Genelec 8050B Speaker Review (and how to read speaker measurements)
Summary
TLDRビデオでは、オーディオサイエンスレビューのアミルが、プロフェッショナルモニターのジェネレック8050bをレビューしています。このスピーカーは、波導の統合、8インチのウーファー、クラスABアンプ、そして背面のポートなどの特徴があります。測定方法として、近接場スキャナーを使用し、スピーカー自体の動作を測定し、部屋の反射を排除しています。スピーカーの周波数応答、ディストーション、および水平・垂直方向性の評価を行い、その性能を詳細に解説しています。
Takeaways
- 📢 このビデオは、Genelec 8050Bというスピーカーのレビューであり、メソッドロジーと測定方法について説明しています。
- 🔍 Genelecはフィンランドの会社で、プロフェッショナルなモニターを長い間作っており、世界でトップ2、3位のブランドの一つです。
- 📏 8050Bは8インチのウーファーで重く、キャストアルミニウムで作られた非常に-solidなボックスです。
- 🌐 統合されたウェーブガイドは、ツイーターとウーファーのレスポンスをクロスオーバー領域でスムーズに合わせるためのものです。
- 🔊 このモデルはアクティブモニタで、ウーファーとツイーターそれぞれにアンプがありますが、従来のクラスABアンプを使用しています。
- 🔧 背面には大きなポートがあり、低周波応答を拡張する役割を果たしていますが、周波数が低いと急激に減衰するというトレードオフがあります。
- 🎛️ スピーカーには周波数応答を調整するための複数のコントロールがありますが、エレクトロニックイコライゼーションの方が好ましいと述べています。
- 📈 スピーカーの測定では、ルームリフレクションを排除するために近接場スキャナーを使用し、非常に正確な結果を得ることができます。
- 📊 スピーカーの測定結果は、SPINデータ(スピノラマ)として示され、これは標準化された測定方法です。
- 👂 聴覚テストでは、スピーカーのトーンの良さを判断する上で重要な要素をいくつか確認しますが、聴覚テストは測定値を補完するものであり、完全に置き換えるものではありません。
- 💡 このスピーカーは低頻帯の力を求める場合、サブウーファーが必要で、非常に大きな音量で再生する場合に限界があります。
Q & A
ゲネレック8050Bスピーカーの特徴は何ですか?
-ゲネレック8050Bはフィンランドの会社が製造するプロフェッショナルモニターで、8インチのウーファーを備えており、鋳造アルミニウムで作られた非常に-solidなボックスです。また、統合されたウェーブガイド技術を採用しており、ツイターとウーファーのレスポンスを滑らかに一致させるように設計されています。
ウェーブガイドとは何で、どのような役割を持っていますか?
-ウェーブガイドは、ツイターのレスポンスを狭める仕組みで、クロスオーバー領域でウーファーのレスポンスと一致するように設計されています。これにより、ツイターとウーファーの間で不連続さが生じるのを防ぎます。
このスピーカーはアクティブタイプですか、それともパッシブタイプですか?
-はい、これはアクティブタイプのモニターで、ウーファーとツイターそれぞれにアンプが搭載されています。ただし、これは古いデザインのため、最近のクラスDアンプではなく、従来のクラスA/Bアンプを使用しています。
スピーカーのバックサイドにはどのような特徴がありますか?
-バックサイドには大きなポートがあり、これは小型のエンクロージャーで低周波応答を拡張するために使用されています。また、周波数応答を調整するためのいくつかのコントロールがあります。
スピーカーの測定方法として特に重要なポイントは何ですか?
-スピーカーの測定では、部屋の反射を排除することが重要で、実際にスピーカー自体の動作を知るためにアンエコイックチャMBERを使用することが伝統的な方法です。しかし、低周波ではアンエコイックチャMBERでも正確ではありません。
クリッパブル近場スキャナーとは何ですか?
-クリッパブル近場スキャナーはロボットシステムで、スピーカーの近くで円筒測量を行い、反射を排除することでスピーカーからの直接音と反射音を区別することができます。
スピーカーのスピンデータとは何で、何が重要ですか?
-スピンデータとは、スピーカーの周波数応答を視覚化したグラフであり、標準化された測定方法です。重要なのは、オンアクス(直接音)がフラットであることと、反射音も同様にフラットであることです。
スピーカーのディストーション測定はどう行われますか?
-ディストーション測定では、86dB SPLと96dB SPLの2つのレベルで測定され、これはスピーカーが音量を上げた時にどの程度ディストーションが増えるかを示します。
水平方向性とは何で、スピーカーの性能にどのような影響を与えますか?
-水平方向性はスピーカーから音が水平方向にどの程度広がるかを示し、良い水平方向性は音のトノロジーが周囲で一貫性を保ちます。これは特にミキシングや録音の際に重要です。
ウォーターフォール分析とは何で、スピーカーの測定においてどの程度の役割を果たしますか?
-ウォーターフォール分析は、スピーカーのインパルス応答を可視化したもので、共振やディストーションを特定するのに役立ちます。しかし、解釈が難しいため、他の測定値と組み合わせて使用されます。
スピーカーのリスニングテストとはどのようなものですか?
