Whatever else you do, never underestimate the amount of havoc a moronic user can wreak upon perfectly good gear. Unfortunately, I speak from personal experience, the things some of my customers did are beyond the wormhole and a quasar.
One of them connected a welding arc to the output of a power line filter of mine, expecting to get "a cleaner, tidier welding point", and was irritated that the fuses kept blowing. Such dummies are rare, but they do exist and create.
One of them connected a welding arc to the output of a power line filter of mine, expecting to get "a cleaner, tidier welding point", and was irritated that the fuses kept blowing. Such dummies are rare, but they do exist and create.
In many forms of music, realistic programme content has transients at and beyond such SPLs. So if one is serious about it, the playback system needs to be capable of reproducing it, and doing it well. Not all content is designed to listened to at such levels, but some is, including classical. Typical symphony orchestra peaks about 106dBC near the conductor, for example -it's loud!
I know what this means in terms of power headroom. Normal programme material has mean level about -12dB below peak, as a rule of thumb. The system should not run out of headroom at realistic levels, however the vast majority of audiophile systems do..........
It's difficult to achieve. Big, good amps and speakers and well designed rooms. Limiters........pah
well no one other than the conductor (who has hearing damage based on surveys) hears at that level. back in the row 5 sweet spot you'll be down a bit. If KHM has designed for 107dB at 1m that's close enough for me and louder than I am likely to go. BUT some idiot will turn it to 11 if he can.
Where do you get the -12dB mean from? I don't trust the DRM plugin for foobar, but that gives RMS of around -16 for rock and -20 to -30 for classical. I have no idea how it gets its DR number though.
Where do you get the -12dB mean from? I don't trust the DRM plugin for foobar, but that gives RMS of around -16 for rock and -20 to -30 for classical. I have no idea how it gets its DR number though.
The reality is that the cone driver is only the equivalent of a 6 1/2" speaker so 107db is a lot to ask of most any other speaker of that size that has the range this speaker does. It is a long gap motor so it doesn't have high efficiency, it is made for low distortion not maximum output. The actual limit of the speaker is the surrounds size really, that is the limiting factor. If one of my ideas is fruitful I can make a surround that will allow more of the actual excursion possible with the motor but right now I can't so won't say I have done it. It is not a power handling problem, it is not the voicecoil slamming into the back plate, it is only the limitation of the surrounds dimensions. At 107db that is about half the rated power handling of the voicecoil so headroom from the amplifier isn't an issue nor the thermal limit of the speaker. I have some outlandish ideas for a new surround design but I can't say it will work until I make it and test it. I have high hopes that it will work and then perhaps this small little speaker could get up to 115db, if that isn't enough well to bad, and I wouldn't want to be in the room anyway. As it is two of these speakers running flat out send most people running for the door, they are more than loud enough for anyone I have demo'd them for.
When you start pushing that hard (105 dB) you run into dynamic compression. Everything from the voice coil heating up to the cone warping and the surround inverting. With the smaller drivers these are real issues. I have had to sit through tuning sessions that were at an average of 105 dB (Thank god for hearing protection) so don't think its not something people will do.
Most products in this class (self amplified speakers) have DSP, digital amps and digital crossovers. Usually limiters are done in the chain driver by driver. They do work pretty well once you understand how to tune them.
However you need an open mind to go down this path. If you think high quality audio requires passive crossovers and linear amps this is not the road to take. its curious that what was the holy grail 20 years ago, electronic triamplification, has surfaced in the under $300 self powered speaker market before becoming common in the premium market.
Most products in this class (self amplified speakers) have DSP, digital amps and digital crossovers. Usually limiters are done in the chain driver by driver. They do work pretty well once you understand how to tune them.
However you need an open mind to go down this path. If you think high quality audio requires passive crossovers and linear amps this is not the road to take. its curious that what was the holy grail 20 years ago, electronic triamplification, has surfaced in the under $300 self powered speaker market before becoming common in the premium market.
