Crossover Basics - Impedance

In my first post on the subject of Crossover Basics we talked about the electrical response. How various types of high and low-pass filters changed the voltage that the speaker drivers experience. We glossed over the issue of impedance entirely. Understanding this will help you better grasp what's going on.

Introduction to Speaker Impedance

Anyone who has examined speaker specifications has seen an impedance rating, such as 4 or 8 Ohms. If you see speaker measurements from Stereophile or SoundStage you would have come across charts showing complicated details about impedance and phase. Here is the impedance chart from the ported LM-1. This first chart shows the combined effects of the drivers, crossover components as well as the cabinet and port.

Ignore the thin blue line, and focus on the thick one instead. Notice the chart axis. Across the X (bottom) we have the frequency scale, across Y (left side) we have Ohms.  So the first thing you learn is that the nominal 4 or 8 Ohms you are used to seeing are kind of a summary, and sometimes an outright lie. No speaker is exactly 4 or 8. These are guides to help match speakers and speaker counts to amplifiers. Personally I would rate the LM-1 as an 8 Ohm speaker, though it dips below this above the bass. The triple hump you see is indicative of a 2-way, ported loudspeaker. The two humps at left are caused by the port, while the hump at right, just below 2kHz  is where the high and low-pass filters meet.

You might have picked up on the fact that we are using Ohms to measure impedance (sometimes called Z). What makes impedance different from resistance is that it is frequency dependent, and it has an angle. This angle is the difference between the voltage and current. We'll ignore this for now and focus on the magnitude in Ohms. Impedance is what we use to understand the effects of capacitors, coils and speaker drivers which are electrically complicated little beasts.

One very important point to note is that even though the tweeter and woofer sections are arranged in parallel, the impedance of the drivers does not actually appear in parallel. That is, a pair of 8 Ohm drivers wired in parallel would normally result in an impedance of 4 ohms, but in this "perfect" system the actual impedance stays very close to 8 Ohms at all frequencies.

Let's talk a little more about the LM-1, above. From left to right, the first two humps are classic indicators of a ported loudspeaker. From the bottom up to around 1,000 Hz the impedance is entirely that of the woofer. The peak around 1,800 Hz occurs where the low and high pass filters overlap, and after about 5kHz, the impedance is entirely a result of the tweeter.

In the sections that follow we'll go over each section in detail.

Effects of High and Low Pass Filters

We'll use the original example of a first-order crossover for all of our discussions.

To refresh your memory, this chart shows us how first order filters behave. The red trace is the electrical signal a tweeter would see, while the yellow is the woofer (1 kHz is actually a very low frequency for most tweeters, let's pretend it's OK). Black is the summed response of the tweeter and woofer. While it is smooth and straight we don't care about it right now.

This is strictly the voltage and electrical view of what would be measured at the inputs of the "ideal" 8 Ohm tweeter. We are, of course, missing the acoustical results, but we cover in the post Crossover Basics - Driver Response.

The question we answer in this post is: "How do we achieve this frequency dependent change in response?" Let's take a look at the original crossover first:

There is a capacitor in series with the tweeter, 20uF. We talked about it "blocking" low frequencies. This is achieved by an increased impedance. That is, as the frequencies drop, the Z (impedance) of C1 goes upwards. Let's take a look at the impedance of the tweeter circuit, with the driver and the driver + capacitor.

Series Voltage and Impedance

Keep this rule in mind: The voltages across each component in a series circuit is proportional to the resistance of each component. We're dealing with impedance, and complicated but for the sake of simplicity we ignore it here. This simple rule-of-thumb is enough to explain the behaviors without delving into reactive impedance calculation.

High-Pass Impedance

In the chart below the red line represents our tweeter impedance. Again, this is theoretically perfect 8 Ohms. No drivers are like that in rea life but we need it simple so we can better see what the high-pass filter is doing.  The blue line however represents the entire tweeter section. That is, C1, S1.