-リスニングテストは、スピーカーの音質を主観的に評価する方法で、測定値から得られない情報を得るために行われます。リスニングテストでは、スピーカーの音の広がり、深さ、または特定の周波数帯での表現力を評価します。
スピーカーの価格帯はどの程度ですか?
-ゲネレック8050Bの価格は高めで、1台あたり1400ドル程度で、2台セットで3800ドルになります。これはヨーロッパ製の高品質な製品であるためです。
スピーカーのパフォーマンスを向上させるためのアドバイスはありますか?
-スピーカーのパフォーマンスを向上させるために、低周波範囲を補うためにはサブウーファーを組み合わせることが推奨されます。また、音量を極端に上げず、リスニングテストで感じた不満をエクイлайザーで補正することも有効です。
Outlines
🔊 オーディオレビューのスピーカー紹介
アミルがオーディオサイエンスレビューでスピーカーのレビューを行い、特にスピーカーに興味がない視聴者にもメトリクスとリスニングテストの方法論を学べる良い機会であることを紹介。レビュー対象はジェネLECの8050Bというスピーカーで、プロフェッショナルなモニタとして世界的に有名。このスピーカーは8インチのウーファーで重く、キャストアルミニウムで作られた-solid box。また、インテグレーテッドウェーブガイド技術を採用しており、ツイーターとウーファーのレスポンスを滑らかに合わせる。
🔧 スピーカーの測定方法と機器紹介
スピーカーの測定方法としてアネコイックチャMBERの使用が伝統的であるが、実際の測定では低音域では反射が生じる問題がある。そこで、CLIPPABLE NEAR FIELD SCANNERというロボットシステムを使用し、スピーカーから反射を排除した測定を行う。このシステムは非常に高価であるが、最も正確な測定結果を得られる。
📊 スピーカーの周波数特性とリスナーの好み
スピーカーの周波数特性は、リスナーの好みに大きく影響する。研究によると、ほとんどのリスナーがニュートラルな音の特性を好む。スピーカーの軸上の特性は平坦で、小さなノッチが存在してもほとんど聴衆には問題ない。しかし、スピーカーは音を全方位に放出するため、反射音も聴衆に届く。理想的なスピーカーは軸上と反射音の両方で同様の周波数特性を持つ。
📉 スピーカーの予測された室内応答
スピーカーの軸上音と早期反射音を組み合わせることで、一般的な室内での音の応答を予測することができる。プロモニタはフラットな応答を好むが、このスピーカーでは軸上音に欠陥があり、低音域での発振が限界があることが示唆される。
📝 スピーカーの個別ドライバー応答とポートの役割
スピーカーのツイーターとウーファーの個別ドライバー応答を測定し、それぞれが周波数帯でどのように動作するかを分析する。また、ポートの応答も測定され、低周波数での助けを提供する一方で、共振現象が発生する可能性があることも示されている。
📊 音響空間におけるスピーカーの特性
スピーカーの水平および垂直方向の指向性は、音響空間での音の広がり方を決定する。理想的なスピーカーは周波数によらない一定の音束を放出し、左右の音色差がないように設計されている。また、反射音の同調性も重要で、空間の広がりやリスナーの位置に応じた音の変化が少なく、音の質が一定であることが望ましい。
👂 オーディオサイエンスレビューの聴覚テスト
アミュラーはスピーカーの測定結果を聴覚テストと比較し、聴覚で確認できない微妙な違いや問題点を特定する。また、スピーカーの失真を測定し、聴覚で検出されるかどうかを判断する。聴覚テストは、測定値から予測される音質と実際の聴覚体験のギャップを埋めるために重要な役割を果たしている。
🔊 スピーカーの配置と設計の重要性
スピーカーの配置は音質に大きな影響を与える。特に、スピーカーを耳の高さに配置し、聴衆が少しでも動いた場合でも音色差がないようにすることが重要である。また、コーアクシャルドライバーやツイーターとウーファーの配置がスピーカーの指向性に影響を与えることを説明する。
💡 オーディオサイエンスレビューの総括
アミュラーはスピーカーのレビューを総括し、スピーカーの周波数特性、指向性、失真、および室内応答の測定方法を解説した。また、聴覚テストの重要性を強調し、スピーカーの配置や設計の選択肢についても触れる。最後に、スピーカーの好みや評価の主観性についても言及し、聴衆に対して理解を深めることを促す。
Mindmap
Keywords
💡ジェネLEC 8050B
💡ウェーブガイド
💡パワードモンイター
💡クロスオーバー
💡近場スキャナー
💡スピンデータ(SPINorama)
💡ディストーション
💡ディレクティブ
💡ルームフレンドリー
💡ウォーターフォール分析
Highlights
Georgie 8050B是由芬兰Genelec公司生产的专业监听音箱,该公司长久以来一直制造专业监听音箱。
Georgie 8050B音箱采用8英寸低音单元,箱体由铸铝制成,非常坚固。
集成波导技术用于缩小高音单元的响应范围,以匹配低音单元的响应,实现平滑的声音过渡。
该音箱为有源监听器,配备低音单元和高音单元各自的放大器,采用的是传统的AB类放大器而非现代的D类放大器。
音箱背面设有尺寸较大的端口,用于扩展低频响应。
音箱提供了多种控制选项,用于调整频率响应,但作者倾向于使用电子均衡器进行调整。
介绍了测量音箱的正确方法,包括在消声室中进行测量以避免房间反射的干扰。
Clippable近场扫描仪的使用,这是一种机器人系统,可以在近距离测量音箱并消除房间反射。
即使在消声室中,极低频率的测量也可能不准确,因为仍然存在反射。
介绍了如何通过测量来预测音箱在普通房间中的声音表现。
Georgie 8050B的轴向频率响应非常平坦,接近理想状态。
音箱的早期反射和声功率表现良好,有助于形成平滑的预测室内响应。
通过测量各个驱动单元的响应,可以了解音箱频率响应不完全平坦的原因。
音箱的端口共振得到了很好的控制,没有对频率响应产生负面影响。
讨论了失真测量的重要性以及如何通过测量来确定音箱的极限。
Georgie 8050B在高音量下表现出可闻的失真,尤其是在低频区域。
水平指向性表现良好,使得音箱在不同位置的听感保持一致,这对于近场监听非常重要。
垂直指向性分析表明,音箱在不同高度的听感也相对一致,但存在一些频率上的凹陷。
讨论了音箱的房间友好性,以及如何通过控制反射来优化听感。
水瀑布图(Waterfall display)虽然美观,但在非活跃音箱上可能会产生误导。
尽管测量数据非常重要,但听感测试同样关键,可以揭示测量无法展现的音质特点。
Georgie 8050B在听感测试中表现出色,但在极高音量下出现了可闻的失真。
尽管音箱价格不菲,但考虑到其性能和制造质量,它仍然具有很高的价值。
Georgie 8050B的偏好得分非常高,表明它是一款处于艺术状态的音箱。
Transcripts
hello everyone this is amir from audio
science review
uh today i have a speaker review for you
um even if you're not interested in
speaker i thought it would be a good
video for you all to watch because it
goes through the
methodology for measurements and and
listening tests
so today's uh speaker in question
actually let me go back full screen um
this is genelec
8050b um finnish company
been in business making professional
monitors for a long time
probably top two or three brand in the
world
this specific sample is quite large
eight inch woofer
it's quite heavy actually it may not
look it but it's made out of cast
aluminum and it's quite a solid box
uh if we go to the review you'll be able
to see a better picture of it
it's got this integrated waveguide
waveguide is a way to uh narrow the
response of a twitter
to get it to match the woofers response
as it the frequencies go up the woofer
starts uh
step back woofer starts with basically
omnidirectional sound
but as frequency go up it becomes
narrower narrower and narrower
um the inverse is true of the twitter if
you just cross over to a twitter
twitter starts off also quite wide and
you get a discontinuity there
so what a waveguide wave guide does
is that it narrows the response of a
twitter at crossover region
so that it smoothly matches uh that of
the woofer
uh if you have a mid-range that's
another way to mid-range drivers another
way to solve the problem but
if if you don't have one trying to get a
tiny little uh
um oops a tiny little uh
twitter to uh blend in properly with a
woofer is next impossible so it's nice
to see this in there
you know a number of companies have
these wave guys in there
this is a powered monitor has spy
amplified
so it has an amplifier for the woofer
and has an amplifier for the tweeter
it's an older design so unlike the new
speakers where they all use class d
amplifiers this uses traditional class a
b amps
and they tend to not be as powerful as
the class d's are because they take a
lot of space in
in with heatsinking what have you
uh on the back you have a pretty good
size uh port
the port is there to extend the low
frequency
response you know especially in a small
enclosure like this
but then they tend to drop off like a
rock uh
afterwards um so there you know it's
almost there's no
free lunch there if you will but it's
you know 95 of speakers now in this
bookshelf size will have that
as a number of controls uh for uh
adjusting the frequency response
i tend to not want to use any of those i
use electronic equalization
in in my software because these things
are very static
and i like to tune the sound fully but
uh if you need to
just plug it in and use it then it's
fine it's got input sensitivity
uh set it to the minimum and still was
