A CLOSER LISTENOne song.
New possibilities.OpenSEP / Journal
New possibilities.OpenSEP / Journal
A better way to listen to your stems.
What to listen for after separation, and how to find the take that works for your project.
A clearer space for your audio
Make more of
what you hear.
Separate the sounds. Find the words.
Turn any recording into your next great piece of work.
YOUR SOUND, UNFOLDEDOne recording.
One recording.
New possibilities.
VoiceMusicEverything else
01 / SEPARATION
Hear it for yourself Interactive demo
Two voices. One transcript.Real speech · Unedited model output
EnglishNo speaker count to enter. OpenSEP finds the voices automatically.
00:00 / 00:212 detected
About this transcript
Four passages read by two narrators, taking turns. Speaker labels and wording come from the recorded OpenSEP engine run. Review important passages against the audio.
Recording sources & changes: Mini LibriSpeech · CC BY 4.0
Five voices. No headcount.Real speech · Unedited model output
EnglishNo speaker count to enter. OpenSEP finds the voices automatically.
00:00 / 01:095 detected
One track. Four overlap levels.5 recorded voices
Input arrangement · A–EMeet all five voices.
About this transcript
Five real readers in one constructed recording: solo turns, brief interruptions, longer overlaps, then several voices at once. Speaker count was not supplied to the engine. The transcript is its unedited output. This run detected 5 speaker labels. Review wording and speaker boundaries against the recording. Review important passages against the audio.
Recording sources & changes: Mini LibriSpeech · CC BY 4.0
Actual model outputs. Switch tracks without losing your place.
Less friction. More making.
Good tools get out of your way.
Everything you need to get from a recording
to something you can use.
Keep the part you need.
Pull vocals from a song. Bring dialogue to the front. Choose the result, and let us handle the separation.
Find your soundGive every word a place.
Turn speech into readable text with timestamps and speaker labels. Review, find a moment, and export.
Find your wordsYour work stays yours.
Files are used for your task, never for model training. Uploads are deleted after 1 day. Results stay available for 3 days.
A thoughtful defaultMade for your kind of work
Where will you take it?
Small tasks. Big projects.
A useful place to start.
01
FOR MUSICIANSHear the arrangement.
Hear the arrangement.
Make it your own.
Isolate a part for practice, a reference, or your next edit.
Voice, up front.
02FOR STORYTELLERSLet the story
Let the story
come through.
Separate dialogue from the sounds around it.
Interview notesEvery idea
starts with
a conversation.00:08 · Speaker 2
03starts with
a conversation.00:08 · Speaker 2
FOR CURIOUS MINDSListen once.
Listen once.
Come back to any word.
Give interviews, research, and voice notes a written home.
A simple rhythmStart with a file.
Start with a file.
Leave with more.
No timeline to learn. No settings to decipher.
Just a clear path to the result you need.
- 01
Choose what you want to make.
Separate audio or transcribe speech. The next step is always clear.
- 02
Drop in your recording.
We check the file and show your place in the queue.
- 03
Listen, review, and take it with you.
Preview your result, then download the files you need.
The OpenSEP journal
Visit the blog A few things worth sharing.
WORDS, IN ORDERGood stories
start with listening.OpenSEP / Journal
start with listening.OpenSEP / Journal
From a conversation to a working draft.
A practical workflow for turning an interview into a transcript you can actually use.
BUILT WITH CAREHere to help.
Then gone.OpenSEP / Journal
Then gone.OpenSEP / Journal
Your files should have a checkout time.
Why uploads and results are temporary, and what that means for your creative workflow.
A little clarity goes a long way
Start creating Bring your next idea to life.
A song, an interview, a voice memo. Start with what you have.