Calorie Counting for Indian Food

Three structural reasons, none of which are about willpower:

Mixed dishes. A chicken curry is not an ingredient, it is a recipe — and its calorie count depends on a cooking process you did not standardise. Western tracking apps are built around discrete foods with labels. Most Indian meals have neither.

Non-standard portions. A katori is not a unit. Household katoris vary from about 100 ml to 200 ml, which is a two-fold difference in everything you serve in one. “One bowl of dal” is not a measurement.

Invisible oil. This is the big one, and it is covered below on its own.

One tablespoon of any cooking oil is roughly 120 kcal, and it is pure fat — no protein, no fibre, no satiety to speak of. A single sabzi cooked with three tablespoons of oil has around 360 kcal of oil in it before you count the vegetables.

This is why the same dish can be 90 kcal or 250 kcal per serving and why restaurant food wrecks otherwise careful tracking. Nobody logs the oil, because nobody sees it once it is absorbed.

The practical fix is upstream of tracking: measure the oil going into the pan rather than estimating it afterwards. Changing from a free pour to a measured two tablespoons per dish is often a 200–300 kcal daily difference on its own — larger than most of the substitutions people agonise over.

Precision on everything is unsustainable and unnecessary. Accuracy where it matters is both.

Measure the three that move the number: cooking oil, protein portions, and grains (rice and roti). Between them these account for most of the variance in an Indian meal.

Standardise your own katori once. Fill yours with water, pour it into a measuring cup, and write the number down. You now have a real unit for every subsequent meal, which is 80% of the benefit of weighing for 2% of the effort.

Estimate vegetables and leave them estimated. Being 20% wrong on bhindi does not matter. Being 100% wrong on the oil it was cooked in does.

Photographing a plate is far faster than searching a database for “dal” and picking between forty entries, and for Indian food it removes the naming problem entirely — you do not need to know whether your app calls it chapati, roti or phulka.

MetaFit's Snap does this inside the ₹199 plan. For comparison, HealthifyMe charges around ₹999/month for photo logging and MyFitnessPal around ₹6,500/year for premium features.

Where it does not help, and where any app claiming otherwise is overselling: a photo cannot see the oil. It cannot tell a lightly sautéed sabzi from one finished with three tablespoons of ghee, because that difference is absorbed and invisible. Photo tracking solves identification and portion estimation. It does not solve the largest single source of error, which is why the measuring-the-oil habit still matters.

For many people, yes — mostly because it makes intake visible rather than because the arithmetic is exact. Self-reported intake is systematically underestimated, often substantially, and tracking narrows that gap.

But it is a means, not the goal, and for a minority it is actively unhelpful — anyone with a history of disordered eating should approach it carefully or not at all. If tracking makes you anxious rather than informed, the habits that matter (protein at every meal, measured oil, consistent portions) work perfectly well without a log.