OK, I didn’t know if there was a way to theoretically calculate a voltage level from knowing the SPL and specs of the test tone to arrive at a comparison between your way and the near century of proven calculated SPL way.
TD’s the one to answer this as I’ve never taken the time to learn much about recording, though the > [clipped] +30 dB transients that some CDs have implies that it be at least at –30 dB if I’m understanding how it works and while pink noise most closely mimics music AFAIK, ergo seems the proper signal, I don’t have enough experience in these types of measurements to know other than to try it if I could.
Again, I have no functioning HIFI or HT system for years now to measure, hence the hypothetical problem and don’t care about the computer 2.1 way system; I know it’s a gutless POS and could never handle much in the way of dynamics no matter how powerful, etc., the signal chain.
Plus, I’ve nothing to measure with unless the mic from my fried dbx RTA will work accurately enough with the motherboard’s sound card. Can you even use it to measure with while it’s driving the source?
Anyway, as you can see, I’m way behind the times on some of this stuff and my response to 5th covers my concerns/reservations that based on experience I don’t believe I need a measurement to question.
GM
Well, apparently you didn’t understand what/why I was attempting nor my first response to you, so not much point in us wasting any more time with each other.
GM
Thank for the reply GM. I'm starting to see where the confusion comes from. Maybe a good number of people just don't have much experience with digital recordings.
Digital recordings have absolute values because the recording is really only abstract. It's just a pile of numbers - no magnetic tape, no grooves, no wire recorders. Of course those numbers are stored on something, but that "something" could be just about anything, even a stack of punch cards. It's still exactly the same recording.
Those absolute values mean that there is an absolute maximum signal - it's called 0dB (full scale). All other values must be below that. That 0dB becomes our reference - we know absolutely where all other values are in relation to the 0dB level. My test tone has an RMS value of 12dB below that absolute maximum. Because it's a sine wave, its peaks are 9dB below 0dB. (3dB higher than the RMS value).
CDs always have an absolute value of 0dB, the peaks can never be higher that than, no matter how it's recorded. The peaks may be clipped, but those clipped peaks are all at 0dB. CDs can be mastered with average levels just about anywhere (unfortunately). It does not matter if the music is recorded 18, 10, 35, or 16dB below 0dB, 0dB is always the maximum level. Not every track hits 0dB on the peaks, but many do. And I've never seen a CD without 0dB peaks somewhere, on one or more tracks.
Once we are in the digital realm it becomes easy to measure and calculate other values from that.
Digital recordings have absolute values because the recording is really only abstract. It's just a pile of numbers - no magnetic tape, no grooves, no wire recorders. Of course those numbers are stored on something, but that "something" could be just about anything, even a stack of punch cards. It's still exactly the same recording.
Those absolute values mean that there is an absolute maximum signal - it's called 0dB (full scale). All other values must be below that. That 0dB becomes our reference - we know absolutely where all other values are in relation to the 0dB level. My test tone has an RMS value of 12dB below that absolute maximum. Because it's a sine wave, its peaks are 9dB below 0dB. (3dB higher than the RMS value).
CDs always have an absolute value of 0dB, the peaks can never be higher that than, no matter how it's recorded. The peaks may be clipped, but those clipped peaks are all at 0dB. CDs can be mastered with average levels just about anywhere (unfortunately). It does not matter if the music is recorded 18, 10, 35, or 16dB below 0dB, 0dB is always the maximum level. Not every track hits 0dB on the peaks, but many do. And I've never seen a CD without 0dB peaks somewhere, on one or more tracks.
Once we are in the digital realm it becomes easy to measure and calculate other values from that.
GM, it is as Pano has said before, you haven't quite understood the concept of the test and how it functions. You can use it for any system with a digital source be it for HT or music. All you have to be aware of is any gain that the system could add outside of the digital 0dBfs maximum. In this way it's similar to running the test for an actively crossed pair of loudspeakers, in that you need to run the test tone at the frequency that corresponds to the point of maximum gain on your active filter.
Here's the problem. Pano's test bares no resemblance whatsoever to the 'century proven calculated SPL way', the two methods aren't even competing with one another because they are set out to do two completely different things. Pano's test is not about calculating an end SPL figure, nor is it a method of coming up with a way to see if you've got enough amplifier voltage swing given a target SPL figure based on your loudspeakers sensitivity.
