Before You Set Up Your New Computer, De-Bloat It

I recently replaced my laptop, and as I was getting ready to set up the new computer, I realized that I was about to do what I always do: immediately start installing everything I use for work.

For me, that can turn into quite a list. There’s Microsoft Office, Trados Studio, Dropbox, browsers, dictionaries, utilities, communication tools and all the other software that has gradually become part of my everyday setup.

This time, however, I decided to do something different and clean up the computer before I started adding my own software to it.

In other words, I decided to de-bloat it first.

What is bloatware, anyway?

Bloatware is one of those terms that gets thrown around a lot, but I think the important thing to remember is that not everything that comes preinstalled on a computer and that you don't recognize is necessarily bloatware.

Manufacturers add their own applications and utilities to Windows, and some of those can actually be useful. Others are trials, promotional software, games, duplicate utilities or applications that you simply don't need.

This became especially obvious to me because my new computer is a Lenovo, while my old laptop, the one I'm replacing, is a Samsung. The software that comes with one isn't necessarily the software that comes with the other, so a list of "programs you should uninstall from a new computer" isn't necessarily going to be very helpful, and this is where I found AI particularly useful.

Let AI help you figure out what you're looking at

Instead of searching for every unfamiliar application individually, you can give an AI chatbot the list of applications installed on your computer and ask it to help you sort them out.

In Windows 11, go to Settings > Apps > Installed apps. You can then give ChatGPT, Claude, Gemini, Copilot or another AI assistant the names of the applications you don't recognize and explain exactly what computer you have.

For example, I could say:

I have a new Lenovo Yoga 83TE running Windows 11. These applications came preinstalled on it. Help me identify which ones are Windows components, which ones are Lenovo utilities that may control hardware or provide driver and firmware updates, which ones are optional third-party applications, and which ones are likely bloatware that I can safely uninstall. Don't recommend removing anything unless you can explain what it does and what functionality I might lose.

That last sentence is important. I don't want AI simply telling me, "Yes, uninstall that." I want to know what the program does, whether it is tied to my hardware and what I might lose by removing it.

For example, Lenovo Vantage might look like just another manufacturer application if you've never used a Lenovo computer, but it can be used for things such as system and driver updates, hardware settings, battery settings and diagnostics. That's very different from a 30-day trial of an application I already use another program for.

If I ran the same exercise on my Samsung Galaxy Book4, I would change the prompt accordingly because I would expect a different set of manufacturer-specific software.

You can also take screenshots of your Installed apps list and give those to an AI tool that can analyze images. That's probably the easiest method if you have a long list and don't feel like typing or copying all those application names.

I would still verify anything the AI identifies as safe to remove, particularly manufacturer utilities and anything involving drivers, firmware, security, the touchpad, keyboard, display, audio, battery or other hardware. AI is very useful here as a research assistant, but I wouldn't give it the final say over what gets deleted from a brand-new computer.

In my Lenovo Yoga, all the inking apps are candidates for removal, as I don't plan on using a pen, and I wouldn't lose other functionality.

Don't forget about startup apps

After looking at what was installed, I also checked what was configured to start automatically with Windows.

You can do this under Settings > Apps > Startup, or through Task Manager. I actually like Task Manager for this because Windows also gives you information about the startup impact of each application.

Again, this is a good place to use AI if you don't recognize something. Take a screenshot or provide the list and ask which applications actually need to start with Windows and which ones can simply be opened when you need them.

There is an important difference here: disabling an application at startup does not uninstall it. It simply prevents it from launching automatically every time you sign in, which makes this a much less drastic place to start if you're not sure whether you need something.

Then I can start installing my own software

Once I've gone through what came with the computer, then I can start installing the applications I actually use.

I like this order much better because otherwise I'm putting my entire collection of software on top of another collection of software that I haven't even looked at yet. By the time I'm done, I have no idea which things came with the computer, which ones I installed myself and which ones are running in the background for no particularly good reason.

