Apodizing filter

[bigshot]

Those of us in sound science are forbidden to go out into the other head Fi forums and discuss the importance of controlled tests, but anyone who wants to crap their placebo all over us here in our forum is free to do so because “it’s only fair”.

 

[gregorio]

it shifts the problem, so you can actually improve it with better filters...

Firstly, the “problem” was entirely theoretical, GoldenSound could not provide any examples in practice. Secondly, the “problem” occurs in the ultrasonic range, between about 20kHz - 22.05kHz, so how would shifting that “problem” to an even higher frequency “actually improve it”? Certainly it wouldn’t actually audibly improve anything because 20kHz - 22.05kHz is just as inaudible as shifting it even further into the ultrasonic range!

G

 

[bigshot]

The old “better numbers always mean better sound” myth. Interesting that someone who “trusts his ears” would argue theory based purely on numbers.

 

[Ghoostknight]

Firstly, the “problem” was entirely theoretical, GoldenSound could not provide any examples in practice. Secondly, the “problem” occurs in the ultrasonic range, between about 20kHz - 22.05kHz, so how would shifting that “problem” to an even higher frequency “actually improve it”? Certainly it wouldn’t actually audibly improve anything because 20kHz - 22.05kHz is just as inaudible as shifting it even further into the ultrasonic range!

im unsure if linear phase filters are similar to minimum phase in this regard

but to give an example: a minimum phase low pass at 20khz alters phase way below 20khz (probably starting at 15khz or even lower), probably the same thing happens with pre/post ringing

imo thats why its audible ..... i know what a -0,5db point at 20-22khz sound like, also audible but this is not what makes filter "sound nasty", its subpar transient or phase response

 

[Ghoostknight]

The old “better numbers always mean better sound” myth. Interesting that someone who “trusts his ears” would argue theory based purely on numbers.

some red line you have to follow to make sense of things... specially if you read all this nonsense about it

 

[bigshot]

i know what a -0,5db point at 20-22khz sound like, also audible

That’s beyond just wrong. It’s delusional. You probably don’t hear anything at all above 16kHz, and you certainly can’t discern half dB differences anywhere near those frequencies. You’d be lucky to hear -0.5 dB at 3kHz.

You throw around numbers without any clue as to what they represent. You’re so far into the weeds, you’d actually do better to steer clear of facts altogether and just go by your bias and perceptual error. At least then you’d be getting some sort of benefit from placebo and you wouldn’t stray quite that far off the mark.

 

[Ghoostknight]

You throw around numbers without any clue as to what they represent. You’re so far into the weeds, you’d actually do better to steer clear of facts altogether and just go by your bias and perceptual error. At least then you’d be getting some sort of benefit from placebo and you wouldn’t stray quite that far off the mark.

kinda the same thing i think about objectivists no worrys

That’s beyond just wrong. It’s delusional. You probably don’t hear anything at all above 16kHz, and you certainly can’t discern half dB differences anywhere near those frequencies. You’d be lucky to hear -0.5 dB at 3kHz.

i was talking about a slow roll off, so you still get -0.1-0.2db or such at 15 khz

it makes a difference, you guys can keep arguing it doesnt but yea...

im not saying lose your sleep over 0,2db, im just saying its audible if you actually know what to listen for...

 

[bigshot]

You don’t know what you’re talking about.

 

[Ghoostknight]

You don’t know what you’re talking about.

kinda the same thing i think about objectivists no worrys

 

[bigshot]

Tag, you’re it.

 

[BS5711]

i was talking about a slow roll off, so you still get -0.1-0.2db or such at 15 khz

it makes a difference, you guys can keep arguing it doesnt but yea...

im not saying lose your sleep over 0,2db, im just saying its audible if you actually know what to listen for...

It makes a difference to you only because you believe it does.

Disagree all you like but you are still wrong.

 

[bigshot]

You also can’t hear a 0.5 dB variation between 20-22kHz on headphones that have 20-30 dB variation in the same range. We’re talking princess and the pea.

 

[Ghoostknight]

You also can’t hear a 0.5 dB variation between 20-22kHz on headphones that have 20-30 dB variation in the same range. We’re talking princess and the pea.

the system you use doesnt make a difference if it can provide resolution up to 20khz

you are comparing A vs B .... 0db vs -0.5db , what difference does the headphones make? unless your argument ist " headphones sound so different to eachother, everything doesnt matter, get other headphones " kind of true in the grand scheme of things, since frequency response variations are the most audible but not if you simply compare A and B on the same system

these kind of arguments are the real strawman arguments...

 

[71 dB]

kinda the same thing i think about objectivists no worrys


i was talking about a slow roll off, so you still get -0.1-0.2db or such at 15 khz

it makes a difference, you guys can keep arguing it doesnt but yea...

im not saying lose your sleep over 0,2db, im just saying its audible if you actually know what to listen for...

If you take over the ears headphones off your head and put them back on, you probably get a change of SEVERAL dBs at 15 kHz because you can't put the headphones back in the exact same place and the interaction of high frequencies with pinna. Or if you are listening to speakers, even a very small movement of head will change the frequency response above 10 kHz more than 0.2 dB. Considering these factors present in audio, worrying about -0.2 dB gain at 15 kHz due to a slow roll off filter is silly. I don't even know why you insist as if slow roll filters were the only option. Use standard filter and you won't have attenuation at 15 kHz!

 

[bigshot]

I’d show you what a typical response curve in headphones looks like above 15kHz, but it would be a complete waste of time.

With speakers, it would be even worse. 20kHz probably doesn’t even make it from the speaker to the listening position in a good sized room.