Unfortunately at 100 Hz the impedance is cut off but it's close to 80 Ohms. If C1 is 80Ohms we can then estimate the relative voltages across C1 and R1. It's only an estimate because we're not taking phase into account, but it's going to be very close. So we estimate the voltage at S1 to be 8 / 80 = 0.1 of the total. So at 100 Hz the tweeter gets approximately 10% of the voltage. At 10 kHz Z ~= S1, meaning our tweeter gets nearly all the voltage, and C1 is acting like a short circuit.

With those two points (100 Hz and 10 kHz) in mind, now the following graph should make sense. It plots the voltage across C1 and S1 as the frequency varies. Assume a 4V input voltage.

As we calculated, at 100 Hz the voltage across S1 is about 4V x 0.1 = 0.4V

This shifting of impedance and voltage between the filters and the drivers is similar regardless of the order of the filter. It's also mirror-imaged for the woofer.

From Voltage to Decibels

Let's calculate the dB drop at the crossover point of 1 kHz. The peak tweeter input is 4V. At the crossover point it is approximately 2.8V. So, we calculate using the rule:

dB = 20 log ( Vout / Vin ) = 20 log ( 2.8/4 ) = -3 dB

In other words, at the crossover point the input to the tweeter is 3 dB less than the amplifier output. Simple, right? let's do this again for 100 Hz:

 20 log ( 0.4/4) = 20 log ( 0.1 ) = -20 dB

So at 100 Hz, the voltage to the tweeter is -20 dB below the amplifier's output. A very good thing since tweeters are easily damaged by low frequency signals.

Low-Pass Impedance

The same and complementary effect is happening at the woofer side of this example:

In the woofer side, by around 100 Hz L1 has no meaningful contribution to the impedance magnitude (i.e. Ohms) so nearly 100% of the input voltage is at the woofer. By 10 kHz however the impedance of the coil is approximately 80 Ohms. Estimating 8/80 = 10%.

Additional Exercises

Using XSim, and "ideal" drivers, create 2nd order filters and compare the impedance of the high and low pass sections. Compare the electrical response and impedance curves. Combine them and examine the impedance bump where they meet.

Using a second order filter, examine why the 2nd component works without shorting out the entire speaker. For instance, if a high pass filter, examine why the coil (L) works. What would the impedance be if the coil alone was there, but not the capacitor?

Crossover Basics - The Zobel

Introduction to the Zobel

A Zobel network is used to flatten a driver's impedance (usually a woofer or mid-range), therefore making the filter (usually low-pass) more effective. There are many kinds of Zobel networks, but for speakers the simplest and most common Zobel circuit consists of a capacitor and resistor which are wired in parallel to a driver.

C1,R1 on the left are an example of the Zobel network. Their values are chosen to minimize the impedance rise of a driver above the resonant frequency.

Speaker Impedance

If you aren't quite sure what this means, please visit my blog post on Crossover Basics - Impedance before reading more.

Introduction to Series Circuits

Take a look at a very simple series circuit. It consists of two resistors in series (one after another) with the amplifier. The circuit is closed by the ground points (the downward facing triangles).

As before, this may be something you like to play with, so I encourage you to grab a copy of XSim and try this out along with the Peak Voltage chart.

What's important to understanding here is that the voltage across the resistors will be proportional to the resistance offered.  So

Vr1 = Vin * R1/(R1 + R2)
Vr2 = Vin * R2/(R1 + R2)

And of course:

Vr1 + Vr2 = Vin

That is, the voltage across R1 and R2 must add up to the input voltage.

Think of Ohms as elephants that eat volts. The more elephants, the larger portion of the incoming voltage they eat.  Yes, this is a very silly analogy.  Still, we can do some quick math. The total resistance is 100 Elephants (hah!) or 100 Ohms. R1 has only 10 elephants, so it gets 10% of the incoming voltage, whatever that voltage may be. R2 has 90% of the Elephants, so it takes 90% of the incoming voltage. There's a lot more to circuit analysis, but this is the bare minimum to understanding Zobel networks. Hopefully you'll be intrigued and learn more on your own.

Woofer Impedance

So with a little background under your belt you are now ready to look at a typical woofer. We'll use the LM-1 woofer, the Peerless 830991. You should know that impedance graphs will change once a driver is in the cabinet, especially if the cabinet is ported, so the data I present here will be different than a specification sheet which measures the driver in free-air.