pretty sensitive i had to keep the
volume pretty low
anyway that's the speaker um
we're going to get into the measurements
the measurements here
for speaker a proper way to measure a
speaker is to make sure that the room
reflections are not there
you don't want to see a frequency
response that is specific to my room you
want to know how the speaker itself
behaves and that traditionally has meant
that you take the speaker
and you go to an anechoic chamber that
gets rid of the reflections and you
measure there
but even anechoic chambers anything of a
reasonable size is still not anechoic
meaning that as frequencies get lower
and lower it actually has some
reflections that come back to the mic
measurement microphone and screw up the
measurements
there are calibrations for that
and they mostly work but just know that
in very low frequencies
even anechoic measurements are not very
accurate i use a system called
clippable near field scanner it's a
robotic system
let me open the link uh hopefully
there's a good picture of it there
and uh you can see in here
enlarge this you put the
speaker in the middle and uh
the system will then use this uh boom
and it will scan the speaker and
take a cylindrical measurements from top
to bottom
and then uh if you instructed to get rid
of the
reflections in a room it will do a
second pass
and using the those two measurements you
can determine if the sound is coming
from the
speaker or it's coming from an outside
the wall and come back in if it comes
back from another direction
it will take it out of the system
because of that it has two benefits one
is that it measures very close distance
to the speaker
and the benefit of that is that if
there's ambient noise it doesn't matter
that's one of the benefits of a
well-done anechoic chambers that's very
quiet
this system does not require a quiet
room because the microphone is very
close to the speaker
and as a result the noise is dwarfed and
doesn't show up in the measurement
there's a problem with doing that in
that if you measure very close to the
speaker you're measuring what is called
the near field response
not the far-field response for the sound
has gotten integrated and
but this system mathematically will
calculate
what that sound field is at any distance
so even though it measures at a short
distance
he can project the sound at 1 meter 10
meter 100 meters whatever you want
excuse me
um so putting it all together it gives
you a
basically a measurement that is uh
more accurate than just about any other
method that you could come up with you
could go spend a million dollars in
another koi chamber
and not get this kind of result
nothing's free in life and this one
literally is not free
this that system that i show you with
all the bells and whistles that i use to
make the proper measurements
cost about a hundred thousand dollars so
it's not for faint of heart but i you
know when i decided to get into speaking
measurements i just didn't know how else
to do it you have to do it this way
to get proper measurements there are diy
techniques that you can put things on
the ground and measure and combine
measurements and that's what
some of the other magazines and the
websites use it's very time consuming
and to do it right takes a lot of you
know meticulous attention
uh i do a speaker review every other day
every third day
and there's just no way i could
measure speaker like that you know
using the other people's methods and
you'll see why it's not just one
measurement
by default i have the clipable nfs
system measure 1000 points
so it is sampling the sound field a
thousand points around
it can go higher and can go lower lower
can lose
can cause accuracy errors higher takes a
lot longer to run but i have tested some
speakers
with two thousand points um two thousand
points takes about four and a half hours
for measurements
uh one thousand points half of that
about two and a half hours
and then another 15 minutes a number
crunching afterwards so
just the raw measurement takes me about
three three and a half hours
on the system once there we don't just
get a simple frequency response
measurement
we get this beautiful graph and this
graph and the measurements is in the
standard nc standard
and also called cea or cta 2034
but the common name for it we call the
spin data or spinorama
and i highly encourage you to use the
word spin data because this sounds
like you know more than you do so don't
spell it out when
you go to somebody try to sell a speaker
just say you have spin data
and the guy says what you want to walk
out because they don't
know how to measure this um anyway
dr tool when he was a national research
council in in canada that was set up to
try to promote an audio industry in
canada
set up a in 70s uh 1970s try to
figure out you know do peop is there
some common theme to what speakers
people like
or is it just a free-for-all that you
know there's no commonality you build
different frequency response different
tonality and x group of people like
it and the other people hate it and
surprisingly it turned out
to be that vast majority of people like
the same thing
and not only do they like the same thing
but they like a neutral
sound when it comes to frequency
response
so if you look at this response over
here there's this top line says on axis
which is the i don't know if you can see
this black line on top here
and the research says that we want this
thing to have to basically
be flat and have no variations in it
now this speaker is extremely good it
comes within 80 90
of of that target of uh looking flat
it's got little notches up and down
there are more perfect speakers
professional speakers