There's another advantage to doing this when setting up a new computer: it makes me look critically at my own software too.

Do I still need every utility I had installed on my previous computer? Am I installing something because I actually use it or simply because it has been on every computer I've owned for the past ten years?

Apparently, manufacturers aren't the only ones capable of creating bloatware. We can do a pretty good job of it ourselves.

But what if you've already set up your computer?

This was one of the questions I had when I started looking into all of this. What if you've already spent hours installing your software, connecting Dropbox, setting up Office, configuring Trados and getting everything just the way you like it? Do you really have to wipe the computer and start again?

Fortunately, no.

You can do essentially the same cleanup after the fact. Go to Settings > Apps > Installed apps, review what's there and research the applications you don't recognize. Then check your startup apps and disable anything that doesn't need to run automatically.

This is another situation where AI can make the process considerably easier. Tell it the make and exact model of your computer, give it your list or screenshots, and ask it to categorize what it sees. If you've already installed your own applications, tell it that too so it doesn't assume everything on the list came from the manufacturer.

You could even ask it to create three categories for you: keep, probably safe to remove, and research further. I would much rather see something unfamiliar land in "research further" than have an AI tool confidently tell me to remove a utility that turns out to control some feature of my laptop.

There is usually no reason to uninstall all of your own programs first, and a factory reset seems like overkill if the only problem is that you didn't de-bloat the computer before setting it up. A reset or clean installation makes more sense if you're having significant software problems or genuinely want to start from scratch, not simply because you discovered a few unnecessary programs after the fact.

In fact, doing this a little later may have an advantage because by then you have a better idea of what you're actually using.

My new new-computer routine

I used to think that setting up a new computer meant reproducing my old computer as closely as possible. Now I'm thinking it's a good opportunity to be a little more selective about both what the manufacturer has put on the machine and what I put back on it myself.

So, the next time I set up a computer, I'll first see what's already installed, use AI to help me investigate anything I don't recognize, check what is starting automatically and remove only the things I'm confident I don't need. Then I'll start installing my own software.

And if I forget and install everything first, at least now I know I don't have to start over.

Who Let the Bots In? How to Keep AI Notetakers Out of Your Zoom Meetings

AI notetaking bots like Otter.ai, Read.ai, Fireflies.ai, and Fathom can join your Zoom meetings automatically, often without your knowledge. They do this by connecting to a participant's calendar, scraping the meeting link, and auto-joining as a separate account. Once in, they transcribe the audio in real time and store it on a third-party server, under that service's own terms of service and privacy policy.

In plain terms: allowing a participant to bring one of these bots into your meeting is the same as allowing them to record it, except the recording goes directly to someone else's platform, outside of your control. To make matters worse, meeting hosts often won't even know who the bot belongs to or which participant's account it is linked to. The bot simply sits there in the participant list, as a separate attendee.

For meetings that involve confidential, legally privileged, or otherwise sensitive information, that is a serious problem. And because no single setting in Zoom blocks every AI transcriber, a layered approach is the only reliable solution to keep them out of the room.

Here is how to set it up.

Step 1: Block third-party domains

This prevents AI assistant accounts from joining at the account level, before they ever reach your meeting.

1. Sign in to your Zoom Web Portal.
2. Go to Settings > Meeting > Security.
3. Turn on the toggle labeled "Block users in specific domains from joining meetings and webinars."
4. Click the pencil icon and add the domains you want to block, separated by commas. Known AI notetaker domains include: otter.ai, read.ai, fireflies.ai, fathom.video.
5. Click Save.

This list requires manual maintenance. New AI transcription tools appear regularly, so review it periodically and add domains as needed.

Step 2: Disable third-party apps

Domain blocking alone is not enough. A participant can still invite a bot through the Zoom App Marketplace. To close that door:

1. Sign in to the Zoom App Marketplace as an administrator.
2. Click Manage in the upper right corner, then select Admin App Management > Apps on Account.
3. Find any third-party transcription apps and click Manage App to disable or remove them.
4. To enforce this organization-wide, uncheck the option that allows users to install unapproved apps.

