Logging food by voice: what it actually fixes
The slow part of tracking is choosing between database entries, not typing. Why speaking a meal removes that step, and where voice logging still fails.
4 min read
What separates calorie apps from each other, which one suits which person, and how to move your data when you switch.
Every calorie app makes the same promises and differs in the places you only discover in week three: how it handles food you cooked yourself, whether servings are household units or grams, what it costs after the trial, and whether you can export what you logged. These articles compare them on those grounds rather than on feature lists.
The database decides accuracy, speed decides whether you keep using it, and export decides whether you can leave. Everything else is decoration.
The slow part of tracking is choosing between database entries, not typing. Why speaking a meal removes that step, and where voice logging still fails.
4 min read
Photo estimation identifies foods well and estimates quantity poorly, and quantity is where the calories are. Where it helps and where it misleads.
5 min read
Specialist apps do each job better. A combined one answers questions neither can. Which trade-off suits you, and what a combined app has to get right.
5 min read
Not a feature table. The four axes on which calorie apps actually differ, and how to work out which alternative fits the food you eat.
4 min read
A food log is health data: what you eat, what you weigh, how it changes. What apps collect, what to check in a policy, and the settings to change.
5 min read
What free tiers typically withhold, which of those features change behaviour, and how to decide without paying to find out.
4 min read
Six checks that take ten minutes and tell you more than any review. Where entries come from, why duplicates matter, and what a good result looks like.
5 min read
Two years of logs are worth something. What a real export looks like, the questions to ask before committing, and what to do when there is no export button.
5 min read
Five requirements Indian food puts on a tracker that Western-built apps rarely meet, and how to test any candidate against them in ten minutes.
5 min read
Scanning is fast and accurate for packaged food, which is a small part of most Indian diets. Where it helps, where it fails, and what fills the gap.
5 min read
Paper is faster to start, private by default and hopeless at arithmetic. An app does the sums and adds friction. Which failure mode matters more to you.
4 min read
The database decides accuracy, speed decides whether you keep using it, and export decides whether you can leave. Everything else is decoration.
6 min read
Every calorie app made me choose between eleven entries for the same roti. So I stopped logging, like everyone does. This is the app I wanted instead.
5 min read
A data export is a file containing everything you have logged in an app, in a format you can read elsewhere; the ability to take it with you is what stops a food log from becoming a hostage.
A food database is the set of foods and nutrition values a calorie app searches; its quality — whether entries are verified, sourced and deduplicated — matters more to logging accuracy than any feature in the app.