← Back

Guide

How to get your session recording into Scrivorn

Whether you play in person or online, here is exactly how to capture your session audio and upload it for transcription.

In-Person Sessions
Record on the phone already sitting on your table, then upload the file
You already own the equipment. Every phone ships with a voice recorder that handles a four-hour session and saves to a file. Record on that, upload the file afterwards.
Steps
1
Open your phone's voice recorder
On iPhone it is Voice Memos. On Android it is usually Recorder or Voice Recorder. Both save a file you can send to yourself when you are done.
💡 Do a ten-second test recording and play it back before session one. Discovering the mic was covered is much cheaper at the start of the night than at the end of it.
2
Put the phone in the middle of the table, face up
The mic is usually on the bottom edge, so keep that clear, not under a character sheet, not in a dice tray. The closer it sits to the people talking, the better the transcript.
💡 Plug the phone in. Four hours of recording with the screen waking up will flatten a battery.
3
Stop when the session ends, then send the file to your computer
AirDrop, email it to yourself, or drop it in Google Drive or Dropbox. Voice Memos exports .m4a; Android recorders usually export .m4a or .mp3. Scrivorn takes all of them.
4
New Session › Upload Audio
Give the session a title, drag the file in, and leave it. A four-hour session takes about 5-8 minutes to transcribe, and you can close the tab while it works. Scrivorn picks it back up when you return.
Tips for better recordings
Reduce background noise
Turn off music, close windows, put phones on silent. Even small sounds accumulate over 4 hours.
Encourage one person at a time
Crosstalk is the hardest thing for speaker labeling to handle. Brief pauses between speakers dramatically improve accuracy.
Say names at the start
Have everyone say their name at the beginning of the recording. This helps Scrivorn label speakers correctly throughout.
External mic is worth it
A $30 USB desktop mic placed in the center of the table is a significant upgrade over a built-in laptop mic for large groups.
Discord
Use the Craig bot to record your voice channel, then upload to Scrivorn
Why Craig? Discord does not have built-in recording. Craig is a free, widely-trusted bot that records each speaker on a separate audio track, which gives Scrivorn better speaker identification than a single mixed file.
Setting up Craig (one time)
1
Add Craig to your Discord server
Go to craig.chat and click "Add Craig to Discord." You need Manage Server permissions.
2
Test it before game night
Join a voice channel and type :craig:, join in any text channel. Craig will join and confirm it's recording.
Recording each session
1
Start recording when your session begins
Type :craig:, join in your text channel. Craig joins the voice channel and starts recording.
2
Stop recording when done
Type :craig:, leave or kick Craig from the channel. You will get a DM with a download link.
3
Download the mixed file
Open the download link Craig sent you. Choose Download as FLAC (mixed) or Download as MP3 (mixed) for a single combined file.
💡 The "mixed" file combines all speakers into one track. This is what Scrivorn needs. The individual tracks are for advanced audio editing.
4
Upload to Scrivorn
In Scrivorn, create a new session, choose "Upload File," and drag in the file Craig gave you.
Craig download links expire after 7 days. Upload to Scrivorn before then or you will lose the recording.
Zoom
Record your Zoom call locally and upload the audio to Scrivorn
Free Zoom accounts can record locally. You do not need a paid Zoom subscription to record to your computer. Cloud recording requires a paid plan but local recording is free.
Before your session
1
Enable local recording in Zoom settings
Open Zoom > Settings > Recording. Make sure "Local Recording" is turned on. Set the save location to somewhere easy to find like your Desktop.
During your session
2
Start recording when the session begins
Press Alt+R (Windows) or Cmd+Shift+R (Mac), or click the Record button in the toolbar. A red dot confirms you are recording.
💡 Only the host can record by default. If you are not the host, ask the host to give you recording permission.
3
End the meeting when done
When you end the Zoom call, it will automatically convert and save the recording. This can take a few minutes for long sessions.
Uploading to Scrivorn
4
Find your recording
Zoom saves recordings to Documents/Zoom/[date] by default. Take the file called audio_only.m4a, Zoom writes it alongside the video for you, and it is the one Scrivorn wants.
💡 If only the .mp4 is there, cloud recordings hide the audio version behind an "Audio only" download on the Zoom web portal. Failing that, see I have a video file.
5
Upload to Scrivorn
Create a new session in Scrivorn, choose "Upload Audio," and drag in the m4a.
Supported file types from Zoom
.m4a .mp3
Google Meet
Record via Google Meet or use a browser extension as a workaround
Google Meet's built-in recording requires a paid Google Workspace account. Free personal Gmail accounts cannot record directly. Use the workaround below if you do not have Workspace.
Option A: Google Workspace recording (paid)
1
Start recording in Meet
Click the three-dot menu in the bottom right > Record meeting. Everyone in the call will see a notification that recording has started.
2
Stop and find your recording
Click the three-dot menu > Stop recording. The file saves to Google Drive automatically in a folder called Meet Recordings, and it will be an MP4.
💡 Meet only gives you video, so pull the soundtrack out before uploading, two minutes in VLC or QuickTime. Option B below skips this by recording audio in the first place.
Option B: Free workaround using OBS
1
Download OBS Studio (free)
Go to obsproject.com and download OBS Studio. It is free and open source.
2
Add an Audio Capture source
In OBS, click the + under Sources > Audio Output Capture. This captures all audio playing through your computer including your Meet call.
💡 You may also want to add a Mic/Auxiliary input so your own voice is included in the recording.
3
Set the recording format to audio-only first
In Settings › Output › Recording, set the Recording Format to MP3 or M4A. OBS then writes sound and no picture, which is all Scrivorn needs and a fraction of the size.
💡 Worth doing once, before your first session. If you have already recorded an MKV or MP4, extract the audio instead.
4
Hit Start Recording when your session begins
OBS saves to your Videos folder by default. When the session ends, drag that file into Scrivorn's Upload Audio option.
Microsoft Teams
Record your Teams meeting and download from OneDrive or Stream
Teams recording availability depends on your plan. Microsoft 365 Personal and Family plans include meeting recording. Free Teams accounts have limited recording options.
1
Start recording in Teams
Click the three-dot menu (...) in the meeting toolbar > Start recording. A banner confirms everyone that recording is active.
2
Stop recording and find your file
Click the three-dot menu > Stop recording. The recording processes and saves to OneDrive or SharePoint automatically.
3
Download as MP4
Find the recording in OneDrive under Recordings folder. Click the three-dot menu next to it and choose Download. It downloads as an MP4.
4
Extract the audio, then upload
Teams only offers video, so pull the soundtrack out first, two minutes in VLC or QuickTime. Then create a new session in Scrivorn, choose "Upload Audio," and drag in the result.
Pulling the audio out of a video file
Zoom, Teams, Meet and OBS all hand you an MP4. Here is how to get the soundtrack out of it.
Scrivorn does not accept video files. Nothing in a transcript comes from the picture, and the picture is roughly 95% of the file, a four-hour MP4 is several gigabytes that would take an hour to upload and be thrown away on arrival. Extracting the audio first takes a couple of minutes and produces a file about the size of an album.
Option A: your platform may have already done it

