To set audio levels for an Indian FM radio feed streamed to YouTube, first identify exactly where the stream takes its audio from, then measure and adjust that online signal at the encoder input. Keep the FM transmission processor, the streaming encoder’s input level and YouTube’s viewer-side playback behaviour separate; none is a substitute for the others.
There is no YouTube-published integrated loudness target for live audio in the encoder guidance. Use a clean stereo feed, prevent clipping, meter loudness and true peaks, and test the complete chain with the programme material you actually broadcast.
Map the FM-to-YouTube signal path
Start by tracing the signal from the station’s programme output to the encoder and then to YouTube. The encoder may receive a clean programme output before the FM processor, a post-processor output intended for transmission, or audio captured from an off-air receiver. These points can sound similar in a short check but behave differently across speech, music and transitions.
Write down each stage and its connection: source output, any distribution amplifier or mixer, audio interface, encoder input, and the YouTube player. Note where volume controls and meters sit. A meter on the FM processor does not tell you what level reaches the encoder after a cable, interface, or separate gain control. Likewise, a healthy-looking encoder meter cannot tell you whether the station’s transmission chain is already making the programme sound harsh or compressed.
If you can choose the feed point, ask the station engineer whether a clean programme output or a separately conditioned digital-distribution output is available. EBU radio guidance recommends conditioning the digital distribution path independently of FM processing. That is useful engineering context, not a rule for an Indian YouTube stream. If the only accessible source is the FM processor output, you can still stream it, but listen critically before applying more processing.
Do not infer electrical compatibility from a connector alone. Confirm whether the output is balanced analogue or AES/EBU digital, and whether the receiving device expects line level, microphone level, or a digital input. Prasar Bharati’s January 2023 audio processor specification, for example, describes an analogue stereo output adjustable from −4 dBu to +20 dBu peak at 100% peak modulation, alongside digital interface details. Those are equipment interface specifications, not a stream loudness target. Check the actual station and interface manuals before connecting or setting gain.
A useful first exercise is to compare the station’s intended online output with the signal that actually appears at the encoder. Listen at a conservative level and watch the encoder’s input meter while the station plays representative material. If the encoder is silent, distorted or unexpectedly low, stop and check the route and input selection before changing processing. A wiring or reference-level mismatch will not be fixed by an aggressive limiter.
Choose a clean stereo online output
For a radio feed, the preferred starting point is a clean stereo source that preserves the programme without adding another layer of FM transmission processing. If the station has a separate distribution output, make clear which processing is applied to it and whether its level is fixed or adjustable. Ask before changing station equipment settings; a control that appears to affect only the online feed might also affect the transmitter.
If you are receiving a post-FM output, do not assume it is unsuitable. It may already be the station’s normal programme sound, and duplicating its processing could make it worse. Listen for pumping, flattened dynamics, brittle high frequencies, and transient distortion. A short loudness meter reading cannot tell you all of that. Keep an unprocessed comparison, if possible, so you can hear what any adjustment actually changes.
YouTube’s live encoder guidance recommends AAC or MP3 audio, with 44.1 kHz sampling and 128 kbps for stereo audio. These are encoding recommendations, not a loudness specification. Set the encoder to match the source and the platform guidance, then verify that its audio input is the intended stereo pair rather than a mono channel or the wrong interface input. Consult YouTube’s current live encoder settings before the broadcast in case the guidance changes.
If the station supplies only analogue audio, use a suitable interface and correct input type rather than plugging a line output into a microphone input and compensating with software gain. If it supplies AES/EBU, use a compatible digital input and confirm the receiving equipment’s reference settings. The goal is a stable route with gain controls you understand, not the largest possible meter reading.
For a stream assembled from scheduled content rather than a continuous station output, keep the audio route and level consistent across items. A playlist can expose mismatched recordings or abrupt transitions even where the live station feed sounds steady. The practical checks in this guide to scheduling songs for a 24/7 YouTube radio stream are useful when programme changes and hand-offs are part of your stream.
