# Visual or acoustic inspection for cracks: what does each one detect?

> Visual inspection, by camera or by the human eye, detects visible cracks, leakers and dirt on the shell. Acoustic inspection detects hairline cracks, which do not show under normal light. Neither detects everything: dirt and interior defects are outside acoustic testing, and fine cracks are outside the camera. That is why the two methods complement each other.

Updated: 2026-10-09 · URL: https://vigiovo.com/en/resources/visual-vs-acoustic-crack-detection · Português: https://vigiovo.com/recursos/inspecao-visual-vs-acustica.md

## Key takeaways

- Visual inspection detects visible cracks, leakers and dirt on the shell.
- Acoustic inspection detects hairline cracks, through the change in the shell's resonance.
- Hairline cracks do not show under normal light, so a camera alone cannot guarantee their absence.
- Combining the two methods covers what each one would otherwise let through.

## What does visual inspection detect?

Visual inspection, done by an operator or by a camera with computer vision, detects what appears on the surface: visible cracks, leakers and dirt. It looks at the shell as it presents itself on the conveyor, and it can run continuously, egg by egg.

In Vigiovo, a visible crack is measured by length and depth, as described in [crack detection with black-hat and top-hat filters](https://vigiovo.com/en/resources/crack-detection-with-black-hat). The limit is the same as for any surface camera: what does not appear in the image is not detected.

## What does acoustic inspection detect?

Acoustic inspection detects hairline cracks by observing the change in the shell's sound response when it receives a light tap. It relies on a property the camera does not have: the shell responds to a tap, and a crack changes that response.

The method is not aimed at dirt or interior defects. For the detail of the method, see [acoustic crack testing](https://vigiovo.com/en/resources/acoustic-crack-testing).

## Which method detects each defect?

The table summarises the reach of each method in practice. It is a practical assessment, not a standard.

**Reach of each method (practical assessment)**

| Defect | Visual camera inspection | Acoustic testing |
| --- | --- | --- |
| Visible crack | Yes | Not the focus of the method |
| Hairline crack | No | Yes, through the change in resonance |
| Leaker | Yes | Not the focus of the method |
| Dirt on the shell | Yes | No |
| Interior defects | No | No |

Interior defects, such as the air cell, yolk and blood spots, are examined by [candling](https://vigiovo.com/en/resources/glossary#candling), not by either method in this comparison.

## Why combine the two methods?

Because each method covers the part the other cannot reach. The camera handles dirt and visible cracks continuously, with measurements recorded for each egg; acoustic testing looks for the fine cracks that the camera lets through.

The order depends on the line's flow. MAPA requires a pre-selection step, which removes cracked eggs, only in flows with continuous-line washing before candling (MAPA, question 24). That step is a requirement of the flow, not a rule about cameras.

## How do you choose the combination for your line?

The choice depends on which defects your line needs to control and on how much each one matters to your customer and to your production flow. If the line only needs to remove dirt and visible cracks, visual inspection covers the need. If hairline cracks also count, the line needs a method that finds them, such as candling or acoustic testing.

Before deciding, check the lighting and the grading rules, because they affect the camera's results. See the [camera lighting setup checklist](https://vigiovo.com/en/resources/camera-lighting-setup-checklist) and the comparison of [tuning rules vs retraining a model](https://vigiovo.com/en/resources/tuning-rules-vs-retraining-a-model).

## Frequently asked questions

### Can visual inspection find hairline cracks?

Generally not. Hairline cracks do not show under normal light, so a surface camera or the human eye cannot see them reliably. Visual inspection alone therefore does not guarantee that fine cracks are absent.

### Does acoustic testing replace visual inspection?

No. The acoustic method is not aimed at dirt and does not assess interior defects, and visual inspection covers exactly what it leaves out. The two complement each other, and the choice depends on the defect the line needs to control.

### Does Vigiovo perform acoustic inspection?

No. Vigiovo is visual camera inspection and grades eggs as clean, dirty, cracked or unverified. It complements the methods for hairline cracks by removing dirty eggs and visibly cracked eggs before them.

## Sources

- [Perguntas e respostas sobre a Portaria MAPA nº 612/2022 (versão de 05/06/2023)](https://wikisda.agricultura.gov.br/manual_ovos/perguntas_e_respostas_port612_2023.pdf), Ministério da Agricultura e Pecuária (MAPA)
- [Como o Vigiovo classifica ovos sem IA / How Vigiovo grades eggs without AI](https://vigiovo.com/como-funciona), Vigiovo

## Read next

- [Tuning rules vs retraining an AI model](https://vigiovo.com/en/resources/tuning-rules-vs-retraining-a-model.md) (Comparison, 2026-10-09): A comparison of maintaining an egg grader: tuning rules and thresholds or retraining an AI model, covering cost, explanation, change control and validation.
- [Camera lighting setup checklist](https://vigiovo.com/en/resources/camera-lighting-setup-checklist.md) (Checklist, 2026-10-09): Checklist to set up and test the lighting of a camera above the egg conveyor: diffuse light, no glare, stable exposure and a test with clean eggs.
- [Candling vs computer vision: what each one detects](https://vigiovo.com/en/resources/candling-vs-computer-vision.md) (Comparison, 2026-10-09): Candling compared with camera inspection in egg grading: which defects each detects, speed, record keeping and why the two methods work best together.

## How Vigiovo helps

Vigiovo grades eggs on the conveyor as clean, dirty, cracked or unverified, in real time, with deterministic computer vision: every decision comes with the measurements behind it, on an ordinary CPU. https://vigiovo.com/en/how-it-works.md

3 min read.