Here's the problem. Pano's test bares no resemblance whatsoever to the 'century proven calculated SPL way', the two methods aren't even competing with one another because they are set out to do two completely different things. Pano's test is not about calculating an end SPL figure, nor is it a method of coming up with a way to see if you've got enough amplifier voltage swing given a target SPL figure based on your loudspeakers sensitivity.
Thanks for the test Mooly. 
From the two pink noise at 85dB tests so far we see that I hit the same volume setting and you actually set yours 2.6dB lower. I'm beginning to be a lot less worried about the "you won't set it loud enough" complaints.
That contrasts with your 5V RMS measurement when you set the level by ear, right?
From the two pink noise at 85dB tests so far we see that I hit the same volume setting and you actually set yours 2.6dB lower. I'm beginning to be a lot less worried about the "you won't set it loud enough" complaints.
the test is intended to provide a figure with one's current system. who claimed and where that it's supposed to give an absolute maximum power level regardless system, room and preference? it's the most power people use in their current homes, with the current speakers. it's a statistic basically.
it's made under the assumption that the amp is not clipping. if one turns the volume all the way up to get to the so called "party level" obviously there's a chance that clipping occurs. obviously, using a scope or sound board to ensure that clipping really doesn't occur would be nice but the test would take too long and get too complicated for most people and they would be put off by it.
I for one actually measured and I know where my amp starts clipping. the "party level" is well below that point. now if the speakers start clipping is another thing and there's no way to tell unless by ear or with measuring equipment but for the zillionth time: this is not the point. my speakers are rather good and I can listen at very high levels without fatigue so I think it's fair to say that they don't run the into compression too much. maybe that's not the case with everyone, but this being a statistic, averaging applies and hopefully there will be a sufficiently large number of people with good amps and speakers that voted. it's not like the only people with good systems are the one that voted >20V (6.5%). I would find that hard to believe.
again, totally irrelevant. provided that clipping and current limiting doesn't occur, the test frequency is irrelevant. the test frequency is supposed to tell you what's the largest voltage the amp outputs during listening at the maximum listening level. no matter how dynamic the music is, the output voltage would not exceed the measured one. it's not like some music miraculously causes the D/A converter or CD/DVD/SACD player to exceed digital maximum (0 dBfs).
I think it goes without saying that whoever reads this or is interested in it know what to listen for.
the test is supposed to be made with music. what you're saying is irrelevant. again, why do you keep repeating those 105/85 numbers? AFAIK there isn't any law that forces anyone to listen at 105 dB peak level. "realistic level" is IMO an illusory thing. most people (including myself) would not want/need the SPL of a full scale symphonic orchestra as measured in front row in their living room. for me that doesn't add enjoyment once past a certain threshold and there's no point in torturing the neighbors (not everyone lives in a house) just so I can rest assured that the SPL is realistic.
we can keep hanging on to the space shuttle launch example ad nausea but TBH I'm never interested in reproducing that SPL either. anyone can invent extreme examples like that but it's absolutely pointless.
later edit: I read the last replies only after posting mine.
Pano, 5th, this is getting almost ridiculous. like a previous poster said, this topic is filled with repetitive posts explaining obvious things. and in the end we find that the confusion comes partly from such a basic thing as lack of understanding that a D/A converter can't output levels above a certain level no matter what. what really bothers me is that someone who doesn't understand this basic thing is at the same time so bold in stating that everyone else (including you two) is wrong. I admire your patience but I think it's time to give up.
Last edited:
Correct, it is a measurement of what is, not what should be. It's meant as a snapshot of voltage levels that people are actually using in their homes. It is not any kind of setup, calibration or target level test. Not at all.
It just so happens that when you measure the voltage at your speakers, you can use Ohm's law to figure out of your amp might be clipping. That's a benefit of actually measuring what's going on.
It is NOT a test to tell you "how little power you need". Some folks have read that into it, but it is not stated in the test. It's just a real voltage test that allows you to know if you are clipping your amp when playing loud.
You could set the volume control anywhere else and the test would be valid - at that volume setting. I just asked folks to turn it up to see what voltages they get when running as hot as they ever do.
That's about it.
