speakers for classical music

06-24-08: Shadorne said:

"So combine performance levels around 110dB with speakers that easily reach 105 to 110dB and there's no 20dB of lost dynamic range.

I don't think I am exaggerating. I don't think you will find many dynamic speakers that easily do what you say (maybe two or three?). It is extremely rare to find a dynamic non compression horn consumer audio speaker that will do 110 db SPL comfortably and without any distortion, stress or serious compression at 8 feet back ( typical listening position )."

I said 105-110dB and there's a lot of difference between 105 and 110, BUT I routinely measure 105dB peaks at my listening position, 7-feet back from the speakers' plain. They're only $3500 Vienna Acoustic Beethoven Baby Grands. My Rowland Continuum 500 is capable of 1000 clean watts RMS into their 4 ohm load and much higher peaks, so I'm not hamstrung for power, given their 92dB sensitivity.

Because I'm a musician and chose with my ears, maybe I gravitated toward a pretty dynamic speaker, BUT I don't really think that the VA's performance is all that rare in this regard.

Please realize that I'm not putting 105dB sine waves through my speakers. I suspect that'd be a recipe for disaster. Orchestral peaks, IME, tend to only last fractions of a second, then settle quickly down around 100dB and diminuendo back down to 80-something dB quickly.

Guido, BTW, I only take my SPL meter to rehearsals. I'd never think of sullying an actual performance with such techno-dweeb activity. ;-)

Dave

When I recorded Canto General, I made sure that we had the best bass drum in the Twin Cities. The score calls for nearly 40 different percussion instruments and a 60-voice choir.

The piece is spectacular, very passionate, with hints of Carl Orff. We recorded it without compression and we were at the limits of what the media could do for dynamic range.

Most people have no idea how loud an orchestra can play because there are no artificial harmonic artifacts. But having been there when it was recorded, I find it crucial to have as much efficiency as possible so long as bandwidth and resolution are not compromised. This allows the home system to play without loudness artifact also.

Dave,

Sure most dynamic speakers can probably achieve 105 db SPL peaks at 7 feet that with 1000 watts behind them. The issue is how clean and uncompressed are those peaks.

This was my point about giving a link to what many audiophiles regard as a relatively dynamic speaker with good slam compared to others - the Watt Puppy 8's are certainly no slouch when it comes to dynamics. No doubt they too can easily reach 105 db SPL peaks when driven by 1000 watts.

However, the question is how much compression must be endured to achieve this? As Ralph has pointed out 1000 watts is all very well but that is a lot of heat (since more than 97% becomes heat) and heat means thermal compression. (Ok for a subwoofer with massive 4 inch voice coil perhaps - but that tweeter and midrange can only take so much. Lower impedance drivers just mean that more current flows through the coils and create even more heat)

So I must ask what use is it to achieve the 105 db SPL peak if in reality it should have been 111 db SPL because the speaker voice coils are running hotter than a toaster oven - something to consider!

I don't think many people know about this - so you would not be alone in believeing that 105 db SPL is so "easily" achieved by most speakers at the listening position - when it is not. It is obscure and speaker manufactuers and driver manufacturers and reviewers do not appear to want to educate us about this issue. Several manufacturers add a second midrange driver on larger dynamic models for a good reason - it gets them an extra 3 db before compression hits - very little benefit in reality - and of course it includes a heavy compromise due to the lobbing effects from driver interference - but the alternative is to design drivers that play louder without compression - of which you can probably only sell a few of - so you can see which is easier ...just double up on the popular drivers also used in the lower models (and bought in large quantities) to get a modest 3 db gain in dynamics for your highest end model. They aren't likely to tell you that the added driver is largely to eek out an extra 3 db SPL, as all the marketing will speak to the beautifully "controlled vertical dispersion"...

Shardorne, I'm actually measuring those peaks, with my speakers, in my room. I don't think the peaks are lasting long enough to lead to significant compression when listening to actual music in my system.

Those graphs you showed a few days ago of some top flight speakers really didn't display much compression when compared to the frequency response graphs for the same speakers. Adding less than 1% due to compression on top of a curve that's flucuating several percent seems small and perhaps not all that obvious.

Believe me, I'll be listening to horn speakers at the next RMAF to see what I hear. Assuming that horns now produce a natural midrange, then I'll be very interested to hear the true impact of this added dynamic capacity. If I hear something I like, then I'll be following with my wallet. I already know that Duke will be there and I hope that a few others from the genre will be represented.

What a speaker can do with a sine wave for several seconds is not of great interest to me, if it doesn't correlate with my actual listening experience. The compression shown on the graphs your referenced took seconds to start building up. I think that's way longer than typical musical peaks above 105dB (live or recorded).

Dave

I don't think the peaks are lasting long enough to lead to significant compression when listening to actual music in my system.

What a speaker can do with a sine wave for several seconds is not of great interest to me, if it doesn't correlate with my actual listening experience.

I don't expect anyone to agree with me. Especially those with dynamic speakers. I think you bring up some very good points on thermal compression, such as the lengths of the peaks. For sure some music will be more forgiving than others.

Have you considered though that driver non-linearities above Xmax are another form of compression (as the voice coil travels further outside the linear region of the magnetic field with large cone excursion). Again horns tend to have an advantage there too.

A demo at RMAF is a great idea - if they will allow you to play that loud. Sheffield Labs Drum track is a good stress test.