Audibility of cut-off frequency, roll-off rate and amplitude ripple

[witwald]

The following conference paper by Soren Bech (Bang Olufsen a/s) may be of interest to users of subwoofers.

Søren Bech. Quantification of subwoofer requirements, part II: the influence of lower system cut-off frequency and slope and pass-band amplitude and group delay ripple. Preprint 5199, 109th Convention of the Audio Engineering Society, 22–25 September 2000, Los Angeles, California, USA.

During testing, subjects evaluated the magnitude of Lower Bass and Upper Bass in relation to a fixed reference condition. The test signals consisted of different programme material, chosen to ensure that there was sufficient energy content at the frequencies of interest. The influence of filter order (2nd, 4th, and 6th) and lower cut-off frequency (20, 35, and 50 Hz) were studied. The influence of amplitude and delay ripple corresponding to four different reverberation times was also studied.

Interestingly, Bech's experiments utilised two reproduction methods: 1) a real loudspeaker in an anechoic chamber, and 2) a simulated system reproduced via headphones. The results showed that there were no significant differences between the data produced by these two methods of reproduction.

As might be expected, Bech found that the lower cut-off frequency has a significant influence on the perceived level of Lower Bass and Upper Bass reproduction, independent of reproduction levels. The effect of filter order was not found to be of significant importance. The amplitude ripple was also found to have a significant influence on the perceived level of Lower Bass and Upper Bass reproduction.

With regard to the audibility of amplitude ripple, Bech found that the threshold of audibility corresponded to a ripple of +3 dB relative to the reference transfer function. The threshold level was further found to be independent of reproduction level, whereas the influence of higher-amplitude ripples increased with increasing reproduction level. The audibility of the ripple also depended on the signal.

How do the above findings sit with the broad experiences of subwoofer users in this forum? Have people found that the low-frequency roll-off rate (e.g. 2nd-order for a closed box system, 4th-order for a vented box system, 6th-order for a filter-assisted vented box system) has little or no affect on the perceived quality of subwoofer bass reproduction?

 

[Member 639844]

For me, 4th and 6th order design are off the menu. They are usually employed at the cheaper end of the market to try get some extra bass from a product, but its all fake boom and sounds pants. Further up the scale we have the like of Bose. While their design is much better (you dont hear the cabinet like you do in the cheap designs) I feel the same 'fake bass' effect is still present, albeit it less apparent at lower volumes.

Ported vs sealed designs is a hot topic, but I always felt a sealed design gives an ever so sightly purer sound. Ive found them better for music as a general rule, but its by no means set in stone. A high end, well designed ported effort will be better than a poor underpowered sealed effort after all.

I currently use a ported sub and I find it excellent, it has more grunt than any other sub Ive used and there is an effortlessness that comes across in its deliver. Whether I can actually tie this in with sound quality is too difficult for me to honestly answer, which should be an answer in itself. The sealed subs Ive heard or ran in the past always just seem to sound a bit better musically than their equally priced counterparts, but you give a little in that they start to sound a bit underpowered if you try run very loud.

IMO its a fine balance, I think my feeling is that ported design are more of an effect device, while sealed ones actually are the better sounding on the whole. This doesnt mean I think ported subs sound bad, I love the sound from mine after all. I just feel from experience that sealed subs just have a little something you cant quite put your finger on that makes them sound a little better at any given level. For whatever reason that may be.

I cant honestly answer about lower cut off frequency effects as its not something I have honestly sat down and given any time to analysing. As for ripple effects, Ive noticed them easily in frequency sweeps, and in some music where bass lines are a bit more drone-y, and at any given level. This has always been very apparent to me from a young age, even before I bought my first sub before the internet existed. For me it depends on the source material, its never a problem on films but it can influence music enjoyment. Not to critical levels but it does detract from the listening at times so I'd say its definitely detectable, and a definite issue, but its not a system destroying one. Its also an across the board issue that only EQ starts to address, I guess if you end up getting into EQ its because you detect ripple and it is an issue, for whatever reason.

 

[witwald]

SVS appears to be a progressive company, in that they publish anechoic frequency response curves for their PB13-Ultra subwoofer. Separate response curves are provided for the sealed, 20 Hz, 15 Hz and 10 Hz operating modes.

