Psychoacoustic bass enhancement

Hi

For a new project of mine, I'd like to know if a psychoacoustic bass enhancement plug-in exists for the miniDSP similar to SRS TruBass or AM3D PowerBass plug-ins for Sigma Studio. Or if it is indeed possible to make such with miniDSP.

Given that the original patents for the technique expired between 2006 and 2008, there should be no problems implementing it.

To give a brief overview, the technique uses the psychoacoustic effect of missing fundamental to generate a perceived bass response that goes deeper than the speakers are capable of reproducing.

Technically, it isolates the incoming signal below that which the speakers can reproduce. Generates harmonic frequencies of these. At least 2nd and 3rd harmonic frequencies are needed but including 4th harmonic frequency will greatly enhance the effect. Even 5th harmonic is used in some implementations. The harmonic frequencies are then attenuated to -6dB (2nd), -12dB (3rd), -18dB (4th), and -24dB (5th) of the original filtered signal. The fundamental frequencies are filtered away from this signal with at least 24dB/octave filter before being mixed into the original signal.

It's important to note that the original signal is not filtered at all at this stage. The generated signal is just added to the original signal.

Any ideas?
 
Last edited:
Interesting idea but I am puzzled.

Say you want to hear an organ pedal note but the fundamental note is below the range of your subs. The harmonics are already present in the recording (duh) and you brain is already "hearing" fundamental note that your subs aren't playing.

So why would you need to add anything?

Of course, lotsa fun boosting the bass and this method may just be a wisely configured way to do it compared to just torquing the whole range.

Ben
 

Bob Brines

Member
2003-01-31 10:11 pm
Psychoacoustic bass enhancement is dependent on the first overtone being present, and with the very small and poorly designed computer speaker that SRS etc were designed for often couldn't go that deep. Then again to the young and impressionable, bass = kick drum = ~80Hz. That's more doable. Put a 6dB hump around 80Hz and suddenly you have "great bass" and lots of "slam". Do that on decent speakers and organ music and it is going to sound truly weird.

Bob
 
In Wikipedia, under "bandwidth extension," I found this:

"Bass enhancement of small loudspeakers

Most often small loudspeakers are physically incapable of reproducing low frequency material. Using a psycho-acoustical phenomenon like the missing fundamental, perception of low frequencies can be greatly increased. By generating harmonics of lower frequencies and removing the lower frequencies themselves the suggestion is created that these frequencies are still remaining in the signal. This process is usually applied through external equipment or embedded in the speaker system using a digital signal processor.

High frequency response can also be enhanced through generation of harmonics. Instead of mapping frequencies inside the reproducible region of the speaker, the speaker itself is used to generate frequencies outside the normal reproducible region. By boosting high frequencies and overdriving the speaker or amplifier slightly, higher harmonics can be generated."

Seems to explain it all. Except for my earlier question. But to answer my own question, if you are playing "computer" music (as compared to simulated real music), you may well have only a fundamental tone present and no naturally occurring partials. In that case, a DSP introducing partials of the unreproducible fundamental can do the trick.

Except for the sternum shaking thrill of real organ music, you can enjoy the full bass-like experience of organ music on speakers with little output below maybe 60 Hz.

Ben
 
Last edited:
So why would you need to add anything?

Because you need to add harmonics in a specific pattern when you remove the fundamental. This added harmonics is on top of the existing harmonic content.

The specific pattern require the 3rd and 5rd harmonics to be slightly emphasized over a normal harmonic decay. In practice this is most often realized by limiting the harmonics to 5th as the highest and using a lower harmonic decay of h=0.707 instead of the natural h=1.000 of a sawtooth-like signal (which is what a sinus signal overlaid with harmonic content most resemble).
 
Because you need to add harmonics in a specific pattern when you remove the fundamental. This added harmonics is on top of the existing harmonic content.

The specific pattern require the 3rd and 5rd harmonics to be slightly emphasized over a normal harmonic decay. In practice this is most often realized by limiting the harmonics to 5th as the highest and using a lower harmonic decay of h=0.707 instead of the natural h=1.000 of a sawtooth-like signal (which is what a sinus signal overlaid with harmonic content most resemble).
What research underlies your belief that you need that "specific pattern" that is different than what is already present on recording? Wouldn't you think a recording of a sound ought to have the harmonics already present that correctly help your brain imaginatively recreate the fundamental?

Ben
 
Have you even tried googling? There's a lot of research and patents on the subject. It's worth noting that most (all except one) of the patents have expired as I mentioned earlier.

Would I think a recording of a sound ought to have the harmonics already present that correctly help your brain imaginatively recreate the fundamental? No, and it's not a question of what I think or believe, it has already been empirically proven that it is not the case.

To be clear. This technique is completely different from traditional bass boost techniques in that it does not increase the amount of low frequency energy content. It actually substantially limits it. It increases the low frequency phon content, i.e. the amount of perceived sound.

