Harmonic Distortion is overrated.
Summary
TLDRこのビデオでは、音響における歪みについて解説しています。歪みはハーモニック歪み、インターモジュレーテッド歪み、マルチトーン歪みの3つのタイプがあります。ビデオではリアルタイムでの測定と音声を組み合わせて説明し、スピーカーXを使用した測定セットアップとリスニングセットアップを紹介しています。また、測定マイクとボイスマイクを用いてテスト音を録音し、測定プロセスとその結果をスクリーンショットと共に分析しています。異なる音量レベルでテストトーンを再生し、その違いをヘッドホンで聴くことをお勧めします。歪みテストにはハーモニック歪み、インターモジュレーテッド歪み、マルチトーン歪みの3つのテストを行い、それぞれのテストで異なるトーンを再生します。歪みテストの結果をグラフィックに示し、スピーカーの性能を理解する上で重要な情報を提供します。
Takeaways
- 📊 この動画では、ディストーションの3つの主要なタイプ:ハーモニックディストーション、インターモジュレーションディストーション、マルチトーンディストーションについて説明しています。
- 🎵 動画では、リアルタイムでの測定とオーディオの同時再生を通じて、ディストーションの種類とそのサウンドについて理解を深めます。
- 🔍 測定セットアップには、スピーカー、測定用マイク、ボイスマイクが使用されており、これらのデバイスはクリッパー近場スキャナーとディストーションアナライザに接続されています。
- 📈 ディストーションテストでは、70dBから96dBまでの4つの音量レベルで、100Hzから6kHzまでの周波数範囲でテストが行われます。
- 🎧 ヘッドホンを推奨し、テスト前に音量を下げることをお勧めします。これにより、テストトーンが非常に大きな音量で再生されることを避けることができます。
- 📉 ハーモニックディストーションテストでは、1つのトーンが再生され、そのハーモニックが測定されます。しかし、人間の耳ではハーモニックディストーションを特定するのが難しいとされています。
- 🔁 インターモジュレーションディストーション(IMD)テストでは、2つのトーンが同時に再生され、音量が上がるにつれて奇妙なワブリングサウンドが聞こえます。
- 🎶 マルチトーンディストーションテストは、音楽そのものと似ています。複数のトーンが同時に再生され、スピーカーの非線形性を引き出すことで、より現実的なディストーションの理解が得られます。
- 🔊 バンドリミットを使用することで、ディストーションと圧力を減らすことができます。これは、低音をサブウーファーに譲る場合に、スピーカーのディストーションが減少することを示しています。
- ⚙️ ディストーションテストは、スピーカーの性能を理解するための重要なツールであり、リスナーがスピーカーをどの程度まで使用できるかを判断するのに役立ちます。
- ❗ ディストーションテストの結果は、測定された距離に基づいています。距離と音量は常に一緒に考えなければならず、片方だけでは意味がありません。
Q & A
ディストーションとは何ですか?
-ディストーションとは、音の波形が元の形から変化することを指します。通常、音楽や音声が再生される際に、再生装置の性能不足や過剰な音量などによって生じます。
ディストーションの種類にはどのようなものがありますか?
-ディストーションにはホーミックディストーション、インターモジュレーションディストーション、マルチトーンディストーションの3つの主要なタイプがあります。
ホーミックディストーションとは何ですか?
-ホーミックディストーションは、基本周波数の倍数になる周波数の音が発生することを指します。例えば、100Hzの音が鳴らされた場合、200Hz、300Hz、400Hzなどのハーモニックが生成されます。
インターモジュレーションディストーションとは何ですか?
-インターモジュレーションディストーションは、2つの異なる周波数の音が同時に鳴らされたときに生じるディストーションです。基本周波数と呼ばれる低い周波数の音と、高い周波数の音との相互作用によって、新しい周波数の音が生成されます。
マルチトーンディストーションの重要性は何ですか?
-マルチトーンディストーションは、音楽そのものが多数の異なる周波数の音を一度に再生するため、実際の音楽再生におけるスピーカーのディストーションを再現するのに役立ちます。これにより、スピーカーがどのように歪むのか、そしてどの程度の音量まで安全に再生できるかを理解することができます。
ディストーションテストの際にリスナーにはどのようなアドバイスをするとのことですか?
-ディストーションテストの際には、ヘッドホンを装着し、テスト前に音量を下げること、そしてテスト中には音量を上げすぎないこと、また最初の音が非常に大きいように聞こえる場合は音量を下げることなどをアドバイスしています。
スピーカーのテストで使用された測定装置は何ですか?