Thanks Demian. I have thought myself that 107db is plenty myself, it just gets to the stupid level at some point. The I can turn mine up to 11 crowd. At least I don't have to think beyond a bi-amp system, only two devices. And even the argument about dynamic range comes to the question of how loud are the dynamic peaks of even a loud passage of a piano or other instrument? When someone is tickling the ivories in a normal level sitting in your home is the piano really that loud. I have a stand up piano in the house and I just can't imagine that I can hit the keys so hard as to hit some of the levels that I hear. Perhaps a grand piano with the sound board up can get fairly loud but what is a real number for normal music level. Now when you start talking about a trumpet or other horns that i have more respect of how loud that can be but that is usually when you are very close to the end of the horn.
Member
Joined 2009
Paid Member
Funny routine with those Marshall amps being modded to "go to 11." But I've recently been researching vintage Fender tweed amps, and discovered that those old things could go beyond 11 and all the way up to TWELVE!!!
An externally hosted image should be here but it was not working when we last tested it.
se
I think you might be surprised at how large natural SPL transients can be.
It doesn't seem loud during natural listening because it doesn't distort. Distortion is the main cue to perception of unnatural levels. This is one reason we might consider such levels to seem unnaturally loud during playback of recordings, and not in row 5 of a classical concert. Another is the cues from room reflections which can be unnatural.
A crest factor of about 12dB is just a studio rule of thumb for what to expect in real audio material. So, for metering, rms 0dBFS rms needs at least a level headroom of +12dB, or put it another way 0dBFS peak requires rms programme material not to exceed -12dBFS. IME this is what one finds in commercial music production, sometimes less due to pre-compensation for limitations of real playback systems or 'loudness wars'.
This is not the 'No 11' brigade, it is about proper playback and presentation of recorded programme material at realistic levels. In the absence of which, realism is compromised. And yes, one always needs to be careful about hearing health, especially during extended listening for rock and other content at realistic levels. Though often studio rock content is not mixed or intended as though to be presented at natural levels IME.......nevertheless, some is and so a decent playback system should cope without compromise IMO.
But what are the ways of us mere mortals calculating that? It seems way too low based on digital VU on foobar, but I don't know how much I can trust those either. Maybe I just have to splash for Jriver.
Why use rules of thumb from the 70s when we must be able to accurately calculate it for ourselves based on real music?
Why use rules of thumb from the 70s when we must be able to accurately calculate it for ourselves based on real music?
107 dB SPL at 1m works out to about 101 dB SPL at 3 meters. For me, even 96 dB SPL 3 m is too loud to listen to the whole day. My wife faints at less than 93 dB SPL 3 m, her hearing is quite e bit more sensitive than mine (which is why I say she hear the paint dry), but on the other hand, she doesn't hear half of what I hear. Difficult to have both.
It hasn't changed. 0dBu always has been, and still is, based on a true rms reference level. So proper metering and peak levels should still be right. A quick trawl of the web seems to cite values for crest factor of between 12dB and 20dB, and I'm sure one could find exceptions even so. However, one should allow at least 12dB margin, as a rule of thumb.
It's important not to confuse transient peak and rms listening levels.
Easier said than done.
Bill, you have to verify the operation of your meters (type and integration time) so you know what they are registering. Then, what do you mean with “real music” ?
http://www.av-iq.com/AVCat/images/documents/pdfs/MeasDigAudLvls.pdf
https://mikeriversaudio.files.wordpress.com/2010/12/meter-madness_1-2_revised.pdf
Planet Analog - Articles - Signal Chain Basics #73: Audio metering, enjoying the VU
Notice the distinction btn “Headroom” and “Peakroom”:
http://www.rane.com/pdf/ranenotes/No_Such_Thing_as_Peak_Volts_dBu.pdf
http://www.diyaudio.com/forums/minidsp/230470-signal-level-minidsp-x-over-3.html#post3412715
Calibraton using noise is complicated
http://www.meyersound.com/pdf/cinema_technical_papers/cinema_calibration_tech_report.pdf
George
Easier said than done.
Bill, you have to verify the operation of your meters (type and integration time) so you know what they are registering. Then, what do you mean with “real music” ?
I mean I have a digital music file and want to know what the real RMS to peak ratio of that file is. Nothing harder than that.
- Status
- Not open for further replies.