You have heard the term "coil" used interchangeably with "inductor." Which is correct, but you may not have thought about the term "voice coil" in the same context. The voice coil is the part of the speaker that will electrically connect to your amplifier and produces the magnetic force which moves it against the magnetic field of the permanent magnet. We won't get too much into this, but suffice it to say that above the resonant peak the voice coil behaves like that of any other inductor, specifically it has both a resistive element (DC Resistance, or Re) and an inductive element (Le). These combine to give us the woofer's electrical impedance (Z) at any given frequency.

In the chart below we will compare the impedance of the woofer in a sealed cabinet (red) with a woofer that has a Zobel network applied (blue).

Let's ignore what happens below 200 Hz. That's a topic beyond this posting, and it's also far below our likely filter points. From about 200 Hz to 400 Hz the woofer is purely resistant, and we have the minimum around 6.6 Ohms, but what happens to the right? That's correct, suddenly more voltage-eating elephants arrive! The inductive qualities of the voice coil become more and more important and overwhelm the well behaved resistance. Compare the peak difference of the two impedance charts, all the way at the right. The blue line represents a woofer compensated with a Zobel network. The impedance never goes above 7 Ohms, while the normal un-compensated woofer goes to over 30! That's more than 4:1 difference. This increase impedance is going to compete with the low pass filter and make it behave in ways we probably don't want.

For clarity we'll leave behind the LM-1 schematic and create a new one, with two identical woofers and 2nd-order low-pass filters set to 4 kHz. Of course, this is not how a real speaker would be designed, we are just using this to see exactly how a Zobel circuit works. The Zobel consists of C1 and R1. 

The first thing we should do is examine the transfer function of the two filter sections:

As you can see, S2 is behaving like we expect a low pass filter to work. S1 however is having a very difficult time getting to the right slope. At 5kHz the output is almost 10 dB higher than we want it to be. That's a big deal. Eventually the impedance (elephants) on the low pas filter take over, but they don't reach our desired behavior until past 20 kHz, definitely not good enough for us. Let's take a look at the final outcome, below:

How Does a Zobel Work? 

Above we discussed how serial components work in a circuit. You may feel a little tricked because while we learned enough to understand why coil inductance needs to be compensated for, we never talked about how a parallel circuit works, which is what a Zobel is. A parallel circuit has one unique property:

The apparent impedance of a parallel section is never more than the smallest impedance.

If we imagine C1 as a short, then no matter what S2 rises to, the impedance will never go above R1, or 8.2 Ohms. It can be less than that, but never more.  In a parallel circuit you calculate the apparent impedance like so:

Rtotal = 1 / (  (1/R1) + (1/R2) + (and so on and so forth)  )

Things are more complicated because we are actually calculating impedance, but you get the picture.

Let's do some quick, Dr. Leach style of analysis on the components in a Zobel. At very low frequencies, C1 behaves like an open circuit, essentially removing R1 and C1 from meaningful contributions to the system impedance. Remember we mentioned that impedance cannot rise more than the smallest value? So at low frequencies C1 is so large that S2 becomes the limit on impedance. You can see the impedance below 500 Hz or so is barely affected. At high frequencies, C1 rapidly decreases until it evectively becomes a short, putting R1 in parallel with the driver, S2 and limiting the absolute maximum impedance to 8.2 Ohms. In this case we don't reach 8.2 until well-past 20kHz but it would eventually reach that point once the woofer's impedance was high enough. Also past our point of concern. If we can limit the impedance from 6.6 Ohms to 7  Ohms then we have a much more stable impedance curve than before, and that's good enough.

Do I Need a Zobel?

That's a tricky question! So let's examine this woofer and it's output. If you wanted to cross it over at 4kHz I would think the Zobel was mandatory, however if you were going to set your crossover frequency at 2kHz or lower I would say not really. The LM-1 uses it, but the effect of the Zobel is small, and benefits the phase response so I leave it in. It is possible a very similar sounding LM-1 could be built without a Zobel and with different choices in the filters left behind. The best chart to look at to see if a Zobel matters in your circuit is the transfer function chart.