-スピーカーのテストで使用された測定装置には、測定用マイク、ボイスマイク、クリップ_near field scanner、そしてクリップable distortion analyzerが接続されたコンピュータが使用されました。これらの装置は、クリップalソフトウェアと連携して動作し、テスト結果を報告します。
スピーカーのテストで使用された音楽の周波数帯は何ですか?
-テストで使用された音楽の周波数帯は、400Hzから6kHzの範囲です。この範囲は、人の声の周波数帯をカバーし、ディストーションを検出しやすいです。
テストで使用された音量レベルは何db単位で、どのように変化しましたか?
-テストで使用された音量レベルは70dbから開始し、4つの段階で96dbまで増加しました。具体的には、70db、78db、86db、96dbの4つのレベルが使用されました。
ディストーションテストの結果をどのように読むことができますか?
-ディストーションテストの結果は、総ホーミックディストーション、2番目のホーミックディストーション、3番目のホーミックディストーションなどのグラフで表示されます。また、マルチトーンディストーションテストでは、周波数ごとのディストーションレベルが示されます。これらのデータを元に、スピーカーの性能と制限を理解することができます。
スピーカーのテストで気をつけるべきポイントは何ですか?
-スピーカーのテストで気をつけるべきポイントは、テストの距離、使用された音量レベル、再生された周波数帯、そしてディストーションテストの結果の解釈です。また、テスト結果は、特定の条件下でのスピーカーの性能を反映しているため、一般化しないでください。
Outlines
🔊 オーディオの歪みについて
このビデオではオーディオの歪みについて解説します。特に、3つの主要な歪みタイプ:ハーモニック歪み、インターモジュレーション歪み、マルチトーン歪みに焦点を当てます。ビデオではリアルタイムでの測定とそれと同期した音声を提供する予定です。測定と聴き取りのセットアップが紹介されており、様々なボリュームレベルでのテストが行われます。また、聴くためにヘッドホンの使用が推奨されており、テスト前にボリュームを低く設定するよう注意喚起されています。
📊 ハーモニック歪みの測定
ハーモニック歪みは、単一の音をテストし、その音によって生成されるハーモニクスを測定します。ビデオでは400Hzから6kHzまでの20ステップの単純な音を再生し、それぞれのステップでボリュームを70dBから96dBまで4つのレベルまで増やしていきます。歪み成分が記録されると、画面に表示されます。総ハーモニック歪みは約1%で、400Hzから1kHzまでの高アウトプットボリュームでは2番目のハーモニック歪みが目立つことがわかります。
🔁 インターモジュレーション歪み(IMD)のテスト
インターモジュレーション歪みは、2つの音を同時に再生し、それらが相互作用して新たな歪みを生成することをテストします。ここでは50Hzの基本トーンと400Hzから6kHzまでのVOICEトーンのスweepを行います。ボリュームが上がるにつれて、音の変化が聞こえ始め、特に2kHz周辺で顕著な歪みが発生します。歪みは振幅変調とドップラー歪みに分けられており、ここでは振幅変調が主な要因とされています。
🎵 マルチトーン歪みの重要性
マルチトーン歪みは、音楽のような複数の音を同時に再生し、スピーカーシステムの非線形性をテストします。これはハーモニック歪みやインターモジュレーション歪みとは異なり、実際の音楽再生における歪みを再現するのに役立ちます。ビデオではマルチトーン歪みのテストを再開し、その重要性と実際の音楽再生における歪みの理解を深めることを目的としています。
📉 マルチトーン歪みテストの結果と分析
マルチトーン歪みテストでは、フルバンド幅、100Hzからカットオフ、200Hzからカットオフの3つの異なるテストを行います。それぞれのテストでは、スピーカーが異なる周波数帯でどの程度歪みが発生するかを比較します。結果は歪みと圧力損失を示しており、バンドリミットをかけることで歪みと圧力損失の両方を減らすことができることがわかります。
👂 マルチトーン歪みと圧力損失の比較
フルバンドからのテスト結果とバンドリミット後のテスト結果を比較することで、スピーカーの歪みと圧力損失を理解することができます。フルバンド再生では歪みが高く、周波数帯を狭めることで歪みと圧力損失の両方が減少することが示されています。これはスピーカーを極端なレベルまで再生しないようにするべきという教訓も提供しています。
🔊 音響テストの意義とリスニングテスト
ビデオの最終部では、スピーカーの歪みとリスニング体験について話します。スピーカーを近距離で聴くことの重要性と、距離と音量の関係について説明します。また、スピーカーの評価指標として音量と距離を組み合わせる必要性を強調しています。さらに、リスニングテストを行い、スピーカーの歪みを実感することを促します。
📚 オーディオテストの複雑性と学びのコミュニティ
ビデオの締めくくりでは、オーディオテストの複雑性と学びの重要性について触れています。スピーカーの歪みを理解することの難しさと、コミュニティを通じて学びを深めることを提案しています。また、今後のライブストリームやゲストを招待する可能性についても言及し、オーディオ専門家の知識を共有し、より深い理解を目指す意向を示しています。
🔧 スピーカーのテストと評価
ビデオでは、スピーカーのテストと評価に関する洞察を提供しています。スピーカーを極端なレベルまで再生することの意義と、リスニング距離が音の音量と質に与える影響について解説しています。また、スピーカーの性能を正確に測るためには、リスニング距離と音量の両方を考慮する必要性を強調しています。
Mindmap
Keywords
💡ディストーション(Distortion)
💡ハーモニックディストーション(Harmonic Distortion)
💡インターモジュレーションディストーション(Intermodulated Distortion)
💡マルチトーンディストーション(Multi-tone Distortion)
💡スピーカー(Speaker)
💡周波数(Frequency)
💡音量(Volume)
💡圧縮(Compression)
💡測定マイク(Measurement Microphone)
💡マルチバンド制限(Band Limiting)
Highlights
介绍了三种主要的失真类型:谐波失真、互调失真和多音调失真。
视频将通过实时测量和音频展示来解释失真类型。
展示了整个测量和听音设置,包括扬声器、测量麦克风和声音麦克风。
解释了测试麦克风通过Clipper近场扫描仪连接到Clippable失真分析仪的过程。
将测试音频和测量过程的截图同步,以展示整个测试过程。
测试将从70分贝逐步增加到96分贝,以展示不同音量下的失真情况。
建议使用耳机进行测试,以便更容易区分不同的失真。
在播放测试音之前,提醒观众降低音量以保护听力。
谐波失真测试是通过播放单一音调并捕捉其产生的谐波来进行的。
在400赫兹到6千赫兹的范围内进行谐波失真测试,重点关注声音区域。
总谐波失真在70分贝到87或88分贝之间大约为1%。
第二谐波失真在高输出音量下显著增加,可能导致听到失真。
互调失真(IMD)测试通过同时播放两个音调来揭示非线性失真。
多音调失真测试通过播放多个音调来模拟音乐中的复杂声波。
多音调失真测试结果表明,限制频率范围可以减少失真和压缩。
视频最后强调了在测量中考虑音量和距离的重要性。
提出了未来可能进行直播讨论失真主题的想法。
Transcripts
distortion i know you've heard that term
before but what does it mean what does
it sound like that's what we're going to
talk about in today's video
[Music]
in order to keep this video short i'm
going to try to keep things very concise
and to the point there are three main
types of distortion harmonic distortion
intermodulated distortion multi-tone
distortion and in this video what i'm
going to try to do my best at is to
provide you with
real time measurements
synced up with
audio of those measurements so what
we've got here is the entire measurement
setup and listening setup
we've got the speaker which is speaker x
now i may block this out because i don't
want people focusing on the speaker i
want them focusing on the content of
this video
we've got my measurement microphone
and then we've got my voice microphone
now normally the voice microphone would
not be out here it's only out here for
my testing
we normally have the test microphone
only which runs through the clipper near
field scanner out the bottom
and then runs into the clippable
distortion analyzer which is tied into
my computer which is tied into the
clipal software and depending on what
module or what test i'm trying to run at
the time the software knows what to do
with the stimulus collected from the
microphone and then it will report that
to me on the screen
what i'm going to do in this video is
i'm going to actually be recording the
stimulus sent out from the speaker
captured by the measurement microphone
as well as captured by the voice
microphone now this voice microphone is
connected to my computer that's inside
so what i'm going to be doing is
collecting audio from the test while i'm
also grabbing a screenshot of the
measurement process and for each one i'm
going to talk to you about what the
