Why Does an Objective Lens with NA 0.8 Collect Less Than 50% of the Light?

Understanding the Real Relationship Between Numerical Aperture and Collection Efficiency

A high-NA microscope objective is often expected to collect a large fraction of the available light.

So when an objective is specified with:

NA = 0.8

it may seem intuitive to expect that it collects something close to 80% of the emitted light.

In practice, however, the optical power reaching the detector can easily be less than 50% of the incident or emitted light, and in some fluorescence experiments the final detected fraction can be far lower.

This does not mean the NA specification is wrong.

The misunderstanding comes from treating numerical aperture as an efficiency percentage.

It is not.

Numerical aperture describes the angular acceptance of an optical system. The actual optical throughput depends on several additional factors, including:

1. The collection solid angle

2. Coverslip-induced spherical aberration

3. Transmission losses inside the objective

These losses are often not obvious from the headline NA specification.

---

1. What Does NA = 0.8 Actually Mean?

The numerical aperture of an objective is defined as:

NA = n · sinθ

where:

• n is the refractive index of the medium between the specimen and the objective

• θ is the maximum half-angle of light that can enter or leave the objective

For a dry objective in air:

n ≈ 1

Therefore, for:

NA = 0.8

we have:

sinθ = 0.8

which corresponds to:

θ ≈ 53°.

But this only tells us that the objective can accept rays within approximately a 53° half-angle.

It does not mean that the objective collects 80% of the available light.

That distinction is critical.

---

2. Why NA 0.8 Does Not Mean 80% Collection Efficiency

Consider an isotropic point source, such as a fluorescent molecule.

An isotropic emitter radiates light uniformly into a full sphere:

4π steradians

An objective collects only the fraction of this sphere that falls within its acceptance cone.

For an acceptance half-angle θ, the collected solid angle is:

Ω = 2π(1 − cosθ)

The fraction of total isotropic emission collected is therefore:

ηcollection = Ω / 4π

or:

ηcollection = (1 − cosθ) / 2

For:

θ ≈ 53°

the theoretical collection fraction is approximately:

20%

for isotropic emission.

So even before considering transmission losses:

An NA 0.8 dry objective does not collect 80% of isotropically emitted light. Its geometrical collection fraction is only around 20%.

This is purely a consequence of solid-angle geometry.

Numerical aperture and collection efficiency are therefore related, but they are not linearly proportional.

---

3. The First Major Loss: Coverslip-Induced Spherical Aberration

High-NA objectives are particularly sensitive to the optical properties of the coverslip.

A common microscope coverslip is:

#1.5

with a nominal thickness of approximately:

0.17 mm

However, real coverslips have manufacturing tolerances.

The source material notes that thickness variations of approximately:

±0.02 mm

are common.

At low NA, this may not be a major problem.

At high NA, it can become extremely important.

---

4. Why Does Coverslip Thickness Matter So Much?

A high-NA objective collects rays over a wide angular range.

Paraxial rays travel close to the optical axis.

Marginal rays enter the optical system at much larger angles.

When these rays pass through a coverslip whose thickness differs from the design value, they experience different optical path lengths.

The result is spherical aberration.

Instead of all rays converging to the same focal point:

• Central rays focus at one position

• High-angle rays focus at another

This reduces the intensity delivered into the intended image or detection region.

The effect can become surprisingly large.

---

5. A 0.02 mm Coverslip Error Can Cause Major Intensity Loss

The source cites microscopy data for an approximately NA 0.85 dry objective.

With a coverslip thickness error of:

0.02 mm

the intensity reaching the detector can decrease by as much as:

57%

Even a thickness error of only:

0.01 mm

can cause an intensity loss of approximately:

19%.

This is why high-NA microscopy is much more sensitive to coverslip thickness than many users expect.

The number:

0.17

printed on many microscope objectives is not decorative.

