Headphone Bass booster idea

Chester01 · Jan 16, 2005, 08:05 PM

Please read http://headwize.com/projects/showfil...meier5_prj.htm or atleast browse through it, it talks of creating harmonics of bass tones that headphones normally do quite poorly reproducing. Basically you create the harmonics that the headphones can play and the brain then "creates" the lower frequency. Apparently it works, would it be possible to make a dsp effect that did something similar to this? Thanks!

-Matt

ROBSCIX · Jan 16, 2005, 09:48 PM

Actually it's a very well know effect and it is used in alot of DSP's....I think the idea cam from the middle ages..I am not kidding, it does work...but you will not get the physical vibrations you only hear the lower notes....Check Waves.com they have an effect called Maxxbass or rennsaince bass that does this very thing, I am pretty Sure I still have the proof of Concept Equation for the effect...hmm, Check www.Maxx.com..you should be able to find all info you need..I never considered adding a Bass harmonic modeler....it does work though...post back

Chester01 · Jan 17, 2005, 12:01 AM

well i made a sort of bass extender, the APS Pitch plugin doesnt sound umm clean enough i dont think... it sounds a bit muddy right now. try it, i am using the 3538's, it needs a 4th order crossover, 3 aps pitches and two stereo mixes. it sorta works though

sounds a bit like its being played underwater lol maybe if we had a better pitch shifter it would sound better (no offence to the guy who made the pitch shifter)

http://n.1asphost.com/Chesteta/bass%20booster.JPG

Chester01 · Jan 17, 2005, 12:43 AM

I believe I have made a bit of progress now, it appears you should not run the modified stuff through a subwoofer and after closer examination of how they do things, I have removed one aps pitch and gotten it to sound pretty good except on REALLY low stuff will keep posted

ROBSCIX · Jan 18, 2005, 05:49 PM

If we are both talking about the Same thing..It's called Bass Harmonic modelling....IF the human ear hears the Higher order harmonics of a root note the ear hears the root note..is that what your talking about? I don't think you could use a pitch shifter to create this effect....You need a algorithm to add the higher order harmonics that the speakers can produce to get the lower roots the speakers cannot..I'll look around for the equation..actually it might be a good strating point for me to delve into DSP......I would probably need some help though..when I find the equation I will post it...

ROBSCIX · Jan 18, 2005, 05:53 PM

Now that I look at your setup...I say well done, I think your on the right track...did you bother looking at the info for the Maxx bass Processors?....It performs this exact audio "illusion" we are speaking of....The general purpose for the plugin was to add bass to low end speaker systems..or add bass to speakers with a sub-but draw less power..It's a nice effect and I used it exstensivley for quite a while..I had crappy speakers..you can hear the bass plain as day but UNFORTUNATLEY..you cannot feel it...but it is still very cool...

Chester01 · Jan 18, 2005, 06:04 PM

yes, actually i downloaded their powerpoint on how it works (the presentation thing) and that is what my dsp is based upon. the pitch shifter seems like it makes a whiring effect (like its under water) though otherwize it would work perfectly. the presentation is where i got the ideas for the frequency of the crossover, and the drop offs for the resonances (first -3, then -9.) thanks for the link! if you have any good ideas how to make it better please share

ROBSCIX · Jan 18, 2005, 06:16 PM

your welcome, I never actually thought about using this algorithm..now I was wondering if maybe it would be easier just to wirte a DSP effect from scratch based on this idea..that link you gave when I looked through that Hardware processor looks like it functions exactly the same as the Software MaxxBass-meaning you can set crossover...etc..I am gonna try and build one from scratch..instead of doing piecework...Are you good with programming?..I would hope some programmers would chime in on this collaboration.....you may be getting alot of mud because of your Particular setup...your speakers can drop down very low..so what you may be trying to achive with this is lower power comsumption..our more control over your bass signal..
this idea could be a good addition for KX...adding a realistic bass effect in real time for people who don't have subs...

Chester01 · Jan 18, 2005, 10:16 PM

about my setup; yes it does go quite low, actually the subwoofers that you see have now been put into a sonotube subwoofer, 10 cubic feet each with a port tuning of 20 hz. it theoretically should be 197.77 through 18 hz plus or minus 1.3 decibles

anyways, i was intending to use it for headphones (when its late and my parents dont want the floor shakin

) I am not very good at programing, I am currently a sophmore in highschool and am in FST, functions, statistics and triginometry; im good at mathematics. I would be willing to work on this with you though, we could both learn off eachother... it would be nice to have some of the big league kx programmers interest though