Step 3: Enforce the Waiting Room

If a bot slips through, for example, because it joined from a calendar invite before your domain block was in place, the Waiting Room is your safety net.

1. Enable the Waiting Room in your meeting security settings.
2. Before admitting anyone, review the waiting room list.
3. Look for names like "Otter.ai Meeting Assistant" or "Read.ai Bot," or any participant that does not correspond to an actual person you are expecting.
4. Do not admit them.

Step 4: Remove bots already in the meeting

If a bot has already joined, you can remove it manually.

1. Open the Participants panel.
2. Find the bot's name, click the More button (the three dots) next to it, and click Remove.

One critical caveat: once a bot has been in the meeting, anything it already transcribed may not be retrievable or deletable from that service's servers. That is why the Waiting Room matters so much. Removing a bot after 20 minutes of sensitive discussion does not undo those 20 minutes.

When AI transcription is fine and when it is not

These tools exist for legitimate reasons. Internal team meetings, study sessions, training calls: in many contexts a transcript is useful and uncontroversial. The issue is not the technology itself but the lack of visibility and consent. A bot that joins without announcement is a problem not just for privacy but for basic trust.

As a meeting participant, it also makes sense to scan the list of participants to identify any bots and report them to the host as needed for removal when appropriate, as they are easy to miss in very large meetings.

The practical rule is simple: everyone in the meeting should know a bot is present, what it records, and where that data goes. If that transparency is not there, the controls above exist for exactly that reason. 

Add Auto Copy to Your Triple Clicks

If you're thinking "Triple clicks? Auto copy? What's that?", then this post is definitely for you, and I hope by the end of it you'll have a new favorite habit and a useful trick to help your work smarter.

Triple-clicking is one of those small, simple tricks that can make a surprisingly big difference in your daily workflow. It's been around forever, and yet many people never use it.

What Is Triple-Clicking?

Simply put: clicking the left mouse button three times in quick succession selects the entire paragraph (or segment, or line, depending on the program). That's it. Three clicks, whole paragraph selected. No need to click and drag, no need to reach for the keyboard.

It works natively in most text environments you probably already use every day: Microsoft Word, Google Docs, Trados Studio, PDF readers like Adobe Acrobat, and most web browsers. Try it right now in any Word document: click three times rapidly on a paragraph and watch the whole thing light up. Compare that to the old click-hold-drag approach, especially when a paragraph is long and spans multiple lines. Triple-clicking is faster, more precise, and eliminates the risk of accidentally missing the first or last character of a selection.

Taking It To The Next Level: Triple Click + Auto Copy

Once you get used to triple-clicking, the natural next question is: what if selecting the paragraph also copied it to the clipboard automatically? That single extra step, pressing Ctrl+C after selecting, is small, but it interrupts your flow when you're doing it dozens of times a day.

This is where AutoHotkey comes in. AutoHotkey (AHK) is a free, open-source scripting tool for Windows that lets you automate keyboard and mouse actions. The script below intercepts your triple click and adds one thing: it automatically copies the selected text to the clipboard. Select and copy in a single gesture.

; AutoHotkey v2 ; Triple Click to Select + Copy to Clipboard ; Shows tooltip after copying #Requires AutoHotkey v2.0 clickCount := 0 ~LButton Up:: { global clickCount clickCount++ if (clickCount = 3) { Sleep 120 ; allow triple-click selection to finish A_Clipboard := "" Send "^c" if ClipWait(1) { ToolTip "Selection copied to clipboard" SetTimer RemoveTooltip, -1500 } clickCount := 0 } else { SetTimer ResetClickCount, -500 } } ResetClickCount() { global clickCount clickCount := 0 } RemoveTooltip() { ToolTip }

The script monitors left mouse button clicks. When it detects that a third click in quick succession after the previous one, it sends Ctrl+C , copying whatever text was just selected by the triple click. The result: one fluid mouse action that both selects and copies.

