How to Add Subtitles to Recorded Zoom Meetings with AI

# AI Subtitle Workflow for Zoom Marketing Standup

## Workflow

1. **Prepare the file**: Confirm the MP4 is finalized and rename it descriptively (e.g., `Q4_Marketing_Standup_2024-11-05.mp4`). Move it to a dedicated project folder.

2. **Choose a transcription tool**: Upload the MP4 to an AI service that supports speaker diarization. Recommended options:
   - **Descript** (best for speaker labels + editing)
   - **Otter.ai** (fast, accurate for meetings)
   - **Rev AI** or **Happy Scribe** (high accuracy, paid)
   
   Enable **speaker identification**, set language to **English (US)**, and toggle **timestamps on**.

3. **Generate and label speakers**: After AI processing (usually 5–10 minutes), relabel generic tags ("Speaker 1") with real names (e.g., Maya, Jordan, Priya, Alex) based on the intro segment.

4. **Export as SRT**: Export captions in **.srt format** with settings: max 2 lines per caption, ~42 characters per line, 1–6 seconds per cue.

5. **Review and edit**: Open the SRT in a text editor or Descript/Subtitle Edit. Correct brand names, acronyms (CPC, ROAS, CTR), and fix overlapping timestamps.

6. **Burn subtitles into video**: Use **HandBrake** (free) or **FFmpeg** to hard-code captions. FFmpeg command:
   ```
   ffmpeg -i input.mp4 -vf "subtitles=captions.srt" -c:a copy output.mp4
   ```

7. **Quality-check the final MP4**: Play the exported file at 1.5× speed to confirm sync, legibility, and speaker label accuracy.

8. **Share with stakeholders**: Upload to your preferred platform (SharePoint, Google Drive, Vimeo) with a clear title noting "Captioned Version."

## Sample SRT Output

```
1
00:00:04,120 --> 00:00:07,480
[Maya]: Good morning, everyone. Let's kick
off the Q4 standup.

2
00:00:07,600 --> 00:00:11,250
[Maya]: Jordan, can you start with campaign
performance highlights?

3
00:00:11,800 --> 00:00:15,900
[Jordan]: Sure. Our Q4 paid social spend
is tracking 12% above forecast.

4
00:00:16,100 --> 00:00:20,300
[Jordan]: ROAS on the holiday creative set
came in at 4.2, which is strong.

5
00:00:20,800 --> 00:00:24,600
[Priya]: That's great. Should we reallocate
from display to social?

6
00:00:24,900 --> 00:00:29,100
[Alex]: I'd suggest shifting 15K before the
product launch on November 20th.

7
00:00:29,400 --> 00:00:33,200
[Maya]: Agreed. Let's finalize the budget
shift by Friday.
```

## Review Tips

- **Accuracy**: Focus proofreading on jargon, product names, metrics, and numbers—AI often misrenders "ROAS," "CTR," or dollar figures. Listen at 1× speed for dense sections.
- **Speaker labels**: Verify the first 2–3 lines per speaker; diarization errors usually cluster when participants interrupt or talk over each other. Merge misattributed segments manually.
- **Timing**: Ensure no caption exceeds 6 seconds on screen; split long sentences. Avoid gaps under 100ms between cues.
- **Readability**: Keep lines ≤42 characters; use sentence case, not ALL CAPS. Position subtitles in the lower-third safe zone.
- **Export settings**: Use H.264 codec, 1080p, bitrate 8–10 Mbps for crisp text. Test playback on mobile and desktop before sharing—stakeholders on mute depend entirely on caption legibility.
- **Archive**: Keep both the original MP4 and the SRT file separately in case edits are needed later.

# Video Captioning Workflow for Zoom Meeting Recording

## Workflow

1. **Extract and Prepare Audio**
   - Upload your MP4 file to an AI captioning platform (Descript, Rev, or Kapwing). These tools automatically process video files and generate timestamped subtitles with speaker detection.
   - Alternatively, use free open-source tools like Whisper (OpenAI) for audio-to-text conversion if budget is limited.

2. **Generate Initial Subtitles with Speaker Labels**
   - Select English language and enable speaker identification in your chosen tool's settings.
   - Process the 45-minute recording (typically takes 5-15 minutes depending on platform).
   - The AI will automatically segment dialogue and attempt to label speakers (Speaker 1, Speaker 2, etc.).

3. **Manual Speaker Identification**
   - Review the generated subtitles and identify speakers by name (e.g., "Sarah," "Mike," "Jennifer," "David").
   - Use the Zoom participant list or meeting notes to assign correct names to speaker labels.
   - Update the SRT file to replace generic labels with actual names for clarity.

