When running ~ 2K watts + to a subwoofer - what gauge speaker wire should be used? - Page 5 - AVS Forum | Home Theater Discussions And Reviews
Baselworld is only a few weeks away. Getting the latest news is easy, Click Here for info on how to join the Watchuseek.com newsletter list. Follow our team for updates featuring event coverage, new product unveilings, watch industry news & more!



Forum Jump: 
 13Likes
 
Thread Tools
Old 04-16-2015, 02:28 AM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
not quite a sine wave. this one at about 23hz. with only a few breaks, it runs the length of the whole track. watching the spectrum, it appears to be mixed in about 30db hot relative to the other content (encouraging the listener to keep the volume turned up). with a sub tuned to 23hz, it could be a coil cooker.





(spectrumlab - Psyph Morrison - Dope)
i don't want to post the link because the video image contains "bad words" youtube video: lK7WN7yD1-4


this kind of stuff is increasingly in what folks listen to, so other than a "smart amp" that somehow governs average current over time, what to do? is there anyway to fuse a sub without castrating it?
Attached Thumbnails
Click image for larger version

Name:	spectrum - Psyph Morrison - Dope.png
Views:	216
Size:	167.4 KB
ID:	669377  

Listen. It's All Good.

Last edited by LTD02; 04-16-2015 at 02:32 AM.
LTD02 is offline  
Sponsored Links
Advertisement
 
Old 04-16-2015, 03:26 AM
AVS Special Member
 
jaddie's Avatar
 
Join Date: Jan 2007
Posts: 2,541
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 1254 Post(s)
Liked: 710
Quote:
Originally Posted by LTD02 View Post
not quite a sine wave. this one at about 23hz. with only a few breaks, it runs the length of the whole track. watching the spectrum, it appears to be mixed in about 30db hot relative to the other content (encouraging the listener to keep the volume turned up). with a sub tuned to 23hz, it could be a coil cooker.

(spectrumlab - Psyph Morrison - Dope)
Spectrumlab is the wrong analysis tool, it won't show you RMS levels.

Here's what I get:
Left listed first, then Right
Peak Amplitude: -0.17 dB, -0.10 dB
Total RMS Amplitude: -7.41 dB , -7.48 dB
Maximum RMS Amplitude: -4.77 dB , -4.88 dB
Average RMS Amplitude: -8.01 dB , -8.07 dB

The RMS "window" is 1 second, file normalized to -0.1dBFS before analysis.

Quote:
Originally Posted by LTD02 View Post
this kind of stuff is increasingly in what folks listen to, so other than a "smart amp" that somehow governs average current over time, what to do?
Don't clip the sub amp, or anything else, and all will be well. With the average RMS level in a 1 second window at -8dBFS, if the 2800 watt sub runs just short of clipping, the voice coil will see the heating effects of 444 watts. Clip the sub amp so it sounds distorted, like about 2dB worth, you're still well below melting anything.
Quote:
Originally Posted by LTD02 View Post
is there anyway to fuse a sub without castrating it?
I don't know, but I don't see a need either.

The piece has a surprising number of rather long "cooling off" periods in it.
jaddie is offline  
Old 04-16-2015, 03:36 AM
AVS Special Member
 
jaddie's Avatar
 
Join Date: Jan 2007
Posts: 2,541
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 1254 Post(s)
Liked: 710
Ok, I thought I might have messed up here. The analysis was broad band, not sub only. So I applied a subwoofer-type filter, 24dB/octave at 80Hz, then re-normalized. Guess what? That increased the average RMS by a whole 1dB, so we're still at -7dBFS, and total RMS is at -5.5dBFS.