It is very rare, but not unheard of, that a tweeter needs a Zobel because their voice coils are relatively tiny and therefore don't have a lot of inductance. The most common exception to this rule is with ribbon tweeters. The ribbon itself is not inductive but the entire assembly often include matching transformers. Transformers are coils .... and coils are inductive... see where this goes? :)

The real point to the Zobel is to make things better in the area you need the filter to behave at it's best. That's usually up to about -20 to -30 dB. Beyond that if your slope isn't perfect we no longer really care. There's no audible difference between -60 and -67 dB for example.

An important consideration in choosing a Zobel or not is that they are not free. The more parts in a system, the more expensive, the more chances of failures or parts being out of specification. If this is a personal project, no problem, it's all experience. However if you are building for mass production eliminating unnecessary components is the final stage before committing a design to the factory. 

Placement

In most cases, you want to place a Zobel closest to the driver. Put anything else such as padding resistors, filters, etc. before it. While the order of serial components does not matter, the order of parallel components does. Leaving the Zobel last prevents unexpected consequences.

There is a rare exception, when you must equalize a driver (usually a tweeter) by adding inductance. In which case you want the equalizing circuit closest to the tweeter, then the Zobel, so the Zobel can also control the EQ's impedance. I'll write more about this later in a section on handling difficult tweeters. 

The Secret Uses of the Zobel

Many will rely on on-line calculators to determine the right values for a Zobel network, and that's fine, but be aware that the absolute values can be tweaked. The main benefit of this tweaking is to gently nudge the phase charts one way or the other, helping you get near-perfect matching between two drivers you otherwise might not have. This is where having a tool like Xsim to simulate your tweaks comes in super handy.

Exercises

The data for the Peerless 830991 is contained in the LM-1 XSim files here.  Feel free to take it and modify it to help you complete these exercises.

The Peerless 830991 has an Re of around 6.6 and Le of around 0.330mH. Try simulating this in Xsim using a resistor and coil in series. Compare your impedance curve with the red impedance curve, above. What's the biggest difference you see?

Try using an online-calculator to create a Zobel for this driver. Tweak the capacitor and resistor values. Can you do better than the on-line calculator?

Using the complete LM-1 schematics, compare the woofer response with and without the Zobel. Is it a big difference? Can you fix the LM-1 so it no longer needs a Zobel? What difficulties did you encounter?Pay attention to the phase matching as well as the frequency response.

Use XSim to calculate the power through the Zobel resistor if the amplifier is set to 100 Watts. This is wasted power. Is it worth it?

Stereophile - Data Part II

This is a follow up to a previous article on what makes a speaker great to the editors at Stereophile.

I thought that was the end of that discussion, but thanks to an article written in 2008 but recently republished by the good Dr. Joseph D'Appolito there is more. Those who don't follow speaker design and measurement will not know D'Appolito literally wrote one of the most cited books on speaker measurements, in addition to having a configuration named after him.

In the article published by audioXpress D'Appolito shares an interaction with Stereophile head honch John Atkinson (JA). JA did something I though was pretty interesting mathematically, but I call bullshit on his message. He claims he analyzed a number of speakers and compared them to those which would make the recommended components according to frequency response and that most were perfectly neutral. D'Appolito states:

[John Atkinson] defined the standard deviation (SD) from flat response over the frequency range of 170Hz to 17kHz as a criterion for judging flatness of frequency response.

Further:

Of the 15 speakers with an SD of 1dB or less, 14 were added to the list by Stereophile reviewers.

But take a look at two speakers Stereophile raves about in my previous post. FAR from neutral as defined above. Then take a look at the hatchet job they did to the Crystal Cable Minissimo Diamond here.

So, bunk. I personally don't care what John Atkinson likes. If he likes the B&W diamonds above all others that's fine with me. But to call them neutral, or try to sell them as the reference against which other speakers are too dull or bright is shilling.