EDIT: after a bit more through reading of the link i posted, i noticed this: "The input signal, Vin(t), is buffered by two parallel buffer stages. Input buffer 1 leaves the input signal unaltered. Input buffer 2 filters the signal and removes the high frequency (and the very low frequency) components." perhaps this is why it sounded "whirry" or "underwater" to me, I will work on removing the low low bass and post back

ROBSCIX · Jan 18, 2005, 11:47 PM

Ok, SOunds cool..I am reading through the programming guide right now..shouldn't be too hard I have been programming since it was just C..lol..but I haven't did that type in awhile..more basic compiled machine code for micro controllers but the language at which it's programmed is not important, getting the algorithm right IS..Now give me a bit to find the actual equation that produces this effect..If not then I will have to figure one out...then we can add some input values and output values...should be able to get it to function..I understand headphones or sub..but if we produce this plug-then the design would be universal-just different cutoff points for speakers or headphones... we should be able to get a working model..then who knows if it functions the way it's supposed to maybe it would get added to the project..but you are right, maybe a couple other people will give input on it..

ROBSCIX · Jan 19, 2005, 12:03 AM

Hey Chester, did you come across the formula? Can you piece one together..my brain seems to be in programming mode right now?..but I will keep looking for it..

Chester01 · Jan 19, 2005, 12:05 AM

i believe it is listed on the headwize site... ill check and post back, btw, check ur e-mail

or atleast the one u listed on these forums

EDIT: here is the maxx bass "presentation" (i hope you have powerpoint)
http://www.maxx.com/objects/MaxxBass_Overview_Jan03.ppt

nyquility · Feb 1, 2005, 03:01 PM

Hi Chester and ROBSCIX!

It would be good to know if this idea is making any progress, if not maybe you (Chester) could post a shot of your newest, tweaked setup for headphone users like myself to emulate.

Thanks in advance

Chester01 · Feb 1, 2005, 10:58 PM

hey, we have not made any progress on the programming; actually ROBOSCIX had been looking into creating a plugin that would allow for realtime VST plugin processing through the soundcard, that did not work out very well. I have been looking more and more into the dsp setups, if you read the links I have posted (especially the powerpoint) you should be able to figure out how things work. In a bit I will post a picture (I need to fix a few things)

-Matt

Chester01 · Feb 2, 2005, 12:58 AM

Well, here is an image of what I have found to work.
http://n.1asphost.com/Chesteta/Perfe...ss%20Boost.JPG
I find that I do not even need a EQ with this setup; it bumps (as much as headphones can bump). I have set the levels by using the WaveGenerator 3.0 and running varying square waves at 0.0 db (a sweep from 0-300 hz, also some checks at higher frequencies) and have found with the current settings I can get +0.1 db through the entire frequency range; this is before cutting the frequencies below the crossover.

Principle
1.reduce gain so that things do not clip
2.divide high and low frequencies (I use 50 hz, I wouldnt go above 60, you cannot go below 40 (plugin limitation))
3. send the higher frequencies to be mixed with processed audio
4. create two resonances; one up one octave and then take that up half an octave (this gives you multiples of the frequencies)
5. send the multiples outputs through a gain plugin, first multiple down three db, seccond one is down nine db
6.mix your three stereo sources into one output
7. cut out anything below the crossover frequency, this is VERY IMPORTANT otherwise you will get a whirring effect and the bass will not kick, remember; if you everything to 50 hz, you should be able to achieve 25 hz of bass (divide the frequency of the crossover and highpass filter by two)

Tell me your expleriences please! Good Luck!
-Matt

EDIT: you may want to use another 4th order crossover instead of the EQ Highpass, or if you understand the width properly you could tell me a good setting to have it cut off low frequencies fast without spiking at the crossover frequency. sorry for my error

nyquility · Feb 2, 2005, 09:23 AM

Thanks for the screenshot, I will take a look at both the setup and the powerpoint and see what I get. BTW Chester after seeing some shots of your subwoofers... I feel sorry for you neighbours

nyquility · Feb 2, 2005, 11:07 AM

OK, Chester after an hour of testing and research here's what I came up with:

First of all I am by no means a through and through audiophile so some of my assumptions and analysis may be faulty. Please don't jump on my head, I am still trying to think on my feet and learn as I go along.

Lets start with the theoretical:

As far as I can tell, your setup of APS Pitches and Gains seems to do the job of emulating the MaxBass setup very well, inserting the missing fundamental nicely. I still have to do some maths to verify the actual numbers you used but they seem sound. Your question/comment at the end pertained to the high pass you were using and at which value you want it set. The best explaination and examples (with graphs) I could find at short notice was here:

http://www.arboretum.com/support/man...c_filters.html

First of all there is no telling how the values of Q used in this proprietary product relate to the values used in the high pass filter in kX but I am assuming that there is some relation if not a matching set here.