How to Run the Script

1. Download AutoHotkey V2 from autohotkey.com and install it. The free version is all you need.
2. Download the script from here, OR open Notepad, paste the script above, and save the file with an .ahk extension, such as, TripleClickCopy.ahk. In the Save As dialog, set "Save as type" to "All Files" so the file doesn't get saved as .ahk.txt by mistake.
3. Double-click the file to run it. A small green icon will appear in your system tray (bottom-right corner of your taskbar) to confirm it's active.
4. Triple-click on any paragraph in any program. The text will be selected (provided the program supports triple-clicking) and copied to your clipboard automatically.

To stop the script, right-click the green tray icon and choose Exit. To pause it temporarily without closing it, choose Suspend Hotkeys from the same menu.

If you want the script to run automatically every time you start Windows, press Win+R, type shell:startup, and press Enter. Copy your .ahk file (or a shortcut to it) into that folder. From that point on, it will run silently in the background whenever you start your computer.

Final Thoughts

Triple-clicking is one of those tiny habits that, once you pick it up, becomes completely invisible: you just do it automatically and your work feels a little smoother. Adding auto-copy on top of it with AutoHotkey takes it one step further. Give the script a try and let me know how it works for you.

And if you've found other creative uses for triple-clicking or AutoHotkey in your translation or interpreting workflow, I'd love to hear about them in the comments!

From Screen to Clipboard: The Power of Win+Shift+T

 

If you've ever needed to copy non-selectable text, for example from an error message in a dialog box, text embedded in an image, a scanned PDF, and had to end up typing it out by hand, this post is for you.

Windows 11's Snipping Tool has a built-in shortcut that solves this problem in seconds: Win+Shift+T.

What It Does

Pressing Win+Shift+T launches the Snipping Tool directly into text extraction mode. Here's the full sequence:

1. Press Win+Shift+T
2. Your screen dims and a crosshair cursor appears
3. Click and drag to select any area of the screen that contains text
4. The tool reads and extracts the text using OCR (Optical Character Recognition)
5. The extracted text is instantly copied to your clipboard

Then just paste wherever you need it, be it a document, a browser search bar, a CAT tool, anywhere.

How Is This Different from Win+Shift+S?

You may already use Win+Shift+S, the standard Snipping Tool screenshot shortcut. The difference is simple: Win+Shift+S captures an image. Win+Shift+T skips the image entirely and goes straight to the text. If all you need is the words, there's no screenshot involved at all.

Previously, to get text out of a screenshot, you had to: press Win+Shift+S, capture the region, wait for the Snipping Tool window to open, and then click the Text Actions button. Win+Shift+T eliminates all of those intermediate steps.

Do You Need to Install Anything?

No additional software is required. The feature is a part of the Snipping Tool. If you don't see it, here's how to make sure you have it:

1. Open the Microsoft Store
2. Search for Snipping Tool
3. If an Update button appears, click it

Once updated, the shortcut works system-wide immediately.

What You Can Extract Text From

The OCR works well with both images and scanned PDFs. In practice, this covers a wide range of everyday scenarios that come up in translation and interpreting work (and honestly, in everyday computer use):

• Error messages and dialog boxes that don't allow text selection
• Images containing text — infographics, screenshots, photos of printed material
• Locked or scanned PDFs where no text layer exists
• Paused video frames — grab a subtitle, a quote, or an on-screen label
• App interfaces that block standard copy-paste

A Note on Accuracy

For clean, high-contrast screen text, such as user interface elements, document content, standard fonts, the results are reliable and fast. Accuracy can drop with very small font sizes, decorative or handwritten fonts, low-contrast color combinations (light gray on white, for example), and low-resolution images.

For day-to-day use, you'll find it more than adequate without any extra configuration.

One more thing worth noting: the tool processes everything locally on your device. Nothing is sent to external servers, which matters when you're working with confidential documents.

