@mark100 I remember that track very well and think its a good test. Unfortunately when you just even kiss the back plate with the VC. It may be too late. I tend to be more conservative with sub levels at live gigs. Canned music is easier to deal with due to its predictable nature, but some bass players can be the worst offenders if they play slap style. If the comp isn't set carefully, the subs can be bottomed with just a few hits of low B or E. Sometimes you have to side chain kick and bass for the best headroom. Very few bassists can play in a given dynamic range and most will bump their DI send after sound check. DJs are often guilty of this too. Ask me why I say this... bass is my main instrument and I used to be a club DJ as well. Lol
I use these 2 tracks alot for setting low end parameters. The low end is deep and tight with lots of detail and definition. Most systems will turn this into a garbled mess
I use these 2 tracks alot for setting low end parameters. The low end is deep and tight with lots of detail and definition. Most systems will turn this into a garbled mess
Hey, thx for the test track recommendations. Quite clean and steady bass, for establishing levels etc.
And yeah, getting close hitting the backplate ain't my idea of where to set limiters. I really try to make xmax the limit.
Only subs I've ever let do that, or even get close, are some Labhorn subs, the ones TDanley designed years ago for DIY prosound.
The eminence drivers in them, Lab-12's, are some kind of tuff. https://eminence.com/products/lab_1...d=4543434031174&pr_seq=uniform#specifications
130Vpk was below striking on any kind on transient i've tested. I think I remember limiting them around 100Vpk when used for live or DJ
Hi b_force, don't mean to be confusing or unclear.
I strongly believe having uncompressed, unclipped linear SPL that includes +18dB headroom for transients, thoughout the entire spectrum...all the way down to f-3....is a major factor in sound quality.
Major to the point, I think it might be second only to freq response...maybe even outranking smooth directivity.
That kind of SPL capability just sounds so damn clean/clear/right to my ears....
...compared to the usual home audio's lack of capability at unclipped, uncompressed SPL. And which seldom has any breathe left for headroom, especially down low.
So yes, I offer it as general advice.
D
Deleted member 375592
Old PSTN provides 24 dB of headroom.
EU TV standards prescribe to set RMS below -23dBFS.
Analyzing live acoustic music CD tracks will often show RMS as -28dBFS with peaks at -0.1dBFS. Some go even lower.
30dB headroom is a piece of very reasonable advice.
@mark100 @mikets42
I agree that headroom/dynamic range is a nice thing to have. 🙂
But the average SPL you're talking is pretty high in general I think. (for most rooms and people I know)
It's also a bit debatable how important it is that these very short high SPL peaks need to be so "compression free".
But that is an entirely different story and debate lol 😀
Again, there is nothing inherently wrong with it at all! 🙂
But in my experience those numbers (especially also with those low frequencies) are a bit of a niche.
That's not just said from a technical point of view but also practical point of view.
I agree that headroom/dynamic range is a nice thing to have. 🙂
But the average SPL you're talking is pretty high in general I think. (for most rooms and people I know)
It's also a bit debatable how important it is that these very short high SPL peaks need to be so "compression free".
But that is an entirely different story and debate lol 😀
Again, there is nothing inherently wrong with it at all! 🙂
But in my experience those numbers (especially also with those low frequencies) are a bit of a niche.
That's not just said from a technical point of view but also practical point of view.
Mark,
The displacement (excursion) Hornresp predicts past Xmax won't be correct.
Repeating what I wrote 10 years ago when David added filters to Hornresp:
post # 4298
https://www.diyaudio.com/community/threads/hornresp.119854/page-215
Although Hornresp is quite good at predicting excursion within the linear range of operation of a loudspeaker, that is displacement at or below Xmax, it is not accurate above Xmax, where most users are concerned regarding placement of filters to protect drivers from over excursion. Since it predicts more excursion than usually will be experienced, it is a benign problem, but an inaccuracy nonetheless.
Above Xmax the driver's BL (magnetic force) drops off, as the voice coil progressively leaves the magnetic gap there is no longer the force left to push a stiff suspension past a certain excursion. Hornresp does not take the dropping BL into account, and it appears to consider the suspension to have near infinite elasticity and linearity, not the case for most high displacement drivers which "put the brakes" on excursion before the voice coil can hammer the back plate.
Attached are sine wave excursion tests on some Dayton PA385S-8 drivers conducted in free air.
Measured results are quite close to the Hornresp simulation up to a bit past Xmax, but then diverge more and more, the simulation predicting as much as three times the measured displacement at lower frequencies.
To recap, two points:
1) Most modern drivers voice coils will not hit the back plate unless the suspension is torn.
That said, when the suspension hits it's limits, it can make some nasty sounds.
2) Most drivers excursion won't be nearly as high past Xmax as simulations would indicate.
Art
Art, great info. Thank you.