As you can see by the graphs with the high pass, 1 gives you an accurate ~0 db response at your cutoff frequency with a smooth dropoff below that value, anything higher gives you a boost of the cutoff point but unfortunately also a boost of some lower frequencies especially close to the cutoff point. As by the definition in the powerpoint we want to achieve a sharp cut at 50Hz we would probably need to use a 4th order crossover instead of the highpass. Even a highpass set to Q=1 is going to give you some trickle of lower frequencies.

But then again, listening to music is a very subjective things in many ways (RIAA testing be damned) and I am guessing any setup will still be dependant on equipment used in a large way. To get thumping basses without whirring will depend on the headphones you use, for myself (I use a set of midrange Phillips Clarity headphones) I have found that without your filter setup I get very snappy, kicking bass lacking some depth and with some distortion, adding your booster setup (with a 4th order crossover as the final stage) added some depth to the bass (making it sound warmer), removed the distortion but also removed the "snap". This was especially noticeable with modern trance music where I was getting less of the spine tingling effect with your booster setup and in classical music with heavy vibrations being almost distant. So it seems while my brain does a good job of actually inserting the bass tone, the physical sensation of vibration is missing.

I am gonna spend the next few hours listening to a variety of music to see how it shapes up with and without the booster setup and get back to you.

Edit: A couple hours later and I have to admit the results at the moment aren't too impressive

I've reached R&B and with that style of musics multiples of long, clean bass thumps the setup is showing some weakness. I am getting a lot of rolling basses, which means that the bass thumps are drifting into each other creating an almost monotone bass staccato, lacking definition. I tend to agree with your analysis that the APS Pitch may be causing this. There are some delay lines in the microcode which may soften the attack and decay on the notes.

Chester01 · Feb 2, 2005, 07:29 PM

do you think that there is some way we can remove those delay lines? or delay some other part of the music to match the APS Pitch delay (so everything would be delayed with the same amount)? Thanks for your input! I have also found that a 4th order crossover works well also, I think the reason it was "bumping" as I said before was because of a Q of 20

. we may need to boost the bass before we process it seems. btw, my neighbors houses are atleast 20 feet from my room so they shouldnt hear much

just the really low notes

Tril · Nov 25, 2005, 01:47 AM

I wanted to make one for months but I was having problems figuring out a nice algorithm. I finally got something working. It's not bad at all. It produces less artifacts than using APS Pitch plugin.
It's an harmonics generator. I called it HarmonicsGen. It leaves the fundamental frequency in the output signal.

I recommend putting a Lowpass filter set at around 100 to 150 Hz before HarmonicsGen.

Connect the non-Lowpassed signal and the output of HarmonicsGen to a StereoMix plugin and adjust the Volumes of the Stereo Mix plugin. If the volume of the harmonics generator is too high there will be some distortion so don't set it too high.

I get the harmonics by multiplying the signal by itself. For the even harmonics, this creates a DC offset. I use an algorithm that corrects this offset. It's especially this part that gave me trouble in the past when I was less experienced with kX plugins programming. Now, it's different. It only took me an evening to write this plugin.

The problem I see with multiplying the signal by itself to mutiply it's frequency by two is that the quieter a signal is, the quieter the output will be (it's not linear).

Some examples ;
input amplitude = 1, output amplitude = 1^2 = 1
input amplitude = 0.8, output amplitude = 0.8^2 = 0.64
input amplitude = 0.6, output amplitude = 0.6^2 = 0.36
input amplitude = 0.4, output amplitude = 0.4^2 = 0.16

Even with that problem, it still works well.

Code:

; New microcode
name "HarmonicsGen";
copyright "Copyright (c) 2005. Tril";
created "11/24/2005";
engine "kX";
; comment "";
guid "251a33a7-412d-4ee4-a8b6-c6503c1bd23b";
; -- generated GUID
 
;Both
input inL, inR;
output outL, outR; 
control Vol = 1;
temp test;
; Left
temp h1L, h2L, h3L;
temp h4L, h5L;
static amplitude1L = 0;					
temp camplitude1L, uamplitude1L;
static amplitude2L = 0;					
temp camplitude2L, uamplitude2L; current amplitude and test and used amplitude
; Right
temp h1R, h2R, h3R;
temp h4R, h5R;
static amplitude1R = 0;					
temp camplitude1R, uamplitude1R;
static amplitude2R = 0;					
temp camplitude2R, uamplitude2R; current amplitude and test and used amplitude
 