Guess I didn't screw up that bad.
jaddie is offline  
Old 04-16-2015, 04:17 AM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
Quote:
Originally Posted by jaddie View Post
Spectrumlab is the wrong analysis tool, it won't show you RMS levels.

what tool do you suggest?


spectrumlab shows a "long term average". my assumption was that is rms. i don't know how it is defined.
the green line represents the peak value and the red line represents the long term average (whatever that may be).
i was taking the difference to be an indication of the intensity of the signal.
just eyeballing it, the difference is about 7db or so in the crazy bass track.




here is a sine wave for comparison:



eyeballing it suggests about 3db or so for the sine wave.


so the difference between a sine wave at -3db and the crazy track at 7db is 4db of voltage, that is 2.5x power, so the 2800 watt sine wave drops to 1120 watts. but, with a little bit of clipping, which would not be unexpected in a signal recorded 30 db hot relative to the rest of the content...well, that's going to run the power right back up there.
Attached Thumbnails
Click image for larger version

Name:	s1.png
Views:	195
Size:	190.7 KB
ID:	669425   Click image for larger version

Name:	sine1.png
Views:	193
Size:	42.7 KB
ID:	669433  

Listen. It's All Good.

Last edited by LTD02; 04-16-2015 at 04:20 AM.
LTD02 is offline  
Old 04-16-2015, 01:18 PM
AVS Special Member
 
jaddie's Avatar
 
Join Date: Jan 2007
Posts: 2,541
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 1254 Post(s)
Liked: 710
Quote:
Originally Posted by LTD02 View Post
what tool do you suggest?


spectrumlab shows a "long term average". my assumption was that is rms.
Spectrumlab and most if not all spectrum analyzers use an averaging level detector per band, not RMS. This is done for two reasons, it's much simpler to calculate, and it means they will agree well with other spectrum analyzers.
Quote:
Originally Posted by LTD02 View Post
i don't know how it is defined.
Not sure if you mean how the level is defined, or what the difference is between average level and RMS level (they are not the same), but I won't insult you with an explanation or link, you can google if necessary.
Quote:
Originally Posted by LTD02 View Post
the green line represents the peak value and the red line represents the long term average (whatever that may be).
Careful, the red line represents a peak per frequency group or band, not the overall peak level, and certainly not the peak level of a filtered signal sent to a sub. Same for the average level line, it's average per group/band, and does not represent the composite average at all.
Quote:
Originally Posted by LTD02 View Post
i was taking the difference to be an indication of the intensity of the signal.
just eyeballing it, the difference is about 7db or so in the crazy bass track.
The lines represent something that will only be meaningful to another spectrum analyzer with the same configuration. At no time do these levels have meaning to any amplifier or transducer used in a sound system.



here is a sine wave for comparison:

eyeballing it suggests about 3db or so for the sine wave.
[/quote]
Again, wrong tool. The reference in this case should be a sine wave at OdBFS. If that references then equals maximum power from the amplifier, then the only signal capable of hitting that figure will be another 0dBFS sine wave for a given integration time window. We must deal with integration time when we deal with any time-variant signal.

So for proper analysis, you want to determine the heating effect the signal has on a load, so what you want is a true RMS detector (not an average or peak detector) calibrated for a 0dBFS sine wave. Then, not making the same error I initially did, you apply a subwoofer type filter so we look at only frequencies below crossover. Then since we are looking for the difference between a steady state sine wave at amp clipping, we normalize the filtered waveform so that it peaks just shy of 0dBFS. So far, this means we have a system calibrated so digital clipping is matched to amplifier clipping, which permits maximum signal to the amp without distortion. Then using a true RMS detector we look at the wave form in time chunks, and find an oddly named "average RMS" reading, which each integration time window considered in the average. We can also include the entire file in a single integration time window for the Total RMS figure. They differ slightly, but the Average RMS figure will represent the average heating power dissipated in a load over time. We can then pick out the highest RMS value in any integration time window and call that Maximum RMS. That would be the very highest RMS value...heating value...presented to a load. If the time window matches the way a voice coil heats and cools, and a voice coil is rated for continuous dissipation equal to the maximum output of our amp, we can see that our program material is mostly 7 or 8 dB below maximum power, and at worst hits 5dB below maximum power.

