How to use Fix Results in Cal.AI
When to tap it, what each correction does, and a worked salad example that shows how the totals move.
What does Fix Results do in Cal.AI?
Fix Results lets you correct a food scan without starting over. After Cal.AI analyzes a photo, the result screen lists each item with calories, protein, carbs and fat, plus a meal total. If any line is wrong, Fix Results lets you rename the food, swap an ingredient or adjust the portion. The totals update the moment you make the change.
The correction is also remembered. Corrected foods and recipes are saved for one-tap reuse, so a meal you eat every week only needs fixing once.
When should you use Fix Results?
A photo gives the AI a lot of information, but not everything. It cannot see the oil in the pan, taste whether the yogurt is nonfat, or know that the "rice" is cauliflower. Use Fix Results in these cases:
- Wrong name. The scan says "beef stir-fry" and it was tofu, or it says "white rice" and it was quinoa.
- Wrong ingredient. The salad is right but the dressing is not. Ranch and olive oil look similar in a photo and differ a lot in fat.
- Wrong portion. The chicken is listed as 85 g (3 oz) and you know the piece was closer to 120 g (4.2 oz).
- A hidden item. Butter on the vegetables, sugar in the coffee, or the second slice you took after the photo.
If the whole plate is simply bigger or smaller than estimated, you do not need Fix Results. The − / + serving control on the result screen scales the item up or down in one tap.
Step-by-step: correcting a scan
- Scan the meal with Scan Food, or pick a photo from the Library.
- On the result screen, read each item and its calories. Compare the listed portions with what is on the plate.
- Tap Fix Results.
- Rename any food that was misidentified.
- Swap an ingredient where the type matters more than the amount, such as a dressing, a cooking fat or a dairy product.
- Adjust portions for items that are the right food in the wrong amount.
- Check the new meal total, then save. The corrected version is now available for one-tap reuse.
The whole process usually takes less time than typing the meal in by hand, because you are editing an estimate rather than building an entry from nothing.
Worked example: fixing a chicken salad scan
Say you scan a lunch salad. The result screen comes back with romaine, grilled chicken, tomatoes, feta and a creamy dressing. Two things are off: the chicken portion is too small, and the dressing is olive oil, not ranch. The numbers below use USDA per-100 g values.
| Item | As scanned | After Fix Results |
|---|---|---|
| Romaine lettuce, 100 g (3.5 oz) | 17 kcal | 17 kcal |
| Chicken breast | 85 g (3 oz): 140 kcal, 26 g protein | 120 g (4.2 oz): 198 kcal, 37 g protein |
| Tomatoes, 80 g (2.8 oz) | 14 kcal | 14 kcal |
| Feta cheese, 30 g (1.1 oz) | 80 kcal | 80 kcal |
| Dressing | Ranch, 30 g (2 tbsp): 129 kcal, 13 g fat | Olive oil, 10 g (2 tsp): 88 kcal, 10 g fat |
| Meal total | 380 kcal | 397 kcal |
The arithmetic for the chicken: USDA lists cooked chicken breast at 165 kcal and 31 g protein per 100 g. At 85 g that is 165 × 0.85 = 140 kcal; at 120 g it is 165 × 1.2 = 198 kcal. The ranch swap goes the other way: 430 kcal per 100 g × 0.3 = 129 kcal becomes 884 × 0.1 = 88 kcal for the oil.
The total barely moved, up 17 kcal, but the macros changed a lot. Protein went from about 33 g to about 43 g for the meal. That matters if you are tracking a protein target, and it is exactly the kind of difference a photo alone cannot catch.
Serving control versus Fix Results
The two tools answer different questions. The − / + serving control deals with how much: it multiplies the whole item. Fix Results deals with what: it lets you change the identity, the ingredient or the exact portion.
- Ate one and a half portions of the same pasta: use the serving control.
- The pasta was whole-wheat, not white: use Fix Results to rename it.
- The sauce was cream, not tomato: use Fix Results to swap the ingredient.
- The portion was 180 g (6.3 oz), not 250 g (8.8 oz): use Fix Results to adjust the portion.
When in doubt, open Fix Results. It can do everything the serving control does and more.
Common mistakes
- Accepting the scan without reading it. A two-second check catches most errors.
- Fixing the calories but not the food. If you rename "ranch" to "olive oil", the macros correct themselves; if you only lower the number, the protein and fat stay wrong.
- Forgetting that saved foods carry their corrections forward. If you fixed a portion last week, check it still matches this week's plate.
- Chasing precision on low-calorie items. Lettuce at 90 g versus 110 g is a 3 kcal difference. Spend the effort on oil, cheese, nuts and meat.
- Not using the scan at all for mixed meals. See how accurate AI calorie counting is and where corrections matter most.
Frequently asked questions
Does Fix Results change my saved foods?
A corrected food or recipe is saved for one-tap reuse, so the next time you log the same meal you get the corrected version, not the original scan. If you fix the same food again later, the newest correction is what you log. Check the portion each time, since your plate may differ from last week.
Can I use Fix Results on a barcode or search result?
Fix Results is designed for the photo scan result screen, where the AI has identified items that may need correcting. For barcode and search entries, the serving control handles portion changes. If a packaged product's numbers look wrong, manual entry lets you type the label values directly.
What if the scan missed an item entirely?
Open Fix Results and add the missing item, or log it separately with search or manual entry. Hidden items such as cooking oil, butter and sugar are the most common omissions. Adding 10 g (2 tsp) of olive oil adds 88 kcal, so small additions are worth the few seconds they take.
Why did the meal total change only a little when I fixed the dressing?
Because corrections can cancel out. In the salad example, increasing the chicken added 58 kcal and swapping ranch for olive oil removed 41 kcal, so the total rose by only 17 kcal. The macros still moved a lot: protein rose by about 10 g. Look at protein, carbs and fat, not just the total.
Sources
- U.S. Department of Agriculture. FoodData Central. USDA