; code
;Left
; Reduce amplitude (by a factor every sample)
macs amplitude1L, amplitude1L, 0.00001, -1;
macs amplitude2L, amplitude2L, 0.00001, -1;
macs h1L, inL, 0, 0;
; Find current amplitude
macs camplitude1L, h1L, 0, 0;
skip	ccr, ccr, 0x180, 1 ;>0
	macints camplitude1L, 0, camplitude1L, -1;
 
; if (camplitude > amplitude)
; amplitude = camplitude			 
macints test, amplitude1L, camplitude1L, -1;
skip	ccr, ccr, 0x180, 1 ;>0
	macs amplitude1L, camplitude1L,0, 0; 
macs uamplitude1L, 0, amplitude1L, amplitude1L;
 
macs h2L, 0, h1L, h1L;
macs h3L, 0, h1L, h2L;
 
macs h2L, h2L, uamplitude1L, -0.5;
macints h2L, 0, h2L, 2;
macs h4L, 0, h2L, h2L;
 
; Find current amplitude						
macs camplitude2L, h2L, 0, 0;
skip	ccr, ccr, 0x180, 1 ;>0
	macints camplitude2L, 0, camplitude2L, -1;
 
; if (camplitude > amplitude)
; amplitude = camplitude
macints test, amplitude2L, camplitude2L, -1;
skip	ccr, ccr, 0x180, 1 ;>0
	macs amplitude2L, camplitude2L,0, 0; 
macs uamplitude2L, 0, amplitude2L, amplitude2L;		 
 
macs h4L, h4L, uamplitude2L, -0.5;
macints h4L, 0, h4L, 2;		 
macs h5L, 0, h3L, h2L;					
macints h5L, h5L, h4L, 1;
macints h5L, h5L, h3L, 1;
macints h5L, h5L, h2L, 1;
macints h5L, h5L, h1L, 1;
; Right
;Left
; Reduce amplitude (by a factor every sample)
macs amplitude1R, amplitude1R, 0.00001, -1;
macs amplitude2R, amplitude2R, 0.00001, -1;
macs h1R, inR, 0, 0;
; Find current amplitude
macs camplitude1R, h1R, 0, 0;
skip	ccr, ccr, 0x180, 1 ;>0
	macints camplitude1R, 0, camplitude1R, -1;
 
; if (camplitude > amplitude)
; amplitude = camplitude			 
macints test, amplitude1R, camplitude1R, -1;
skip	ccr, ccr, 0x180, 1 ;>0
	macs amplitude1R, camplitude1R,0, 0; 
macs uamplitude1R, 0, amplitude1R, amplitude1R;
 
macs h2R, 0, h1R, h1R;
macs h3R, 0, h1R, h2R;
 
macs h2R, h2R, uamplitude1R, -0.5;
macints h2R, 0, h2R, 2;
macs h4R, 0, h2R, h2R;
 
; Find current amplitude						
macs camplitude2R, h2R, 0, 0;
skip	ccr, ccr, 0x180, 1 ;>0
	macints camplitude2R, 0, camplitude2R, -1;
 
; if (camplitude > amplitude)
; amplitude = camplitude
macints test, amplitude2R, camplitude2R, -1;
skip	ccr, ccr, 0x180, 1 ;>0
	macs amplitude2R, camplitude2R,0, 0; 
macs uamplitude2R, 0, amplitude2R, amplitude2R;		 
 
macs h4R, h4R, uamplitude2R, -0.5;
macints h4R, 0, h4R, 2;		 
macs h5R, 0, h3R, h2R;					
macints h5R, h5R, h4R, 1;
macints h5R, h5R, h3R, 1;
macints h5R, h5R, h2R, 1;
macints h5R, h5R, h1R, 1;
; output
macs outL, 0, h5L, Vol;
macs outR, 0, h5R, Vol;
end

Maddogg6 · Nov 25, 2005, 04:55 AM

I did this:

Open KX editor - Pasted the above post.. copied a unique GUID and pasted that over top your GUID above...
Saved as dane file - and then dbl clicked on that .da file to register it, and got 'invalid plugin library'. or something like that.

Is this dane source supposed to replace other dane source with accompaning DLL??

Tril · Nov 25, 2005, 05:21 AM

It's a simple dane plugin. It does not need any dll and it's compatible with all driver versions. Your mistake is probably that you need to add an empty line at the end of the file. Go at the end of the code and press enter a few times and save that.

I'll host the file for some time (right-click and save) : http://pages.globetrotter.net/samaus...armonicsGen.da

Maddogg6 · Nov 25, 2005, 08:14 AM

That was it ... I did not know about the extra line thing

Thank you sir...