It turns out, Adobe Audition has the tools you need. There may be others, I like Audition because the results are available as text you can copy/paste. If you had a means of pre-filtering the signal and normalizing it, or at least scaling the analyzer properly, you could use anything with good RMS detection and a controllable time window. TrueRTA probably can do that, though I haven't actually tried it, since it has no way to pre filter.
jaddie is offline  
Old 04-16-2015, 02:12 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
thanks for the info. i will look into adobe audition.


i'm having difficult seeing the differences that you are talking about.

first, level. in spectrumlab, the difference between the peak energy and the average energy for a sine wave is 3db. the crest factor representing the difference between peak energy and rms energy in a sine wave is 3db. so that lines up. when i think look at the peak to average for the angry sinal, it looks to be about 7db. if 3db of that would represent a full scale sine wave, then 4db of that represents the additional crest average crest factor. so the rms of the content at that frequency is running about 4db less than a sine wave. isn't that roughly right?


second, window width. with the settings that I am running, there are about 7 windows in from 20-30hz, so that would make each window about 1.5hz. that should be sufficiently narrow to capture what i am trying to measure in that band, no? it would seem that using a window with a width of about 1.5hz would paint a more accurate picture than calculating peak/rms over the entire bass range, but i'm not sure how rms calculators work when dealing with such content. if for example, you had a sine wave at 23hz, the peak to average would be running 3db. so rms would be -3db. add in a whole bunch of other bass content at other frequencies that is 30db lower in level and it would seem the peak would still be provided by the sine wave, but the rms would now be much lowever because it is averaged over all the other frequencies. but again, i don't know how rms averaging works with such content.


third, reference level. in the ideal case, i was simply assuming that one would turn the gain up until the peak remained unclipped. that would represent full power. in this case 2800 watts with -0dbfs peaks and -3dbfs of rms, which would of course = 2800 watts. since the average is -4db less than the -3db rms of the sine wave, i calculate that power would be reduced by just over half (10^P1/P2 = 10^4/10 = 2.51, 1/2.51 = ~40%) 40% of 2800 is 1114 watts.


fourth, most folks aren't calibrated for bass content that is 30db hot. levels are set by wag guess method using tracks with hot bass and "quick hand limiters". that's just kind of the reality for most guys. when this track comes crashing through the system like an electrical wrecking ball, the typical system will experience some clipping, if not outright full hard clipping. i'm not sure off hand how average power increases with a -7dbfs signal that is 3db clipped (for example). is there a rule of thumb for this? as you know, the rms of a clipped signal is higher, so the power will be higher too. so that could get power (and current from the amp to the sub) back up into the danger zone, or so it would seem.

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 02:56 PM
Member
 
Join Date: Dec 2014
Posts: 179
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 70 Post(s)
Liked: 7
Audacity is easier to use to find what frequencies peak at what dbfs
MagnumMafia05 is offline  
Old 04-16-2015, 03:19 PM
AVS Special Member
 
jaddie's Avatar
 
Join Date: Jan 2007
Posts: 2,541
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 1254 Post(s)
Liked: 710
Quote:
Originally Posted by LTD02 View Post
thanks for the info. i will look into adobe audition.


i'm having difficult seeing the differences that you are talking about.

first, level. in spectrumlab, the difference between the peak energy and the average energy for a sine wave is 3db. the crest factor representing the difference between peak energy and rms energy in a sine wave is 3db. so that lines up.
Actually, no. My peak figures are simply to show that the file has been properly normalized. The reference level we are concerned with is the RMS value of a sine wave at 0dBFS. All RMS readings I've provided reference that level, so the crest factor has already been accounted for.
Quote:
Originally Posted by LTD02 View Post
when i think look at the peak to average for the angry sinal, it looks to be about 7db. if 3db of that would represent a full scale sine wave, then 4db of that represents the additional crest average crest factor. so the rms of the content at that frequency is running about 4db less than a sine wave. isn't that roughly right?
Not quite. But I'm going to come back to this when I have more time to write.
Quote:
Originally Posted by LTD02 View Post
second, window width. with the settings that I am running, there are about 7 windows in from 20-30hz, so that would make each window about 1.5hz. that should be sufficiently narrow to capture what i am trying to measure in that band, no? it would seem that using a window with a width of about 1.5hz would paint a more accurate picture than calculating peak/rms over the entire bass range, but i'm not sure how rms calculators work when dealing with such content. if for example, you had a sine wave at 23hz, the peak to average would be running 3db. so rms would be -3db. add in a whole bunch of other bass content at other frequencies that is 30db lower in level and it would seem the peak would still be provided by the sine wave, but the rms would now be much lowever because it is averaged over all the other frequencies.
You really cannot use a spectrum analyzer this way. When you change bandwidth you also change response time, and all your figures move too. The other issue is that when looking at spectrum analyzer levels you don't ever see what the sub gets, which is a composite sum of all energy below the crossover frequency. Just have to get you away from the spectrum analyzer, it's the wrong tool.
Quote:
Originally Posted by LTD02 View Post
but again, i don't know how rms averaging works with such content.
Same way it works for any content. Not trying to be curt, but you may have to brush up on what RMS means to get this. Average is not RMS, and average for a single band of a spectrum analyzer isn't at all like RMS for the entire bandwidth of the sub.

