Transcriptive is the new way of automatically transcribing your video within Premiere Pro! It’s a Premiere panel that uses Artificial Intelligence and Machine Learning to turn all the speech in your video into text. You can use this for transcriptions, captions, subtitles and for searching your video. It’s never been easier to find where something was spoken in your hours of footage.
Transcriptionists are about to go the way of the Scribe
The printing press brought about a revolution in how periodicals are printed. Artificial Intelligence (AI) is about to do the same for transcripts. Since the dawn of television, the transcribing problem has always been the same story. It's slow, it’s expensive and it’s prone to human error. AI is about to change that. How? By creating accurate transcripts quickly and economically.
Need to find where something was said in your video? Easily search the transcript and jump to that point in your video.
With Transcriptive being fully integrated into Adobe Premiere, you can now search the audio using text as if you were using Google. Whether it’s in a sequence or a clip, simply type in the word or phrase you’re looking for and find it. Don’t know where in the hour long video your talent said that funny quip? No worries, just use this amazing tool and search for the phrase. Perfect for working on documentaries, news or even that never ending cache of video clips! With Transcriptive, you can find it expeditiously.
Get Accurate Transcripts in Minutes for Pennies
Pressed on time? Transcribe 60 minutes of video in 10 minutes for under $4.00. It’s easy, fast and [best of all] economical. There are two great choices when it comes to having your speech converted to text.
First, Speechmatics has up to a 95% accuracy rate and works well even on noisy audio. It’s priced at $0.07 a minute. The second choice is Watson, which is a little less accurate but you get 1000 minutes (over 16 hours!) of FREE transcriptions per month. It’s only $0.02 a minute, after you go through your free minutes.
Captions, subtitles and more
The accuracy of Transcriptive makes it easy to create captions or subtitles to meet legal requirements or to enhance your search engine rankings via metadata on YouTube (or any other online platforms). There’s no longer an excuse for not having subtitles. The accuracy and the low cost makes it possible for any video editor or producer to easily add captions to all their projects.
Get Accurate Transcripts in Minutes for Pennies
Transcriptive's use of AI is truly a revolutionary way of getting transcripts
Transcribing 60 minutes of video in 10 minutes for less than $4.00. It’s fast, easy and economical. Artificial Intelligence (AI) and machine learning make it possible to analyze audio and quickly turn it into text. Furthermore, the AI has up to a 95% accuracy rating, and it’s truly an amazing solution to getting transcripts all within Premiere Pro.
Two AI services make getting accurate transcripts easy and economical
Choose which is the better fit for you between the two amazing AI speech services:
Speechmatics is about 95% accurate and even works well with audio with a lot of background noise. It costs $0.07 per min. Use Speechmatics if you want the highest accuracy and least amount of text cleanup.
Watson gives you 1000 minutes (over 16 hours!) free per month ($0.02 per minute after the 1000 min. are used up), but is less accurate than Speechmatics.
You sign up with the speech services directly. Transcriptive doesn’t add any fees, so you only pay the speech services listed price. You get the wholesale cost of the minutes, making Transcriptive the most cost effective solution for transcriptions available.
Search Your Video
Don’t scrub through hours of video looking for a single quote. The most powerful component of transcribing is searchability. Find a word or phrase in hours of video or caches of clips in seconds. Quickly jumping to the point in the video’s timeline when the words were spoken.
Sync up an existing transcript to your video, making it searchable
Already have a transcript or a script? Transcriptive has the ability to sync it up to your audio, adding timecode to every word. This makes searching the existing script, and then jumping right to the exact spot something was said in the video a breeze.
Export Captions as .srt, .stl, .vtt, .xml or to Premiere metadata
Creating captions is a common and important use of Transcriptive. Easily do this all within Premiere; or any other captioning program. Transcriptive exports file formats such as: SRT, STL, VTT, or SMPTE-TT (xml). As an added bonus, having subtitles help YouTube videos get better rankings on Google!