Crossover Basics - Driver Response

The Decibel or dB

Decibels (dBs) are a curious way to measure electrical and acoustic energy. Curious, and terribly convenient! For us, we use relative electrical dBs to discuss how filters work, and absolute acoustic dBSPL to measure speaker output.

When discussing the effects of a filter on a signal, we'll use relative dBs. That is, there's no set standard, but we talk about something being +4dB or -18 dB. This is useful because we can map this to speaker outputs no matter the volume settings. It is how we will discuss how a filter works, without worrying about the absolute output levels.

On the other hand, when we discuss the acoustical outputs we'll use dBSPLs which are in absolute terms, but using a set input level. Don't worry too much if this is confusing, we'll make it more clear as we go along.

The LM-1 Crossover Revisited

As mentioned in my first installment on Crossover Basics, the effects of a crossover filter are additive to the speaker driver.

We are going to use the LM-1 crossover and focus on the tweeter response in detail. Let's refresh your memory about the crossover, here it is on the left.


We'll focus on the tweeter filter section. This includes C1, L1, R1, and R2. The woofer section will seem neglected by comparison, but we cover it in more detail in other blog posts, including the Zobel.

Let's go over the transfer function. That is, how the voltage at the tweeter is different from the amplifier output because of the crossover. 0 dB means there was no change, the input and output are the same. The woofer response (in red, below) is almost exactly 0 dB until around 700 Hz when the low-pass filter kicks in. The tweeter on the other hand is more complicated. Let's discuss.

Anytime you see a chart this clean, you can be sure you are NOT looking at acoustical measurements. The blue line is tne tweeter filter's response. Except for the level shifting, this seems like something straight from my previous post on Crossover Basics. First, notice the level of the tweeter. It has been "padded" or "lowered" 6 dB below input. This is accomplished by the 4.2 Ohm R1. R1 is effective at all frequencies. Everything gets shifted down about 6 dB because of it. It's not exactly always constant, but let's pretend it is for right now, which is very close to true.

In addition to the padding there is a high pass filter reducing the midrange and bass at about 8 dB/octave below 2kHz. We discuss pads by an absolute number, like "6 dB" because it's effect is constant at all frequencies but we talk about high and low-pass filters with rates. In this case, 8 dB/octave means every time you cut the frequency in half, you will loose 8 dB. This is the actual "high-pass" section at work. This is C1,L1,R2. Notice that after about 4 kHz the high pass filter effectively stops working. It's as if it wasn't there anymore.  Above this level the only parts still involved in the high frequency response are the tweeter and R1. 

Putting it All Together

The point of this post is that these changes are not in isolation, but rather in combination with the driver so let's take a look at how the padding resistor and thigh high pass filter combine withe the acoustical response of the driver to produce the final outcome.

Notice the scale is now different. We are now looking at dBSPL, or sound pressure dBs. It is most common to take the frequency response measurements of a driver at 2.83 Volts input with the microphone at 1 meter distance. As you can see, below, this particular tweeter outputs about 90 dB at 2.83 volts above 4kHz or so.  2.83V is a common reference standard because at 8 Ohms this is about 1 Watt.

The top black line represents the tweeter with no filter at all. The green line represents the tweeter with just R1 added. It's not exactly 6 dB down everywhere due to the tweeter's impedance curve, but it's close enough for us! You'll learn more about this in the next post which covers the Zobel. The red line represents the addition of the high pass filter section, C1, L1 and R2. You can see it pivots around 3 kHz.

By carefully selecting the filter knee (-6dB point) and it's Q, or steepness we can get a little bit of EQ thrown in for free. Take a look at the original response (black) at around 2 kHz. You see the broad bump centered there? The bump is pretty much gone thanks to the high pass filter. We have not only added the high pass filtering, but we also tamed a little over-activeness int he tweeter without increasing the part count.

Padding

In the chart below you can see the final LM-1 design in red, vs. the a redesign without R1:



It may not be obvious from this, especially since this author likes to use far-field as his reference, but the LM-1 without padding would shriek.

In designing a crossover, I find it easiest to start low and work my way up. The low pass filter will reduce the sensitivity of the woofer at the crossover point. After this, we must adjust the tweeter to match and then add the high pass filter.