Quote:
Originally Posted by LTD02 View Post
third, reference level. in the ideal case, i was simply assuming that one would turn the gain up until the peak remained unclipped. that would represent full power. in this case 2800 watts with -0dbfs peaks and -3dbfs of rms, which would of course = 2800 watts. since the average is -4db less than the -3db rms of the sine wave, i calculate that power would be reduced by just over half (10^P1/P2 = 10^4/10 = 2.51, 1/2.51 = ~40%) 40% of 2800 is 1114 watts.
The correct analysis technique will put the average RMS value at -8dB and the total RMS value at -7dB re: a 0dBFS sine wave.
[/quote]
fourth, most folks aren't calibrated for bass content that is 30db hot. levels are set by wag guess method using tracks with hot bass and "quick hand limiters". that's just kind of the reality for most guys.
[/quote]I must admit to being a bit confused by this. In home theater there is more calibration going on than any other application other than pro audio. There's no guessing, there are auto cal systems that get it right, or at least pretty darn close. Who are these guys guessing and getting it that far off?
Quote:
Originally Posted by LTD02 View Post
when this track comes crashing through the system like an electrical wrecking ball, the typical system will experience some clipping, if not outright full hard clipping. i'm not sure off hand how average power increases with a -7dbfs signal that is 3db clipped (for example). is there a rule of thumb for this? as you know, the rms of a clipped signal is higher, so the power will be higher too.
The RMS relationship is not much unlike just turing up the level without clipping, aside of course from the gross distortion.
Quote:
Originally Posted by LTD02 View Post
so that could get power (and current from the amp to the sub) back up into the danger zone, or so it would seem.
Sure, if taken far enough, but there are lots of warning signs, like the bass sounds horrible number one... Yeah, someone could ignore that, but really not all that many would. In a calibrated system everything should distort at more or less the same level, and main speakers distorting are really obvious. Perhaps those that go beyond that point deserve to have their subs melted.
jaddie is offline  
Old 04-16-2015, 03:34 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
Quote:
Originally Posted by MagnumMafia05 View Post
Audacity is easier to use to find what frequencies peak at what dbfs

thanks. i've got audacity. it is 23hz and as can be seen, that signal is about 30db hot relative to the next highest bass peak.


Attached Thumbnails
Click image for larger version

Name:	23hz.png
Views:	173
Size:	18.7 KB
ID:	669825  

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 03:58 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
"fourth, most folks aren't calibrated for bass content that is 30db hot. levels are set by wag guess method using tracks with hot bass and "quick hand limiters". that's just kind of the reality for most guys."
Quote:
Originally Posted by jaddie View Post
I must admit to being a bit confused by this. In home theater there is more calibration going on than any other application other than pro audio. There's no guessing, there are auto cal systems that get it right, or at least pretty darn close. Who are these guys guessing and getting it that far off?
ah! that may be part of the wide chasm here. there is no reference level for music or whatever one wishes to call the content to what we are discussing here. as you know, 105db is all you get in a theater calibration from the mains in seat. in music, it is wherever the listener chooses to put the dial. i can't think of anybody on the entire forum who has 30db of headroom at 23hz above their maximum music listening level--and that is what we are talking about here.

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 04:32 PM
Member
 
Join Date: Dec 2014
Posts: 179
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 70 Post(s)
Liked: 7
Quote:
Originally Posted by LTD02 View Post
"fourth, most folks aren't calibrated for bass content that is 30db hot. levels are set by wag guess method using tracks with hot bass and "quick hand limiters". that's just kind of the reality for most guys."


ah! that may be part of the wide chasm here. there is no reference level for music or whatever one wishes to call the content to what we are discussing here. as you know, 105db is all you get in a theater calibration from the mains in seat. in music, it is wherever the listener chooses to put the dial. i can't think of anybody on the entire forum who has 30db of headroom at 23hz above their maximum music listening level--and that is what we are talking about here.
Your confused here. Let me explain in my car system. I have my 2 saz-3500v2 set @ -5db gain overlap. So anything recorded at -5db will be under 1% distortion.