It answers a nagging impression I've had that hornresp's displacement past xmax, had to be based on some math that drivers obviously can't really do.
Your PA385 test and explanation has helped confirm that.
I'm an admitted headroom / clean SPL junkie.......and a phase-o-holic too.
Iow, just another fellow audio nut 🙂
Cheers.
Phase-O-holics... that sounds like a viable name for a 10 step self help group for audiophiles who listen to test tones and pink noise instead of music and have stacks of hifi review magazines next to the bathroom throne.
haha...oh lord no ! I simply think flat phase makes MUSIC come alive 😀
I think it helps transients, rhythm, timbre, and vocals' intelligibility & separation of singers
So don't go givin' my addiction a bad image haha 🙂
D
Deleted member 375592
I do not understand what "practical point of view" means, especially on this forum. Let's do a spectral / histogram analysis of the popular song "Time" by Pink Floyd.
As you can easily see, 22dB of dynamic range for f < 35Hz.
Let's also see what peak-to-average you need for the woofer for a chosen crossover frequency (and order). About 18dB, if I am not mistaken.
This is the reality, not fantasies.
How loud a lot of people play on average basically, especially in smaller rooms.
The majority sits somewhere around 70-85dB(A), so peaks around 100-105dB or thereabouts.
I don't find CD albums a good reference btw, just go to a live big orchestra and try to record and playback the dynamics.
Good luck, and don't get sad when everything clips and blows up to pieces. 😎😉
Btw, you also won't hear those silly artificial "details" 😉
I agree strongly....as i guess has been obvious from my posts in this thread (and many others)
I think the problem...for sound qualities' sake.... is how we find ways of rationalizing away from the truth of that reality.
Last edited:
That is much more a philosophical question than a technical one.
Not less interesting though 🙂
Don't get me wrong, used to work on and design double 21 inch subwoofers, so I know the feeling 😀
I've been working up a reply to @hifijim, after measuring the peak voltage going to my sub on the Joe Walsh test track i recommended....vs average SPL.
Here's some of the data.
At 71 dB(A) for the track, Vpk to sub was 15.7V. This is with a sub that has a 101 dB, 2.83V efficiency.
Knock sub sensitivity down 10 dB, to say 91 dB more in line with normal home sub levels, and we need 50Vpk....for 71 dB(A).
85 dB(A) (which is damn loud IMO)..is gonna take another 14 dB voltage gain to stay linear for sub peaks. Who's kidding who if anyone thinks that's gonna happen with a 91 dB sens sub.
I know the rebuttals......that this forgets room gain, quarter & half space placement options, and multiple subs (which is itself another big compromise.).
True. My analysis ignores those...cause I've come to think/believe they are all notions of smoothing out reality
Tyr and find some of TDanleys recordings of fireworks, and trains.
Eye/ear opening when it comes to dynamics.
Since the simulation has no input for the suspension limits, it considers it limitless 😳
Your example of the SB34NRX75-6 with an Xmax of 11mm reaching over 30mm excursion is obviously impossible when you look at it's half roll surround, and do the math of it's "unrolled" length at a diagonal:
Somewhere around 15mm excursion, the half roll has straightened out completely, reaching it's mechanical limit, (Xmech or Xlim) it can't go further without stretching or ripping.
The difference between 11mm and 15mm is under 3dB:
Jim,
When William Shatner played Captain Kirk, 40Hz seemed "low enough".
So around the low "E" on a four string bass hits Xmax at 107dB, there is less than 3dB left before Xlim, when the SB34NRX75-6 goes from clean to making rubber band snapping noises 😉
You did suggest using a 32Hz +6dB boost with this speaker, so when a 5 string bass hits a low "B" (31 Hz) at ~105.5 dB, the speaker has hit Xmech.
At any rate, when you play songs like Pink Floyd's "Time" with dynamic range of ~22dB of dynamic range below 35Hz, it's hard to keep track of 3dB.
103dB SPL at 30Hz would read only 63dB (-40dB) on an "A" scale weighting...
Art
Attachments
Hey man, I'm with you 100 percent on striving for linear phase response. My system uses no EQ, subsonic filters or the like. I also can tell if something is reversed in absolute phase.
I used to be a die hard electrostat guy. That 300 hz to 10k slice of range is critical to get right. I stay away from filter slopes steeper than 2nd order and cross in the linear portion of FR whenever possible where most people think its ok to use the existing driver rolloff, which worsens timing and phase. Under 300 hz has to be crossed 1st order. Everything acoustic event has to be in sync and in time with another so it sounds like one piece cohesive piece of audio, as close to the original source as possible.
The guys who record drums with multiple mics really don't understand how much they're mutilating the whole drum sound. I understand the issue of having control over all the separate components, but it never ever ends up sounding convincing after pasting it all together in the mix. Its just as bad as micing a piano under the lid, making it sound fake and 10 ft wide.