The total amount of padding (dB loss) depends on a number of things, including:

• Innate woofer efficiency
• Woofer low pass filter and baffle step compensation
• Innate tweeter efficiency
• Tweeter high pass filter

Unfortunately there is no simple, accurate way to go from a manufacturer's sensitivity specs to appropriate filter design.

The crossover designer must balance all four of these issues at the same time which is why in-cabinet measurement and simulation are so important. I encourage you to grab the LM-1 simulation files and attempt this for yourself.

Also note, that doing the reverse, padding the woofer, is generally discouraged because the power dissipation needed to lower a woofer a few dB is pretty large and requires big resistors and will waste a larger amount of amplifier energy. If your tweeter is too insensitive you probably need to change tweeter or woofer. It is pretty rare to find any design that does not require any tweeter padding.

Summary

With this posting, you now have learned:

• How crossover filter's add to driver output to create the combined effect of both. 
• How you can use leverage a high pass filter to also work as an EQ for you. 
• Why tweeters usually have resistors to pad them down. 

In my next post, Crossover Basics - The Zobel,  we'll go over the LM-1 woofer response but spend particular attention on the often misused or misunderstood circuit, the Zobel.

Cheers! 

LM-1 2.1 Subwoofer System

Work is under construction!

I've come across a couple of incredible deals over at Parts Express I hope to have the time and money to take advantage of. As you may know, I've been listening to the LM-1 as desktop speakers. They do a fantastic job with just 20 watts, but.... they really have no bass. It shows up more on games than movies.

The LM-1's have amazing bass in a bookshelf with very good rear wall reinforcement though! 

Parts Express is selling a powerful but small 10" subwoofer/cabinet bundle for under $200, plus, they have a 2.1 channel plate amplifier for another $100. The combination makes the foundation for a perfect multimedia system that would still be small, and couple very well with the LM-1 kit.

I've been using the LM-1 with a full-range 20 Watt digital amplifier and they really sound great. I expect the 2.1 amplifier to sound even better. The plate amplifier adds a high-pass to the amp. So the speakers will still have 20 Watts, but dedicated to 80 Hz on up. A separate 50 Watt amp drives the subwoofer itself. See where this is going? Of course, everything depends, but you could end up with the same volume and power of a s150 Watt/channel system. Bi-amplified systems are more power efficient than single amp systems, generally speaking and with music so this idea makes mathematical sense.

The combination is the perfect size for a desk or dorm-room. I just hope the cash magically appears so I can build it before the sale ends.

The Dayton iMM-6 Calibrated Microphone

When designing and building speakers I use OmniMic. It's perfect for that but when it comes to basic Blu-ray or Home Theater setup I rely on a much less expensive and much more convenient device, the $30 Dayton iMM-6 calibrated microphone. 

Dayton provides calibration files via their website which make this almost a laboratory piece of equipment. It works in your cellphone or tablet. I personally use it on an Android along with Audio Tool which will read the calibration file and adjust the levels accordingly.

One of the main benefits of this device is how deep and flat it goes and is therefore more accurate than the old SPL meter method. It's much easier to set your speaker levels, including your subwoofer using it and your tablet than any other way I know.  It also has outputs for a mini-jack so you can use test signals from your iPhone/Android device simultaneously. You can use a cable like this one to connect it to your stereo inputs.

If you are a DIY hobbyist and want to build your own speakers or want to do detailed acoustical analysis of your listening environment this is also the perfect front end for Room EQ Wizard but you will need an adapter cable from it to your PC or laptop.

The only real downside of this is that it's tiny and I'm constantly misplacing it after I use it.

Fix Ground Loops Quickly, Safely and Easily

What is a Ground Loop?

It is video noise or audible hum that plays through your speakers that occurs when multiple ground points with slightly different potentials are connected together through your equipment. More on this is available from Jensen Transformer's web site. It's called a loop because it's actually that, a closed circuit.  All we have to do is either connect everything to the same ground, or break the loop in the safest and most convenient location.

A clue to a ground loop is that the hum stays constant whether you play music or video or not. Sometimes equipment such as a TV doesn't even have to be on to help cause in the loop.

