BBC Reply to complaint about limiting on Prom 1 sound

[Chris Newman]

I am not a hi-fi expert and can only base my comments on what I have listened to over some forty five years of Promming and have seen dangling around the performing area. When I first attended there were two microphones visible above the head of the conductor and a couple of mikes above the arena for ambience. Listen to old recordings on the BBC Legends label and my memory of radio and the sound and balance are and were excellent. Before the Verdi Requiem began last night after three tries I counted 51 visible microphones (there were probably more but that was an average). These mikes are supposed to pick up particular parts of the ensemble which of course they do but if each microphone is also picking up a vestige of the ambient sound (as it will) the final result becomes nearer to a sonic blancmange than the older recordings. The old radio broadcasts had a far wider range of volume and a more realistic sound than the overworked bland digital sound we are given today. I guess the middle-of-the-road tastes of pop/pap recordings have affected the preferences of sound engineers these days.

[johnb]

Might a better comparison be to other massive choral works recorded in RAH, instead of this year's Proms considered as a whole? And what about last night's Requiem; are you disgruntled about that too?

OK, you asked for it:

Prom 13, Verdi Requiem


Prom 4, Gothic Symphony


Prom 1, First Night


Draw your own conclusions.

My problem with Philip Boyce's response is that he quite evidently did not investigate my complaint but either had a cursory listen or just drafted a dismissive reply without having done anything at all. The facts speak for themselves. (I complaint was about Prom No 1 partly because I feared that this could set the pattern for the proms as a whole.)

The graphics for the proms are derived from the captured aac Radio 3 HD Sound streams. (The HD Sound 'flv' stream was captured, demuxed to extract the aac stream which was then decoded to a wave file which was then opened in Sony's "Sound Forge" audio editing software. The graphics were created by "Print Screen" and then cropped.)

As I said before on another thread, the likely explanation is that the sound engineers believed they had to implement significant dynamic range compression and limiting to be able to cope with the massive forces employed in Prom 4 (The Gothic) and that Prom 1 was a trial run to test how it performed - a wise procedure in the circumstances. No other Proms have been affected. (It would have been nice if Mr Boyce had taken the trouble to find that out, of course.)

The criticism I have about the actual dynamic range compression and limiting employed is not that it was there at all (as something similar was obviously required for The Gothic) but that, for what-ever reason it seems to have been crudely implemented, e.g. why the limiting was set to -10.5 dB for the Gothic (once again the 320kbps aac HD Sound stream) is a mystery to me. Surely the purpose of a trial run is to be able to make adjustments - but then it is possible that they had significant problems getting everything working properly and, once set up, they were afraid to make further adjustments.

[doversoul1]

Dear Administrator
Please, can the word ‘whinge’ be censored from this Forum?

[Simon B]

My problem with Philip Boyce's response is that he quite evidently did not investigate my complaint but either had a cursory listen or just drafted a dismissive reply without having done anything at all. The facts speak for themselves.

I quite agree and this (along with Cavatina's unnecessarily combative responses evidently informed by little understanding of the most basic technicalities even at a totally non-anorak level) is what grates with me, more than the rather inadequate broadcasts themselves.

Those graphs are glaring in the obviousness of the conclusion anyone capable of interpreting them would have to come to.

So the problem with this:

Besides, in what sense is it a "brush off" if Boyce took the time to listen to the broadcast, evaluate what you said, and flat-out, straight-up tell you that YOU'RE WRONG? No weaseling about there.

is that he did indeed "evaluate" and flat-out, straight-up declare "YOU'RE WRONG" but unfortunately, the evaluation couldn't have been based on any actual material evidence because this flatly contradicts the conclusion.

Oh. I didn't realise meter readings off an individual radio could be in any sense considered authoritative.

Yes they can, at least in the case of all the digital delivery methods (DAB, streaming, terrestrial and satellite TV). For these purposes, everyone will receive an essentially identical copy of what the BBC sends out, and they can be measured (as above) easily and unambiguously.

And anyway, "limiting" compared to what?

Relative to what the broadcast system is, by design, capable of and normally used for. (Which by the way, is miles short of anything even an easily pleased audiophile could possibly get worked up about. We're talking about a huge chunk of the dynamics being flattened out and thrown away for no good reason, not some esoteric minuscule refinement). It's not unreasonable to ask why the BBC in 1959 with a couple of microphones and a few yards of bell-wire could capture Mahler 8 in the Albert Hall more convincingly than similar works 50+ years later.

