PocketSmiley
100+ Head-Fier
- Joined
- Dec 20, 2012
- Posts
- 173
- Likes
- 104
So,
months after trying to simulate the built-in Windows bass boost enhancement, specifically at 50Hz +24dB, I measured its frequency response again, zoomed in, and saw that it was peaked at 41Hz lol. But then I saw that my 25Hz peak was shifted to 35Hz this time. Then I saw my soundcard was using a 192KHz sample rate and was causing audible intermodular distortion. Anyway ...
@milk asked for advice on EQ for Mac OS X, so I'll just share what I said in here as @hummel is looking for better Mac EQs as well. I've never used OS X my entire life--UNIX yes--and my last attempt at a Hackintosh required buying a new wireless card, so I'm not aware of important Mac apps. What you want is a Parametric EQ (PEQ). For starters, try searching for "Parametric EQ" for OS X. I'm very, very sure PEQs are available for iOS as well as OS X since they're basic stuff. Here is a head-fi google search so you can look for similar threads before starting your own--I found this thread to be the most helpful so far. I was able to simulate the Windows bass boost using only a parametric EQ (Equalizer APO), so my settings should be usable in other parametric EQs:
Filter 1 : ON PEQ Fc 50 Hz Gain 24.00 dB Q 0.329
Filter 2 : ON PEQ Fc 25 Hz Gain 21.00 dB Q 1.000
Filter 3 : ON PEQ Fc 3000 Hz Gain 21.00 dB Q 0.266
Click here for the Windows tutorial.
RMAA FR measurements:
Notice how the green (APO w/ WBB) and white (APO only) curves are very close by -2dB from the 25Hz peak up, and only different in the less audible lower frequencies. Also note Cowon iAudio's BBE effect visible as an increasing slope starting from 600Hz towards infinity, which purports to "refine sound quality into clearer and more vivid sound," and see how Filter 3 is similarly curved up to the magic 3KHz peak frequency. You can definitely simulate other JetEffect technologies using only EQ (except reverb and stereo effects such as crossover and soundstage), even the $250 digiZoid ZO3 (which maxes at +14dB at 25Hz PF measured from the midrange floor; the ZO2.3 reaches +15dB at 25Hz PF):
Now, these settings should be a bit too extreme even for the extreme basshead, but tweaking them is as easy as lowering the gain of each filter by the same amount. Hawaii shared some EQ tricks in here. As a general solution, it works for most of the tracks that I listen to, and I even find it enjoyable with delicate piano tracks!
These settings, simple as they are, doesn't work properly with RockBox as RockBox clips the signal even with a -50dB precut. Let us know if you know any app with parametric EQ for Mac and Android!
Why do this?
Straightforward answer: Listen for yourself:
Wiz Khalifa - On My Level
Before: (Listen to normal volume.)
After: (This track is preamped to -24dB, so raise normal volume +24dB. This is an older version of my EQ settings with a much weaker Filter 3, so it does not sound as clear. This track with the new EQ sounds better and much clearer without losing bass impact.)
Judas Priest - Painkiller
Before: (Listen to normal volume.)
After: (This track is preamped to -24dB, so raise normal volume +24dB. This is an older version of my EQ settings with a much weaker Filter 3, so it does not sound as clear.)
Click here for more song comparisons.
Detailed answer:
Most people are slaves to what they unknowingly internalize as "original," "authoritative," "canonical." They grow a system of beliefs based on unchecked foundations, and within the hi-fi community, one such foundation is that there IS an "original" recording of music that needs to be reproduced accurately. Philosophers like de Man and Baudrillard have long freed us from this tethered thinking since last century, with catchphrases such as "They kill the original, by discovering that the original was already dead" and "The author is dead" graffitied in the blood of tyrants. Poststructuralism has been powerful in destroying absolutist frameworks that determine systems of oppression; however, some of its theories have also been used to justify nihilism and to discredit history. In the context of signal reproduction, true, there is such a thing as accuracy, where signal "A" must cross a journey to arrive at its destination in "C_T" as "A" and not "U" or else the resulting message becomes "CUT" and not "CAT." However, the issue is not accuracy, but whether YOU intend to send the signals "U" and "E" to "C_T" to make it "CUTE." It is not about killing the author as it is about bringing YOU, the audience, to life. You have agency. You are free. Most importantly, you are ALIVE.