Now i listen to decaf, chopped and screwed, dubstep....etc where they are at 0db at frequencies. Now.. if i play them i will have distortion/clipping depending on where i set the volume. But i know if i set my sub output down 3 im @ 0db gain overlap.

Just because the bass is recorded close to 0db doesn't meann you need headroom. Morrison music is INTENDED for BASS. Its like decaf music.

What i mean at overlap gain, i use a 0db 40hz sinewave and turn up the gain till i get 1% distortion and stop. My car i used a -5db 40hz sinewave.

My FB14K was at 134.? Volts when it hit 1% distortion which 134V is ~4489 watts? I use 0db for my FB14K because i dont sit dowm and find what the recordings are at like i do with music for my system. I have folders with song lists with the db recording levels. Then by peak frequencies.

I won't ever have to worry about clipping using the 0db overlap. The down side to using this is i lose output if the recording is lower. Regular music would be fine some where from -5db to -10db overlap.

Last edited by MagnumMafia05; 04-16-2015 at 04:37 PM.
MagnumMafia05 is offline  
Old 04-16-2015, 04:46 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
just thought of another complicating factor...folks could be running more than 1 sub off an amp.
so if we start with speaker coil at 16 awg for a sub and double the number of subs, the current carrying capability of the two subs drops by AWG 3--which means the 16 awg coils together have the ampacity of a 13 awg line.
double the number of subs again to 4 drivers and the ampacity drops by 3 awg again--which means the 4 x 16 awg coils together have the ampacity of a 10 awg line.
double the number of subs again to 8 drivers and the ampacity drops by 3 awg again--which means the 8 x 16 awg coils together have the ampacity of a 7 awg line.
double the number of subs again to 16 drivers and the ampacity drops by 3 awg again--which means the 16 x 16 awg coils together have the ampacity of a 4 awg line.
so if relying on the drivers to "blow" before the in wall speaker line fails, one might want to keep this in mind. :-)

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 05:02 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
Quote:
Originally Posted by MagnumMafia05 View Post
Your confused here. Let me explain in my car system. I have my 2 saz-3500v2 set @ -5db gain overlap. So anything recorded at -5db will be under 1% distortion.

Now i listen to decaf, chopped and screwed, dubstep....etc where they are at 0db at frequencies. Now.. if i play them i will have distortion/clipping depending on where i set the volume. But i know if i set my sub output down 3 im @ 0db gain overlap.

Just because the bass is recorded close to 0db doesn't meann you need headroom. Morrison music is INTENDED for BASS. Its like decaf music.

What i mean at overlap gain, i use a 0db 40hz sinewave and turn up the gain till i get 1% distortion and stop. My car i used a -5db 40hz sinewave.

My FB14K was at 134.? Volts when it hit 1% distortion which 134V is ~4489 watts? I use 0db for my FB14K because i dont sit dowm and find what the recordings are at like i do with music for my system. I have folders with song lists with the db recording levels. Then by peak frequencies.

I won't ever have to worry about clipping using the 0db overlap. The down side to using this is i lose output if the recording is lower. Regular music would be fine some where from -5db to -10db overlap.

thank you magnum. working backwards from -0dbfs and limiting your maximum listening level based on that is good practice.


134 db is much easier to achieve in a car than a large room and 134db might not even be that "loud" when you kick in the 23hz +30db content. that means with 134db potential, your upper bass is going to be coming in at around 104db, which isn't that loud at all.


because of this dynamic, most folks don't set their -0dbfs bass level to anything with respect to music content. subs are level matched to mains, adjusted (typically up) for personal eq preferences and then content level is adjusted to whatever is desired. the +30db peak at 23hz is going to hit the headroom ceiling long before most folks stop turning up the volume knob.