and they all use dsp to filter out these
notches in and out this speaker is older
it doesn't do that
newer genelecs do that and they have
flatter response
but from audibility point of view a
little bit higher a little bit lower
it's just not material because the music
is recorded with no standards
so who knows if the track you listen to
has a little dip over here versus this
little peak
so if you like you can try to filter
these things and correct them but
once you get to this level it's good
enough
in my book now that's just on axis
meaning the direct sound
coming at your ear so that's one sound
it's a very important sound is this sets
the stage and the tonality that your
brain
perceives of a speaker but the speaker
also sends sound to
in all directions and he hits the
ceiling hits the floor
hits the wall behind you then goes back
it's the front wall
and all these sounds also come to you at
different arrival times
and the brain has this challenge that
not only does it
hear this the redirect sound but it also
hears the reflected sound and
hears the reflected sound differently in
the right ear versus the left ear
so if there is reflection on the right
it will come to my
right ear faster and also stronger than
it comes to my left ear
and then my face also will cast the
acoustic shadow
where some of the high frequencies won't
go to this ear as well as they go in
this ear
now if the ear try to just present both
of those
images to your to you you go nuts right
it's like
you know frogs think frogs see you know
uh images that are you know ghostly
images with their eyes
that would be cooler than the sound
where your ear will be hearing
many many sounds simultaneously that
would make us nuts so
from a point of view of of how we've
evolved that species our brain had
decided to
um
filter out all those other secondary
sounds
now the best job that he can do is
if those secondary sounds have similar
tonalities to
what the direct sound is if that's the
case the brain says oh
these are just reflections of the same
thing
and therefore they're not
informationally as important
if you will and that i'm not going to
present them as secondary data to you
and
it is for lack of better word it's
happier
so an ideal speaker will have a direct
sound that comes to you with a certain
frequency response
but also it's indirect sound that hits
the walls and come to you
and through other reflections also have
similar frequency response now they
don't have to have
identical frequency response and indeed
it's usually the case where their
frequency response tilts down
because that twitter at high frequency
just like that woofer that explained it
also starts to narrow down
and when it narrows down there's less
energy hitting the walls
and coming at you so uh
you know some of some high frequency
droop is fine
now not all reflections are the same
some reflections are stronger than the
others first reflections are the
strongest so reflection of the side
while coming to my ear stronger than a
reflection hitting a back wall hitting
the sidewall then coming to me
there's a group of reflections that
let me jump ahead a little bit called
early reflections
and involve a research that was done at
harmon which was they sampled a few
different listening rooms
and determined which are the most the
strongest reflections that combine
and arrive at you they're dominant in
signature
and they're the floor obviously there's
a ceiling
front wall sidewalls rear wall and then
some of all of those together
and as you can see for this genelec if i
sum all those together i get a pretty
nice smooth
sloping down curve as i mentioned we
would however we also have a dip
here which is at the crossover region so
this speaker is not as perfect as it
could be
again newer genelecs will not have this
kind of dip in them
or you know most perfect speaker
wouldn't have any
the uh if we take the uh
uh step back we can take this
information that i've talked
so far that is captured in an anechoic
chamber
and actually predict what sound you
would get in a typical room
and that brings us down to this
estimated
or predicted in-room response
so how do you do that well it's a mix of
different things but
predominantly is the on axis sound which
i said is
extremely important this early window
which we know is tilting down so as a
result when we add the two together the
final thing will tilt down
and then some of the base area comes
from what is called sound power which is
this
um going back this uh
dash red line and uh sound power is just
sound everywhere and bass is everywhere
so if you combine all of those things
together and make this do
you get a predicted in-room response and
a predicted interim response in here
says we have it tilt down
although pro monitors tend to have less
of a tilt because they believe in a
flatter response
for somebody using a pro monitor for
make some music
then you would listen for hi-fi although
i find this one it's
got decent amount of tilt but the
formula of how much tilt you need
that's to taste you know you may like
less
high somebody else may like more highs
but as a general rule monitors tend to
have
less of a tilt in there and here we see
that our main flaw is this little dip in
here
so if it didn't have this this is
extremely good
response okay let me go back to a graph
that i
skipped over which is these um
individual driver responses so
my fancy system with a thousand
measurements generated the graphs that
you saw
but i also do a simple test where i use
a robotic boom
and i bring it point blank in front of
each element that radiates sound in a
speaker
and just measure what comes out of it so
in case of a twitter is this uh