Set and meter encoder input levels
Set gain in stages. Begin with the source output and interface input, then check the encoder’s input meter. Adjust the stage that is actually causing a problem rather than raising or lowering every control at once. Leave headroom so ordinary peaks do not reach digital full scale; do not chase a particular universal LUFS or peak number that YouTube has not specified.
Use a loudness meter to understand perceived programme level over time, and a true-peak meter to identify peaks that can exceed the sample peaks shown by a basic meter. Peak readings alone do not describe whether a programme sounds consistently loud; an average loudness reading alone does not show whether individual transients are at risk. EBU R 128 uses programme loudness and Maximum True Peak Level descriptors, providing a useful measurement framework. It is not a YouTube requirement.
The EBU’s R 128 version 5.0 recommends average programme loudness of −23 LUFS with Loudness Range and Maximum True Peak Level descriptors. Its radio supplement discusses distribution loudness choices: where distribution differs from −23.0 LUFS, it generally advises a range between −20.0 and −16.0 LUFS. These are European broadcast references, not binding targets for an Indian FM simulcast or requirements published by YouTube. If your station follows EBU practice, label that clearly as your chosen technical reference and confirm which distribution guidance it has adopted. See the EBU R 128 guidance for the source documents.
The YouTube live encoder page gives audio encoding settings but no incoming integrated LUFS target. Avoid claims that a livestream must be set to −14 LUFS or −23 LUFS for YouTube. A sensible online level is one that remains clean and consistent through the full chain and passes representative listening checks. If your feed is clean but quieter than you intend, make a cautious output adjustment, retain headroom, and test again rather than applying a limiter simply to make the meter busier.
Check for overload at more than one point. If the source output is already distorted, reducing gain later may make the distortion quieter but will not undo it. If the interface input clips, reduce its input gain or correct the source reference. If only the encoder output is clipping after a software gain increase, lower that gain. Change one control, replay the same passage, and note what changed so you can reverse a bad adjustment.
Distinguish FM processing from stream processing
FM broadcast processing is designed for a transmission chain; encoder input gain is simply the level at which the online chain receives audio. FM modulation is not a measurement of integrated loudness at the streaming encoder, and an equipment specification expressed relative to peak modulation is not an online target. Treat the station’s processor output level and the encoder’s digital level as different measurements at different points.
A post-FM signal may already have substantial compression or limiting. If you apply another aggressive compressor or limiter without listening and measuring, you can increase pumping, flatten musical dynamics, make speech tiring, or emphasise distortion that is already present. Avoid stacking processing by default. First identify whether a fault is caused by the source, a gain mismatch, or the encoder configuration.
Where routing allows it, use a digital-distribution feed conditioned separately from FM transmission processing. EBU R 128 s3 advises that radio streaming follow its streaming guidance and highlights the value of separate conditioning for digital distribution. That separation gives the online path room for choices appropriate to its output without altering the FM chain. It does not mean that a particular preset or loudness value is right for every station.
When a separate feed is not available, keep any additional processing modest and purposeful. For example, if speech level varies because of a genuine source inconsistency, investigate the programme source and station settings before using stream-side compression to mask it. If the stream clips only on a particular input, correct that input’s gain. If it sounds harsh throughout, compare against the station source before changing equalisation or limiting. Make one adjustment at a time and preserve a way to restore the previous sound.
If your stream is a pre-recorded loop rather than a direct station feed, the source files may have different levels from one another. A practical guide to streaming a pre-recorded video on repeat can help you think about the programme sequence and hand-offs; for audio, still measure the assembled output rather than assuming each file will match. The same principle applies to a live radio source: the full route matters more than the loudest isolated clip.
Check playback-level changes on YouTube
YouTube may alter how audio is presented to a viewer. Its Stable volume feature balances quiet and loud portions on supported videos, is on by default, is not available on every video, and is off for YouTube Music and official music videos. This is viewer-side playback behaviour, not a correction to the signal you send from the encoder. Read YouTube’s Stable volume help page for the current availability details.