Here at Head-Fi, digital/analog filters as simple as an equalizer are generally considered taboo that you see veteran head-fi members prefacing a review to justify/apologize for their use of any equalizer. But most music in the industry are in fact mixed in the studio for optimal reproduction with particular categories of sound systems. This optimization ensures that the music would sound its best to as many of the target audience as possible. Since most domestic sound systems notoriously roll off the bass at around 80Hz, mastering studios bump the bass a bit higher than that, which really sucks if your preferred bass kick peak frequency is a deep 25Hz.
There are so many who genuinely love bass, and would love to have more of it, but are limited by a purist, antiEQ mentality. To get their desired bass level, they try to raise the volume enough until the more sensitive frequencies--the 2KHz-5KHz range for example breaches your perceived loudness threshold earlier than the other ranges--get too loud. They are stuck with this method, so when testing the same volume with the SZ2000, they find the bass could not deliver compared to other headphones with stock FRs that have the bass boosted higher than the SZ's.
With this reasoning, they solve their problem by relying on stock FRs. Hey, many, if not the majority, are averse to audio processing of any kind and rely on stock FR. I respect that. That is fine. And costly.
You need EQ. Because it's cheap. And it can bend your perceived FR however you want it (within an inaudible margin of error). For example, consider having headphones that are harsh on the 2KHz-5KHz range and lacking in the sub 100Hz range. You can use EQ to attenuate the 2KHz-5Khz range and emphasize the sub 100Hz range to make the FR flat. In fact, Golden Ears has launched an EQ product that simply uses an inverse linear filter customized to match the stock FR of your headphones, basically cancelling out the dips and spikes for a resulting flat FR.
However magical a tool it is, EQ cannot correct room reverberations. Excessive room reverb is when sound gets echoed off walls (such as the headphone cups and earpads) and create standing waves that interfere with the direct sound (sound coming directly from the speaker). This is why it is ideal to have an anechoic chamber (basically a chamber with walls that absorb sound to mimic an infinitely open space, and at the same time prevent external noise), or a simulation of it when it comes to headphones. To minimize room reverb in headphones, most of the big name manufacturers use an open design, which you will notice is the most implemented design among Summit-Fi cans. For closed headphones, some line the interior with sound absorbent material. Room reverb is the bane of acoustic engineers. For bassheads, however, room reverb is often beneficial especially for the lower end, as it adds room gain to the lower frequencies and sometimes lowers the peak frequency, making the bass deeper. Basshead cans such as the old XB series have these huge, roomy leather pillows for earpads that enhance the bass experience. The SZ2000 earpads are covered in a kind of absorbent leather, and the isolation isn't strong enough that it leaks sound--this design possibly was an attempt at minimizing room reverb, but to most of the owners that was a mistake.
Can room reverb be corrected with audio signal processing? Perhaps; I've seen a few solutions already, but they aren't widely used. There IS a method called convolution, which takes the impulse response of a room (the room reverb of an actual room or inside a headphone can), and "convolves" any audio signal using that impulse response to simulate how that signal sounds like in that room. In Foobar2000, you can get the Convolver plug-in and use the impulse response of a basshead headphone to enhance your bass experience. EQ can't do this. Now, if you want the other way around, such as recording the impulse response of a room, and creating an inverse filter function to counter its effects, it is called deconvolution and you can find some solutions in Google, but I personally haven't tried them.
One irony with people averse to audio processing of any kind is that the volume control itself is a form of audio processing. Increasing the gain of all the signals equally, with little distortion, either requires amplification of that signal to add gain, or an attenuator that decreases resistance if you want to increase volume. My simple point is: Don't be afraid to use tools.
months after trying to simulate the built-in Windows bass boost enhancement, specifically at 50Hz +24dB, I measured its frequency response again, zoomed in, and saw that it was peaked at 41Hz lol. But then I saw that my 25Hz peak was shifted to 35Hz this time. Then I saw my soundcard was using a 192KHz sample rate and was causing audible intermodular distortion. Anyway ...