blue line that you see or t online uh
brought the microphone right in front of
a twitter
and ran the full 20 to 20 000 sweep
you see that it shows the high
frequencies of a twitter and it drops
but then it goes back up and appears to
show that there's bass response
there isn't so when i have the
microphone up here this guy is also
playing because
there's no way for me to isolate these
drivers
without tearing up the uh speaker so the
base is still goes up and pollutes that
so you want to ignore when you look at
my graphs in here
you want to ignore when it lifts back up
and
for twitter that responses is not coming
out of twitter so just ignore
everything from this 1.5 kilohertz down
for the twitter
i then do the same thing with the woofer
a good thing about a woofer
is that when i measure that this guy is
pretty directional the tweeter is pretty
directional so when i'm measuring the
woofer
the twitter doesn't pollute it hardly at
all so for the woofer we see this red
line
comes down during the crossover region
and down here
you know response gets you know messy
because the woofer isn't designed
to produce good response at high
frequencies but luckily our level is way
below that of the twitter so twitter
sound dominates
but some of these peaks and valleys that
you see in here maybe the reason our
frequency response isn't
perfectly flat in these things these
little ticks that you see
you know it is the sum total of these
peaks and stuff
the third radiating element is the port
so i
put the microphone all the way behind it
and do the same measurement
and here we see the port and we see how
it helps the low frequency response so
this is the woofer
it loses its response base as the
frequencies go down
but then the port kicks in and gives it
a lift
up here around 40 hertz
okay so these two sum together
and basically give you a flat response
that then rolls off
down here um all ports unfortunately
have a artifact which we call resonances
so sound can bounce up and down inside
there
and amplify so a port normally would
have this response and you can see that
it keeps going down which is what we
want
and then all of a sudden it perks back
up
and starts to generate a lot more energy
it goes back down and has some other
because you have modes that are
horizontal and vertical and in all
different dimensions
and it has yet another one up here and
these are actually fairly controlled in
some cheap speakers i find resonances
coming all the way up
very close to the response of the woofer
or the twitter
and as a result you get peaking in the
frequency response due to that resonance
which doesn't sound good
so um here there's their control but
they're enough
in here little peaks in here but to
again cause these little
humps that you see in the frequency
response and usually you can tell the
resonances when the humps are repeated
in all three curves you can see these
three
are repeated and if we go down here yeah
we can see them over here this is the
resonance
so typically pulling those down helps
with clarity
i find these resonances are little
sinking
you know the birds inside the speaker
they get activated and all of a sudden
they sing along and you know if they're
highs you can pull them now again this
is just one or two db so
i would not worry about it so
far the assessment is quite good but
we're not done this
science sort of ends here and says that
like tonality
is everything in the speaker tonality is
everything
in a speaker in that it dominates
preference
so if i have two speakers one has boomy
bass the other one doesn't have boomy
bass
you're not going to like the one with
the booming bass it doesn't matter if it
has lower distortion and twitter
somewhere else
if it's going pro or there's too much
bass or it's too bright
you're not going to pay attention to
anything else about that speaker because
it's just too bright
this isn't controlled blind test so
but if we have two excellent speakers
with similar frequency response similar
you know early windows similar
everything
my opinion then distortion matters and
we want to measure that and we want to
find the limits of a speaker
know that distortion measurements are
somewhat tricky
in that they can be impacted by
reflections which i can also fix and
you know the science there of how you
measure distortion is
you know it's not the most perfect thing
so don't be horrified about some of
these graphs that you see
so i make two measurements for
um distortion one of them at 86 db spl
the other one is 96. this 96 is similar
to
soundstage network that uses nrc's
anecdotal chamber
for distortion measurements they
reported at 90 i reported at 96 but it's
exactly the same level
um but then i thought when i first
started doing this that
not all speakers do well with
uh are able to produce 96 db spl
so i do one at 86 db so we have both
what's nice about both is that we can
see
whether distortion is proportional to
level or
all of a sudden we have elements that go
nuts
if you will here we actually have very
good
uh distortion in that now you might say
whoa distortion is shooting up above
that but at very low frequencies our
hearing is not very sensitive
so those harmonic distortions that we're
showing in here are not
audible certainly and not at 86 db spl
when we increase volume to 96 db which
is the one on the right we can see the
bass gets worse and that's always it's
like physics
speakers most perfect one is not moving
the moment you move it is voice coil and
magnet geometry changes
and the response is just not as linear
it's the more you push it the less
linear it gets
it's just happier when happiest when you
stop producing sound
so we increase this and so on we can see