As mentioned above the fall down is with analogue sources, but it can be used with any number of channels in any situation where there's a digital source and you know what the gain structure/proccessing, is like of the system.
You don't know if it's already clipping when you're running the test. Perhaps it is, but if it is clipping for your given volume setting then the test will identify this and show what size of amplifier you need for it not to clip.
Well Tom said himself that on dynamic music one doesn't hear the flattened peaks as sounding bad as per say and that you need to upgrade to a more capable amplifier to hear what's gone wrong. If the user has used reasonably dynamic material, it is likely that the average levels aren't clipping and that it's only the peaks that are. This test will show this and you will then know, roughly speaking, what size of amplifier you should be looking at.
Also as per the results of the test, I think it shows that most peoples amplifiers are suitable for their listening habits and that most people probably aren't listening as loud as you might expect them to.
Well this is dependent entirely on the music and how hard you loudspeakers are to drive. It has already been mentioned that this test isn't focusing on the current the amplifier might have to provide, merely the voltage swing necessary to reproduce the transient without clipping. It is taken for granted that the amplifiers current limiting wont come in.
Well indeed, if it's that hard to detect then one could say its largely irrelevant to where someone will set their volume control and it will not have that much, if any baring on the outcome of the test.
I spec'd my system to be able to do this, all the way down to 20hz but the truth is, I never set it that loud as it's painful to listen to (with movies) and certainly not what someone wants to endure for the full length of a movie.
At no point in this thread has a 4x voltage 'margin' ever been mentioned. Will people please stop mentioning this and read the thread until they understand where this 4x is coming from, it does not represent some preconceived idea about a reasonable average to peak level and has been explained many times over. Specifically post 346.
This is an important point that bears further discussion. This is how I see it:
If you turn up the volume until you find it really loud - but your amp is clipping - how do you know you would not have turned it up louder if the amp didn't clip?
You don't. But you might be able to figure it.
If you set your volume, measure the tone and do the math, you might see that your amp is clipping - the maximum voltage your amp can supply is less than 4X the test tone voltage. You might conclude that you could play louder without hurting your ears with an amp that doesn't clip. Myself, I used an amp with so much more voltage than needed that clipping was not a problem.
Of course if the limit is your speakers, or your room, or both - you're out of luck. It's not the amp that limits you, it's speaker or acoustics. At least you'll know your amp isn't clipping.
the word is statistic.
if for instance one makes a statistic of average income, it would make sense for me to fill in my current actual income, not the income I wish or my income from 10 years ago. no?
Yes.The test is starting to move beyond that, tho. For those of you who own an SPL meter you can set the industry standard of 85dB on a -18dB pink noise signal (mine or another). You can accept that as your maximum volume or see how it compare with what you set by ear with music.
Digiphobe
I faked it. Beethoven Ninth last movement has some sustained moments which read around 5 V. System is NAD 2600 Amp, about 150 wpc, driving Fostex RM 765 coax monitors in 4pi mode, and my homebrew TL subs. Room is 9.5 H 11 W 19.5 D. I can borrow a spl meter from my prosound buddy to back this up, or join the 21st century. I confess to being a Loudite.
I am very interested in preserving these Fostex monitors so getting louder means something else for mids/highs and I'm spoiled. So it goes.
I faked it. Beethoven Ninth last movement has some sustained moments which read around 5 V. System is NAD 2600 Amp, about 150 wpc, driving Fostex RM 765 coax monitors in 4pi mode, and my homebrew TL subs. Room is 9.5 H 11 W 19.5 D. I can borrow a spl meter from my prosound buddy to back this up, or join the 21st century. I confess to being a Loudite.
I am very interested in preserving these Fostex monitors so getting louder means something else for mids/highs and I'm spoiled. So it goes.
not sure I understand what you mean.
if you did the test the way Pano described and you got a lower voltage, it can only mean that you made a mistake somewhere.
if "digiphobe" means that a digital source wasn't involved, well, we found the mistake.
just when I thought things couldn't possibly get any worse...I think he measured the amplifiers output when listening to Beethoven
musical instruments have very complex spectra and are not steady signals but have complex envelopes too. DMM's give useful readings only when supplied perfect sines and as said before cheaper ones are frequency sensitive too.