It's not about more SPL Pull, it's about size and realism of sound ..... !!!!
OK, mini symphony sound ....
Yep, but that high sensitivity speaker better be multi-driver , multi-way and not the typical 2 way or single driver deal ...
Single drivers lol don't get me started on those, I wont touch them with a 10 foot barge pole. Okay they do have their uses, but not for any main listening system of mine.
Can you get typical 2 ways that hit the sensitivity figures we're talking about? The best candidate for this as far as I've seen is the 18 sound 6.5" with 92dB sensitivity and excellent other stats. That'd fall back to ~87dB after BSC though, decent, but still not quite there imo. A pair of them in a 2.5 way with a wave-guide loaded tweeter would do okay though and would probably represent the best you can really do for a 'standard' type design. Of course if you cut out the lows and use a sub I think that'd be rather capable.
Ha! I see that this has degenerated into the "I use more power than you" and "I listen louder than you " thread. It was inevitable, but took longer than I thought.
What does it matter? Shouldn't you listen to what you like, how you like? And if you want some sort of standard, use the SMPTE/EBU pink noise at 85dB SPL. Set your level by ear, or by SPL meter, this test will tell you what the peak voltages will be. It will NOT tell you about power compression or distortion in your speakers.
This test was just to give you an idea of where you are with your system using a simple, easy measurement. If you want to discuss what others "should be" doing, or ultimate design goals, or brag about how much power you do (or don't) use - go start another thread.
Me? I'm happy to know what voltage it takes to hit my goal of ~105dB peaks, even tho I rarely play that loud. Your system and your goals are probably different. Why not measure and find out? That's what this thread is all about.
What does it matter? Shouldn't you listen to what you like, how you like? And if you want some sort of standard, use the SMPTE/EBU pink noise at 85dB SPL. Set your level by ear, or by SPL meter, this test will tell you what the peak voltages will be. It will NOT tell you about power compression or distortion in your speakers.
This test was just to give you an idea of where you are with your system using a simple, easy measurement. If you want to discuss what others "should be" doing, or ultimate design goals, or brag about how much power you do (or don't) use - go start another thread.
Me? I'm happy to know what voltage it takes to hit my goal of ~105dB peaks, even tho I rarely play that loud. Your system and your goals are probably different. Why not measure and find out? That's what this thread is all about.
Just a few comments; first let me say I like the test and the OP has come up with a creative way to find ballpark figures for most people. One of those things where it's obvious but that doesn't mean you think of it until it's pointed out to you.
RE: the -12V RMS level
About 7 years or so ago I started digitizing albums so that I could enjoy music that isn't available on CD in the usual digital places. I also experimented with normalizing CDs to a level that matched the digitized albums perceived average level, avoiding things like "sound check" on an iPod (which didn't sound that good then, still doesn't, and eats up iPod battery life if enabled).
To probably no-one's surprise, the CDs were mastered pretty hot and generally had to be reduced in level if the goal was a consistent average level across a library of music.
The albums were recorded at 96/24 with an RME soundcard. I kept the hirez files on the drive, did any processing on a copy, and downsampled to 16/44 with dither for convenience, eg if I wanted to burn a CD for use in a car player, or if I wanted to minimize file size on an iPod (I have always used 16/44 uncompressed files on the iPod, no mp3's). Using iTunes it was simple to make a playlist with sample rate criteria to use as a library if I wanted to move a copy for use on another device or burn to disk.
Not necessarily the process you or I would use today, but this was 2004. Naturally, I still have both the original and normalized 24/96 files and they are playable over a much wider variety of digital gear now.
For albums and CDs to match in average volume, I found I had to normalize down to at least -13dB RMS. This was of course a limit set by the albums I transposed, not the CDs. Most albums would generate digital clipping at -12dB. Some albums needed -14dB RMS to avoid digital clipping, and a few need to go lower still.
Generally speaking the older the recording, the more headroom required. Also since the goal was average volume consistency, there were occasions where a few songs on relatively new albums had more dynamics so the entire album was normalized to match at lower than -13dB.
Depending on your listening habits, some might consider calculating your maximum necessary power to a higher multiple; -12dB (4x power) may not be sufficient in all cases for all kinds of music. It almost certainly wouldn't be sufficient if you have non-digtal sources (the difference is related to mastering practices adopted with digital recording).
For reference -14dB would mean 5x (1v @ -14dB = 5.011872v @ 0dB). If you want to add a "fudge factor" (and I think it's wise to do so) you could calculate your needs with different values. A recent post suggested 18dB or 8x (1v @ -18dB = 7.943282v @ 0dB) which seems reasonable to me. None of this addresses the question of LF energy power requirements (eg below 220 Hz) but that is beyond the scope of this "quick-and-dirty" but none the less interesting test.
RE: the -12V RMS level
About 7 years or so ago I started digitizing albums so that I could enjoy music that isn't available on CD in the usual digital places. I also experimented with normalizing CDs to a level that matched the digitized albums perceived average level, avoiding things like "sound check" on an iPod (which didn't sound that good then, still doesn't, and eats up iPod battery life if enabled).
To probably no-one's surprise, the CDs were mastered pretty hot and generally had to be reduced in level if the goal was a consistent average level across a library of music.