Final Words

Win+Shift+T is a small shortcut with a surprisingly large impact. Once it's in your muscle memory, you'll reach for it constantly, whether you're capturing text from a non-editable PDF, capturing an error message, or lifting text from a video.

Give it a try the next time you see text you can't click on!

Quick High-Quality Subtitles with Subtitle Edit and Whisper: A Step-by-Step Guide

This is a practical guide to generating quick, clean subtitle drafts using Subtitle Edit paired with Whisper, an AI speech recognition engine. The goal is a smart balance between speed and quality. Not perfection, but a solid draft that doesn't take forever to clean up.

If you'd like to watch a tutorial, you can find one here, where the same steps listed below are demonstrated.

Steps 1 and 2 (and part of Step 4) below are one-off procedures you need to complete only once to get ready to start subtitling efficiently. Once they're ready, you will find the actual subtitling process, which consists of loading your video, transcribing it with Whisper, and doing a quick, systematic clean-up of the output, to be quite fast. So, let's get started!

Step 1: Download and Install Subtitle Edit

Head to the Subtitle Edit website and download the latest version. The installation is straightforward. follow the default prompts and you're good to go.

Step 2: Initial Setup

Before you start any actual subtitle work, it's worth spending a few minutes configuring the program. This upfront investment saves a lot of manual work down the line.

General Settings

Go to Options > Settings and take a look at the General tab. One of the key settings here is the single line maximum length, which is set to 42 characters by default. If that works for your project, leave it. If not, adjust it here. There are also a few profiles available in the dropdown that are worth exploring.

Tools: The "Do Not Break After" List

This is one of the most impactful settings for subtitle quality. Still inside Options > Settings, go to the Tools section and scroll to the auto-break settings.

Make sure the option "Use Do Not Break After List" is checked. Then click Edit to open the word list.

This list tells Subtitle Edit which words should never appear at the very end of a subtitle line. The idea is that if a line breaks after a word like "for," "a," "and," or "the," the reading experience feels awkward, as those words belong with whatever comes next.

To add words to the list:

1. Type the word in the input field.
2. Click Add (or press Enter) and it gets added immediately.

You can also use a regular expression to add multiple words at once, which is useful if you have a long list of conjunctions, articles, or prepositions to include. Below is a possible regex you can use. You can adapt it according to your needs and paste it after selecting Regular expression.

(?i)(?:\b(?:a|an|the|of|to|in|on|at|by|for|with|from|and|but|or|so|that|if|because|as|than|is|are|was|were|be|been|being|have|has|had|do|does|did|can|could|will|would|should|may|might|must|not|this|these|those|my|your|his|her|our|their|about|into|onto|over|under|after|before|between|among|through|during|without|within|around|across|along|although|though|while|unless|until|since|whether)\b|\b\w+(?:'m|'re|'ve|'d|'ll|'s)\b|n't)\s*$

Video Player Appearance

This is optional, but worth mentioning: the default subtitle appearance in the built-in player may not be comfortable to read. In my case, I don't like the default font size and shadow. To change it to suit your preferences, go to the video player settings and adjust the following:

• Primary color: Yellow works well for readability.
• Outline: Disabled.
• Shadow: Set to 0.0.
• Opaque box: Selected.

Again, this is purely a matter of personal preference, but it's nice to know that we can adjust it to whatever is most comfortable for us during review.

Shortcuts

Go to Options > Settings > Shortcuts. You don't need to configure everything right now — the shortcuts will make more sense once you're actually working with subtitles. We'll come back to the specific ones as they become relevant throughout this guide.

Step 3: Load Your Video

Go to Video > Open Video File and load your file. Once the video is loaded, the waveform should appear at the bottom of the screen. If it doesn't generate automatically, just click in that area and it will be created. You'll use this waveform later for fine-tuning timing.

Step 4: Transcribe with Whisper

This is where the AI does the heavy lifting. Go to Video > Audio to Text.

