Confocal microscopes are sold as complete instruments, so the laser inside them rarely gets discussed on its own. That changes when you are building an instrument, replacing a failed source or adding a line to an existing system. Then the laser becomes a component you have to specify, and the numbers that matter are not the ones on the front of the datasheet.

The common laser lines

Commercial dye panels cluster around a small set of excitation lines. Four of them cover a large share of routine work, and most systems offer one or two more for cases where the standard four sit badly on the absorption peak.

Line	Common fluorophores	Note	From RGB
405 nm	DAPI, Hoechst, Alexa Fluor 405	Sits on the shoulder of the DAPI band, not its peak	405 nm module
445 nm	CFP	Usually excites CFP more efficiently than 405 nm	yes
488 nm	GFP, FITC, Alexa Fluor 488	The most used single line	yes
514 / 515 nm	YFP	Where YFP and GFP must be separated	yes
561 nm	mCherry, TRITC	DPSS wavelength	no
594 nm	Texas Red, Alexa Fluor 594	Often a better match than 561 nm	no
633 / 638 / 640 nm	Cy5, Alexa Fluor 647, APC	Close together but not interchangeable with a given filter set	633 nm module

Check the actual excitation and emission spectra of your dyes against your dichroics and emission filters before fixing a line. Two wavelengths a few nanometres apart can behave very differently once a filter set is in the path. For how excitation wavelength, Stokes shift and dye choice interact, see which fluorophore needs which line.

Power at the sample, not at the aperture

The number on a laser datasheet is measured at its output. What matters is what arrives at the focal plane, and the path between them loses a lot. The figures below are typical ranges for the individual stages. They are not a specification for any particular instrument.

Stage	Typical transmission
Fiber coupling	50 to 80 %
AOTF or filter wheel	60 to 90 %
Scan head and dichroics	50 to 80 %
Objective	70 to 90 %
Multiplied end to end	roughly 10 to 50 %

Worked through: at 10 to 50 % end to end, 1 mW at the sample needs roughly 2 to 10 mW from the source. Add margin for component ageing and for the drop in diode output over service life, and you arrive at a sensible specification of around 5 to 20 mW per line for that target. The honest version of this calculation is to measure the transmission of your own system rather than to inherit a rule of thumb.

As a starting point for the target itself, many confocal applications work in the region of 1 to 10 mW at the sample for point scanning. Treat that as orientation. The real figure depends on sample, objective, detector sensitivity, dwell time and how much bleaching you are prepared to accept.

What the source should deliver

Parameter	Why it matters	How to specify it
Power stability	Intensity differences between images become artefacts rather than biology	With a stated period and measurement point. The laser is one contributor among AOTF, fiber and polarization
Noise	Appears as pixel noise and forces longer dwell times	Separate fast intensity noise from long term drift, and state the measurement bandwidth
Beam quality	Decides how tightly the beam focuses and how well it couples	Ask for M² and the beam mode, not just the label TEM00
Polarization	Polarization sensitive optics in the scan head cause throughput to vary	Specify at the relevant output, after the fiber if there is one
Modulation	Blanking during scanner retrace avoids dose that produces no signal	Only needed from the laser if the microscope has no AOTF or AOM doing the job

That last row is worth a sentence. Most commercial confocal systems already contain an AOTF that handles blanking and intensity control, and its switching time and synchronisation matter more than the laser’s own modulation bandwidth. Direct modulation of the source becomes a requirement mainly in custom builds without such a modulator, which is the usual situation when an OEM laser module goes into a new instrument.

Built for microscopy Lambda Beam Pigtailed Polarization maintaining single mode fiber · temperature stabilized · 405 to 1550 nm View range

Point scanning or spinning disk

	Point scanning	Spinning disk
Illumination	one focused spot, sequential	many spots in parallel
Total power needed	lower	higher, the factor depends on pinhole count, disk and camera
Modulation	retrace blanking is useful	less critical
Speed	limited by the scanner	can be considerably faster

Spinning disk systems spread the same job over many simultaneous foci, so the required total laser power is higher. How much higher is not a fixed multiple: pinhole count, disk design, camera sensitivity, exposure time and the sample all enter. Work it out for the specific head rather than applying a factor.

Why confocal systems are usually fiber coupled

• Reproducible beam geometry. A single mode fiber spatially filters the beam, so what reaches the scan head is similar across units regardless of diode to diode variation.
• Alignment leaves the production line. You align the fiber output into the scan head rather than the laser itself in every unit.
• Thermal separation. The laser and its heat sit away from the optical head.
• Combining lines. Several wavelengths arrive on a common axis instead of each needing its own alignment.

It is not free. Coupling costs optical power, the fiber has to be single mode across every wavelength you send through it, and polarization maintaining fiber adds launch alignment and connector requirements. For multi line systems the benefits usually win. For a single line bench setup, free space may be the simpler answer.

Frequently asked questions

Does a confocal microscope need a laser?

Confocality comes from the geometry: a point focus paired with a pinhole that rejects out-of-focus light. That does not strictly require a laser, and early Nipkow disk instruments used arc lamps. In practice almost all commercial point scanning systems use lasers, because they deliver high brightness into a small, narrowband, well defined focus, which lamps cannot match.

How many laser lines do I need?

One per fluorophore channel you want to separate cleanly. Four lines around 405, 488, 561 and 640 nm cover many commercial panels. Add 445 nm for CFP or 594 nm for the Texas Red family if those appear in your work. Specialised instruments often manage with one or two.

How much laser power does confocal microscopy need?

For point scanning, often in the region of 1 to 10 mW at the sample, with large variation by application. Because the optical path typically transmits only 10 to 50 %, the source is usually specified several times higher per line. Measure your own path rather than relying on a general figure.

Can I add a wavelength to an existing microscope?

Often yes, if the system has a free fiber input and the dichroics and emission filters pass the new band. The laser is usually the easy part. Check the filter set first.

What are the downsides of confocal laser scanning?

Point scanning builds the image sequentially and is therefore slower than widefield. Spinning disk architectures scan in parallel and can be fast enough for live cell work. Concentrating the dose into a scanned focus tends to increase photobleaching and phototoxicity, though the outcome depends on total dose, scan strategy, frame rate and the sample rather than on the architecture alone.