So you’ve got your transcript, either using Transcriptive’s AI or from a service, now what?
These are examples of a wide variety of features Transcriptive can do. Hopefully, it’ll give you an idea of how easy it is to use. This is not an exhaustive list of examples, it certainly works great on all types of footage, but we wanted to demonstrate how it works.
Search Your Timeline
You can search in the Transcriptive panel. Once you find what you’re looking for, just click on the word and you’ll jump to where it’s spoken on the timeline. There is no easier way of finding what you want in a video.
Search Individual Clips
If you batch process a bunch of clips, the entire transcript for each clip gets put into the metadata for that clip. You can then search each clip’s metadata to find exactly what you want. Need to find the right pull quote for tonight’s news? Just type it in and jump to it.
Fully Integrated into Premiere Pro
What makes Transcriptive so powerful is that everything works from a Premiere panel. This makes it easy to communicate with other parts of Premiere and use all of Premiere’s tools. There’s no extra application, you don’t have to pass around XML files to get the transcript into Premiere… Everything lives in Premiere.
Exporting Closed Captions
You can use SMPTE-TT (xml) to turn your transcript into Closed Captions. Premiere’s captioning tool supports this as do many other captioning tools. Support for .SCC is coming soon.
Exporting to YouTube
YouTube supports VTT or SRT files and Transcriptive can export out both. If you want to increase how highly your videos rank in Google, adding subtitles is an important way of doing it.
Transcriptive is currently solely on Adobe Premiere Pro CC 2015.3 [or newer], (Premiere Pro 13.0.2 or higher is recommended) and supported by both Windows and Mac Operating Systems.
With so many new versions of OSes and applications coming out sometimes we forget to add something. All of our products are kept up to date and run in the latest versions of the OS or Premiere Pro.
Jim says that your serial number should automatically be recognized. To try the demo, click here for a demo serial number or email email@example.com for an extended trial. The biggest new feature is the ability to merge transcripts. This allows you to work on different sequences and then combine them in a final edit to get the final transcript. It's something that's been requested for a while and it also helps with the bug on Premiere Mac. If you have a long transcript you can work on it in 30-40 minute chunks and then combine them. To also help work around that bug, the panel now monitors memory and will restart itself if memory gets low. You don't need to restart Premiere to solve the memory issue, just close and open the panel. So we try and do this automatically.
What's new in Transcriptive 1.5?
The biggest new feature is the ability to use the Transcriptive A.I. service. You'll need to create an account on Transcriptive.com to use it, but it's then accessible in the panel. This tends to be more accurate than Speechmatics, but it's also more expensive at .12/min. However, Transcriptive A.I.'s Align function is only .02/min.
We've removed Watson from the panel. The main reason we included Watson was because they offered 1000 minutes free per month. This discontinued that in December. That combined with their other constant changes which regularly broke the login process and made it inaccessible until we fixed it, made it the right decision to remove it. We realize that removes the lowest cost option but we hope to have some announcements after NAB that will reduce the cost of the other services.
Performance on long transcripts should be improved. We fixed some bugs and made a bunch of optimizations that should make overall usability much snappier.
A few important UI changes:
The Import button is now front and center
The Status Field has been replaced with a Notification menu
There's an Account menu to help manage your Transcriptive.com account.
The Jobs dialog shows you all your transcripts regardless of what service you used.
The Lock Transcriptive button (padlock) now changes color.
Of course, as mentioned we worked around the Adobe bug that is crashing Premiere and corrupting projects. For Transcriptive to work in 13.0.2 or 13.0.3 you MUST go to the Transcriptive Main Menu and select Update Project. We highly recommend you do this in 13.0.1 as large projects may crash Premiere before Project's metadata can be updated.
This completely addresses the bug on Windows.