In the dialog that opens, configure the following:

Engine: Select Purfview Faster Whisper XXL. There are several engine options listed, but this one is significantly faster than the original Whisper and works well on both GPU and CPU.

Language: Choose your source language.

Model: This is important. If you're opening Subtitle Edit for the first time, there won't be any models listed in the dropdown yet. Here's how to get one:

1. Click the three dots (…) next to the model selector.
2. Select a model from the list.
3. Click Download.

As a rule, use at least the medium model. If your computer can handle it, go for one of the large models, large or large-v2/v3. The reason: a larger model produces more accurate transcriptions, which directly translates to less time fixing errors during cleanup. And because everything runs locally on your computer, nothing is sent to any server. Once downloaded, the model stays on your computer and doesn't need to be downloaded again.

Once the model is downloaded, it will appear in the dropdown. Select it and continue with the settings below.

Auto-adjust timings: Make sure this is selected.

Use post-processing: Also select this. Click the Settings button next to it and verify that at least the first two post-processing options are checked.

Generating Subtitles for a Single Video

Click Generate. Depending on video length and your hardware, this may take anywhere from under a minute to several minutes. When it finishes, the transcribed subtitles will appear in the subtitle list, next to the video player.

Batch Mode: Processing Multiple Videos at Once

If you have several videos to subtitle, you don't need to process them one by one. In the same Audio to Text dialog, switch to the Batch tab. Click Add to select all the videos you want to process, and they'll appear in a list. Click Generate, and the tool will run transcription for all of them in sequence, producing one SRT file per video. This is a significant time saver for larger projects. 

Step 5: First Cleanup Pass — Unbreak and Auto Break

Once your transcription is ready, you'll notice the subtitles may have inconsistent or awkward internal line breaks. Whisper handles the segmentation automatically, but it doesn't always break lines in the best places. This is where the "Do Not Break After" list you set up earlier becomes useful.

The fix is a two-step process applied to all subtitles at once:

1. Press Ctrl+A to select all subtitles.
2. Click the Unbreak button — this removes all line breaks, collapsing every subtitle into a single line.
3. Click the Auto Break button — this reapplies line breaks according to your configured rules: the maximum characters per line, and the "Do Not Break After" word list.

The result is a consistently formatted set of subtitles where, for example, a line that previously ended in "a" will now have that "a" correctly moved to the beginning of the second line within the same subtitle.

Step 6: Fix Transcription Errors

Before moving on to segmentation cleanup, it's efficient to address systematic transcription errors first.

For example, if Whisper transcribed "Claude" as "Cloud" (which can happen with proper nouns), there may be multiple instances of that error throughout the file. Use Find and Replace to fix them all at once:

1. Press Ctrl+H to open Find and Replace.
2. Type the incorrect word in the Find field (e.g., "Cloud").
3. Type the correct version in the Replace field (e.g., "Claude").
4. Click Replace All.

It's a good habit to do this early, since fixing one systematic error in bulk takes seconds, while catching individual instances during manual review takes much longer.

Step 7: Fix Dangling Words with Shortcuts

Now it's time to do a pass through the subtitle list looking at the end of each subtitle, specifically to identify any "dangling weak words." These are words like conjunctions, prepositions, or articles ("and," "or," "for," "a," "the") that have ended up stranded at the end of a subtitle line. While the auto break pass handled many of these, some will still appear between subtitles, i.e., at the end of the second line, rather than within them.

Setting Up the Shortcuts

Go to Options > Settings > Shortcuts and search for "move." You'll find several relevant actions:

• Move last word to next subtitle: Takes the last word of the current subtitle and moves it to the start of the next one. I like using Alt+W for this, as it's a left-hand shortcut that you can press easily while keeping your right hand on the mouse to navigate the subtitles.
• Move first word to previous subtitle: The reverse, as it sends the first word of the current subtitle up to the end of the previous one. A good shortcut is Alt+D.
• Move last word from first line to second line (within the same subtitle): Moves a word from the top line to the bottom line inside the current subtitle. You can set a shortcut like Ctrl+W for this.
• Move first word from second line to first line: The opposite internal move. To keep things simple, I like Ctrl+D for this.