Russ · Nov 25, 2005, 03:06 PM

Pretty good Tril. Personally, I find settings between 60-80 for the lowpass to work best, but a lot depends on the music you are listening to. I have to play around with it some more, and make some presets for the lowpass filter so that I do not have to keep adjusting the settings for different music. If I use settings in 100-150 range, than I find that what I hear is mostly due to the lowpass filter (i.e. bypassing this plugin has little or no effect), but again, it probably depends on the type of music that you are listening to.

BTW: I only quickly glanced over your code, but it could be optimized a lot. i.e. You can probably eliminate half of the temp registers by simply re-using them for both channels rather than having seperate ones for each channel, etc.

Russ · Nov 25, 2005, 03:44 PM

BTW:
I would also recommend adding a GainHQ plugin between this plugin and the stereo mix plugin, as the stereo mix plugin cannot boost the signal enough (for my taste anyway) by itself.

Tril · Nov 25, 2005, 06:30 PM

Quote:

Originally Posted by Russ

BTW: I only quickly glanced over your code, but it could be optimized a lot. i.e. You can probably eliminate half of the temp registers by simply re-using them for both channels rather than having seperate ones for each channel, etc.

That's probably true. I did not take the time to optimize. I wanted to make something that works before trying to optimize.

Also, the four macints at the end of both channels can be changed to two acc3, saving 4 instructions.

Maddogg6 · Nov 25, 2005, 07:19 PM

Quote:

Originally Posted by Russ

Pretty good Tril. Personally, I find settings between 60-80 for the lowpass to work best, but a lot depends on the music you are listening to.

Id think it depends on external amp/speakers too..

My headphones start dropping off at 60hz - so for headphones I have:

low-pass (80hz/ Q: +3)

then

gainHQ (+3, cuz source is pretty hot, ~0db peaks)
thanks for idea Russ - btw..)

then

HarmonicGen (set to 100)

then

Stereo mix (100% - original) + 80% hamonics added - Ill call the 'ASS Channel')

For satriani's 'Souls of Destruction' its sounds phenominal...

This is accomplishing what I was using EQs and e-compressors to compensate for with these headphones. (well, mostly because I could, too

)

Point is - its saving me some resources even if its not optimized yet...
Also as Russ was pointing out - depending on what Im listening to I tweak em - and just a little makes a pretty big difference configured this way..
I think Ill play with it more too to get something that will be more 'set and forget' for me.

Thanks again Tril. (if I had your brain, Id throw mine away.. :S)

Russ · Nov 25, 2005, 07:59 PM

Maddogg, that is similair to the settings I am using, except I set the HarmonicGen lower and increased the GainHQ instead. I also added a 10 Band EQ to the end of the chain, and got some pretty sweet sound out of my MS Game Voice headset.

Jan 16, 2005, 08:05 PM	#1
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	Headphone Bass booster idea Please read http://headwize.com/projects/showfil...meier5_prj.htm or atleast browse through it, it talks of creating harmonics of bass tones that headphones normally do quite poorly reproducing. Basically you create the harmonics that the headphones can play and the brain then "creates" the lower frequency. Apparently it works, would it be possible to make a dsp effect that did something similar to this? Thanks! -Matt __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands...

Jan 17, 2005, 12:01 AM	Thread Starter #3
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	well i made a sort of bass extender, the APS Pitch plugin doesnt sound umm clean enough i dont think... it sounds a bit muddy right now. try it, i am using the 3538's, it needs a 4th order crossover, 3 aps pitches and two stereo mixes. it sorta works though sounds a bit like its being played underwater lol maybe if we had a better pitch shifter it would sound better (no offence to the guy who made the pitch shifter) http://n.1asphost.com/Chesteta/bass%20booster.JPG __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands... Last edited by Chester01; Jan 17, 2005 at 12:19 AM.

Jan 17, 2005, 12:43 AM	Thread Starter #4
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	I believe I have made a bit of progress now, it appears you should not run the modified stuff through a subwoofer and after closer examination of how they do things, I have removed one aps pitch and gotten it to sound pretty good except on REALLY low stuff will keep posted __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands...

Jan 18, 2005, 06:04 PM	Thread Starter #7
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	yes, actually i downloaded their powerpoint on how it works (the presentation thing) and that is what my dsp is based upon. the pitch shifter seems like it makes a whiring effect (like its under water) though otherwize it would work perfectly. the presentation is where i got the ideas for the frequency of the crossover, and the drop offs for the resonances (first -3, then -9.) thanks for the link! if you have any good ideas how to make it better please share __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands...

Jan 18, 2005, 10:16 PM	Thread Starter #9
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	about my setup; yes it does go quite low, actually the subwoofers that you see have now been put into a sonotube subwoofer, 10 cubic feet each with a port tuning of 20 hz. it theoretically should be 197.77 through 18 hz plus or minus 1.3 decibles anyways, i was intending to use it for headphones (when its late and my parents dont want the floor shakin ) I am not very good at programing, I am currently a sophmore in highschool and am in FST, functions, statistics and triginometry; im good at mathematics. I would be willing to work on this with you though, we could both learn off eachother... it would be nice to have some of the big league kx programmers interest though EDIT: after a bit more through reading of the link i posted, i noticed this: "The input signal, Vin(t), is buffered by two parallel buffer stages. Input buffer 1 leaves the input signal unaltered. Input buffer 2 filters the signal and removes the high frequency (and the very low frequency) components." perhaps this is why it sounded "whirry" or "underwater" to me, I will work on removing the low low bass and post back __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands... Last edited by Chester01; Jan 18, 2005 at 10:22 PM.

Jan 16, 2005, 09:48 PM	#2
ROBSCIX DriverHeaven Senior Member Join Date: Mar 2004 Posts: 1,732 Rep Power: 0	Actually it's a very well know effect and it is used in alot of DSP's....I think the idea cam from the middle ages..I am not kidding, it does work...but you will not get the physical vibrations you only hear the lower notes....Check Waves.com they have an effect called Maxxbass or rennsaince bass that does this very thing, I am pretty Sure I still have the proof of Concept Equation for the effect...hmm, Check www.Maxx.com..you should be able to find all info you need..I never considered adding a Bass harmonic modeler....it does work though...post back

Jan 18, 2005, 05:49 PM	#5
ROBSCIX DriverHeaven Senior Member Join Date: Mar 2004 Posts: 1,732 Rep Power: 0	If we are both talking about the Same thing..It's called Bass Harmonic modelling....IF the human ear hears the Higher order harmonics of a root note the ear hears the root note..is that what your talking about? I don't think you could use a pitch shifter to create this effect....You need a algorithm to add the higher order harmonics that the speakers can produce to get the lower roots the speakers cannot..I'll look around for the equation..actually it might be a good strating point for me to delve into DSP......I would probably need some help though..when I find the equation I will post it...

Jan 18, 2005, 05:53 PM	#6
ROBSCIX DriverHeaven Senior Member Join Date: Mar 2004 Posts: 1,732 Rep Power: 0	Now that I look at your setup...I say well done, I think your on the right track...did you bother looking at the info for the Maxx bass Processors?....It performs this exact audio "illusion" we are speaking of....The general purpose for the plugin was to add bass to low end speaker systems..or add bass to speakers with a sub-but draw less power..It's a nice effect and I used it exstensivley for quite a while..I had crappy speakers..you can hear the bass plain as day but UNFORTUNATLEY..you cannot feel it...but it is still very cool...

Jan 18, 2005, 06:16 PM	#8
ROBSCIX DriverHeaven Senior Member Join Date: Mar 2004 Posts: 1,732 Rep Power: 0	your welcome, I never actually thought about using this algorithm..now I was wondering if maybe it would be easier just to wirte a DSP effect from scratch based on this idea..that link you gave when I looked through that Hardware processor looks like it functions exactly the same as the Software MaxxBass-meaning you can set crossover...etc..I am gonna try and build one from scratch..instead of doing piecework...Are you good with programming?..I would hope some programmers would chime in on this collaboration.....you may be getting alot of mud because of your Particular setup...your speakers can drop down very low..so what you may be trying to achive with this is lower power comsumption..our more control over your bass signal.. this idea could be a good addition for KX...adding a realistic bass effect in real time for people who don't have subs...

Jan 18, 2005, 11:47 PM	#10
ROBSCIX DriverHeaven Senior Member Join Date: Mar 2004 Posts: 1,732 Rep Power: 0	Ok, SOunds cool..I am reading through the programming guide right now..shouldn't be too hard I have been programming since it was just C..lol..but I haven't did that type in awhile..more basic compiled machine code for micro controllers but the language at which it's programmed is not important, getting the algorithm right IS..Now give me a bit to find the actual equation that produces this effect..If not then I will have to figure one out...then we can add some input values and output values...should be able to get it to function..I understand headphones or sub..but if we produce this plug-then the design would be universal-just different cutoff points for speakers or headphones... we should be able to get a working model..then who knows if it functions the way it's supposed to maybe it would get added to the project..but you are right, maybe a couple other people will give input on it..

Jan 19, 2005, 12:03 AM	#11
ROBSCIX DriverHeaven Senior Member Join Date: Mar 2004 Posts: 1,732 Rep Power: 0	Hey Chester, did you come across the formula? Can you piece one together..my brain seems to be in programming mode right now?..but I will keep looking for it..