To give a primer here's the AES paper for Waves Ltd. MaxxBass implementation.

http://www.maxx.com/objects/PDF/MaxxBassAESPaper.pdf
 
Last edited:
Saturnus -

Thanks for the link. But for a guy who reacts so tartly by being asked for evidence as you seem to be, couldn't you find a link with:

1. references that aren't 18 years at the youngest

2. empirical testing on humans (which gave me an opportunity to reflect on the variety of people who post on the web considering your flaming retort about, "... not what I think.... empirically proven....")

I found your response needlessly offensive. BTW, "...wouldn't you think..." is the way a polite Canadian says, "... I have a suspicion that what you are saying is full of ..."

Ben
 
Would I think a recording of a sound ought to have the harmonics already present that correctly help your brain imaginatively recreate the fundamental? No

Well, not no, I would expect this to have already been done by the engineer at the recording studio (using the afore mentioned Maxxbass plugin, tube compressor, analog tape machine or the bass player's tube amp :) ) if he thought the music required it.

The easiest way to implement it yourself would be using a crossover to split off the bottom octave, running it through a distortion algorithm to generate harmonics, then mixing it back in with the original signal. I don't see any distortion algorithms for the MiniDSP, so maybe just use an undersized tube amp to drive your subwoofer.
 
Thanks for the link. But for a guy who reacts so tartly by being asked for evidence as you seem to be, couldn't you find a link with:

I'm sorry if it come of as snide but psychoacoustic research is a very big area in modern acoustic research. And it should be well-known to anyone in the audio engineering circle. I just figured it would be common knowledge.

Here's a paper from the 2006 AES conference that described a practical low complexity implementation for headphones. It is just a plug-in similar to this that I request for the miniDSP.

http://cseweb.ucsd.edu/~marora/files/papers/13868.pdf
 
Well, not no, I would expect this to have already been done by the engineer at the recording studio (using the afore mentioned Maxxbass plugin, tube compressor, analog tape machine or the bass player's tube amp :) ) if he thought the music required it.

By that you mean the recording engineer, or actually mastering engineer as it would be, should dictate or accurately predict the type of speakers you are using for reproduction?

And I have no problem creating virtual bass enhanced material, there are numerous plug-ins and other simpler techniques that can be used. I merely request such a function for the miniDSP.

I might also add that the addition of virtual bass is not only relevant for recorded music playback. It can also be used for a compact portable bass guitar monitor. Especially for practice use. As we know lower frequencies travels through buildings more easily than higher frequency content it would be a great benefit to be able to replace that very low frequency content with a higher frequency virtual content without affecting the sound.
 
Last edited:
I'm sorry if it come of as snide but psychoacoustic research is a very big area in modern acoustic research. And it should be well-known to anyone in the audio engineering circle. I just figured it would be common knowledge.

Here's a paper from the 2006 AES conference that described a practical low complexity implementation for headphones. It is just a plug-in similar to this that I request for the miniDSP.

http://cseweb.ucsd.edu/~marora/files/papers/13868.pdf
Apology accepted.... on the other hand, is it really necessary for you to insert a nasty put-down like "should be well-known to anyone in the audio engineering circle" as if those* lacking this information are boobies.

Thanks for the link. But isn't there any research with humans? There are lots of ways to augment (and over-augment) bass and to implement "loudness" compensation; but I'm curious about the special theory you seem to be endorsing about certain partials.

Ben
*I wouldn't be surprised if there are people on this thread who have taught university courses in perception yet are unfamiliar with "... should be well-known...", ahem, ahem
 
Last edited:
While there's a lot of research going on, unfortunately precious few articles are publicly available as everyone studying the field seem to be snatched up by companies with interest in the area.

AAU (Aalborg University Center), DTU (Danish Institute of Technology) and MIT are the primary places to find relevant research papers into the field. (Unfortunately much of the research is Danish language only).
 
Last edited:
While there's a lot of research going on, unfortunately precious few articles are publicly available as everyone studying the field seem to be snatched up by companies with interest in the area.

AUC (Aalborg University Center), DTU (Danish Institute of Technology) and MIT are the primary places to find relevant research papers into the field. (Unfortunately much of the research is Danish language only).

Must be a conspiracy.

B.
 
By that you mean the recording engineer, or actually mastering engineer as it would be, should dictate or accurately predict the type of speakers you are using for reproduction?

Yes. Studios commonly use small monitor speakers that are supposed to be representative of the average hi-fi system, and tailor the material to sound good on those.

I might also add that the addition of virtual bass is not only relevant for recorded music playback. It can also be used for a compact portable bass guitar monitor. Especially for practice use.

An underpowered cheap practice amp with an open-backed speaker cabinet will do exactly the same thing. :)
 
Yes. Studios commonly use small monitor speakers that are supposed to be representative of the average hi-fi system, and tailor the material to sound good on those.

Not the point.

An underpowered cheap practice amp with an open-backed speaker cabinet will do exactly the same thing. :)

No.

Since you obviously don't grasp what the concept and use is, why comment?