there are hot keys to stop and start recording.

Who needs audio-to-text transcription

Audio transcription is usually required for those who work with courses, lectures, conferences, and similar events.

A situation may arise that after the conference there is only an audio recording that needs to be translated into text and then posted on the Internet.

This is exactly the kind of activity that requires audio-to-text transcription. The recordings themselves come in various qualities, so you may sometimes have to listen to the recording for a long time to make out what it says. Most likely, in the assignment, the customer will ask to arrange punctuation marks, separate dialogues for different participants (man, woman, lecturer # 1, etc.). Therefore, machine translation will most likely fail here.

Job exchanges to search for orders

If you are just starting your way in freelancing, then you are probably wondering where to get orders for this type of work on the Internet. The fastest and most reliable way would be to search for jobs on all kinds of freelance exchanges - FL.ru, freelance.ru, work-zilla.com and many others.

On large exchanges, orders appear almost every day. It is usually difficult for beginners to take on assignments at first, since the customer may feel that they have little experience. Therefore, I would recommend starting with such orders, where, even if they pay little, but with a well-executed task, your level will grow and thereby the trust of customers in you as a reliable performer.

In addition to freelance exchanges, you can also advertise your skills on various message boards and post typical services on the Internet. This will not bring instant results, but over time you will be able to receive additional orders. The main thing is not to write standard phrases like "negotiable price", but a specific cost and skills.

The SendPulse service is a marketing tool for creating a subscription base and converting random visitors to your site into regular ones. SendPulse brings together the most important functions for attracting and retaining customers on one platform:
● e-mail newsletters,
● web-push,
● SMS mailing,
● SMTP,
● messaging in Viber,
● sending messages to facebook messenger.

Newsletter email

You can use various tariffs for conducting e-mail newsletters, including free ones. Free tariff has limitations: a subscription base of no more than 2500.
The first thing to start with, when working with the e-mail mailing service ok, is to create your own address book... Set a title and upload a list of e-mail addresses.

SendPulse makes it easy to create subscription forms as a pop-up window, inline forms, floating and fixed in a certain part of the screen. With the help of subscription forms, you will build a subscriber base from scratch or add new addresses to your base.
In the form designer, you can create exactly the subscription form that most fully meets your needs, and the service tips will help you cope with this task. It is also possible to use one of the available ready-made forms.

When creating subscription forms, it is mandatory to use an e-mail with a corporate domain. Read how.
Message templates will help you to design your letters to subscribers beautifully. You can create your own letter template in a special constructor.

Autoreports... Content managers actively use automatic mailing. This helps automate the customer service process. There are several ways to create an auto-mail:
● Consecutive series of letters... This is the simplest option, when, regardless of the conditions, several letters are written, which will be sent to recipients in a certain order. There may be your own options - message series(simple message thread), special date(letters are timed to specific dates), trigger letter- the letter is sent depending on the actions of the subscriber (opening a message, etc.).
● Automation360- mailing with certain filters and conditions, as well as taking into account conversions.
● Ready chains by template. You can create a series of letters according to a given template, or modify the template and adjust it to suit your needs.
A / B testing will help you experiment with different options for sending a series of letters and determine the best option for openings or transitions.

Sending Push Notifications

Push-mailings are a subscription in a browser window, it is a kind of replacement for rss-subscriptions. Web-push technologies have rapidly entered our lives, and it is already difficult to find a site that does not use push-mailing to attract and retain customers. The request script for, you can send letters either manually or create auto-mails by creating a series of letters or collecting data from RSS. The second option implies that after a new article appears on your site, a notification will be automatically sent to your subscribers with a short announcement.

New from SendPulse- you can now monetize your site with push notifications by embedding advertisements in them. Upon reaching $ 10, payments are made every Monday to one of the payment systems - Visa / mastercard, PayPal or Webmoney.
Push messages on the service are absolutely free. Payment is taken only for White Label - mailings without mentioning the SendPulse service, but if the service logo does not bother you, then you can use push notifications for free without restrictions.

SMTP

The SMTP function protects your mailings from being blacklisted by using whitelisted IP addresses. DKIM and SPF cryptographic signature technologies, which are used in SendPulse mailings, increase the credibility of sent emails, so your emails will be less likely to end up in spam or blacklisted.

Facebook Messenger Bots

Facebook chatbot is in beta testing. You can connect it to your page and send messages to subscribers.

Sending SMS

Using the SendPulse service, it is easy to send mailings to the database phone numbers... First, you need to create an address book with a list of phone numbers. To do this, select the "Address Book" section, create a new address book, upload phone numbers. Now you can create an SMS mailing list for this database. The price of SMS messaging varies depending on the recipient's telecom operators and is on average from 1.26 rubles to 2.55 rubles for 1 SMS sent.

affiliate program

SendPulse implements affiliate program, within which a registered user using your link who paid for the tariff will bring you 4000 rubles. The invited user receives a discount of 4000 rubles for the first 5 months of using the service.

Transcription Is the decryption audio- and video materials, their content translated into text. What is said in such materials, literally or with a certain degree of censorship, is displayed in text document... In this article, we will talk about ways to simplify transcription on a computer. It would seem, what is difficult in the decryption process? You take a convenient media player, periodically pause the recording and rewrite what you hear.