It indicates that the objective was designed and corrected for a coverslip thickness of approximately:

0.17 mm

Operating significantly away from this value can degrade the optical correction.

---

6. The Second Major Loss: Fresnel Reflection Inside the Objective

A high-NA objective is not a single lens.

It is typically a complex optical assembly containing many individual lens elements.

The source notes that a high-NA objective may contain approximately:

8–12 lens elements.

Every optical interface introduces some reflection.

At normal incidence, an uncoated air-glass interface with a refractive index around:

n ≈ 1.5

can reflect approximately:

4%

of the incident optical power.

A single lens has two surfaces, so multiple lenses create many interfaces.

---

7. Small Reflection Losses Multiply Quickly

Consider an optical system with eight lens elements.

That gives approximately:

16 air-glass interfaces

If each interface transmitted approximately:

96%

then the simplified transmission would be:

T ≈ 0.96¹⁶

which is approximately:

52%.

Real microscope objectives use antireflection coatings, so actual performance depends heavily on the coating design and wavelength.

However, the calculation illustrates an important principle:

A few percent loss at each optical surface can accumulate into a large total transmission loss in a multi-element objective.

Additional losses can come from:

• Absorption in optical glass

• Cemented interfaces

• Internal apertures

• Coating spectral response

The source describes practical objective transmission as potentially falling into the 50–70% range depending on the optical system.

---

8. High-Angle Rays Can Experience More Reflection

The situation becomes more complicated for high-NA objectives.

The marginal rays do not strike every optical surface at normal incidence.

For an NA 0.8 system, high-angle rays can encounter interfaces at angles above approximately:

40°

inside parts of the optical system.

Fresnel reflection depends on:

• Incident angle

• Polarization

• Refractive indices

In particular, the reflectivity of S-polarized light generally increases with incident angle.

Therefore, the simple 4% normal-incidence estimate can underestimate losses for some high-angle rays if the coatings are not optimized for the relevant angular distribution.

This is another reason why theoretical NA and real optical throughput should not be treated as equivalent.

---

9. Putting the Three Losses Together

Now consider a representative fluorescence imaging experiment.

The source gives the following simplified budget:

Step 1: Geometrical collection

For isotropic emission:

NA = 0.8 → approximately 20% collected

Step 2: Coverslip error

Assume a coverslip thickness error of:

0.02 mm

and an intensity loss of approximately:

57%

Remaining fraction:

20% × 0.43 ≈ 8.6%

Step 3: Internal objective losses

Assume approximately:

40% additional loss

through the multi-element optical system.

Remaining fraction:

8.6% × 0.60 ≈ 5.2%

The resulting detected fraction is therefore only around:

5%

in this illustrative example.

This does not mean every NA 0.8 objective has 5% efficiency.

Rather, it illustrates why seeing less than 50% throughput is not surprising once the complete optical system is considered.

---

10. NA Is a Geometrical Limit, Not an Efficiency Specification

This is the core conceptual point.

An objective's numerical aperture describes:

The maximum angular cone accepted by the objective

It does not directly describe:

The percentage of optical power delivered to the detector

The final detected power depends on:

Emission geometry × collection solid angle × aberrations × optical transmission × detector geometry

So the correct relationship is closer to:

Detected Power = Emitted Power × Geometrical Collection × Optical Transmission × System Efficiency

not:

Detected Power ≈ Input Power × NA

This distinction eliminates a large amount of confusion when evaluating high-NA optical systems.

---

11. Practical Recommendation 1: Control Coverslip Thickness

For high-NA dry objectives, coverslip thickness should be treated as an optical parameter, not merely a mechanical specification.

The source recommends that for objectives with:

NA > 0.7

the coverslip thickness should ideally be controlled within approximately:

±0.005 mm

when maximum optical performance is required.

For demanding experiments:

• Use calibrated coverslips

• Check the objective's specified coverslip thickness

• Avoid assuming that all #1.5 coverslips are optically identical

Objectives equipped with a correction collar can sometimes compensate for coverslip-thickness variation over a limited range.

