When many people adjust the aperture, their first reaction is:

The lower the F-value, the brighter the image.

That’s right, but it’s only half true.

Because the F-value not only affects brightness but also influences...Depth of field.

Sometimes, when the image is in focus and the subject moves slightly back and forth, it actually starts to become blurry.

What exactly is the F-value?

Just understand it simply:

A small F-number = a large aperture = more light entering the lens = a brighter image.

A large F-number = a small aperture = less light entering = a darker image.

So, when the image appears dark, lowering the F-number is indeed a very straightforward solution.

But problems also arose along with it.

The larger the aperture, the shallower the depth of field usually is.

To put it bluntly:

The truly clear scope has become smaller.

Why does the image become blurry even when the screen is brighter?

If you're photographing a very flat target in a fixed position, the problem might not be obvious.

But if the workpiece has height differences or its position changes back and forth during operation, a shallow depth of field can easily lead to problems.

The bottom may be clear, while the top feels vague.

It’s also possible that this one is very clear, but the next one, with just a slight shift in position, will have less sharp edges.

Therefore, in machine vision, we can't just focus on “brightness.”

Also see:

Is the clear range sufficient?

The F-value isn't necessarily better the smaller it is.

This is also the pitfall most easily stepped into on-site.

If the image is dark, just keep opening up the aperture.

As a result, the brightness is there, but the depth of field is gone.

Then I started to doubt the lens, the camera, and the algorithm.

In fact, the problem might just be that the aperture is set too wide.

If the target has a height difference, you’ll need to ensure sufficient depth of field.

If the target changes before and after, you shouldn't open the aperture too wide.

So a more practical understanding is:

To make it brighter, you can use a smaller F-number.

To achieve a greater depth of field, the F-number should be slightly larger.

You need to find balance on both sides.

How exactly should we look at the scene?

Don't ask right off the bat:

“What’s the best F-value to set?”

First, take a look at your goals.

Is the target a plane?

Is there a height difference?

Will it move back and forth during runtime?

Do all the multiple locations have to be clear?

If these changes are relatively noticeable, you’ll need to leave a bit more margin for depth of field.

If the target is flat and its position is fixed, but the image is relatively dark, a smaller F-number is usually more appropriate.

Ultimately, there’s no single “correct” F-value.

It has to follow the task.

Just remember one sentence in the end.

A lower F-value results in a brighter image, but typically produces a shallower depth of field.

A higher F-value results in a darker image, but typically offers a wider depth of field.

So in machine vision, what really matters is not:

“Could you make it a bit brighter?”

But:

“Is the brightness sufficient, and can the target remain clear at all times?”

The larger the aperture, the better is not necessarily true.

Bright enough, clear enough, and stable enough—that’s when it’s truly right.