1. Difficulties of transcription

But no, it's not that simple. Otherwise, those interested in such a service would not be looking for remote employees on freelance exchanges.

First, not everyone can do the job. We need a seasoned, diligent person, capable of long-term concentration of attention. For reference: it takes about 5 hours to decipher the 1st hour of the recording.

Secondly, a high-quality source code is not always supplied for decryption, and this, accordingly, complicates the transcription process. In such cases, the performer must also be technically savvy in order to resolve the issue of eliminating the interference of the recording.

Transcription- a resource-intensive process. The constant switching of attention from one task to another, from one program window to another, quickly depletes forces. Is it possible to somehow simplify this process? Alas, there is no perfect way to automate this process yet. Only a person can cope with such a task. But its work can be simplified with the help of separate software tools.

For the convenience of working with two program windows - media player and text editor- in the environment of modern Windows versions you can use the function of efficient use of screen space Snap ... We attach the player window to one part of the screen, the editor window to the other.

Advice: when decrypting audio materials, it is better to use players with large controls, such as an application "Groove Music" shipped with Windows 10.

When transcribing video recordings, you can use the compact mode of the standard Windows 10 video player - applications "Cinema and TV"... In compact mode, the player turns into a mini-window docked on top of other windows with a limited list of controls. But this list will contain all the buttons you need to work - pause the video, steps back and forth.

Advice: if the source is a video, so as not to be distracted by the video sequence, you can extract the sound in audio file... And work with him already. Such operations are able to carry out programs-media converters, for example, free Convertilla and Freemake Video Converter... And programs like Adobe audition or its free equivalent Audacity will help improve the sound quality if the original is poor.

4. Third party video players

Third-party video players for Windows, such as, for example, KMPlayer, GOM, VLC, will offer more functionality, which, in particular, can be used during transcription:

Configurable forward and backward steps;
Adjusting the playback speed;
Setting bookmarks;
Customizable control panel;
Customizable hotkeys.

Advice: some performers prefer not to pause the source, but to slow down its playback for dictation. This is not a good idea for optimizing transcription. When writing a dictation, pauses are important, delimiting even parts of sentences. Plus, our consciousness enters into resonance with slow playback, and this, accordingly, slows down the work of the mind and typing.

5. Express Scribe

- a program specially created for transcription. At the top of its interface there is a field for adding media files, at the bottom - a player and a text editor. Among its advantages:

Single window for working with sources and text;
Support for a large list of media formats;
Functions to slow down and speed up playing files;
Rewind;
Improving sound quality;
Hot keys, as well as the ability to use the foot pedals of the gaming steering wheel for certain functions;
Saving sources and their decryption as program projects;
Export typed text to Microsoft Word.

disadvantages - lack of support for the Russian interface language and the active imposition by the creators of the program of its paid editions when working with a free one.

6. Services of voice data entry

Web services offering the ability voice input data is perhaps the best way to simplify transcription. Especially suitable for those whose printing speed is poor. The principle of operation here is simple: we memorize fragments of text, repeat them into the microphone and along the way carry out text correction - we correct mistakes and insert bickering.

The web service is equipped with a speech recognition engine ... Documents created in this online editor can be dictated using the tool "Voice input".

The tool supports a huge list of languages, including Russian. However, it does not provide for an auto-disable setting for data entry. If you do not speak into the microphone for a couple of seconds, voice input is disabled. And it needs to be manually turned on constantly.

Another web service - "Voice notebook Speechpad.ru"- will be many times more effective, since this is a narrow-profile project, and speech recognition is its main focus. To enter voice data in the service form, you must press the button to enable recording. The text is edited in the resulting field. There is a button at the bottom of this field. "+ Transcription".

It adds playability to the service form audio- and video source codes for work in one program window. For example, to transcribe a video on Youtube, in the service form, you must select its media type and insert into the appropriate field ID (part of the web address after the "=" sign) .

The service supports several other languages besides Russian.

Advice: is able to recognize two punctuation marks - a period and a comma. They can be spoken.

The disadvantage of this transcription method is the imperfection of the human speech recognition mechanism. How many of mobile users having played enough with "Ok Google" when typing search queries, have come to the conclusion that it is easier to find something the old fashioned way, using the keyboard. Whichever service you use, you still have to pause during the work and track the recognized text in small blocks.

7. Automation of transcription

The process of voice input of data during transcription can be simplified to the maximum by setting up the system sound as a sound recording source and, accordingly, removing from the chain of actions the pronunciation of what was heard into the microphone. Naturally, in the case of problematic sources, this option will not work. As a result, we get unreadable rubbish, which will be easier to rewrite than to edit. If the recording is of more or less quality, it's worth a try.

We call the context menu on the sound icon in the Windows taskbar, select. In the window that appears, using the options in context menu turn off the microphone and turn on the stereo mixer.

However, the process still needs to be monitored: to delve into the essence of the information, periodically pause playback to correct the decryption, put down bickering.

8. And most importantly ...

Transcribing on a laptop with its wheezing speaker is problematic. However, in particularly difficult cases, a good audio system will not help either. For constant work on audio decoding, it is better to get headphones. And if the work is done using speech recognition, it might be worth looking into a high-quality headset. Hands Free .