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 05:13 PM
Member
 
Join Date: Dec 2014
Posts: 179
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 70 Post(s)
Liked: 7
Quote:
Originally Posted by LTD02 View Post
thank you magnum. working backwards from -0dbfs and limiting your maximum listening level based on that is good practice.


134 db is much easier to achieve in a car than a large room and 134db might not even be that "loud" when you kick in the 23hz +30db content. that means with 134db potential, your upper bass is going to be coming in at around 104db, which isn't that loud at all.


because of this dynamic, most folks don't set their -0dbfs bass level to anything with respect to music content. subs are level matched to mains, adjusted (typically up) for personal eq preferences and then content level is adjusted to whatever is desired. the +30db peak at 23hz is going to hit the headroom ceiling long before most folks stop turning up the volume knob.
I'm not following you. That 30db difference isnt adding on top of your reference level db on your system. Your bass is going to be louder than the rest. Thats it. Nothing magical here.

Update: look at more songs on audacity and report back on your findings.

I could play that track just fine at 0db MV on my avr with no worries

Last edited by MagnumMafia05; 04-16-2015 at 05:19 PM.
MagnumMafia05 is offline  
Old 04-16-2015, 06:56 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
another factor is the insulation in the wall. apparently in the u.k., wire that is surrounded by the kind of insulation typically stuffed in walls must be de-rated by 50% to meet their code. a 15 amp line only meets code when operated at 7.5 amps continuous when surrounded by insulation. haven't checked ours yet.

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 07:06 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
Quote:
Originally Posted by MagnumMafia05 View Post
I'm not following you. That 30db difference isnt adding on top of your reference level db on your system.

that is the problem. there is no reference level for music level, so any excess like this +30db 23hz sine wave gets added on top, not subtracted from the -0dbfs level as in your system.


in your system, with 134db of peak output, the 23hz signal will limit the other bass to 104db and the lower midrange by another 10db or so (the level where we typically set the volume level). so as you are configured, you could listen to the track at about 94db spl.


for a guy who is not set up the way that you are, he would level match his bass to his mains and then turn up the volume level until the lower midrange is "loud". that might be 100-110db or more depending on a variety of factors. the bass content tends to run +10db, so that gets you up to 120db bass requirement and then there is some personal eq that might bump that another 5db taking it up to 125db. so he would need a system capable of 125db output to listen to the track at high volume BEFORE the 23hz content at +30DB is added to the mix. now our poor user either needs a system capable of 155db output at 23hz (and nobody on the forum has that) or he will be clipping the snot out of the signal or he will be turning the level down. since turning down the level is hardly guaranteed, clipping is almost surely going to result.

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 08:23 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
REALLY good article on wire, ampacity, insulation, and heat dissipation (includes some measured experiments):


http://goodsonengineering.com/wp-con...g-Ampacity.pdf


a few key points:


1. in free air a wire can take twice its rated amperage without significant overheating. example given was 12 gauge rated for 20 amps can handle 40 amps in free air OK (walls without insulation count as free air). [good]


2. per code, breakers may allows up to 20% over current (e.g. 24 amps on a 20 amp circuit breaker) indefinitely. [keep in mind]


3. in free air 20 amps through 12 gauge will dissipate about .66 watts per foot and 14 gauge will dissipate about .59 watts per foot. this is almost all of the heating effect that i had originally calculated. [excellent!]


4. packing wires together and in insulation greatly reduces their ampacity. foam insulation is particularly bad and removes the majority of the "safety factor" for wire that was built into the nec.
Archaea likes this.

Listen. It's All Good.

Last edited by LTD02; 04-16-2015 at 08:26 PM.
LTD02 is offline  
Old 04-16-2015, 08:54 PM
AVS Special Member
 
jaddie's Avatar
 
Join Date: Jan 2007
Posts: 2,541
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 1254 Post(s)
Liked: 710
Quote:
Originally Posted by LTD02 View Post
"fourth, most folks aren't calibrated for bass content that is 30db hot. levels are set by wag guess method using tracks with hot bass and "quick hand limiters". that's just kind of the reality for most guys."