After assigning any new shortcut, click Update to save it.

Important: When you move a word to the next subtitle using one of these shortcuts, Subtitle Edit also automatically adjusts the timing of both subtitles to reflect the change. You don't need to fix the timestamps manually.

The Cleanup Pass

Go to the very top of the subtitle list. Use your arrow keys or mouse to move down one subtitle at a time, looking at the last word of each. When you spot a dangling weak word:

• Press Alt+W once to move the last word to the next subtitle.
• If you need to move two words, press it twice.
• If you'd rather pull words from the next subtitle up instead, move to the next subtitle and press Alt+D as many times as needed.

If a subtitle turns orange, it means it's exceeding the maximum line length. Keep moving words around until the warning clears. If the total character count exceeds the maximum for both lines, you will need to split the subtitle.

This pass can be done either as a dedicated sweep through all subtitles, or on a subtitle-by-subtitle basis as you review timing, whichever feels more natural to you.

Step 8: Split Long Lines

After your dangling word pass, go to Tools > Split Long Lines. This identifies any subtitles that are still too long to display comfortably and proposes splits for them. Review the suggestions and click OK to apply.

After splitting, run the unbreak + auto break combination again on all subtitles (Ctrl+A → Unbreak → Auto Break) to make sure the new line breaks are in the right positions based on your rules.

Step 9: Review Timing with the Waveform

At this point, the segmentation is in decent shape. Now it's time to check the timing of individual subtitles against the actual audio.

Set your playback to a speed that's comfortable to move through the video faster during this pass. You can play/pause with the spacebar or by clicking the play button.

Watch the subtitle list as you play, and stop when you see a timing issue. The waveform at the bottom shows you exactly where each subtitle starts and ends. If you can see that a subtitle's boundary doesn't align with the speech, drag the start or end marker directly on the waveform to fix it.

To zoom out on the waveform for a broader view, use Ctrl + scroll wheel down.

To resume reviewing from a specific subtitle after making a fix, double-click that subtitle in the list to jump the playhead back to it.

Step 10: Manual Splitting and Merging

As you review, you'll occasionally find subtitles that need to be restructured rather than just adjusted. Here's how to handle each case.

Splitting a Subtitle

1. Click into the subtitle text and place your cursor exactly where you want the split.
2. Press Alt+S to split. The subtitle is divided at the cursor position, and the timing is split proportionally.
3. After splitting, click Unbreak and then Auto Break on the individual subtitle (no need to do all) to clean up the line break placement.

Merging Two Subtitles

If two adjacent subtitles contain a phrase that would read better as one unit, such as, for example, "welcome to this module / where we will have an introduction", you can combine them and re-split at a better point:

1. Select the first of the two subtitles.
2. Press Ctrl+Shift+M to merge. The text of both subtitles is combined into one.
3. Inside the merged subtitle, use Enter to place a line break if needed.
4. If the merged subtitle is too long and needs to be split, press Alt+S to split at the cursor.
5. Click Unbreak and Auto Break to finalize the line breaks.

Step 11: Save the SRT File

Once you're satisfied with the subtitle draft, save the SRT file, which is now ready to be used with your video.

The Right Balance

Good subtitles take time. For some projects, the priority is a clean, readable draft that accurately represents the audio, not a perfect broadcast file. 

Whisper's large models are accurate enough that the bulk of the work happens automatically, and that means that your time is best spent on:

• Systematic transcription errors (fix with Find & Replace)
• Dangling weak words at subtitle endings (fix with shortcuts)
• Subtitles where the timing is visibly off in the waveform
• Any splits or merges where the segmentation genuinely affects comprehension

And now, it's time to go create some subtitles! If you give this a try, let me know how it went in the comments below.