Mastering Adobe Sensei: Transcribe, Caption, and Subtitle Your Videos with Ease
Learn how to use Adobe Sensei in Premiere Pro to create accurate transcriptions, captions, and subtitles, enhancing your video content and SEO.
File
Create and Stylize Captions FAST in Premiere Pro Make Your Videos More Accessible to Everyone
Added on 09/30/2024
Speakers
add Add new speaker

Speaker 1: You can transcribe that whole maybe hour long sequence and it gathers all those keywords and you can search here. So I can search DJI and it pops up all the instances. You can cycle through where that keyword appears and find that content in your long form, and then cut that down to your short preview pieces that you want to use to promote it. Hey everybody. My name is Dave Simons. I wanted to create another video tutorial based on a feature I find immensely valuable and I use with every single video. So this feature may have been available in Adobe Premiere CC 2021, but they have seriously upgraded it in 2022. And that is Adobe Sensei. This is top-notch voice recognition software used to create transcriptions, captions, and subtitles automatically. I used to have my company pay Rev.com for captions. They now charge a $1.50 per minute of video. Not too bad, but the voice recognition inside Premiere Pro has gotten so good that you can simply transcribe a sequence and create captions and subtitles from that. You can then export a nice SRT caption file, which is compatible with nearly all video platforms. All right. So here's my sample sequence. This is actually a reel slash YouTube short slash TikTok I made from my most recent trip to Brazil. Creating captions and a transcript is super simple. You just click the new text tab that's next to your effect controls and source panel area. Only option right now is create transcription. And usually I just leave all the settings the way they are by default, except for you can use a mix and it'll still understand. Well, I usually just use the voiceover track of course. So you just change that to audio one. And then I usually have transcribe in point to out point only just in case I'm doing sort of other short video edits on that timeline and then transcribe. So then it, you know, it takes a few seconds, renders the data depending on the length of the sequence. This is all in real time. You can see how fast it goes. And the only thing that got wrong in this transcription is the proper names. So you can just double click into it. This is Baja Grange, which in Portuguese, R's often sound like H's at the beginning and middle of words. They can also sound like D's. And then this is Jericoacoara. I'm trying to use kind of more legit Portuguese pronunciation as I'm trying to learn it. So the way I pronounce those kind of fooled Adobe Sensei, which makes sense. But those are really minor changes to make. Everything else, it got perfectly right except DJI Mini TO instead of two. So I'll just add that and then add a comma. That's it. It even got the comma right here from a moving vehicle for the first time ever. And since that there was a pause there, then you can just click create captions. Here, I pretty much always leave this to its defaults. Since I'm using this for short form vertical video, I'm going to leave it to single lines, so it'll only have one line of captioning. Two lines is usually great for any horizontal video, any slightly longer video, because you have more space to work with in those horizontal videos. I'm going to reduce the maximum length in characters to 20. Normally I would leave it at the default 42 and then, yeah, that'll be all I change. It creates the captions based on that transcription. So there is a little bit of a bug that can happen when you do this. If you notice right here, it's starting at like 10 hours on the timeline. So you'll notice these didn't show up. All I have to do to fix that issue is re-transcribe sequence, uncheck transcribe in point to out point only. I have nothing else in this timeline, so it should be okay. Leave it to audio one. And I just make these few changes. You can see it actually starts at zero now or 19 frames in, because that's when the voiceover starts. So then it'll all work fine. You can see it still has the subtitle track up there. That's because 10 hours later in this timeline is where the captions currently exist now that everything is correct there. So create. And there it is. You can see it automatically times out the captions, which is beautiful. And as you can see, it's easy to read because it's only a couple words appearing on screen at once, which seems to be the trend for TikTok, YT shorts and reels. You have the opportunity to make it an independent SRT caption file, which you can toggle on and off on Facebook. YouTube, LinkedIn are the three most common platforms I use it on. And you can also burn in the captions. If you leave this track on, not only that, you can also click into all these and change the appearance of the captions, which is a game changer. You may have noticed some creators doing that, so you can select all the different individual caption lines with shift and click, and then you can change all of their visual settings at once, and a common thing to do now is to put them in the center of the screen. That's what all of the best YouTube channels do. I don't think I'm going to do that for this. I know it's like a bunch of landscape shots. It might work out well. The good visual stuff in this video is located in the middle of the screen. So I'm going to put it at the bottom. Still, I might raise it just a bit. It's great. You get all this customized control over how your captions look, which is a total game changer. I think I'm going to keep it simple like this. I don't really want much of a stroke right now. I know you're not going to believe it. I'm going to have just, yeah, that little drop shadow, maybe extend the distance. Yeah. I mean, Myriad Pro is fine with me. I think that looks good by default. Let's try 90. Yeah, 90. I think that looks pretty good. If you want to export the SRT file separate, then you can just go to the three dots in the top right of this panel, export to SRT file, and then you just name it whatever you want. Maybe the same as the video. And then you have a nice caption file that you can upload, and it works a little bit differently on all the platforms, but it's really straightforward to upload those attached caption files. It also means the world for your SEO. Unlike the auto-generated captions on YouTube, it'll read all of that in the search engine. So it really gives your videos a leg up for people who are searching for that type of content. So you don't have to rely on your title and description to have all those keywords. Yeah. So I'm going to just do Baha Granji Tejeri Reel, save it there, and then I can export the whole video. Since this is going to be a short, I'm going to export with the caption file activated, use maximum render quality always for better quality scaling. Let's kick it up to 25 megabits per second and export. And not only is it useful for short form video creators, but it is a game changer, maybe even more so for long form content creators, because say you want to cut down like a social media preview video for your longer form content. You want to make a highlight reel. You can transcribe that whole, maybe hour long sequence or longer, and it gathers all those keywords and you can search here. So I can search DJI and it pops up all the instances. You can cycle through where that keyword appears. And find that content in your long form, and then cut that down to your short preview pieces that you want to use to promote it, which is amazing. Oftentimes it's great to just use the transcriptions in your posts, because especially on like LinkedIn and Facebook, other places, they're not reading the SEO of that content as they are on YouTube. That is a game changer for streamlining your workflow from a caption standpoint, subtitles to creating transcription. I hope you found this video helpful. I'll see you next time.

ai AI Insights
Summary

Generate a brief summary highlighting the main points of the transcript.

Generate
Title

Generate a concise and relevant title for the transcript based on the main themes and content discussed.

Generate
Keywords

Identify and highlight the key words or phrases most relevant to the content of the transcript.

Generate
Enter your query
Sentiments

Analyze the emotional tone of the transcript to determine whether the sentiment is positive, negative, or neutral.

Generate
Quizzes

Create interactive quizzes based on the content of the transcript to test comprehension or engage users.

Generate
{{ secondsToHumanTime(time) }}
Back
Forward
{{ Math.round(speed * 100) / 100 }}x
{{ secondsToHumanTime(duration) }}
close
New speaker
Add speaker
close
Edit speaker
Save changes
close
Share Transcript