Speaker 1: Greetings to all, welcome to a new video tutorial. This time I want to share with you an excellent tool that will allow you to transcribe audio to text for free, online and easily. I find it a very interesting tool because I know most of these tools and almost all of them are paid. There are also online transcription services that also have a cost. After a lot of searching, I found this tool that is open source and that allows us to do this task easily. Here I will quickly tell you what are the main applications that I see, the main uses, so that we have clarity in this since it will allow us to work much more agilely in different projects. For example, it will help us to transcribe interviews, that is, people who are journalists who work in the area of ​​interviews and content like that, they will be able to make direct transcription of those interviews they do. For example, an interview in audio, they will be able to make the complete transcription in text. What is this for? To have both formats. There is the audio format and to be able to show the transcription of that interview, because there are people who like to have the ladder with the main ideas of that interview. We will also be able to make podcast transcriptions. The podcast is a very interesting content format, I like it a lot, it is on the rise, and that will also allow us to do this task. For example, when we are recording the podcast, we are automatically doing the transcription. We will see that now when I explain how the tool works. And finally, another application that I see is to create articles much faster. We talk much faster than we can handle the keyboard. So, through the voice, we will be able to write an article or comment on our ideas, and from there build our article. In the end, we will only have to make an edition, possibly it will take us much less time to make this edition of this content, and from there we will be able to create content much faster, which benefits us and will differentiate us from our competition, because we will be able to publish much more high-quality content. Well, that said, I hope it is very useful, and that everyone can implement it in their own projects. Here we are in the tool. Here I leave the URL, I will leave it in the description of the video, because it is a fairly extensive URL. Well, the main domain would be bluemix.net, but before that appears speech2textdemo.ng and there it would be like the complement to this. The name is Watson. Here we have all the explanation and all the detail of the tool. Here it tells us that this is related to IBM, so it seems to me that it is a very interesting project. We are going to put this in Spanish so that there is more clarity, so that it can be understood more easily. IBM Watson speech-to-text dictation service. Here we have the API and documentation references, the code in github, which is very interesting to have, the terms of use appear here, transcribe audio, and here comes the interesting part. What are the files or formats allowed? .mp3, .mpeg, .wav, .flac or .opus. These are only the types of formats allowed. Here we also have playback of sample audio files, that is to say that if you click here, any of these two will appear as playback and it will automatically be transcribed. For example, we can make a brief sample so that you can see how it works. The playback starts there. I have paused it and it is not heard here because at the moment I have ... Well, surely it was heard what was being spoken and there it was being transcribed. We also have sample playback number 2 and the same thing is going to happen. It is detecting the speakers. Here it tells us and indeed the transcription is very good. Then we have here the option to change the model of the voice. In this case we have the different languages. In my case it would be a wide band model, Spanish-Colombian. Here we have the possibility to do the same. Some samples with playback 2 or we can do the audio recording at once. In this case we are going to do the audio recording so that you can see how the transcription is happening here. At this moment I am doing a sample with the text transcription tool and the audio transcription to text so that automatically this happens that I am talking to time while I record. Notice that in this case it seems to me that the transcription was very improvised, what I was dictating, but the truth is that it is very good. Here it appears how all this immediately once we are talking. It is important to have a good microphone, have a good internet system, a good speed for this to work correctly. It is one of the recommendations they make and it is something very simple. We also have the possibility to upload an audio file that we have already recorded and from there you can make the transcription. Let's go to the browser to do it again here from scratch. And from there, load the audio file. How are we going to do it? Let's go directly to Audacity and let's make a comment in audio to be able to download it and from there upload it. At this time I am doing a test with Audacity to be able to make the audio file so that we can upload it directly to the Speech to Text tool. From there we will be able to check how this file looks after we load it and how the transcription looks. Let's go then to download the file. Here it would be to save another, we could also save it as mp3. Let's save it as WAV. Let's go to the desktop and I'm going to leave it here as test speech. I like to save it in 32. We save it, clean it, accept. And we already have our file. Let's upload the audio file. Upload the audio file. Here we have test, open and it is very important the following. Here we forgot, for example, to change the language theme. Let's start again and let's change the language so that it is correct. I almost always forget that part, so that you have it present. We upload the file again and it starts to read. Notice that at this time I am talking while I do the video transmission, while I do the recording. Transcribing the audio that we just uploaded. That is, it does not interfere at all that I am doing this recording and automatically the same thing happens in the transcription. Note that the text is fine. What I read said, at this time I am doing a test with the city. There are some errors, here it was not with the city, but with the Audacity tool to be able to make the audio file in such a way that I can upload the tool directly. Here we have Spiritus to Text and here the question is that according to how it reads, it will transcribe me. In this case, the tool understood this. It is necessary to keep in mind that it is a tool that is not perfect, however, it seems to me that in general it has all the correct information. Here I meant it was Speech to Text and recorded it as Spiritus to Text. I wanted to talk about how this file looks after we upload and how the transcription looks. That's what I meant. If we look in general terms, the truth is that the transcription is very good. It would only have to be some corrections and from there we will have everything ready in our text to be able to correct it. I hope this information is very useful, that you can implement it in your projects and any questions you have, you can leave it in the comments. Thank you very much. Thank you very much. Thank you. Thank you. Thank you.