4. **Edit for Accuracy and Timing**
   - Review the transcript line-by-line for AI errors, especially industry jargon ("Q4 campaign," "budget reallocation," "product launch timelines").
   - Adjust timestamps if dialogue overlaps or timing feels off.
   - Break long sentences into 2-3 readable subtitle chunks (aim for 42 characters per line maximum).

5. **Quality Check and Formatting**
   - Verify speaker labels are consistent throughout.
   - Ensure no subtitles remain on-screen longer than 7 seconds.
   - Check that technical terms related to marketing metrics are spelled correctly.

6. **Burn Subtitles into Video**
   - Export your corrected SRT file from the editing tool.
   - Use FFmpeg, Handbrake, or your platform's built-in burn feature to embed subtitles permanently into the MP4.
   - Command example: `ffmpeg -i recording.mp4 -vf subtitles=captions.srt output.mp4`

7. **Final Export and Delivery**
   - Test the output by playing it on mute to confirm readability and speaker identification.
   - Export at original resolution and share with stakeholders via secure file transfer or cloud storage.

---

## Sample SRT Output

```
1
00:02:15,000 --> 00:02:22,500
Sarah: Alright team, let's kick off with Q4 campaign
performance. We saw a 23% increase in engagement.

2
00:02:23,000 --> 00:02:29,800
Mike: That's solid. Did we hit our conversion targets
for the regional campaigns?

3
00:02:30,200 --> 00:02:38,600
Sarah: Yes, three of four regions exceeded targets.
The Southeast campaign needs optimization.

4
00:02:39,100 --> 00:02:46,900
Jennifer: Moving to budget reallocation—I recommend
shifting $50K from underperforming channels.

5
00:02:47,400 --> 00:02:55,300
David: Agreed. Should we reallocate toward the product
launch timeline? We're 8 weeks out.
```

---

## Review Tips

- **Accuracy**: Playback at 0.75x speed to catch rapid speech or overlapping dialogue.
- **Speaker Labels**: Cross-reference with Zoom's "Participants" panel if recording metadata includes attendee names.
- **Technical Terms**: Create a glossary of marketing-specific terminology beforehand and search-replace errors.
- **Timing Gaps**: Ensure subtitles appear 1-2 seconds before speech starts for better comprehension on mute.
- **Testing**: Watch 5-10 minutes of the final burnt video on a mobile device to simulate stakeholder viewing conditions.

## Workflow

1. **Prepare the file**
   - Confirm the Zoom recording is a local **MP4** and that audio is clear.
   - Rename it clearly, e.g., `Marketing_Standup_Q4_2026-04-21.mp4`.

2. **Choose an AI transcription tool**
   - Good options: **Descript**, **Adobe Premiere Pro Speech to Text**, **CapCut Desktop**, or **Whisper-based tools** like MacWhisper.
   - Best fit for speaker labels: **Descript** or **Premiere Pro** with speaker detection/diarization.

3. **Generate the transcript**
   - Import the MP4.
   - Set language to **English**.
   - Enable **speaker identification** if available.
   - Use subtitle settings such as:
     - **Max 2 lines per caption**
     - **32–42 characters per line**
     - **Caption length: 1–6 seconds**

4. **Edit the transcript**
   - Correct key terms such as **Q4**, campaign names, product names, and budget figures.
   - Replace generic labels like “Speaker 1” with real labels, e.g., **Sarah:**, **James:**.
   - Split long captions so they read naturally.

5. **Export subtitle file**
   - Export as **SRT** first for easy review and reuse.
   - Make sure timestamps match the spoken audio and speaker labels are included in the subtitle text.

6. **Burn subtitles into the video**
   - In Premiere Pro, CapCut, Descript, or HandBrake + subtitle workflow, import the edited SRT.
   - Style for mute viewing:
     - Sans-serif font
     - White text with black shadow/box
     - Bottom-center placement
     - Medium font size for mobile and desktop

7. **Export final deliverables**
   - Export:
     - **Captioned MP4** with burned-in subtitles
     - **SRT file** as a separate deliverable
   - Recommended export: **H.264 MP4, 1080p, AAC audio**.