In that case, I'd be willing to bet the engineers made their recording choices deliberately for damn good reasons.

This is also a non-sequitur, as there's absolutely no reason to limit the dynamic range on a CD to anything like the degree needed for broadcasting, let alone clamping it a whole extra 10dB down for no good reason. If there was any prior intention to make a CD release then the unadulterated output from individual microphones ought to have been recorded to allow subsequent mixing down.

Just to be clear, while I'm very glad I was in the hall for the Gothic (and it could never be remotely the same on any recording) I think the balance engineering on the broadcast is actually very good under the circumstances. It sounds so much clearer and more pleasant than the rather nasty sonics of the (enthusiastically played) Lenard CD. Hopefully, some semblance of dynamics can be restored too in the mooted CD release.

Finally, whatever the reason for the severe dynamic ceiling on those transmissions was, blaming the engineers (we usually end up as the blame sponges for the consequences of decisions taken by managers who are slow to listen to boring facts and quick to blame us for the consequences we correctly predict - not that I'm bitter!) is probably the wrong way to go. I'd imagine that these days, they have about third of the staff, half the budget and quarter of the time they need to do the job properly.

[old khayyam]

the volume settings during this day’s broadcast was capped

The grammar in the first line gave me a good idea of what we are dealing with here.

[PhilipT]

I hate to ask, but could someone explain, in a non-anorak way, how this compression/limiting is applied and how the result sounds? The graphs look horrible, but I'm familiar with the hard clipping you get when my church choir sing too loud for the PA system, and presumably the result wasn't as drastic as that. My non-audiophile DAB hi-fi doesn't ever sound like that. (Ma Vlast on the car FM radio was so bad I had to turn it off, but I don't think that was the BBC's fault!) I get the picture that the system says "ooh! it's getting loud - I'll turn the gain down for a bit". How long is the bit? Or is it more sophisticated than that?

[cavatina]

OK, you asked for it:

The criticism I have about the actual dynamic range compression and limiting employed is not that it was there at all (as something similar was obviously required for The Gothic) but that, for what-ever reason it seems to have been crudely implemented, e.g. why the limiting was set to -10.5 dB for the Gothic (once again the 320kbps aac HD Sound stream) is a mystery to me. Surely the purpose of a trial run is to be able to make adjustments - but then it is possible that they had significant problems getting everything working properly and, once set up, they were afraid to make further adjustments.

Thanks very much for your reply; I apologise for being dismissive. You're right--it is odd, but I'm sure someone out there has a reasonable explanation for you. Why don't you try to find the contact information for an engineer or two and raise your concerns directly?

Believe it or not, this evening as I left the concert, I briefly asked someone emerging from the recording trailers about it in passing and they confirmed that yes, they did use dynamic compression. Maybe tomorrow I'll see if I can get an email address for you without making a complete nuisance of myself.

Those graphs are glaring in the obviousness of the conclusion anyone capable of interpreting them would have to come to.

Of course you're right; I can interpret them just fine. And if I'd seen them before I posted, we all would have saved ourselves a whole lot of hot air, wouldn't we. My fault--sorry.

It's not unreasonable to ask why the BBC in 1959 with a couple of microphones and a few yards of bell-wire could capture Mahler 8 in the Albert Hall more convincingly than similar works 50+ years later.

So why do you think this is, and in what respects do you find yourself less convinced? Are you talking solely about "operator error", or do you have a problem with something inherent in the nature of the new technology itself? Seriously, it's an interesting question.

[PJPJ]

Listening to a selection of Proms recordings this afternoon has reinforced the perception of differences in recording quality. As the "Gothic" is due for release shortly, it occurred to me a good off-air recording wouldn't do much for sales. Nonetheless, the response from the BBC about Prom 1 is otiose, as is its author, perhaps. Or he may be undiscerning. A lollapalooza such as the "Gothic" Prom surely deserved better at source.

This afternoon, I've listened to Bridge (Rebus and Enter Spring), Holloway (Fifth Concerto for Orch) and Holst (Invocation) and the recordings from the HD stream sound much more like the BBC of old.

The greater number of microphones used has one advantage over the Mahler 8 and Shostakovich 8/Mozart 33 both on BBC Legends, for example: there is far less audience noise audible during the performances.

I do hope NMC (or A N Other) releases the Holloway on CD.