measurements are showing you as it's
being conducted or maybe after depending
on how hard this whole video is to put
together
as part of all of this testing
you're going to be listening to a tone
or a series of tones depending on what
the test is which i'm going to cover in
a second now that tone or series of
tones is going to be
swept at different volumes
first volume is going to be 70 db at 1
meter and then the last volume is going
to be 96 db at 1 meter and we're going
to step in four increments so we're
going to go
roughly 70 to 78 to 86 to 96 and yes i
know 86 to 96 isn't the same step but
it's like eight and a half db in between
i'm not going to do that math in my head
and look like a fool in real time
because i can't math
but
regardless of that that's what the video
is going to be showing you and that's
what you're going to hear now for each
test so the harmonic distortion imd
multi-tone test
i'm going to play different tones right
and i'll explain to you what those tones
are going to be before i start each test
but just understand up front that you're
going to hear four different volume
levels for each test that i'm doing
i also suggest that you wear headphones
for this test because while you can hear
the differences to your phone through
your pc speakers through your
loudspeakers in your listening room tv
speakers whatever you can hear the
difference hearing the difference on
headphones is much much easier and also
with that said
since i'm going to be playing test tones
what i'd suggest you do is before i play
each test before i start each test
turn the volume down so start at a low
volume now
the flip side of that is when the first
tone plays it's going to be at that 7 db
volume so it's not going to sound as
loud but you can play around with the
volume you can restart the test when you
figure out what's too loud what i don't
want you to do is to have headphones on
and have your volume cranked up and then
that first tone comes in or that last
tone comes in and it just blows your
ears off so i warn you again
start with the volume low before each
test i'll throw up a graphic to warn you
i'll give you five seconds to turn the
volume down make sure you do that and
then you can go back and listen to
different volumes as you need to now for
what it's worth
i did all this testing one round earlier
to make sure that i'm not going to run
into distortion for the voice microphone
where it is and on my motu m2
the volume bars that give me an idea
where the signal level is it's always
green so it never even hits yellow so no
issues with the voice microphone at all
and then this measurement microphone can
take a lot of spl before it runs into
issues but at the same time i don't want
it too far away because if i have it
much further away then i'm going to
capture a lot of the room environment
and you're going to hear some of the
room from the voice microphone but just
understand that normally this isn't how
i conduct my tests this is just set up
for this video i want to state that
right up front
the first test is going to be the
harmonic distortion test the harmonic
distortion test is just a test of one
tone at a time and then it picks up the
harmonics created by that tone
so you can have a tone of 100 hertz and
then the next harmonic of that is 200
hertz and then 300 hertz 400 hertz 500
hertz so on and so forth meaning that
the fundamental tone 100 hertz has a
second order harmonic two times 100 200
hertz has a third order harmonic three
times 100 the fundamental 300 hertz okay
i don't know why i just did that either
that was really weird um the fourth
order harmonic is going to be four times
the fundamental which is 100 so that's
going to be 400 hertz then 500 600 hertz
so on and so forth that's how you get
the harmonics of the fundamental test
tone now in my opinion hearing harmonic
distortion
i just don't think that we have the ear
for that now i know some people say that
they can hear harmonic distortion and i
think to a degree they can hear
distortion or they can i should say i
think they can hear distortion but i
don't think that they can singularly say
it's harmonic distortion because when
you listen to music you're listening to
a lot of different tones at one time
you're not listening to one tone which
again remember harmonic distortion and
we're going to touch on that in a
subsequent set of
sweeps and you'll see what i mean when
we get to that point but for right now
let's focus on the harmonic distortion
so what i'm going to do is i'm going to
play one tone at a time from 400 hertz
to 6 kilohertz and the reason i chose
that band is because i wanted to focus
on the voice region because it's easier
to pick up
distortions through that region and plus
it'll save us some time
normally we also will typically run the
harmonic distortion as a sweep so it'll
sound like
and you probably heard that before but
for the purpose of this video it's going
to be harder to
pick out the individual tones so that's
why i'm doing individual tones there's
going to be 20 steps from 400 hertz to 6
