Talk:Programme level

Not typical

I take issue with some statements in this article:

"This turns out to be irrelevant though, because a typical ‘hi-end’ loudspeaker is not capable of reproducing anything above 105 dB SPL. Even very expensive professional studio monitors only manage 110 or 115 dB SPL, and then only at 3 m listening distance!"

Maybe it's just the "typical" part that's throwing me but it's not very difficult to find loudspeakers capable of shooting way past 105dB. The Klipsch RF-83 has a rating of 100dB at 1W meaning that at 128W (not unreasonable for someone with high end speakers) they've got a rating of 121dB. I know that's a very efficient speaker, but I'm guessing many other companies blow right past the "typical" 105dB as well. I had some cheap bookshelf speakers rated at 89dB@1W with a max power of 125W meaning they could easily hit 107dB at 64W. If I'm missing or misunderstanding something, please let me know.

The "3m listening distance" line is also throwing me, because it seems such an arbitrary measurement. A one meter distance makes sense, because that's the distance you would actually measure the SPL at. - MordredKLB 16:31, 10 April 2007 (UTC)

I understand your point, clearly 'hi-end' is a vague concept, but my feeling is that what I said it true for the majority of pretty-good speakers in use in the home. This would certainly include classic speakers like the highly regarded Quad ESL63s and even the new bigger ESLs which in fact do not specify any higher level than the original ones. It would also include Celestion SL600s (84 dB/watt I seem to remember), and many of the Tannoy, Wharfdale and similar speakers, many of which are still in widespread use and currently on sale. It is true that there are less common speakers that go to higher levels. The Klipsh, with a rating of 100dB/W has to be exceptional - Klipsh have always made horn loaded speakers, which are uncommon in the home, and of course give very high efficiency by impedance matching to the air, but some would say at the expense of colouration. There are two other issues here though. Yes, speakers are measured at 1m, mostly because this provides a compromise between minimising reflections from the room (even in an imperfect anechoic chamber) in relation to the direct sound, and minimising the 'suck-out' problem that occurs because of different path lengths causing cancellation around the crossover frequency if the measuring mic is kept on the tweeter axis but close up. Lowering the mic to avoid this causes a fall off in tweeter response because it is off-axis. We typically do not listen at 1m though; 3m is more typical, and sound power falls off with the square of distance - ie by a factor of 9 at listening distance, or 10dB. In practise, the exact level will depend on room acoustics, and also on addition from two channels, though it cannot be assumed that the two channels always assist one another - if sound is mostly from one channel that that speaker has to handle it alone. The other issue is with specifications. Just because a speaker says it can handle say, 100W, youi should not assume that it can reproduce cleanly, without distortion, at this level, and over the frequency range. The spec is simply an indication that the speaker will not burn out, overheat, on typical programme material, when driven by a 100W amplifier. I have spent many hours measuring speakers in a semi-anechoic studio(as head and founder of Lindos), and years setting up a personal studio in which I could listen to a live band at one end and then listen to the reproduction from the other - the ultimate test! Most speakers sound like 'a bit dropped off' in this situation, once your ears have adapted to the band!! Even professional monitors struggle because many use an LF unit with too small a cone to achieve required level at 100Hz (another problem - covering the LF end). I say 100Hz because I cross over at this frequency to four 18 inch subwoofers in the wall - the only way I have finally achieved required levels down to 20Hz and below. In the end I chose to use the Mackie active monitors with 8 inch cones, which just manage. Other monitors such as Genelec can of course compete, as well as the bigger studio monitors. As you will see; I know this subject well, though of course this could be called original research here. I would point out though that I was invited to write the measurement section of the 'Audio Engineers Reference Book', in which the bar diagram I used here first appeared, so the work might be considered 'peer reviewed'.

Original research

I just added two tags to the top of this article, one saying that it may contain original research and the other stating that there are no references. Like the poster above, I note that there are arbitrary numbers bandied about in this article, numbers that aren't supported by references and which are easily shown to be untrue in all cases. Binksternet (talk) 17:02, 1 November 2009 (UTC)

Sorry I didn't reply sooner! The numbers are far from arbitrary. As founder and head of Lindos Electronics I have been heavily involved in measurement (see my response above) and in working with broadcasters and studios worldwide (esp the BBC and former IBA) who use my LA100 measuring set as what some refer to as the 'de-facto standard' in broadcasting. The niumbers are based on my original article that you can see at the Lindos Electronics website under 'articles' along with many other topics. Also on measurements made by me that can also be seen at that site on the 'test sheet database'. Yes, this is original research. But I did write on measurement in the 'Audio Engineers Reference Book'. Perhaps you would at this, and the Lindos articles as citations - I am not supposed to, as author, under the rules! As an expert in the field, I don't feel like searching other places for citations - they are hard to find, as manufacturers make big claims for speakers but don't release proper detailed measurements, especially of distotrion versus SPL across the full frequency spectrum!