Do not rely on playback processing to repair clipped or poorly staged audio. It cannot recover clean transients lost before encoding, correct a wrong input connection, or make a station’s online feed consistent on every device. Nor should you judge the encoder level solely from a viewer’s phone volume slider: the device, player state and listener’s own settings all influence perceived loudness.
During a test, listen to YouTube’s actual preview or player on another device, ideally using headphones or speakers you know. Compare it with the source at a sensible listening level, not by turning one device much louder than the other. A small level difference can make a feed seem more impressive without making it clearer. Check whether speech remains intelligible, music stays natural and station IDs do not jump out or disappear.
If the encoder input sounds clean but the YouTube playback sounds different, check the encoder’s selected audio source, encoding settings and the player path before adding processing. Confirm that the stream is stereo as expected and that no software mixer or device enhancement is altering the signal. The OBS versus hosted streaming options for a 24/7 YouTube stream discussion is useful when deciding where your ongoing encoding and monitoring should take place; whichever route you choose, verify the audio at the player.
Test representative programme material
A test built from one quiet song or a single voice announcement cannot reveal how a continuous radio feed behaves. YouTube recommends testing before the event with audio and movement similar to the eventual stream, checking the preview and monitoring quality. For an FM feed, include speech, music, a station ID or jingle, a quieter passage, a louder passage, and transitions between programme sources. Use enough material to hear whether the chain settles consistently rather than reacting badly to one change.
Watch the encoder meter while you listen. Record loudness behaviour over the test and note the highest true peaks; do not treat one brief reading as proof that the entire programme is level. Then check the Live Control Room preview or player on a second device. YouTube’s live streaming tips also recommend monitoring audio and video quality and reviewing messages during the stream.
Keep a short test log: feed point, interface input, gain settings, encoder audio format, meter observations, and anything you heard. If the same passage clips again, you have evidence to locate the stage rather than guessing. If level changes only at a station ID or source hand-off, inspect the switching or the material at that point. If it pumps across all content, compare a clean feed with the post-FM output before reaching for more compression.
Re-test after each meaningful change. If an adjustment fixes one loud passage but makes quiet speech hard to hear, the chain may need a better source route or a more balanced gain structure, not more output gain. If the stream is clean but slightly lower in playback than another channel, resist matching by ear through uncontrolled volume changes; compare at matched listening levels and decide whether your own consistent presentation is preferable.
For a channel that must continue when your own computer is off, StreamNeo can remove the need to keep a local computer running the broadcast after you have prepared the file and channel, leaving you to focus on checking the programme and audio before it goes live. It is for uploaded video streams and is YouTube-only, so a station relying on a live FM input should confirm that this matches its source and workflow before choosing it.
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
What LUFS should I use for a YouTube radio livestream?
YouTube’s live encoder guidance does not specify an integrated LUFS target. Measure loudness and true peaks, preserve headroom and choose a consistent online level through representative listening tests. EBU figures are technical references, not YouTube requirements or Indian legal rules.
Should I use the FM processor output or a separate feed?
Use a clean programme or separately conditioned digital-distribution output if the station can provide one. If you must take the post-FM output, listen for pumping, harshness and clipped transients before adding any stream processing. Confirm the connection and reference level with the station engineer.
Why does my radio stream sound distorted or pump on YouTube?
The cause may be clipping at the source or interface, a gain mismatch, or FM processing followed by another aggressive compressor or limiter. Check each stage and compare the feed before and after the encoder; lower gain where overload occurs rather than trying to hide it later. Test the same representative passage again after each change.
Does YouTube Stable volume set the stream’s audio level?
No. Stable volume is a viewer-side feature that balances playback on supported videos, and it is not available everywhere. It does not replace clean encoder levels, correct routing or a representative end-to-end test.