kilohertz it's going to take a little
bit of time give it time
remember wear your headphones but also
remember go ahead and turn your volume
down you can always turn it back up and
start this section over again if you
want to
play it safe turn the volume down don't
blast your ears off and here we go
first of all i want to state that up
here are the different voltage levels
that represent the output volume
1.13 volts is fed to the speaker that
results in about 70 db at one meter
3.06 volts is about 78 db
this is about 86 87 db and then this is
about 96 db all again at one meter when
the tones play you're going to hear
this voltage this voltage this voltage
then this voltage in that order so it's
going to sound like
it's just going to keep ramping up in
volume when those distortion components
are recorded they'll be displayed on the
screen so it's going to take a little
bit of a second for them to display
there's a little bit of a lag up here in
the top left is the total harmonic
distortion
y-axis is zero to ten percent this is in
percent
second harmonic distortion is over here
third harmonic distortion is down here
and then this is going to show you in
real time the tone that's playing
and the voltage that is playing
and then once i get through playing all
these
we'll come back we'll discuss this a
little bit
[Music]
[Music]
let's talk about the results what we can
see is the total harmonic distortion
is about one percent for the most part
from 70 db to about 87 or 88 db
except for
400 hertz to 1 kilohertz at the high
output volume 96 db
relatively high i mean it's above 3
so that's interesting we can see that
something certainly breaks down between
about 88 to 96 db
what makes that distortion component so
high well we can see if we look over
here at the second order harmonic
distortion it's right here so when 400
hertz is played it's going to sound more
like 800 hertz and when one kilohertz is
played it's going to sound more like 2
kilohertz
at this output level than it does at the
lower output levels
as compared to the third order harmonic
distortion where all this is below
roughly half a percent
so more than likely if you do hear
distortion artifacts it's going to be
born from the second order distortion
and given that the second order
distortion is so close to the
fundamental
it is likely to be masked meaning that
it's going to be covered up by the
fundamental tone because it's still
relatively low in level at 3 percent
it's about 30 db down now depending on
the muse that you're listening to and
depending on how good your ear is
you may or may not hear this distortion
and i'll be honest with you in this
particular test i didn't hear this
distortion if harmonic distortion is one
tone at a time
then intermodular distortion which is
what we're talking about right now
is two tones at a time
meaning that
normally what you do is you fix a base
tone and in this particular case i'm
going to be playing 50 hertz
what i'm also going to do is play
a sweep of voice tones those voice tones
are going to be 400 hertz to 6 kilohertz
and 20 different steps just the same as
it was for the harmonic distortion so
the difference here between imd and hd
testing is going to be now with imd
testing i'm adding that 50 hertz
fundamental tone now what you're going
to hear is that lower volumes it'll
probably sound okay you're going to hear
basically just the two tones playing at
one time
but as i increase the volume higher and
higher through those four different
steps
you're going to hear some weirdness
you're going to hear probably what
sounds like
maybe best described as a warble sound
it's going to sound weird that's
intermodulated distortion and that's
created by playing a fundamental tone at
a lower frequency and then higher
frequencies now you can also do this the
other way around
intermodular distortion can be any set
of two tones but that's what i'm playing
right now 50 hertz and then sweeping a
bunch of different tones through the
voice region
listen for how those tones
sound different as i increase the volume
and what you're going to be picking up
on most likely is going to be the
amplitude modulation there is doppler
distortion which is caused by phase or
timing differences
when you have a small driver and it's
moving a lot of excursion like one way
then that creates doppler distortion but
you're not going to hear doppler storage
and you're going to be hearing amplitude
modulation and for what it's worth i
made a post on my facebook group
overnight last night
and then lars riesbo from purify replied
back now my understanding of what i was
hearing i thought it was doppler
distortion this is just me being
completely candid because i don't know
everything i try to do the best i can
with trying to help people understand
and trying to teach i'm still learning
continually and i think this is a great
opportunity for us all to learn so
that's why i'm telling you this okay
so lars posted a link basically saying
hey aaron you think it's doppler but
it's really amplitude modulation and it
turns out that