ah! that may be part of the wide chasm here. there is no reference level for music or whatever one wishes to call the content to what we are discussing here. as you know, 105db is all you get in a theater calibration from the mains in seat. in music, it is wherever the listener chooses to put the dial. i can't think of anybody on the entire forum who has 30db of headroom at 23hz above their maximum music listening level--and that is what we are talking about here.
But there's always a reference level: 0dBFS. Good home theater systems have 20dB of headroom above 85dBSPL for LCRS, and 30dB sub. That's 105dB at 0dBFS and 115dB @ 0dBFS sub. While a music system might be capable of higher, there are some very good physical reasons not to. And while there are no "official" industry monitoring reference levels in music, they still fall within shooting distance of 85dB SPL, if for no other reason than we want peaks below the threshold of pain, and we want averages to not damage hearing. We're not talking concert/dance systems are we?
jaddie is offline  
Old 04-16-2015, 08:59 PM
AVS Special Member
 
jaddie's Avatar
 
Join Date: Jan 2007
Posts: 2,541
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 1254 Post(s)
Liked: 710
Quote:
Originally Posted by LTD02 View Post
thanks. i've got audacity. it is 23hz and as can be seen, that signal is about 30db hot relative to the next highest bass peak.


Same issue with Audacity: no total RMS readings, no RMS as all. Still the wrong tool.
jaddie is offline  
Old 04-16-2015, 10:04 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
Quote:
Originally Posted by jaddie View Post
But there's always a reference level: 0dBFS. Good home theater systems have 20dB of headroom above 85dBSPL for LCRS, and 30dB sub. That's 105dB at 0dBFS and 115dB @ 0dBFS sub. While a music system might be capable of higher, there are some very good physical reasons not to. And while there are no "official" industry monitoring reference levels in music, they still fall within shooting distance of 85dB SPL, if for no other reason than we want peaks below the threshold of pain, and we want averages to not damage hearing. We're not talking concert/dance systems are we?
i'm not talking about monitoring levels in a studio. i'm talking about music playback. there is no reference there. one guy likes 85db and another wants front-row-metallica-concert-130db-in-his-face spl.

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 10:06 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
Quote:
Originally Posted by jaddie View Post
Same issue with Audacity: no total RMS readings, no RMS as all. Still the wrong tool.

so you are saying that the 30db peak at 23hz isn't 30db or isn't 23hz?


if the level is what you take issue with, how would one go about determining it?

Listen. It's All Good.
LTD02 is offline  
Old 04-16-2015, 10:12 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
here is a summary of some of the things that were discussed with respect to wire type, size, and related. not everybody is going to agree with all points (obviously), so take them as a starting point for performing your own research and forming your own opinions. :-)


Attached Thumbnails
Click image for larger version

Name:	some observations on in wall speaker wire.png
Views:	156
Size:	71.9 KB
ID:	670169  

Listen. It's All Good.
LTD02 is offline  
Old 04-17-2015, 01:32 AM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
found an interesting little free/donation tool called wavosaur: http://www.wavosaur.com/about.php
that can calculate rms power across a track with one click.

first, as a check on the new tool, i created a 50hz sine wave in audacity, then analyzed it in wavosaur.



rms power is -3.01 db right on, max value is 0.00 (so -0dbfs). statistics in wavosaur look good so far.

then i went back to audacity, loaded up the test track, combined both channels into one, low passed it at 80hz (4th order), and normalized it to -0dbfs.

opened it with wavosaur and ran statistics.




max value checks out at -0dbfs and rms power is coming in at 37.8%, or about half of the power of a sine wave (assuming zero clipping of course).

wavosaur also has a spectrum analyzer, but I'm not sure if it will have the same issues that were referred to in audacity and spectrumlab already. fwiw, this is the spectrum (again, combined to mono, low passed 4th order 80hz, and normalized).



still looks like about 30db of 'heat' at 23hz. this is a fft analysis though, so probably has +3db per octave baked in, so that peak should probably be reduced by a little bit to account for that fact.
Attached Thumbnails
Click image for larger version

Name:	50hz sine stats.png
Views:	134
Size:	13.3 KB
ID:	671537   Click image for larger version

Name:	spectrum - LP80hz - combined to mono - normalized - Psyph Morrison - Dope.png
Views:	133
Size:	28.3 KB
ID:	671545   Click image for larger version

Name:	stats - LP80hz - combined to mono - normalized - Psyph Morrison - Dope.png
Views:	131
Size:	13.1 KB
ID:	671553  

Listen. It's All Good.