all those other peaks in here also go up
but still pretty controlled
there is a peak in here that we
discussed in the forum
um that does go nuts over here
and the people thought it's related to a
resonance of that twitter at 27
kilohertz
and that this measurement i can get
confused thinking that's the second
harmonic
of something at 13 and a half kilohertz
it sounds plausible to me um if it is a
27 kilos resonance then there is no real
audible distortion in here it's
happening above our hearing range
now this you're not going to see this
kind of measurement
elsewhere most people don't show it like
this and nrc certainly doesn't show it
like this
but i find that this kind of distortion
uh as shown as a percentage over here
and it's a linear scale is very
revealing because we can see the small
and we can sort of predict that this is
shooting up the more common display
people have
is uh showing the distortion as an
absolute level
uh together with frequency response so
if i put the two together
you can see you know their relative
scales to each other but
i find that this is a hard graph to
remember it's very busy
and so forth um but we still have our
little peak over here and we have our
base distortion
in crappy speakers this base distortion
could actually shoot up above our
frequency response which means the
distortion is above 100 percent
when that happens i guarantee if you
play anything in this low bass region
that speaker distorts like nobody's
business
so it's good when there's a gap in here
between these two
um i have made an arbitrary decision to
have 50
db as the threshold for fantastic
speaker don't ask me to justify
it just is and great speakers come very
close
and you can see in here above 100 hertz
this speaker
essentially uh has no distortion and if
this is not real then
you know from 100 hertz uh on this is
extremely good distortion
distortion comes in base low base and it
just is no way around it other than much
larger speakers many more drivers they
don't have to move as much
so when i've measured tower speakers
they're properly designed
these distortions are lower but this is
a small boss got one driver and it's got
to move and the more it moves
the more uh distortion we get okay so
overall
we're 90 95 good good news over here
the next um display
the graph is a pretty critical one it
says if we keep the
loudness the same what will how
wide what happens to the width of a
of that level an ideal speaker that just
sends a shaft of
sound to you with no variation
based on frequency will will look like
these two lines over here
we don't have an ideal speaker here but
we have a darn good one
and we can see this red curve which is
where
sound drops off 6 db is almost hugging
my two ideal lines in here
they're not symmetrical because when i
measure the speaker i can't always align
it perfectly so
assume that they are symmetrical so this
is a response to a very good speaker
what does this mean in practice means
i can go to the left and i can go to the
right
and have the sound drop off 6 db yet
tonality won't change and that's a cool
thing in a
in especially in a near field setting
where you're you know mixing recording
editing
sound and you're going to move around
left and right the client may be sitting
next to you
and you want everybody to hear the same
sound you don't want all of a sudden the
highs to drop off like a rock
when you move a few and just left and
right so horizontal directivity is very
important it's called directivity which
means
how's the sound being controlled and
that wave guide that i talked about
causes this middle region to be so
perfect and so nice
so it means the handout from woofer to
the twitter is
essentially perfect and the nice thing
about getting this
and if i go out i can also show this as
a heat map contour thing and you see the
same thing
you see this you know red area is the
loudest sound and you can see how smooth
this
is and what's nice about is that when
you get this kind of
sound it means that the speaker
is what i call room friendly because the
reflections have the same tonality as a
direct sound
it's okay to have reflections and side
reflections actually could be beneficial
from a
reference point of view because they'll
take a speaker and broaden
its its location and what that does is
that instead of sound coming from one
little point source
comes out of a sort of a wider region
between it
and and the wall and the spatial quality
is something that people
like in controlled settings in studios
people tend to absorb all the
reflections but just know that you have
the option of not absorbing reflections
unless they're needed and if you go back
to my previous graph in here
you can actually tell which reflections
you want to absorb which ones you don't
a ceiling bounce in here is the red one
you can see this dip that we have we
prefer to not have a dip
what's the cause of the dip is because
the ceiling response is pretty weak in
this area
so sound shooting up from this guy going
up is a little weak
in this region so if you can put an
absorber on your ceiling or have a tall
ceiling
that means you'll have less of this dip
most of the time i find that the floor
bounce
blue is a problem you can see the floor
balance also goes down
and is also responsible for that so in
99 cases
i recommend having a thick carpet on the
floor uh shacky carpet or something with
some depth to it because you're trying
to absorb frequencies from thousand
hertz up
and you need about one or two inches
there for effective absorption there
so most of the time you want to have the
floor reflect
absorbed and if you can with the ceiling
do the same
rest of the reflections on this speaker
are so good