you hear amplitude modulation klipple
has a white paper on this i'll try to
link it in the description below
so in the particular case of this
speaker yes indeed i'm hearing amplitude
modulation and in the data that we're
going to see in real time you're going
to see that it's dominated by amplitude
modulated distortion there's also a lot
of doppler distortion but you're not
going to be hearing that you're going to
be hearing the amplitude modulation
because it is much more dominant than
doppler distortion and i'll put the link
below because they have a really good
link on the purify website to show you
the difference of those two bottom cells
because there's no way for me to
separate those two different
contributions amplitude modulation and
frequency modulation aka doppler
distortion i can't separate that in this
particular test but they've done a
really good job of doing that on their
website i'll link that below you can
follow that so yes put your headphones
on turn your volume down let's kick off
the imd test
before i let this play all the way
through i want to point out a couple
things real fast the second order
distortion is up here and the third
order distortion is down here
but mainly note that i've had to change
the scale from zero percent to 100
percent
rather than zero to ten percent like you
saw with the harmonic distortion the
reason for that is you're about to see
some of these numbers jump drastically
especially when you get to around two
kilohertz
listen for how that third voltage and
that fourth voltage sound it'll almost
sound like
to me it reminds me of like a bird
whistling the pitch of the frequency
changes so much and remember what we're
doing in this test is we're playing
a 50 hertz tone down here but we're
recording
the voice tones up here so we're
basically just trying to see how much a
base tone affects the higher frequency
in the mid frequency tones and this test
clearly shows you
just how
close excursion is tied to amplitude and
doppler distortion in the mid to high
frequencies this graphic up here shows
the amplitude distortion in red and then
it's also going to show the doppler
distortion and that's covered by the
blue gray and black the black and the
gray are the components that make up the
doppler distortion now if you go back to
that screenshot that i provided briefly
a second ago or follow the link to
clipple's i said white paper it's it's
actually a presentation
if you look at this screenshot from
clipple's presentation i'm basically
paraphrasing but essentially what it
says to me
is that amplitude modulation is
detectable at about three percent three
percent would be negative 30 db down in
amplitude they state that doppler
distortion is at about 20 db higher than
that so about negative 10 db down
and that's where those thresholds of
distortion are going to come into play
now at this particular voltage because
i'm only showing one voltage in this
actual pane those all run together so
it's really hard to tell what's
contributing to what other than the fact
that we know based on what i said
earlier that amplitude modulation is
easier to detect so more than likely
what you're hearing
is from amplitude modulation and then
some of course influence from the
doppler distortion
[Music]
foreign
[Music]
[Music]
um
okay now that imd testing is done let's
talk about multi-tone distortion
harmonic distortion was one tone
imd distortion that's redundant was two
tones multi-tone distortion is multiple
tones multiple times multiple turns like
see i just did those are three different
tones that i just did
when you played multiple tones you enact
a lot of non-linearities in the speaker
system or the drivers themselves
and in doing so you're able to highlight
distortions that you wouldn't otherwise
hear with just a harmonic distortion
sweep or just intermodulated distortion
and in my opinion if you're going to
present one bit of evidence for
distortion data that actually makes
sense
then it's multi-tone distortion and
here i am being a hypocrite
i used to provide that up until about a
year and a half ago
and i kind of just
nobody was looking at it nobody was ever
commenting on it i don't think people
understood how to read it and i didn't
really do a good enough job of
explaining it i guess so some of that's
on me and i just quit providing it but
with this most recent speaker test when
i hooked this thing up and was listening
even in the near field
not very high volume you can hear all
sorts of graininess
now that graininess isn't from harmonic
distortion and it's not from imd at
least not singularly
it's from multiple tones being played at
the same time multiple tones
is what music is made up of music is a
whole bunch of sound waves a whole bunch
of tones played at one time and i guess
you could play square waves if you
wanted to synth i love scent music but
understand what i'm saying when i say
that music is just tones it's all it's
what it is it's just made up of tones
different frequencies different
harmonics an instrument has a
fundamental frequency and it has its own