Last edited by LTD02; 04-17-2015 at 04:16 PM.
LTD02 is offline  
Old 04-17-2015, 01:50 PM
Member
 
Join Date: Dec 2014
Posts: 179
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 70 Post(s)
Liked: 7
Quote:
Originally Posted by LTD02 View Post
that is the problem. there is no reference level for music level, so any excess like this +30db 23hz sine wave gets added on top, not subtracted from the -0dbfs level as in your system.


in your system, with 134db of peak output, the 23hz signal will limit the other bass to 104db and the lower midrange by another 10db or so (the level where we typically set the volume level). so as you are configured, you could listen to the track at about 94db spl.


for a guy who is not set up the way that you are, he would level match his bass to his mains and then turn up the volume level until the lower midrange is "loud". that might be 100-110db or more depending on a variety of factors. the bass content tends to run +10db, so that gets you up to 120db bass requirement and then there is some personal eq that might bump that another 5db taking it up to 125db. so he would need a system capable of 125db output to listen to the track at high volume BEFORE the 23hz content at +30DB is added to the mix. now our poor user either needs a system capable of 155db output at 23hz (and nobody on the forum has that) or he will be clipping the snot out of the signal or he will be turning the level down. since turning down the level is hardly guaranteed, clipping is almost surely going to result.
Ah yes forgot about level matching. Im "now" following you!! ;-)

But im not following where you keep adding dbs. There is no way someones system is set at -30db at the amplifier. In that situation your right. Even level matching i doubt anyone is over -10db with gain overlap.

Grab a oscope and test your sub amp. Use sinewaves from 0dbsf to -10dbsf and find which one is before clipping at your highest MV on your avr. That would be yiur gain overlap. I dont use clipping as gain overlap but %1 distortion.

Last edited by MagnumMafia05; 04-17-2015 at 02:01 PM.
MagnumMafia05 is offline  
Old 04-17-2015, 04:08 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
why are all my images now broken?


fixed.

Listen. It's All Good.

Last edited by LTD02; 04-17-2015 at 04:17 PM.
LTD02 is offline  
Old 04-18-2015, 12:06 AM
Senior Member
 
SpinMonster's Avatar
 
Join Date: Apr 2014
Location: Colorado Springs, CO
Posts: 426
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 199 Post(s)
Liked: 90
I moved the amp near the sub and run a longer interconnect and 6 feet of 8guage.
I run Peavey 4080hz amps that are 4000watts at 8ohms so the resistance is less of an issue. If you run 2 ohm loads speaker wire is a much bigger issue
SpinMonster is offline  
Old 04-18-2015, 01:44 AM
 
A9X-308's Avatar
 
Join Date: Mar 2008
Location: Australia
Posts: 7,173
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 1040 Post(s)
Liked: 771
Quote:
Originally Posted by LTD02 View Post
another factor is the insulation in the wall. apparently in the u.k., wire that is surrounded by the kind of insulation typically stuffed in walls must be de-rated by 50% to meet their code. a 15 amp line only meets code when operated at 7.5 amps continuous when surrounded by insulation. haven't checked ours yet.
I'd be very surprised if your doesn't either. Down here AS/NZS 3008 is the standard that covers it and there are dozens of pages in it detailing the various derating factors for how various cables are laid out, where they are laid out and how they are bundled together and also by what type of insulation and conductor material they employ. I have an electronic copy somewhere and I'll post a page if I can find it.
A9X-308 is offline  
Old 04-18-2015, 09:48 AM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 20,134
Mentioned: 330 Post(s)
Tagged: 1 Thread(s)
Quoted: 2184 Post(s)
Liked: 2402
from a 2013 article:





link includes spreadsheet for derating based on surrounding insulation r-value (though I have not checked it out).


http://www.electrician2.com/articles/ampacity.htm
Attached Thumbnails
Click image for larger version

Name:	there are no requirements in the nec for excessive insulation around cables.png
Views:	106
Size:	16.4 KB
ID:	672665  

Listen. It's All Good.
LTD02 is offline  
Sponsored Links
Advertisement
 
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off