Also, on Mac, there's another memory leak that can cause the machine to crash. This does not corrupt the project file, so it's not catastrophic but we realize it's still a pain in the ass. If you start to notice the Mac get sluggish, close Transcriptive immediately. Reopen Transcriptive, you don't need to quit Premiere, and you should be fine for a while. You can also keep an eye on the Activity Monitor and see if Memory is almost maxed out.
We are working on the Mac issue and hope to have some sort of fix soon.
What's new in Transcriptive 1.2.1?
The latest free update to the Transcriptive Premiere panel fixes a lot of bugs (without hopefully adding too many new ones), adds better support for Transcriptive.com, and adds a few more minor improvements.
The highlight tracking is much improved, so if you’re having issues with the highlight going off the screen it’s worth updating.
What's new in Transcriptive 1.2?
Support for human transcript import
If you import Plain Text files, we can detect timestamps with certain timecode formats and will use those to assign timecode to every word. So if you get a file from a transcription service with timestamps, we'll be able to use those to get timecode per word. It's only as accurate as the frequency of the timestamps. More timestamps, better accuracy.
Better SRT import: If you get an SRT file back from a transcription service, we can VERY accurately give every word timecode. The cool thing is that now there's a Merge Lines option in SRT Import. This will combine lines until it finds a period or other punctuation. So if your SRT file has punctuation but is broken up into 32 character lines, this allows very accurate import of human transcripts.
Search and Find & Replace now support "quotes". This means you can search for "teach" and not get returned 'teaching' or 'teacher', only "teach".
Support for UTF-8 and UTF-16 for non-English languages.
Transcriptive can now be locked onto a single sequence. So if you're moving between sequences but only one sequence has a transcript that you're editing, you can now Lock it. This will prevent Transcriptive from constantly trying to load transcripts for sequences you don't care about.
Proxy Server support: If you're on a network with a firewall and have a proxy server set up, we can now use it for sending files out for A.I. transcription.
What's new in Transcriptive 1.1.2?
Added VTT support for YouTube captions - import human created transcripts and assign timecode to every word just like the A.I. transcripts.
EBU-STL files are also now supported for Export.
Plus many bug fixes
What's new in Transcriptive 1.1?
New Power Search Beta Panel
What's new in Transcriptive 1.0.3?
You can now import VTT files from YouTube! YouTube has an extremely good Speech-to-Text algorithm and will automatically generate captions. For free. Just upload the video (you can keep it Private), wait for it to generate the captions (this takes a while), and then download the VTT file. You can import that into Transcriptive for editing and searching your video.
Visual indicator when you're in Edit mode. A blue line will appear around the paragraph you're editing. If you get distracted this will make it much easier to see what mode you're in.
Buttons for Split/Merge Paragraph. Very handy if you forget the keyboard shortcuts for this very critical feature.
Improved the messages in the Status Window.
Added Delete Speech Analysis to the Main Menu. This is helpful if you've made edits to the transcript and want to save those into the Metadata. Occasionally in batch a clip will get the wrong transcript and this will help solve that issue.
Drop Frame/Non Drop Frame: Previously we were not detecting Non Drop sequences. This was causing very long clips to have the timing wrong. There's an option in the Preference file that you can set if you're experiencing this. Unfortunately Premiere does not allow us to detect this automatically.
Undo. Premiere 2018 pretty much broke Undo and it's now working again. It's still possible to get it to misbehave but you have to be really trying.
Better tracking of which sequence is selected. It's no longer possible to edit a sequence's transcript if you've requested a transcript for a different sequence (and Transcriptive is still waiting for it to come back from the Speech Service). In some cases this was causing the transcript to end up in the wrong sequence overwriting that sequence's transcript.
Bunch of small bugs. Just continued maintenance and cleaning up some of the odds and ends.
Jim explains, "You can now download Transcriptive 1.1 and try our new Power Search beta panel! This makes it quick and easy to find the timecode for a phrase in a Premiere Pro project." Here's a tutorial to explain more.