How do you compress a long recording for transcription? Usually you do not have to.
You only need to compress a long recording when the file is larger than 5 GB, because one file can run up to 10 hours or 5 GB here, with or without a pack, and is transcribed at about 2 minutes per audio hour. When you do, convert it to mono AAC at 64 kbps (about 29 MB per hour, so even 10 hours lands near 290 MB). Packs start at $4 for 300 minutes, and credits never expire.
Smaller file, same words
When shrinking helps, and how far to go.
Check the size before anything else
Look at the file size in Finder, File Explorer or the Files app. Under 2 GB it uploads as it is, today. Between 2 and 5 GB it uploads once a pack is on the account. Only above 5 GB, or when the connection is slow, is compression worth the few minutes it takes.
The setting that is enough for speech
Speech recognition does not need stereo or studio bitrates. Mono AAC at 64 kbps keeps every word clear and costs about 29 MB per hour of audio, so a ten-hour day becomes a file of roughly 290 MB. Going lower than 48 kbps starts to blur quiet voices, so stop there.
One command does it
With ffmpeg installed, run ffmpeg -i input.mp4 -vn -ac 1 -c:a aac -b:a 64k output.m4a. The -vn drops the picture, -ac 1 folds the channels into one, and the result keeps the same length, so every timestamp in the transcript still matches the original video. Then upload it from the upload page.
What you get back
A timestamped transcript, a brief with the thesis and takeaways, chapters, the names, numbers and dates said, and a chat to ask the recording questions. Speakers are named in the summary, key quotes and chapters. Exports: TXT, DOCX, SRT, VTT and JSON.
Shrinking it wrong
Where compression goes too far.
✗ Over-processed
Squeezing a file harder than speech needs, or cutting it to meet a limit that is not there, costs words and time.
- Bitrates under 32 kbps that smear consonants
- Noise reduction that clips soft voices
- Speed-up tricks that change the clock
- Cutting a 4-hour file into 25 MB slices
- Re-encoding a file that already fits
- Losing the original after converting
✓ Just enough
Keep the clock, drop what speech does not need, and only when the size calls for it.
- No compression under 5 GB
- Mono 64 kbps when it is needed
- Original length kept, timestamps intact
- About 12 minutes for six hours
- The original file kept as a backup
- One upload, one transcript
Sample output
From a compressed 6-hour workshop file.
Illustrative lines, not a real recording. The compressed M4A kept the original length, so these times match the video.
Export
One transcript. Five clean formats.
Every pack exports all five — no format is held back for a higher tier.
Plain text
De-ummed paragraphs. Ready to paste.
SRT captions
Timed to the speech. Every video editor reads this.
WebVTT
HTML5 player + YouTube uploads.
Show notes
Formatted with chapters and pull-quotes.
Machine-readable
Timestamped text for your own tools.
Pricing
Honest pricing, no surprises.
A pack from $4 (300 minutes) opens the full transcript. After that, no subscription and no auto-renew — every tier is a one-time minute pack, and credits never expire. Buy once, use the minutes whenever you need them.
Casual pack
$4/300 min
5 hours of audio — $0.80 an hour. One-time — credits never expire.
- 300 minutes of audio
- All export formats
- AI chat over the whole transcript
- Executive summary & key quote
Best value
$18/3,000 min
50 hours of audio — $0.36 an hour. Credits never expire.
- 3,000 minutes of audio
- Summary, chapters and key quotes
- All export formats
- AI chat over the whole transcript
Bulk pack
$32/6,000 min
100 hours of audio — $0.32 an hour. Credits never expire.
- 6,000 minutes of audio
- Summary, chapters and key quotes
- API access for batch pipelines
- All export formats
FAQ
Compressing long recordings, answered.
Do I need to compress a long recording before transcribing it?
Only when it is larger than 5 GB, with or without a pack. A typical audio recording of several hours is far below either line. Long video is where size grows: a few hours of 1080p or 4K footage can pass 2 GB, and that is when extracting the audio pays off.
What bitrate should I use for speech?
Mono AAC or MP3 at 64 kbps is plenty for speech and comes to roughly 29 MB per hour. 128 kbps is about 58 MB per hour and gains nothing for transcription. Avoid going under 48 kbps, and skip aggressive noise reduction, which can remove quiet speech along with the hiss.
Is uncompressed WAV a problem?
Only by size. CD-quality stereo WAV is about 635 MB per hour, so a little under eight hours already passes 5 GB, and a bit over three hours passes 2 GB. Convert it to M4A or MP3 first, or upload it as it is when it is inside the limit; the words come out the same.
How long does a compressed 6-hour file take, and what does it cost?
About 12 minutes to process, plus the upload, which is short for a 170 MB file. Packs start at $4 for 300 minutes. Packs are one-time and credits never expire.
Will compressing change the timestamps?
No, as long as you only change the codec, bitrate or channels. Trimming, speeding up or removing silences would shift the clock. Keep the original recording, so you can return to the exact moment a timestamp points at.
I can't install ffmpeg. What else works?
Send a link instead of a file: YouTube, Vimeo, Twitch and Kick VODs, Google Drive share links and podcast feeds are fetched directly, so nothing is uploaded from your machine. Otherwise most audio editors export a mono M4A or MP3. Recordings are transcribed on our own servers and are never used for training.
Related
Related answers about file size.
Compress only if you must. Then about 2 minutes per audio hour.
Upload the recordingOperated by Neugence Technology Pvt. Ltd. · contact@whipscribe.com · Security · Privacy · Terms