Looking at the 20 Hz and 15 Hz operating modes, the low-frequency roll-off rate is about 36 dB/octave. As a standard vented box enclosure has a roll-off rate of 24 dB/octave, it seems that an auxiliary 12 dB/octave filter has been applied to equalise the response. This is a so-called 6th-order filter assisted vented box low-frequency alignment. The published design theories related to this type of system date back to at least the 1970s.

The fact that the PB13-Ultra subwoofer has gained good user acceptance seems to reinforce Bech's conclusion that low-frequency roll-off rate has no significant effect on the perceived sound quality of a subwoofer.

Interestingly enough, the sealed operating mode seems to have a 9 dB/octave roll-off rate. This is somewhat less than the 12 dB/octave roll-off rate of a driver in standard closed/sealed enclosure. It would seem that a small degree of low-frequency bass equalisation has been applied to boost the low-frequency output of the sealed response function.

 

[micb3rd]

Moonfly your getting confused.

The article was about filter orders not enclosure type orders.

There are 2nd, 4th and 6th order Analogue and Digital filters. i.e. 2nd, 4th and 6th order high pass, 2nd, 4th and 6th order lowpass and 2nd, 4th and 6th order bandpass filters.

Oh and nipping back to back onto enclosures types.

Ported enclosures are 4th Order enclosures

There is nothing wrong with a correctly designed 4th or 6th order Bandpass subwoofers any way, it is just cheap 4th order bandpass are rubbish, just like cheap ported and cheap sealed subs are rubbish.

 

[Nimby]

It not always wise to generalise though we all do it most of the time.

I built a pair of 6th order series, passive bandpass subs to a professional design and they sounded fine on all music including organ and percussion. They used 10" SEAS units in 19" cubes and were flat from 20-120Hz with 100dB capability and still had useful output at 10hz. I often used to listen to music at low levels in the evening and was often surprised how deep things sounded regardless of level. Being passive I had no control over the raw output. We used them for watching TV sometimes and they increased realism noticeably over the Linn Kans they underpinned.

My big SVS cylinder had much the same reach in the power band at 16Hz but sounded much more open, deeper, powerful and detailed than the bandpass boxes. It too sounded deep and realistic even at low levels on TV programmes including traffic, trains, gunfire, doors closing, music etc. on high level connections.

The IB made the cylinder sound like a woolly sock. An IB is a very big sealed box. The IB reaches 8Hz at very high levels and sounds noticeably deeper even on a small drum. The detail, power, weight and sense of realism is staggering compared to the cylinder. It is a no contest on organ music with the IB sounding very realistic. The cylinder is incapable of sounding like an organ. I have been doing a lot of listening comparisons recently.

I tried switching in a 25hz subsonic filter on Bass Outlaw's Ultimate Woofer Test at 100dB((C) 105dB(C) and 110dB(C) and found the effect on the IB almost inaudible but the room didn't rattle so much. I have yet to discover the slope of this "rumble" filter on my CX2310 active crossover but it is very useful on vinyl even on organ music. It cuts out warp reading and mechanical excitement of my own movements on the turntable due to the bouncy boarded floor. I really could not swear that the filter was enabled without knowing it beforehand. This really surprised me. I thought it would be very obvious. I have more tests of this filter's effect planned on some very deep CD organ tracks.

 

[witwald]

I tried switching in a 25hz subsonic filter on Bass Outlaw's Ultimate Woofer Test at 100dB((C) 105dB(C) and 110dB(C) and found the effect on the IB almost inaudible but the room didn't rattle so much. I have yet to discover the slope of this "rumble" filter on my CX2310 active crossover...

The User's Manual for the Behringer Super-X PRO CX2310 crossover states that the 25 Hz infrasonic filter has a "side gradient" of 12 dB/octave. Hence, it seems that this "rumble" filter is a 2nd-order design, which I'd hazard a guess is fairly typical for this sort of application.

I'm intrigued by the fact that switching on the 25 Hz infrasonic filter appears to have had an almost insignificant effect on the bass output from your IB system. It would seem to indicate that the test material doesn't have a lot of energy below 25 Hz.

I really could not swear that the filter was enabled without knowing it beforehand. This really surprised me. I thought it would be very obvious. I have more tests of this filter's effect planned on some very deep CD organ tracks.

Please report back on what you heard. I'm interested in learning about what you discover.