The albums were recorded at 96/24 with an RME soundcard. I kept the hirez files on the drive, did any processing on a copy, and downsampled to 16/44 with dither for convenience, eg if I wanted to burn a CD for use in a car player, or if I wanted to minimize file size on an iPod (I have always used 16/44 uncompressed files on the iPod, no mp3's). Using iTunes it was simple to make a playlist with sample rate criteria to use as a library if I wanted to move a copy for use on another device or burn to disk.
Not necessarily the process you or I would use today, but this was 2004. Naturally, I still have both the original and normalized 24/96 files and they are playable over a much wider variety of digital gear now.
For albums and CDs to match in average volume, I found I had to normalize down to at least -13dB RMS. This was of course a limit set by the albums I transposed, not the CDs. Most albums would generate digital clipping at -12dB. Some albums needed -14dB RMS to avoid digital clipping, and a few need to go lower still.
Generally speaking the older the recording, the more headroom required. Also since the goal was average volume consistency, there were occasions where a few songs on relatively new albums had more dynamics so the entire album was normalized to match at lower than -13dB.
Depending on your listening habits, some might consider calculating your maximum necessary power to a higher multiple; -12dB (4x power) may not be sufficient in all cases for all kinds of music. It almost certainly wouldn't be sufficient if you have non-digtal sources (the difference is related to mastering practices adopted with digital recording).
For reference -14dB would mean 5x (1v @ -14dB = 5.011872v @ 0dB). If you want to add a "fudge factor" (and I think it's wise to do so) you could calculate your needs with different values. A recent post suggested 18dB or 8x (1v @ -18dB = 7.943282v @ 0dB) which seems reasonable to me. None of this addresses the question of LF energy power requirements (eg below 220 Hz) but that is beyond the scope of this "quick-and-dirty" but none the less interesting test.
Last edited:
Thanks Johnny for the info on your digitized albums. I've have not done much of that, so it's nice to have a reference. How far above average level did you find the ticks and pops on vinyl?
I'm not sure from your post above, but you may be falling into the "-12dB reference trap", a lot of posters have. In no way is the -12dB level any sort of headroom or level reference. Nowhere is that mentioned in the test. The test tone is at -12dBfs simply by convenience. It could be anywhere from -3dBfs to -82dBfs. You do NOT set your levels by the tone, you use the tone to measure your levels.
Just want to be sure that is clear. Maybe I need to go back and edit the test post to make this clear.
I'm not sure from your post above, but you may be falling into the "-12dB reference trap", a lot of posters have. In no way is the -12dB level any sort of headroom or level reference. Nowhere is that mentioned in the test. The test tone is at -12dBfs simply by convenience. It could be anywhere from -3dBfs to -82dBfs. You do NOT set your levels by the tone, you use the tone to measure your levels.
Just want to be sure that is clear. Maybe I need to go back and edit the test post to make this clear.
Supposing that it's a perfect power amp (flat output voltage at any phase or load), and assuming that 120Hz is the worst load.
"People make an awful lot of fuss nowdays, don't they, about the quality of the sound they listen to, have you noticed? They spend all that time trying to get the exact effect of an orchestra actually playing in their sitting room. Personally I can't think of anything I should hate more than to have an orchestra actually playing in my sitting room." -- Michael Flanders.
Hey a.wayne,
I just measured one of the speakers isolated and got 90db at 2.83v into 8ohms (1w). Which suprisingly is spot on for their spec
How were you calculating 86db/1w/1m? Just curious.
Hey a.wayne,
I just measured one of the speakers isolated and got 90db at 2.83v into 8ohms (1w). Which suprisingly is spot on for their spec
How were you calculating 86db/1w/1m? Just curious.
Hello Rev,
I would have to go back and review, but it was based on the numbers you had posted up , as to db, distance and power being used...
If your speakers are 90db/1w/1m that is pretty sensitive..... so with a typical listening distance of 3M you would produce 81db with that same 2.83v..
-12 db reference trap
Is headroom really so far off topic? It's on the short list of amplifier characteristics, as we actually use them, and has been for decades. And thanks for this test. A perfect starting point for those of us who are ready to move past our ears, and thus our beliefs, to data. I wanna be a meter reader!
Is headroom really so far off topic? It's on the short list of amplifier characteristics, as we actually use them, and has been for decades. And thanks for this test. A perfect starting point for those of us who are ready to move past our ears, and thus our beliefs, to data. I wanna be a meter reader!
Not sure what you mean by the above, but a number of people posting here have assumed that the -12dB of the test tone has something to do with headroom or dynamic range. It does not. It's simply a tone that is easy to measure and then maybe calculate your head room. That's all. Just want to be sure that is clear to everyone.
this is the exact opposite of this thread, but thought I'd mention it. I measured the minimum audible level on my system. I chose a 3kHz sine (that's where the ear is most sensitive according to Fletcher-Munson curves) amplitude-modulated with a triangular envelope of a few Hz (similar to the tone used in medical hearing tests) and got 0.13mV. that corresponds to 4nW (!) into 4 ohm. deriving SPL from there I got about -8dB SPL which seems to be the absolute minimum in the F-M curves.
obviously I could not directly measure that voltage but with my rather twisted method I trust the result to be pretty accurate.
obviously I could not directly measure that voltage but with my rather twisted method I trust the result to be pretty accurate.
Last edited:
