what is complex about it?'
i find it obnoxiously easy.
soundwaves look like peaks and dips.
those peaks and dips get recorded by the microphone.
that data hitting the microphone tells the computer to place the sample tap high or low (because of loud) and top or bottom (because of phase)
as the sample taps happen.. the line from left to right gets longer and longer.
the results recorded get flipped opposite and constantly applied to the audio data.
everybody on this forum talking about audio should know about hooking up the positive and negative wires backwards on one speaker .. because it causes the bass to cancel out (canceled decibels .. as well as canceled sound pressure level)
what do you think has happened the last 30 years with computer software?
sometimes not everything is 100% accurate.
they could take that line you recorded and adjust it (make it broken) on purpose .. and then make you pay more to get the program to stop adjusting it after it was recorded.
..other programs are accurate, but they are simply limited functionality.
bit depth (more chapters adding up for a stronger story)
bit rate (more paragraphs adding up to a chapter)
it doesnt strike me as odd to see one program cost $130
then another program doing the same thing with less limits for $600
whether those programs choose to shrink or expand those sample taps when they get adjusted.. that answer isnt important to me.
whether they are using each sample tap as a 'point' in the filter.. or if they are using every space inbetween those taps as a 'point' in the filter ... its an accuracy thing, its a room to fit thing.
maybe people need to know how to view the math.
the shape is squeezed and compressed tightly as if a piece of clay.
how does the squeezing and compression happen?
a huge pyramid.. like some cups stacked up on top of eachother.
those pyramids can 2 or 4
2 = only one up or down
4 = two up or downs (one for amplitude and one for phase)
quoted from here: http://www.dspguru.com/dsp/faqs/fir/basics
'A FIR "tap" is simply a coefficient/delay pair. The number of FIR taps, (often designated as "N") is an indication of 1) the amount of memory required to implement the filter, 2) the number of calculations required, and 3) the amount of "filtering" the filter can do; in effect, more taps means more stopband attenuation, less ripple, narrower filters, etc.'
so if you start with a single shape like this /
you would see that is the tip of the pyramid.
all the bricks below helped 'nudge' or 'squeeze' the shape to become what it is at the top.
many people dont realize it is possible to do math by trimming or adding to a shape.
but think of this..
if you record with the microphone, you are laying down a line that is a shape.
if your filter doesnt have enough taps.. then the entire percentage of the recording wont make it to the final result and the correction will be less.
if the operating system works on a limited number of taps ... then the convolution filter doesnt get to pick how the trim happens.
the program that recorded the impulse response will do it
the program that you use to edit and save the impulse response file will do it.
the program you used to record the impulse response file has to be responsible enough to step in and take action to gather the necessary data.. it keeps the final result high quality.
because the audio editing program might not have a list of what is and isnt important ... having that list of prioritized data value will keep the trimmed final results sounding as best as possible.
i think every time people communicate .... it is always a chance to stay stuck in the mono lifestyle you are in, or expand your list of interest and value.
often times too much effort goes towards allowing people to do one thing their whole life, which really disables them and cuts freedom short (or even takes it away).
yesterday i was thinking about how a cab driver sits in a vehicle and drives around in a car every single day .. and that person might do that job from the time they are 20 years old up until they retire at age 70 or 80 .. and then dead 5-10 years after that.
think about it..
if that person comes back to life, they've got a whole lifetime of being involved with people out there on the street (think of it as a library of evidence)
..if that person doesnt come back to life, they had a whole lifetime of repeating the same electricity day after day as if there isnt thousands of electrical situations existing on this planet every single day.
some people like to ignore the long list.
other people enjoy life for what it is in all of its complexity as they work to embrace it rather than ignore or hide from it.
you want a shorter version of how the correction works?
setup a microphone
use audio recording software to record a sine sweep
take the recording into the audio editor to trim the empty from the front, then inverse the file and save it.
run the file through the convolver.
want to know how it works in a different area?
the lines of audio data come pouring down like any of the dots you see in the video game 'guitar hero'
the filter is the line of buttons on the bottom.
when the treadmill of dots passes the main line of dots on the bottom ... the dot will change as it goes by.
you could say the filter is changing every single dot as it passes .. just like a factory pours chocolate over every single nut as it passes by (to put the nut inside the candy)
call it painted instead of chocolate
you want to think of it more hard?
look at a list of numbers 0 - 9
if you've got a number on the dot as it is traveling towards the filter ...... then the filter says 'each 50hz tone gets an automatic 3 added to it'
talking about binary.. those numbers can go above 9
so if the music said the data value is 9 .. and the filter said everything right there gets a 3 ... then the soundcard is going to see a 12
basics for recording the impulse, analyzing the impulse (limits included), and applying the filter to the stream of audio data
the math has been translated, i'm not going to sour myself because i didnt hand over a bunch of visual math symbols.
(but if people continue to drag eachother's energy down .... there wont be enough energy to listen to people that translate the math)