Jan 19, 2005, 12:05 AM	Thread Starter #12
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	i believe it is listed on the headwize site... ill check and post back, btw, check ur e-mail or atleast the one u listed on these forums EDIT: here is the maxx bass "presentation" (i hope you have powerpoint) http://www.maxx.com/objects/MaxxBass_Overview_Jan03.ppt __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands... Last edited by Chester01; Jan 19, 2005 at 12:12 AM.

Feb 1, 2005, 03:01 PM	#13
nyquility DriverHeaven Junior Member Join Date: May 2004 Posts: 29 Rep Power: 0	Hi Chester and ROBSCIX! It would be good to know if this idea is making any progress, if not maybe you (Chester) could post a shot of your newest, tweaked setup for headphone users like myself to emulate. Thanks in advance

Feb 1, 2005, 10:58 PM	Thread Starter #14
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	hey, we have not made any progress on the programming; actually ROBOSCIX had been looking into creating a plugin that would allow for realtime VST plugin processing through the soundcard, that did not work out very well. I have been looking more and more into the dsp setups, if you read the links I have posted (especially the powerpoint) you should be able to figure out how things work. In a bit I will post a picture (I need to fix a few things) -Matt __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands...

Feb 2, 2005, 12:58 AM	Thread Starter #15
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	Well, here is an image of what I have found to work. http://n.1asphost.com/Chesteta/Perfe...ss%20Boost.JPG I find that I do not even need a EQ with this setup; it bumps (as much as headphones can bump). I have set the levels by using the WaveGenerator 3.0 and running varying square waves at 0.0 db (a sweep from 0-300 hz, also some checks at higher frequencies) and have found with the current settings I can get +0.1 db through the entire frequency range; this is before cutting the frequencies below the crossover. Principle 1.reduce gain so that things do not clip 2.divide high and low frequencies (I use 50 hz, I wouldnt go above 60, you cannot go below 40 (plugin limitation)) 3. send the higher frequencies to be mixed with processed audio 4. create two resonances; one up one octave and then take that up half an octave (this gives you multiples of the frequencies) 5. send the multiples outputs through a gain plugin, first multiple down three db, seccond one is down nine db 6.mix your three stereo sources into one output 7. cut out anything below the crossover frequency, this is VERY IMPORTANT otherwise you will get a whirring effect and the bass will not kick, remember; if you everything to 50 hz, you should be able to achieve 25 hz of bass (divide the frequency of the crossover and highpass filter by two) Tell me your expleriences please! Good Luck! -Matt EDIT: you may want to use another 4th order crossover instead of the EQ Highpass, or if you understand the width properly you could tell me a good setting to have it cut off low frequencies fast without spiking at the crossover frequency. sorry for my error __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands...

Feb 2, 2005, 09:23 AM	#16
nyquility DriverHeaven Junior Member Join Date: May 2004 Posts: 29 Rep Power: 0	Thanks for the screenshot, I will take a look at both the setup and the powerpoint and see what I get. BTW Chester after seeing some shots of your subwoofers... I feel sorry for you neighbours

Feb 2, 2005, 11:07 AM	#17
nyquility DriverHeaven Junior Member Join Date: May 2004 Posts: 29 Rep Power: 0	OK, Chester after an hour of testing and research here's what I came up with: First of all I am by no means a through and through audiophile so some of my assumptions and analysis may be faulty. Please don't jump on my head, I am still trying to think on my feet and learn as I go along. Lets start with the theoretical: As far as I can tell, your setup of APS Pitches and Gains seems to do the job of emulating the MaxBass setup very well, inserting the missing fundamental nicely. I still have to do some maths to verify the actual numbers you used but they seem sound. Your question/comment at the end pertained to the high pass you were using and at which value you want it set. The best explaination and examples (with graphs) I could find at short notice was here: http://www.arboretum.com/support/man...c_filters.html First of all there is no telling how the values of Q used in this proprietary product relate to the values used in the high pass filter in kX but I am assuming that there is some relation if not a matching set here. As you can see by the graphs with the high pass, 1 gives you an accurate ~0 db response at your cutoff frequency with a smooth dropoff below that value, anything higher gives you a boost of the cutoff point but unfortunately also a boost of some lower frequencies especially close to the cutoff point. As by the definition in the powerpoint we want to achieve a sharp cut at 50Hz we would probably need to use a 4th order crossover instead of the highpass. Even a highpass set to Q=1 is going to give you some trickle of lower frequencies. But then again, listening to music is a very subjective things in many ways (RIAA testing be damned) and I am guessing any setup will still be dependant on equipment used in a large way. To get thumping basses without whirring will depend on the headphones you use, for myself (I use a set of midrange Phillips Clarity headphones) I have found that without your filter setup I get very snappy, kicking bass lacking some depth and with some distortion, adding your booster setup (with a 4th order crossover as the final stage) added some depth to the bass (making it sound warmer), removed the distortion but also removed the "snap". This was especially noticeable with modern trance music where I was getting less of the spine tingling effect with your booster setup and in classical music with heavy vibrations being almost distant. So it seems while my brain does a good job of actually inserting the bass tone, the physical sensation of vibration is missing. I am gonna spend the next few hours listening to a variety of music to see how it shapes up with and without the booster setup and get back to you. Edit: A couple hours later and I have to admit the results at the moment aren't too impressive I've reached R&B and with that style of musics multiples of long, clean bass thumps the setup is showing some weakness. I am getting a lot of rolling basses, which means that the bass thumps are drifting into each other creating an almost monotone bass staccato, lacking definition. I tend to agree with your analysis that the APS Pitch may be causing this. There are some delay lines in the microcode which may soften the attack and decay on the notes. Last edited by nyquility; Feb 2, 2005 at 12:12 PM.