---

12. Practical Recommendation 2: Match the Immersion Medium

For oil-immersion objectives, the refractive index of the immersion oil is part of the optical design.

The source gives a typical design refractive index of approximately:

n = 1.515.

If the immersion oil has a significantly different refractive index, high-angle rays can experience additional refractive-index mismatch.

This introduces aberrations similar in principle to those caused by incorrect coverslip thickness.

Therefore, for high-NA oil objectives:

Use the immersion oil specified for the objective rather than treating all immersion oils as equivalent.

Temperature can also influence refractive index, so high-precision microscopy may require temperature awareness as well.

---

13. Practical Recommendation 3: Do Not Always Maximize Condenser NA

In transmitted-light microscopy, maximum illumination NA is not always desirable.

The source notes that the condenser NA is commonly set to approximately:

70–90% of the objective NA.

This intentionally reduces some light throughput.

Why?

Because microscopy involves a trade-off between:

• Resolution

• Contrast

• Depth of field

• Illumination uniformity

• Optical throughput

Reducing the condenser aperture slightly can improve contrast even though it reduces the amount of light entering the objective.

This is therefore an intentional optical adjustment, not an efficiency failure.

---

14. Do Not Use NA as a Percentage

A useful rule is:

Never interpret NA numerically as the percentage of light collected.

For example:

NA = 0.8 ≠ 80% collection

For isotropic emission in air, the corresponding theoretical geometrical collection is closer to:

~20%

before other optical losses are considered.

The difference comes from the fact that light collection is governed by solid-angle geometry, not by a simple linear percentage.

This becomes even more important when comparing objectives.

For example, increasing NA from 0.4 to 0.8 does not simply mean:

40% → 80% efficiency

The underlying angular and solid-angle relationships are nonlinear.

---

15. What Should You Check When Collection Efficiency Is Too Low?

If a high-NA objective delivers less optical power than expected, check the optical system systematically.

1. Confirm what NA actually represents

Calculate the acceptance angle and, where relevant, the corresponding solid-angle collection fraction.

2. Check coverslip thickness

For high-NA objectives, even small thickness errors can introduce significant spherical aberration.

3. Verify immersion-medium refractive index

For immersion objectives, the optical medium must match the objective design.

4. Check wavelength-dependent transmission

Antireflection coatings and glass transmission vary with wavelength.

An objective optimized for the visible range may have very different throughput in the near-UV or near-IR.

5. Evaluate the complete optical chain

The final detector signal may also be reduced by:

• Filters

• Dichroic mirrors

• Beam splitters

• Tube lenses

• Fiber coupling

• Detector quantum efficiency

The objective is only one element in the measurement chain.

---

16. Conclusion

An objective specified with:

NA = 0.8

does not guarantee that 80% of the available light will reach the detector.

Numerical aperture is defined by:

NA = n · sinθ

and describes the angular acceptance of the objective.

For an NA 0.8 dry objective, the maximum half-angle is approximately:

53°

For an isotropic emitter, that corresponds to only about 20% of the full 4π emission sphere.

From there, additional losses can arise from:

• Coverslip-induced spherical aberration

• Fresnel reflection

• Multi-element optical transmission

• High-angle ray behavior

• Refractive-index mismatch

• The rest of the detection optical path

The most important engineering lesson is therefore:

Numerical aperture is a geometrical acceptance parameter—not a collection-efficiency percentage.

If the collected power is lower than expected, do not immediately conclude that the objective is defective.

Instead, evaluate the full optical budget:

Solid angle → coverslip → aberrations → objective transmission → downstream optics → detector

Once these factors are considered, an objective with NA 0.8 delivering less than 50% of the available light is not necessarily surprising.

The missing light is often not missing at all.

It was simply never inside the usable optical cone—or it was lost somewhere along the optical path.