@milk asked for advice on EQ for Mac OS X, so I'll just share what I said in here as @hummel is looking for better Mac EQs as well. I've never used OS X my entire life--UNIX yes--and my last attempt at a Hackintosh required buying a new wireless card, so I'm not aware of important Mac apps. What you want is a Parametric EQ (PEQ). For starters, try searching for "Parametric EQ" for OS X. I'm very, very sure PEQs are available for iOS as well as OS X since they're basic stuff. Here is a head-fi google search so you can look for similar threads before starting your own--I found this thread to be the most helpful so far. I was able to simulate the Windows bass boost using only a parametric EQ (Equalizer APO), so my settings should be usable in other parametric EQs:
Filter 1 : ON PEQ Fc 50 Hz Gain 24.00 dB Q 0.329
Filter 2 : ON PEQ Fc 25 Hz Gain 21.00 dB Q 1.000
Filter 3 : ON PEQ Fc 3000 Hz Gain 21.00 dB Q 0.266
- Filter 1 is for the hardest bass impact, which turns this kick drum into this (indistinguishable to the effect of WBB).
- Filter 2 is for the deepest bass (lower peak frequency) AND rumble, and works together with Filter 1, as the two filters overlap, add up, and form a steeper curve peaked at 25Hz. Why the overlap instead of just one filter? It's impossible to get the same shape/effect, as you either get a significant attenuation of -5 to -7dB (7 compared to WBB) at 50Hz, or the low midrange floor gets raised thus reducing the perceived impact.
- Filter 3 is for restoring "clarity" by raising the Singer's Formant frequency while preserving the dynamic range, so the bass is still loud. Compare this less clear track to this clearer track, and notice that the bass is not significantly diminished.
Click here for the Windows tutorial.
RMAA FR measurements:
Notice how the green (APO w/ WBB) and white (APO only) curves are very close by -2dB from the 25Hz peak up, and only different in the less audible lower frequencies. Also note Cowon iAudio's BBE effect visible as an increasing slope starting from 600Hz towards infinity, which purports to "refine sound quality into clearer and more vivid sound," and see how Filter 3 is similarly curved up to the magic 3KHz peak frequency. You can definitely simulate other JetEffect technologies using only EQ (except reverb and stereo effects such as crossover and soundstage), even the $250 digiZoid ZO3 (which maxes at +14dB at 25Hz PF measured from the midrange floor; the ZO2.3 reaches +15dB at 25Hz PF):
Now, these settings should be a bit too extreme even for the extreme basshead, but tweaking them is as easy as lowering the gain of each filter by the same amount. Hawaii shared some EQ tricks in here. As a general solution, it works for most of the tracks that I listen to, and I even find it enjoyable with delicate piano tracks!
These settings, simple as they are, doesn't work properly with RockBox as RockBox clips the signal even with a -50dB precut. Let us know if you know any app with parametric EQ for Mac and Android!
Why do this?
Straightforward answer: Listen for yourself:
Wiz Khalifa - On My Level
Before: (Listen to normal volume.)
After: (This track is preamped to -24dB, so raise normal volume +24dB. This is an older version of my EQ settings with a much weaker Filter 3, so it does not sound as clear. This track with the new EQ sounds better and much clearer without losing bass impact.)
Judas Priest - Painkiller
Before: (Listen to normal volume.)
After: (This track is preamped to -24dB, so raise normal volume +24dB. This is an older version of my EQ settings with a much weaker Filter 3, so it does not sound as clear.)
Click here for more song comparisons.
Detailed answer:
Most people are slaves to what they unknowingly internalize as "original," "authoritative," "canonical." They grow a system of beliefs based on unchecked foundations, and within the hi-fi community, one such foundation is that there IS an "original" recording of music that needs to be reproduced accurately. Philosophers like de Man and Baudrillard have long freed us from this tethered thinking since last century, with catchphrases such as "They kill the original, by discovering that the original was already dead" and "The author is dead" graffitied in the blood of tyrants. Poststructuralism has been powerful in destroying absolutist frameworks that determine systems of oppression; however, some of its theories have also been used to justify nihilism and to discredit history. In the context of signal reproduction, true, there is such a thing as accuracy, where signal "A" must cross a journey to arrive at its destination in "C_T" as "A" and not "U" or else the resulting message becomes "CUT" and not "CAT." However, the issue is not accuracy, but whether YOU intend to send the signals "U" and "E" to "C_T" to make it "CUTE." It is not about killing the author as it is about bringing YOU, the audience, to life. You have agency. You are free. Most importantly, you are ALIVE.
