Quick Answer: Plastic optical fibre in a TOSLINK lead is reliable to about 5 m at 96 kHz and often to 10 m at 48 kHz. Glass fibre reaches 30 m and beyond. Length is only half the problem: the budget is transmitter power minus connector loss minus fibre attenuation, and it must stay above receiver sensitivity. A scratched end face or a bend tighter than the cable allows spends more of that budget than several extra metres. Above 10 m, 75 ohm coax carrying S/PDIF is usually the cheaper reliable route.
TOSLINK length is an optical budget problem, and the budget is spent by attenuation and end-face quality rather than by copper resistance. Light leaves a red emitter, crosses two moulded faces, travels through a plastic core and arrives at a photodiode that needs a minimum signal. Every joint, scratch and metre takes a share. When the share runs out the receiver mutes, so the symptom is silence rather than distortion.
What the optical budget really is
A working link needs received power above the sensitivity of the photodiode, with margin left for ageing and temperature. Plastic fibre attenuates far more per metre than glass, and the loss that matters rises with the sample rate because the same total light is stretched over shorter pulses. That is why length figures are usually quoted for 48 kHz stereo.
Plastic versus glass
Plastic core fibre of about 1 mm diameter is cheap, tolerant of handling and lossy. Glass fibre of roughly 200 micrometres needs careful termination but loses a fraction as much per metre. A plastic lead is the correct default for a 1 m to 3 m television to soundbar link; glass earns its cost past 10 m.
End faces and bend radius
Most failures happen at the end face. Dust, a fingerprint or a scratch scatters light before it enters the fibre, and the loss can exceed that of five extra metres. Keep the dust caps on, inspect before fitting, and respect the minimum bend radius, because a tight coil on a plastic lead leaks light along its length.
Working lengths by sample rate
The table gives working lengths rather than guaranteed maxima. It assumes clean end faces, no sharp bends and a receiver of ordinary sensitivity. Cheap leads vary widely, so a 10 m plastic lead from one supplier can behave like a 3 m lead from another.
| Signal | Fibre | Working length | Note |
|---|
| Stereo 48 kHz | Plastic | 10 m | Common room run |
| Stereo 96 kHz | Plastic | 5 m | At the rated limit |
| Stereo 96 kHz | Glass | 30 m | Clean termination needed |
| Stereo 192 kHz | 75 ohm coax | 30 m | Outside the optical rating |
| Compressed surround | Plastic | 10 m | Bitrate is still modest |
An optical audio cable article covers construction and connector types in more detail. The length question is settled by the receiver at the far end, not by the figure on the packaging.
When coax beats optical
S/PDIF over 75 ohm coax carries 192 kHz stereo and is not limited by light. It is a screened electrical connection with a solid centre conductor, so it survives a 30 m run that would mute a plastic lead. The catch is earth loops: joining two mains-powered devices with a coax screen can introduce hum.
Breaking the earth loop
An optical link breaks that earth path, which is the real reason it is specified between a television and an amplifier on different sockets. Use it for isolation first, distance second.
Conversion at the ends of a system is common: a TOSLINK to RCA cable with an active converter serves a soundbar with no optical input, and a S/PDIF cable of 75 ohm construction serves the coax side.
Practical scenarios
- Situation: A 12 m run from a wall-mounted television to an amplifier mutes at 96 kHz. Move to glass fibre or a 75 ohm coax lead, and keep any plastic lead under 5 m.
- Situation: A soundbar drops out when the cable is coiled behind the stand. Straighten the coil to a radius above 25 mm and clean both end faces before ordering a replacement.
- Situation: A desktop system hums through coax but not through optical. Keep the optical lead for isolation and leave the coax for the 192 kHz source that needs it.
- Situation: A 10 m plastic lead works at 48 kHz and fails at 96 kHz. That is the budget behaving correctly; step down to 48 kHz or fit a glass lead.
FAQ
Can two TOSLINK cables be joined with a coupler?
Yes, with a loss penalty. Every coupler adds two extra end faces and a small alignment error, and plastic fibre is sensitive to both. A single 5 m lead usually beats two 3 m leads joined together. Where a joint is unavoidable, keep the total under the working length for the sample rate and inspect all four faces.
Does a longer optical cable reduce sound quality?
Not while the link works. The signal is digital, so the receiver either reconstructs the samples correctly or loses lock and mutes. There is no gradual softening of the sound as light is lost. Dropouts, silence or a failure to lock are the symptoms to expect, and they appear suddenly rather than gracefully.
Why does 96 kHz fail on a lead that works at 48 kHz?
Doubling the sample rate halves the time available for each pulse, so the receiver needs more light to resolve it. The same cable and the same emitter now sit closer to the sensitivity threshold. Shortening the run, cleaning the end faces or moving to glass fibre restores the margin that 48 kHz did not need.