harmonic profile that's why
that's why a bassoon sounds different
than a bass that's why a flute sounds
different than a clarinet you could play
the same fundamental tone
but the harmonics are different
and when you play back music you're
playing vocals you're playing
instruments you're playing synthesized
stuff you're playing a lot of different
things at one time and that's what makes
things get more distorted from a real
speaker and so because of this listening
test that i did last night i decided you
know what ma'am
i need to start doing that again and i
thought if i'm going to do that again i
need to make a video explaining why i'm
doing that and that's this video
basically this entire video is
predicated on me providing multi-tone
distortion testing again because in my
opinion that is the quintessential
testing that needs to be provided for a
loudspeaker harmonic distortion is good
for some reasons intermod distortion is
good for some reasons
but multi-tone distortion simulates
music it provides you with complex
series of tones
and it gives you a more real-world
understanding of
how the speaker is going to react in
terms of distortion and in terms of
compression
now keep in mind the compression that
you're going to see in this data test is
going to be different than the
compression that i already provided and
if i need to i'll make a separate video
for that but hopefully for right now
that's good enough
so what i'm going to do is i'm actually
going to run this multi-tone distortion
testing three different times the first
time is going to be full bandwidth
20 hertz to 20 kilohertz the second time
is going to be 100 hertz to 20 kilohertz
and then the third time is going to be
200 hertz to 20 kilohertz now you may be
thinking aaron why are you doing this
three different times think about it
like this
you know that when you high pass a
bookshelf speaker to a subwoofer you
release the stream from that bookshelf
speaker it doesn't have to play low
frequencies anymore and you probably
know from experience that when you do
that you can increase the volume and
have
more
output from that speaker with less
distortion that's basically what i'm
trying to emulate in this test is a high
pass filter
depending on what i'm testing i'll
probably change the filter i may stick
it 80 hertz i don't know but for this
particular speaker being a full range
speaker three inch
surface area i'll let you try to figure
out what it is if i'm not here showing
it here
i wanted to see what happens if i let it
play full band cut it off at 100 hertz
and then do another test cutting it off
at 200 hertz so what i want you to do
and what i expect you're going to need
to do
is listen to each of these three
different tests but then go back and
forth between them try to compare the
highest output which is going to be
around 22 volts
between the three different tests and
i'll screenshot at the end
the difference in all of those so i'll
show you what the distortion looks like
and what the compression looks like
compared between the three different
sets at the highest output volume and
you'll see that it is different and
you'll think to yourself why is it
different right because
if you're only concerning yourself with
the distortion in the vocal region
what you would expect then is that
that's not going to change
regardless of whether or not i play low
frequencies or not but remember this
isn't one tone distortion testing this
is multi-tone so if i'm not asking the
speaker to play down to 20 hertz or 50
hertz
then there's going to be less distortion
in the mid-range and the high frequency
and i know that may sound weird but
you'll see the data and you'll
understand what i mean and then you'll
understand why multi-tone distortion
testing is like in my opinion it's light
years beyond harmonic distortion and
that's why i'm going to start doing that
testing again because it's important to
do
and
anyway that's kind of enough rambling
for me let me run those three tests and
then we'll talk about the results and
then we'll go on
i've already skipped to the end of this
first test result because i want to show
you what you're looking at here
on the bottom is the distortion
okay and this is relative to the
fundamental when something says relative
to the fundamental that's why you have
negative values rather than percent
i like using
relative to the fundamental because
percent sometimes can jump off the scale
pretty quickly and it's hard to tell the
difference between one percent fifty
percent so when you use relative to the
fundamental it's easier to tell these
differences and this is a good case for
where relative
is easier to show than percentage scale
okay
the black line represents the noise
floor in the room and this is important
because as you get lower in frequency
the output of this speaker is low its
roll-off is is around 150 hertz or so
and so that means that the output of the
speaker is
not easily able to overcome the just
regular old ambient noise of a static
room even if there's nothing going on in
the room itself