 

[Nimby]

witwald

Thanks for saving me a search through the manual.

Since my threshold of audibility/inaudibility on sinewaves is ~22Hz one would expect no audible difference from applying these 25Hz filters. Only the infrasonics would be affected.

I can confirm that I can detect no difference on organ music with the 25Hz filter enabled. I muted the speakers and whacked up the volume on the IB to the point where everything in the room was rattling: Walls, doors, coffee table, windows, my chair, several remote controls, radiator, heavy tile coaster, etc.

Even when I tried to mute what sounded like background traffic noise there was no discernible difference when applying the filters. I know they work because they kill spurious cone movement on both CD and vinyl.

That just leaves film bass to try but now the Head Gardener has come back in due to heavy, persistent rain.

 

[Nimby]

I managed some REW sweeps to show the effects of these 25hz filters.(unsmoothed IB only, Galaxy 140 SPL meter resting on chair at listening position)

The cut seems to be roughly 4dB at 25hz:



"Transformers" battle (Ch21 DVD) was gutless with the filters engaged.

 

[Member 639844]

Moonfly your getting confused.

The article was about filter orders not enclosure type orders.

There are 2nd, 4th and 6th order Analogue and Digital filters. i.e. 2nd, 4th and 6th order high pass, 2nd, 4th and 6th order lowpass and 2nd, 4th and 6th order bandpass filters.

Oh and nipping back to back onto enclosures types.

Ported enclosures are 4th Order enclosures

There is nothing wrong with a correctly designed 4th or 6th order Bandpass subwoofers any way, it is just cheap 4th order bandpass are rubbish, just like cheap ported and cheap sealed subs are rubbish.

My bad, thats what I get for not reading the article I guess. I only read the post

As for enclosures:

I dont like either of the 4th and 6th order designs, basically any where the driver is mounted internally .
Ported I dont mind, but in all honesty I feel sealed subs are best. Recently though Ive heard some fairly good PR designs and I cant really say a bad word about that method.

 

[HiFiRuss71]

Recently though Ive heard some fairly good PR designs and I cant really say a bad word about that method.

A PR is just an expensive mechanical port with most of the same problems, so why the downer on ported subs?

Russell

 

[Nimby]

Whatever the article was about witwald requested opinions on enclosure roll-offs.

Moonfly, you forgot the series 6th order bandpass box.

In a bid to get more output from one box SVS made a very serious bandpass but nobody "trashed" it because of its alignment.

Product Review

 

[micb3rd]

My bad, thats what I get for not reading the article I guess. I only read the post

As for enclosures:


I dont like either of the 4th and 6th order designs, basically any where the driver is mounted internally .
Ported I dont mind, but in all honesty I feel sealed subs are best. Recently though Ive heard some fairly good PR designs and I cant really say a bad word about that method.

Moonfly: Orders.

Sealed is 2nd order, a vent is another 2nd order so 2 + 2 = 4th order, thus vented boxes are 4th Order.

With a 4th order bandpass you have 2nd order sealed plus a series vent so 2 + 2 = 4th Order bandpass.

A 4th order bandpass design *is* a sealed design but with a geometric high pass filter which is a port. Below the vent tune the system acts just like sealed as the low frequencies just pass though.

6th order there are two types off bandpass, series tuned (which I really like) and parallel tuned wich I don't, thus this is 4th order plus a further order of vent so 2 + 2 + 2 = 6th Order BP.

Please don't take this the wrong way but you are wrong about 4th and 6th order designs, if you have not built them and tested you are just generailising on theory not real life experience.

Assuming your 4th order BP design is good the port will not peak the responce or be too small to choke the design, it is a fine alignment.

Whatever the article was about witwald requested opinions on enclosure roll-offs.

Moonfly, you forgot the series 6th order bandpass box.

In a bid to get more output from one box SVS made a very serious bandpass but nobody "trashed" it because of its alignment.

Product Review

Excatly Nimby the B4Plus was very well received by all reviewers and enthuasts, one of the loudest but good widerange (bandwidth) subwoofers ever.

 

[Member 639844]

A PR is just an expensive mechanical port with most of the same problems, so why the downer on ported subs?

Russell

I dont have a downer on ports, after all I swapped from a sealed to a ported design, its just a couple of the ways of implementing them. As for passive radiators, well they dont chuff, now do they