Radio 4 FM Sibilance

[knobtwiddler]

Use of a de-esser for pre-recorded broadcasts is widespread, and many involved in making podcasts can attest to that. In live broadcast, there will be a variety of multi-band compression being used, designed to even out dynamics and improve intelligibility in noisy environments such as cars. The processor will have a de-ess function. I haven't worked at the Beeb and do not know what compression algorithm is used for live speech, so do not know if de-essing is part of the standard preset used. I tend to think that BBC programming has gotten 'louder' in recent times in terms of RMS value - but that might be because I'm more attuned to it. The loudest station (I believe they were fined at one point) I can think of was Kiss FM around 20 years ago. Everything was as flat as a pancake, to the point where there were no dynamics, plosives or otherwise.

[Telleadict]

Hello

I have no tech knowledge but have just thought how other stations from the same transmitter sound to you? Classic, R3 or BBC local might still have some speech content and might prove it was just the one station.

Bye
Will

[TIMTAPE]

Neil Armstrong's famous words spoken from the moon LEM ladder in 1969 carry a lot of distortion in sibilants and other
spoken consonants. But there the distortion may have served a useful purpose. The actual bandwidth of the signal is severely limited such that normal speech consonants which make for intelligibility would not have been very audible. But at some earlier point in the audio chain the speech highs became very distorted. The distortion was so bad that it folded down into a lower audio band which the system could reproduce, and presto, we hear Armstrong's consonants if only as distortion but our brain reinterprets the distortion as the largely missing consonants. Whether the distortion was there by design or accident would be interesting to know. Speech intelligibility and high fidelity aren't always the same thing. In my city you can sometimes tell blindly whether the radio station is one of the majors or a more lowly community station by the announcer's speech. I suspect often the better stations have a more experienced tech with good listening skills and who knows how to correctly set up the processor on the live voice channel. A lot of the knack in audio production seems to boil down to not just having the right tools but knowing how to use them. Sometimes the derisive comment "all the gear but no idea" sums the situation up accurately.

[paulsherwin]

I've no idea if Aussie FM broadcasters are using de-essing on their live speech output, but I'm pretty sure the BBC aren't. It's been around for a long time in the recording industry, but it used to be impossible to do with any subtlety on the fly in real time. That's no longer the case.

The Beeb have certainly pushed the modulation level much higher on FM over the last couple of decades, and that may be having unanticipated consequences.

[TIMTAPE]

It's sometimes said that the mark of good speech processing is when we dont notice it.

Sophisticated processing power has been a blessing for the competent operator but for the incompetent, perhaps poorer and more inconsistent product. More ways to mess it up.

Not radio but here's a glaring example from last week of an audio production trainwreck, which appears to suffer from multiple poor speech processing choices. Strangely the mics are fine as is their positioning. Why the need to take to such audio with a metaphorical sledgehammer?

https://youtu.be/2XMLs-D2lOw?si=gG0pAg0geTeID9Hn

[jamesperrett]

Originally Posted by PaulR

It isn't distorted so maybe not true sibilance , just emphasised "s" sounds and is apparent with Lauren Laverne's and to an extent Clive Oppenheimer's voices. I think I will just have to turn the treble down!

Having just listened to an old recording from Lauren Laverne and a recent one on BBC Sounds I think she just has a slight whistle when she pronounces an "s" which could be annoying to some people but it is just part of how she speaks.

[TIMTAPE]

Sometimes someone with an apparently excessively sibilant voice turns out to have a quiet speaking voice but with sibilants about normal in level, hence the sibilants appear excessive. Sometimes training in using the voice better helps. A trained operatic singer can develop such great power from their voice that excessive sibilants arent an issue, except perhaps when they are speaking very quietly. Almost by definition a whisper is excessively sibilant.

[TIMTAPE]

Originally Posted by TIMTAPE

Neil Armstrong's famous words spoken from the moon LEM ladder in 1969 carry a lot of distortion in sibilants and other
spoken consonants...

I remembered wrongly. It pays to check one's evidence rather just rely on memory! The "one small step" Armstrong speech: has little recorded audible sibilant/consonant distortion. There also seems very little speech audio above maybe 2 kHz. But the engineers made the best of what they had and it's intelligible speech which here was the main game I guess.
I must have been thinking of another example which hopefully will reappear from the failing memory.

https://youtu.be/xSdHina-fTk?si=VXxlrsDdvyXKXqNQ

[PaulR]

In answer to earlier questions I have just been listening to Saturday live on FM, DAB and internet radio all through the same amp and speakers, Armstrong 600 series amp and Fyne Audio F302 speakers.

FM from the Armstrong 600 series tuner sounds the best for clarity and presence but with lots of pronounced sibilants as before. DAB and internet radio were sourced from a Majority Fitzwilliam tuner. They both sounded different from FM and not as "present" somehow - a little bit like talking down a tube. The sibilance on internet radio particularly was less and if I didn't have an FM tuner to compare it with would be quite acceptable.

As a matter of interest I thought I would try Classic FM on internet radio. They were playing Barber's Adagio and it sounded pretty poor. The strings were distorted as you might expect to hear from vinyl with the cartridge a bit out of alignment.

Edit - Thinking of Paul's earlier comment about increased modulation I wonder whether these old tuners are finding it difficult to cope with that

[PaulR]

By contrast R3 on internet radio sounds very good. Rather lass harsh then FM and less sibilance