Feb 2, 2005, 07:29 PM	Thread Starter #18
Chester01 HardwareHeaven Senior Member Join Date: Jan 2004 Location: St. Cloud, MN Posts: 492 Rep Power: 0	do you think that there is some way we can remove those delay lines? or delay some other part of the music to match the APS Pitch delay (so everything would be delayed with the same amount)? Thanks for your input! I have also found that a 4th order crossover works well also, I think the reason it was "bumping" as I said before was because of a Q of 20 . we may need to boost the bass before we process it seems. btw, my neighbors houses are atleast 20 feet from my room so they shouldnt hear much just the really low notes __________________ COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium) STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands...

Nov 25, 2005, 04:55 AM	#20
Maddogg6 Tail Razer Join Date: Jun 2005 Location: Bernyurass, AZ - USA Posts: 4,027 Rep Power: 0	I did this: Open KX editor - Pasted the above post.. copied a unique GUID and pasted that over top your GUID above... Saved as dane file - and then dbl clicked on that .da file to register it, and got 'invalid plugin library'. or something like that. Is this dane source supposed to replace other dane source with accompaning DLL??

Nov 25, 2005, 05:21 AM	#21
Tril S-3D enthusiast Join Date: Sep 2004 Location: Canada Posts: 1,676 Rep Power: 64 System Specs	It's a simple dane plugin. It does not need any dll and it's compatible with all driver versions. Your mistake is probably that you need to add an empty line at the end of the file. Go at the end of the code and press enter a few times and save that. I'll host the file for some time (right-click and save) : http://pages.globetrotter.net/samaus...armonicsGen.da

Nov 25, 2005, 08:14 AM	#22
Maddogg6 Tail Razer Join Date: Jun 2005 Location: Bernyurass, AZ - USA Posts: 4,027 Rep Power: 0	That was it ... I did not know about the extra line thing Thank you sir...

Nov 25, 2005, 03:06 PM	#23
Russ HardwareHeaven Extreme Member Join Date: Jan 2005 Posts: 5,563 Rep Power: 62	Pretty good Tril. Personally, I find settings between 60-80 for the lowpass to work best, but a lot depends on the music you are listening to. I have to play around with it some more, and make some presets for the lowpass filter so that I do not have to keep adjusting the settings for different music. If I use settings in 100-150 range, than I find that what I hear is mostly due to the lowpass filter (i.e. bypassing this plugin has little or no effect), but again, it probably depends on the type of music that you are listening to. BTW: I only quickly glanced over your code, but it could be optimized a lot. i.e. You can probably eliminate half of the temp registers by simply re-using them for both channels rather than having seperate ones for each channel, etc.

Nov 25, 2005, 03:44 PM	#24
Russ HardwareHeaven Extreme Member Join Date: Jan 2005 Posts: 5,563 Rep Power: 62	BTW: I would also recommend adding a GainHQ plugin between this plugin and the stereo mix plugin, as the stereo mix plugin cannot boost the signal enough (for my taste anyway) by itself.

Nov 25, 2005, 07:59 PM	#27
Russ HardwareHeaven Extreme Member Join Date: Jan 2005 Posts: 5,563 Rep Power: 62	Maddogg, that is similair to the settings I am using, except I set the HarmonicGen lower and increased the GainHQ instead. I also added a 10 Band EQ to the end of the chain, and got some pretty sweet sound out of my MS Game Voice headset.