there's no reason to absorb them you
know and indeed if you absorb too much
you wind up building a very dead room
that's
not good for enjoyment of music now
this is horizontally we're measuring the
you know what the reflections are
on each side we can also measure
reflections
up and down and that's how we determine
what the reflections look like for
ceiling and floor
and there we see the pictures still
quite
good for a two-way speaker two-way
speakers tend to be quite chewed up
and this guy is very very smooth except
for this hole that it has
and you know sometimes with all these
crab eyes or lobster eyes
and these you'll see so there's a hole
in here that says look
if you go too far above the speaker or
too far
i mean twitter well actually the
reference axis for this is right at the
edge of the woofer
if you go too far above this or too far
below
this you're you fall in these two
ditches and the moment you fall into
those two ditches
at two kilohertz you'll get a much more
pronounced dip
and you don't want that obviously two
killers super important
uh audible range and you don't want to
have a dip so if we look to the left in
here we see
that if you you can go above 20 degrees
up
and about 30 degrees or 40 degrees
low and you're still okay so general
advice for this thing should be that
it's at ear height
that way if you slouch a little bit and
you go up a little bit it's still
cool if it's placed much higher than
that tilt them down
so that the on axis is sent to you
so know that the on axis is the cleanest
sound and it is for most speakers
so if you're going to tilt the speaker
some other way know that you're not
listening to the best sound of the
speaker
a few speakers of design where the off
axis is better
usual because of some pla flaw on access
and they want you to tilt a little bit
but as a general rule speakers should be
tilted directly at you
and if you don't like the tonality use
equalization to fix it
but here you have as much freedom as you
can get the only speakers that are
perfect
um in this regard are coaxial
versions and indeed general like higher
end ones use coaxial drivers with
sitting inside the woofer so whether you
look at it vertically
or horizontally it makes no difference
we have this problem because as we go up
or down the
time difference between these two
changes
and how they add up and subtract changes
whereas horizontally the two are the
same distance
and we get a much better nice response
when the
twitter's inside the drive woofer then
it's a coaxial thing
and uh you know tracks well now
just making a coaxial driver doesn't
mean the sound is better
there might be other compromises power
capability
and there might be other issues sticking
a twitter inside a woofer that's moving
has can have its own problems general
ones are done extremely well and i've
reviewed those but
don't just assume that you want a
coaxial driver it's uh
it's got a stronger marketing message
than a technical one
the design has to be right i also run a
waterfall
this display can be extremely misleading
and i don't show it for
non-active speakers i know that i can
muck with the parameters of this and
make the display anything i wanted to
show
i do tune it a little bit so in this
case we see the port resonances that
we'd seen earlier but other than hmm
that's all you want to go past this
that's the
this waterfall display people fall in
love with it because it's so pretty
but know that it's prettiness factors
far more than its usefulness
um the hardcore objective is on a sr
forum don't like the fact that i listen
to speakers
they say the measurements say everything
as much as i i've measured on 140
speakers
in in 14 months so i'm quite trained
to what speakers sound like because i
listen to them so often and i'm always
comparing measurements to listening
tests
but i tell you though there are things i
listen for in lesson tests that
not measured and um i always say
electronics are 100 valuable the
measurements are
speakers are about 80 90 percent uh
measurements predict the final outcome
and
headphones are about 60 to 70 or maybe
80
so depending on the headphone uh the
measurements may be
excellent that may be you know not so
excellent
and therefore you got to fill the gap
with something and i filled the gap
with speakers and headphones by
listening
and uh you can ignore it if it really
bothers you and you say
i'm all about objective stuff and all
about the graph
more parity ignore this section but
i want to be the person that does a
sanity check on the measurement say okay
you know we have this frequency response
and it's got these
you know little imperfections in it are
those audible are they not audible do
they need fixing
and if they need fixing i build
equalization for it and then i'm able to
turn the filters on and off instantly
and i can quickly determine whether i
like to sound
with them on or without them you might
say well you're biased maybe you like
something and you don't all right you
can do the same thing you should set up
the same eq you turn them on and off if
you
see that they improve sound then great
you have i've already done the work for
you
and by the way i don't go correct every
little niggling things i correct major
errors
in this case i turn on this genelec and
play my reference list i have a revell
system so on two speakers
and i'll have a long list of tracks that
i think are great
demo material i'll run them against this
just as beautiful two different
companies with two different brands two
different products two different
approaches
yet they both follow the same research
that says hey make on axis flat
and make all faces as similar as you can