Here at Head-Fi, digital/analog filters as simple as an equalizer are generally considered taboo that you see veteran head-fi members prefacing a review to justify/apologize for their use of any equalizer. But most music in the industry are in fact mixed in the studio for optimal reproduction with particular categories of sound systems. This optimization ensures that the music would sound its best to as many of the target audience as possible. Since most domestic sound systems notoriously roll off the bass at around 80Hz, mastering studios bump the bass a bit higher than that, which really sucks if your preferred bass kick peak frequency is a deep 25Hz.
There are so many who genuinely love bass, and would love to have more of it, but are limited by a purist, antiEQ mentality. To get their desired bass level, they try to raise the volume enough until the more sensitive frequencies--the 2KHz-5KHz range for example breaches your perceived loudness threshold earlier than the other ranges--get too loud. They are stuck with this method, so when testing the same volume with the SZ2000, they find the bass could not deliver compared to other headphones with stock FRs that have the bass boosted higher than the SZ's.
With this reasoning, they solve their problem by relying on stock FRs. Hey, many, if not the majority, are averse to audio processing of any kind and rely on stock FR. I respect that. That is fine. And costly.
You need EQ. Because it's cheap. And it can bend your perceived FR however you want it (within an inaudible margin of error). For example, consider having headphones that are harsh on the 2KHz-5KHz range and lacking in the sub 100Hz range. You can use EQ to attenuate the 2KHz-5Khz range and emphasize the sub 100Hz range to make the FR flat. In fact, Golden Ears has launched an EQ product that simply uses an inverse linear filter customized to match the stock FR of your headphones, basically cancelling out the dips and spikes for a resulting flat FR.
However magical a tool it is, EQ cannot correct room reverberations. Excessive room reverb is when sound gets echoed off walls (such as the headphone cups and earpads) and create standing waves that interfere with the direct sound (sound coming directly from the speaker). This is why it is ideal to have an anechoic chamber (basically a chamber with walls that absorb sound to mimic an infinitely open space, and at the same time prevent external noise), or a simulation of it when it comes to headphones. To minimize room reverb in headphones, most of the big name manufacturers use an open design, which you will notice is the most implemented design among Summit-Fi cans. For closed headphones, some line the interior with sound absorbent material. Room reverb is the bane of acoustic engineers. For bassheads, however, room reverb is often beneficial especially for the lower end, as it adds room gain to the lower frequencies and sometimes lowers the peak frequency, making the bass deeper. Basshead cans such as the old XB series have these huge, roomy leather pillows for earpads that enhance the bass experience. The SZ2000 earpads are covered in a kind of absorbent leather, and the isolation isn't strong enough that it leaks sound--this design possibly was an attempt at minimizing room reverb, but to most of the owners that was a mistake.
Can room reverb be corrected with audio signal processing? Perhaps; I've seen a few solutions already, but they aren't widely used. There IS a method called convolution, which takes the impulse response of a room (the room reverb of an actual room or inside a headphone can), and "convolves" any audio signal using that impulse response to simulate how that signal sounds like in that room. In Foobar2000, you can get the Convolver plug-in and use the impulse response of a basshead headphone to enhance your bass experience. EQ can't do this. Now, if you want the other way around, such as recording the impulse response of a room, and creating an inverse filter function to counter its effects, it is called deconvolution and you can find some solutions in Google, but I personally haven't tried them.
One irony with people averse to audio processing of any kind is that the volume control itself is a form of audio processing. Increasing the gain of all the signals equally, with little distortion, either requires amplification of that signal to add gain, or an attenuator that decreases resistance if you want to increase volume. My simple point is: Don't be afraid to use tools.