the room has a noise floor and the lower
frequency you go
the higher that noise floor is
so at some point on the lower end the
results are kind of
i would say they're basically inaccurate
if i'm being honest
but as long as you understand that then
you can still know how to use the data
and i would say generally speaking above
about 50 hertz you're okay to use that
data
okay
so then the gray lines represent each
voltage that has been tested
and then the green line represents the
last voltage that's been tested so when
it gets to this final 22.47 right here
that's this right there and this is
showing us that the last voltage tested
22.47 so that's 96 db out output i
should say at 1 meter
is about negative 10 db down
in terms of distortion well that's
really really high you're definitely
going to hear that
this one is 8.32 volts
3.07 volts 1.13 volts so 70 db
78 db 86 87 something like that db and
then 96 db
now if we go up here this is compression
and this shows us
how much output we are losing as we
increase the volume and this is all
relative to the initial
output
voltage or output spl
so
notice you only have three lines here
right you have the green one and then
two grays why don't you have four grays
like you did down here because the first
gray is your baseline
that's basically this flat gray bar
right through here so the other voltages
are referenced to
this initial voltage right there
everything is referenced to 70 db
as you increase from 70 db
to
96 db well according to this data
you've
lost 3 db in compression i know that the
results right here say negative down
here but the way that this particular
data is represented a positive is a loss
in output so you have lost
about three actually to four db of
output as you've increased the volume so
you wanted to be at 96 db well it turns
out you were actually at around 93 to
92 db because you've lost
output due to some kind of compression
issue probably like heating of the voice
coil or something along those lines and
again i mentioned earlier that i think
when i hear the graininess of sound that
that's due to compression and i recall
reading that somewhere but if i'm being
honest i don't recall where so take that
maybe for now with a grain of salt but
keep this information in mind i'm about
to play back the stimulus starting from
the beginning i'm going to speed it up
so it doesn't quite take as long but
listen for how grainy the sound is as
you get to this 3 volts 8 volts 22 volts
i'm telling you you're going to be like
whoa and it's going to be way more
eye-opening than the harmonic distortion
and certainly even more eye-opening than
the intermodulated distortion so put
your headphones on turn the volume down
a little bit and let's go
[Music]
now let's compare the 96 db
output for each of the three multi-tone
distortion tests
again those three are going to be
full band 20 hertz to 20 kilohertz in
red
100 hertz to 20 kilohertz in blue
and 200 hertz to 20 kilohertz in black
and what you notice right off the top is
in red the full band has higher
distortion throughout and if you band
pass this speaker from 200 hertz in
black and above
then you have about
negative 10 db which is about 33
i believe
down to about negative 30 db and maybe
negative 25 db so i'm ballparking that
at around
three percent to about maybe five
percent thd
so you lose a ton
of distortion
going from
full band to band limited if you were to
put a subwoofer with this system then
you could see that you would save
yourself a lot of distortion but what
happens when you talk about compression
here we are
0db references the
lowest output anything above this is
output that is lost same thing as before
20 hertz 220 kilohertz is in red and
then the other colors represent the
other tests
in red you lose about four three to four
db depending on what frequency you're
talking about but if you band limit this
speaker you save some output so you
actually only lose about
2 db to 3 db and when i say only lose
well
that's not
a little bit but you do see that you
save some compression by band limiting
the speaker so you save compression you
save distortion by band limiting the
speaker but more importantly i think
this data tells us don't play the
speaker that loud and i think anybody
realistically already knows that now
just to recap i'm going to play
these three sound clips
back to back so you can hear what they
sound like all in a row and save you
some time from having to scroll all the
way to the back
[Music]
that's it for this video
what i hope is that you've learned
something from it now what i encourage
you to do is to ask any questions if you
have them i'm going to try to drop a
bunch of links
in the description below to various
sources from clipple the one i mentioned
from purify as well
and what i may do if i have enough
interest is i may just do a live stream
sometime soon and we can talk in real
time about this video because
i understand that it's complex and trust
me it gets a lot more complex because
the stuff that i've learned in the last
decade i don't even remember half of it