to that when you do that the same music
sounds just as beautiful and just as
familiar to me so to me
uh close my eyes i may tell you that's a
reveal speaker is not a generic
that's the beauty of standards that's
the beauty of research
so in this case one of you cases where i
turned it on
sounded great i said hey why even bother
doing an eq now normally that would then
say that i'm going to give it my highest
ranking which will be the golfing or the
soccer
panthers in the review but you didn't
see that you saw the one that says i
like it
why is that right when i was gonna
finish and go take pictures of the
speaker
i got to a bass heavy track and i turned
it up and i was like whoa
what is that what is that i noticed that
the uh
a bass drum hit would be boom at low
volume
then it would go at a little bit higher
volume and if you turn the volume way up
you go
sorry i can't repeat the sound any
better than that but basically if it
felt like a
feedback loop where the bass is trying
to go up and someone's pulling it down
and letting it go up let it go
up instead of one clean bass hit
it would go and the more you pushed it
the louder it got
the worse the distortion was quite
audible this is not some oh it's a
placebo effect no there's nothing
i mean it's just you just crank it up
now granted this was loud volume
and i'm only listening to one speaker
with two speakers would be twice as loud
so i wouldn't have to crank it up
so i would say at low to medium then to
moderately high level
you're not going to have this issue but
it doesn't have limitless power
and that's the number one failure i'll
find in most powered monitors
is that they have this limit i think
it's mostly due to amplification
and limiters they put in there and
indeed the clipping light that
turns red on this and it did exactly
that so
when i heard the distortion except for
the slight amount it wouldn't come on
but the moment i turned it up to even a
little bit past that
that light would blink and during my
testing at 1960 vspl that distortion
thing and low frequency that
light came on solid and you shut off so
in my view the amplifier in it is not
powerful enough
uh you need a subwoofer luckily if you
just roll it off
past 40 hertz i'm below 40 hertz 10 is
happy
so you know you just need a sub to fill
in that sub base area
which i value a ton you may not and
so it is not what i call the ideal
speaker
um i've tested one or two speakers or
ideal actually the
jbl 708p is the first powered speaker
i've tested that has no limit meaning as
much as i like loud
sound fantastic and deep bass
i could not get that speaker to bottom
out um
whereas this one i did so generic makes
larger
speakers more powerful ones they have
some insane looking ones
so obviously they don't want to have
this little guy
be you know everything uh even if they
could
so and part of that maybe this class ab
amps are just
anemic they have the power rating but i
don't trust their power rating i think
they're saying is
150 watts or so 150 what what i mean is
it's at a peak power some fake number
what distortion i don't know
so anyway um if you don't try to go
nuts with this as far as volume it's a
fantastic speaker
beautiful tonality warm sound
great low distortion sound it's just
fantastic i mean i have no complaints at
all
very good bass by the way when it's
producing without distortion
i mean it's just a great speaker it's
not
cheap it's a building made in uh you
know europe there are
1400 each i think
they go up uh nineteen hundred dollars
each so a pair of them will set you back
thirty eight hundred dollars
but other than max power trying to put
together this kind of performance and
passive speakers will be very hard
i mean to get two speakers to have this
kind of performance and amplifiers to
drive them and everything
is very hard so in general powered
speakers just
rule the world from this size down if
you don't need insane levels
trying to you know match them with
passive speakers very hard
um because you know they just have a
more optimal design they have two
amplifiers
electronic crossovers there's no losses
there
and much more optimized and if you have
dsp in them they just get so
perfect that it's very hard to replicate
that
so anyway this has gotten to be a very
long video almost 40 minutes so
i won't keep going um by the way for
speakers there's a preference score that
gets computed by one of the members
the preference score can misfire is also
not for near field
but it's got a very high score for the
speaker of 6.3 anything in sixes
is essentially state of the art i forget
what our highest number is maybe 6.7
it can misfire i can show you two
speakers
with identical score and you wouldn't
say that they're they're the same
is you know the science is good but it's
not excellent so
i don't compute it for that reason but
for speakers you do want to you know
sort of look at this and
maybe rule out anything that's extremely
poor in in this preference score
okay hopefully you got a good sense of
how speaker
measurements and testing is done by
audio science review
text version of this review is available
and for this and
as i mentioned 130 other speakers okay
hopefully uh you didn't fall asleep
during this video
see you in future episodes bye-bye
تصفح المزيد من مقاطع الفيديو ذات الصلة
“Hands On” with Klippel's Nearfield Scanner || How I measure loudspeakers
Why KEF R5 Meta Speaker Is a Must-Have for Audiophiles with a Budget
VERTICAL Vs. HORIZONTAL Studio Monitors (Can They Be Flipped?) | ADAM Audio
Room Correction With Convolution Plugin Room Eq Wizard Tutorial
Harmonic Distortion is overrated.
What Is Frequency Response? || Understanding the Measurements Part 1
5.0 / 5 (0 votes)