Check the recording folder before converting anything. Zoom writes an audio_only.m4a alongside the MP4 for local recordings, and Craig on Discord gives you audio only in the first place. If a matching audio file is already sitting there, you are done. Upload that.

Option B: VLC (free, Windows and Mac)
1
Install VLC and open Convert / Save
Get it from videolan.org. In the menu bar choose Media › Convert / Save on Windows, or File › Convert / Stream on Mac.
2
Add your video, then pick an audio-only profile
Add the MP4, click Convert, and set the profile to Audio - MP3. The word "Audio" in the profile name is the part that matters, it tells VLC to drop the video track entirely rather than shrink it.
3
Choose where to save it and hit Start
Give the output file a name ending in .mp3. A four-hour session takes a few minutes to convert and lands somewhere around 100-200MB.
💡 If the result is still over a gigabyte, the video track came along by mistake, go back and check the profile started with the word "Audio".
Option C: Audacity, if you already have it

Audacity can import most video files with the FFmpeg add-on installed, then File › Export › Export as MP3. It is the slower route, but if Audacity is already on your machine for editing out the pizza break, you may as well use it for both.

Option D: QuickTime, on a Mac

Open the video in QuickTime Player, then File › Export As › Audio Only. It produces an .m4a in one step with nothing to install. It only handles formats QuickTime can already open, so MKV files from OBS need VLC instead.

You do not need high quality audio. Transcription cares about clarity, not fidelity. A 64kbps mono MP3 transcribes exactly as well as a lossless file and uploads in a fraction of the time, so take the smaller option wherever one is offered.
Already have a transcript or notes
Paste any text directly into Scrivorn, no audio needed
This is often the fastest option. Scrivorn works with any text, raw notes, rough bullet points, a Discord chat log, or a formal transcript. It does not have to be perfect.
What you can paste
Raw DM notes
Even rough bullet points work. "Party arrived at Beastheaven. Ran into Ares in an alley. He mentioned Asmodeus." Scrivorn turns that into a full recap.
Discord chat logs
Copy your session's text channel or use a bot like Dyno to export your chat history and paste it directly.
Otter.ai or other transcripts
If you already use a transcription tool, export the transcript as text and paste it here. Scrivorn will handle the rest.
Post-session recap emails
Some DMs write a quick recap email to their players after each session. Paste that in and Scrivorn will structure it properly.
How to paste
1
Create a new session in Scrivorn
Open your campaign, click New Session, and enter a title.
2
Choose "Paste Transcript"
Select the Paste option and drop your text into the box. Any length works, a few bullet points or a 10,000 word transcript.
3
Hit Process with AI
Scrivorn reads your text and returns a narrative recap, bullet summary, extracted quests, NPCs, and locations within about 30 seconds.
Supported formats

Scrivorn accepts audio files up to six hours long:

.mp3 .m4a .wav .webm .ogg .flac .aac

Video files are not accepted, see I have a video file for how to pull the audio out of an MP4 in a couple of minutes. Sessions longer than six hours should be split in two and uploaded as two sessions; Audacity does this for free.