i'd be lucky if i remember like 1 20th
of it and a lot of that goes a long way
toward explaining
the different distortion elements the
non-linearity so
if the
suspension isn't uniform in versus out
if the motor force isn't uniform
inverses out
through excursion
that creeps up in different forms and at
different frequencies
depending on the speaker itself so a
large woofer may behave differently than
a small mid-range and all of this stuff
matters the devil really is in the
details but my hope is
between just a basic harmonic distortion
sweep to give you an idea of how low you
can take a driver
and
the multi-tone distortion which will
give you an idea of the contributing
factors toward maybe high distortion and
when you can expect the speaker to start
sounding really grainy and have an idea
of how loud you can turn it up
between all of that you'll be better
informed and you'll better understand
this topic itself rather than just
saying yeah it's high distortion like i
said if you have any questions please
ask in the comments below i apologize if
you feel like this video has taken too
long but if you want to learn that's
what i'm trying to do is i'm trying to
help you learn trying to help teach and
i'm trying to teach myself at the same
time and as a community we will all grow
in our learning and our understanding
and these won't be just numbers that
somebody put up on a website these will
actually make sense and mean something
to us and that's important because the
data is important
i don't do this stuff trivially and i
truly believe that having more
information is much better than just
having me say
it doesn't get loud
that's useless so with that all said yes
i appreciate you watching if you have
any questions let me know and if you're
interested in a live stream sometime let
me know try to do that sometime soon
talk about this stuff and maybe i can
try to have some guests on that are well
more versed in these different
distortion types and the reasons for
them at some point the future
but for now hopefully this is good
enough gives you a good idea of what the
distortion sound like and how to track
it in the measurements and i will talk
to you all later take care
peace
it's really hot in this garage i gotta
turn the ac on
later y'all
if you're thinking to yourself
you chose the worst possible speaker to
do this with why would you push
this little speaker to such high levels
well that's actually the point
that's exactly why
i'm using this speaker
when i listened to it within like a
meter so like three feet just
a regular pair of speakers
i started to hear
compression i started to hear that
grainy sound that i often tribute to
compression and that's what triggered me
into thinking i need to do some
additional testing of the speaker and
then i need to do a video to explain
what that data means and then i thought
man i need to start doing this again for
future speakers
so while i don't think that i would
necessarily push such a speaker
to the limits that maybe i'm going to
show you in this video i can say that to
some degree i actually did
and that really matters when you're
talking about listening at different
distances so here's the thing
when somebody says
a certain volume level so they say 80db
that number is meaningless unless they
tell you how far away they are from the
speaker
80 db
at
a few feet away is going to sound a lot
louder to you than it would be if you
are
10 feet away because the further you get
away from that speaker the less loud
it's going to seem you already know this
every doubling of distance
drops the amplitude by about 6 db
and depending on which source you read
and depending on what frequency range
you're talking about
research shows that our perception of
output
doubles in terms of it sounds twice as
loud
between 6 to 10 db of volume increase
so
if i take the volume from 80 to 90 db
then it sounds twice as loud conversely
if i take it from 80 to 70 db then it
sounds half as loud
now if i move from maybe a foot away to
three feet away or a little bit further
well that's going to drop the volume
level or the spl level a lot so it could
sound half as loud
that's what matters
if you're listening close you don't have
to turn the volume as loud as you do
when you're listening further away so
when some person says i'll listen to
80db and you're saying well i do too but
i'm having these issues well you may be
listening to different distances number
one that that could be a problem
so when i talk about running these tests
at x volume i'll always specify at one
meter or however far away because that's
important if you're at one meter away
and then you go two meters away so
roughly six feet
then you drop the volume by six db so
it's almost not quite as loud or not
twice as loud anymore just keep that in
mind i see this come up a lot and i i
think a lot of people don't quite
understand the significance of not just
volume but volume and distance those
should always always always
be paired together because
one number is useless without the other
5.0 / 5 (0 votes)