---

## Sample SRT Output

```srt
1
00:00:03,000 --> 00:00:07,200
Sarah: Thanks, everyone. Let's start with
a quick review of Q4 campaign performance.

2
00:00:07,300 --> 00:00:12,000
James: Paid social exceeded target by 18%,
but search was under plan for the month.

3
00:00:12,100 --> 00:00:17,000
Mina: Based on that, I recommend shifting
part of the search budget into retargeting.

4
00:00:17,100 --> 00:00:22,400
Alex: How much reallocation are we talking
about before the product launch window?

5
00:00:22,500 --> 00:00:28,000
Mina: Roughly fifteen percent this week,
so we can support launch traffic in time.

6
00:00:28,100 --> 00:00:33,500
Sarah: Okay, let's confirm timelines.
Creative needs final approval by next Tuesday.
```

---

## Review Tips

- **Check speaker labels carefully**: AI diarization often swaps speakers after interruptions or overlapping talk.
- **Verify business terms**: Review names, acronyms, budget numbers, and launch dates manually.
- **Watch for reading speed**: Keep captions short enough to read on mute; avoid dense 3-line subtitles.
- **Fix timing**: Captions should appear when speech starts and clear before the next speaker begins.
- **Test the burned-in export**: Watch on both desktop and phone to confirm font size, contrast, and bottom placement.
- **Keep both versions**: Share the burned-in MP4 with stakeholders and save the SRT for future edits or platform uploads.

Of course. Here is a complete workflow, sample output, and review guide for adding AI-generated subtitles with speaker labels to your Zoom recording.

### Workflow

1.  **Select an AI Transcription Tool:** For your needs, a tool like **Descript** is highly recommended. It offers transcription, speaker identification, an intuitive editor, and the ability to export video with burned-in captions. Upload your 45-minute MP4 file directly to a new project.

2.  **Generate the AI Transcript:** Once uploaded, instruct the software to transcribe the audio. Select "English" and specify the number of speakers (4). The AI will process the file and generate a time-stamped transcript.

3.  **Assign Speaker Labels:** The AI will likely assign generic labels (e.g., Speaker 1, Speaker 2). Go through the transcript and replace these with the actual names of your team members (e.g., Sarah, David, Maria, Ben). Descript's speaker assignment feature makes this simple.

4.  **Review and Correct the Transcript:** Read through the entire transcript while listening to the audio. Correct any errors in punctuation, grammar, or terminology. Pay close attention to marketing-specific jargon, campaign names, and financial figures, as AI can often misinterpret these.

5.  **Style Your Captions:** Since you are burning the subtitles into the video, customize their appearance for readability. Choose a clear font (like Arial or Helvetica), set an appropriate size, and add a subtle background or outline to ensure the text is visible against any video background.

6.  **Export the Final Video:** Once you are satisfied with the accuracy and styling, use the tool's export function. Select "Video" as the export format and ensure the "Burned-in captions" option is enabled. Export the file at a suitable quality (e.g., 1080p).

### Sample SRT Output

Here is a sample excerpt in the standard SRT format, reflecting your meeting's content.

```srt
1
00:15:22,150 --> 00:15:25,880
Sarah: Okay team, let's look at the Q4 campaign performance. Our

### Workflow
1. Upload MP4 to VEED.io or Kapwing (free AI subtitle tools supporting speaker detection).
2. Select "Auto-transcribe" with English language; enable "Speaker labels" (e.g., Speaker 1-4); set accuracy to "High."
3. Review/edit transcript: Assign names (e.g., Sarah, Mike) to speakers, fix timing errors, add punctuation.
4. Export as SRT file; download.
5. Import MP4 and SRT into HandBrake (free): Add subtitle track, select "Burn In" under Subtitles tab.
6. Adjust font/size (e.g., white Arial 24pt, bottom position); encode to MP4 (H.264, 1080p).
7. Preview full video; re-encode if needed.

### Sample SRT Output
```
1
00:05:12,000 --> 00:05:18,500
[Sarah: Marketing Lead] Q4 campaigns exceeded targets by 15% overall.

2
00:05:18,700 --> 00:05:25,200
[Mike: Campaign Manager] Email open rates were strong at 28%, but social ads underperformed.

3
00:05:25,400 --> 00:05:32,100
[Lisa: Budget Analyst] We have $50K unspent; recommend reallocating to product launch promo.

4
00:05:32,300 --> 00:05:38,900
[Tom: Product Coordinator] Launch timelines: Beta next week, full rollout by Dec 15th.
```

### Review Tips
- **Accuracy**: Play video synced with SRT; check 80%+ word match, fix homophones (e.g., "Q4" vs. "cue for").
- **Speaker Labels**: Verify diarization (VEED auto-labels); manually swap if overlaps; use consistent format like [Name: Role].
- **Export**: Test on mute; ensure subtitles don't overlap speech; use FFmpeg CLI for batch: `ffmpeg -i input.mp4 -vf subtitles=srtfile.srt output.mp4`.
(248 words)