Note that this is not mechanical hum. If you hear the chassis or transformers vibrating and making more of a buzzing sound your problem is more likely DC on the line. I'll cover that in another post, but this is often caused by PCs, light dimmers, and modern compact florescent bulbs.

Ground Loops and Digital Signals

The usual "scientific" belief is that digital signals are immune or at least very resistant to noise. While it is true that digital circuits are resistant to many types of noise these circuits, including digital video, can participate in a ground loop which can cause enough jitter to be audible or visible. This includes HDMI, coaxial and USB circuits. Optical digital connections however are completely immune to all external noise, including ground loops.

Ground loops will not occur in Ethernet cabling unless there was a fault in the switches / routers. In other words, almost never since preventing ground loops was part of the design of the entire Ethernet eco-system.

Diagnosing the Cause

There are a few common culprits:

• Cable TV or Satellite Dishes
• External antennas like FM or television
• Audio or video cables from a Personal Computer
• Laptops! (Problem goes away when you disconnect USB or charger) See USB fixes, below.

The best way to find the root cause is to disconnect each suspect and listen for the problem to go away. Sometimes the problem is related to two devices interacting, which gives you a choice of where to break the loop. This process also works for finding noise sources in general. Turning lights off and disconnecting wall-wart supplies may solve other symptoms.

Lethal Fixes and Myths

One type of fix can be lethal to you and your neighbors. That's a "cheater plug like this one. Any attempt to defeat the ground pins in equipment that has them may be lethal. Do not do it. Do not rely on signal grounds to work the same way. They don't.

Pangea originally sold "high end" IEC cables with removable ground pins. Don't buy them, don't let your friends buy them. They appear to be discontinued, probably due to safety concerns. There is now a 2 conductor C7 version with a removable pin, but that's perfectly safe, and the feature is kind of useless.

I recently heard this argument:

I've been removing ground pins for years and never had a problem.

The problem is these pins are like safety belts. So imagine me telling you this:

I've not had a car accident in 20 years, so I no longer wear a seat belt.

That should sound dangerous to anyone who drives a car. That's how we electrically minded people think when we hear of audiophiles removing ground pins for that last bit of audio nirvana. There are better and safer ways. Another myth, spread by audiophiles who do not understand the safety ground or the life safety issues involved in UL certification and the National Electric Code:

Your system will be grounded by your RCA cables. No problem!

If no problem means dead and on fire, they're right. 

Free Fixes

If your problem is caused by a piece of audio/video gear, try connecting it all to the same power strip or conditioner. This ensures all the ground wires are at the same potential.

Another free fix may be to use XLR cables. XLR cables are not usually quieter in homes than balanced, BUT! there is a difference. XLR cables don't mix the ground and signal together. You avoid this contamination and XLR inputs often have a safe "Ground Lift" switch. It prevents the grounds loop from occurring at all.

Not A Ground Loop

Some issues have nothing to do with ground loops but are caused by induced noise from other sources. This noise can come through the power lines OR be induced by proximity to interconnects and electronics. Things to try turning off, disconnecting, or removing from the environment:

• Compact Flourescent bulbs - VERY noisy! 
• Wall warts - These tricky bastards stay on, and polluting even if the device they are feeding is off.
• Wall dimmer switches
• PC and laptop power supplies (yes, again!). Disconnect your PC or laptop cables to your stereo, TV, etc. If the problem comes and goes with the PC/laptop being plugged in, then you have a noise problem and will need to relocate it.
• WiFi devices, including routers, streamers, receivers, modems, etc. If your Wifi device is part of your stereo, try moving the antenna or putting it on a different power strip/conditioner. 

 

Noiseless Cables

Sometimes the problem is noise our cables pick up. Especially problematic in apartments with a heavy concentration of WiFI signals or near transmission or cellular towers. Some electronics will help pick this up more than others.

Make sure your interconnects are 100% shielded. Most cheap and a lot of expensive RCA cables use a braided ground, which is more of a pick-up antenna than anything else. Regardless of whether you use RCA or XLR cables, the best use 2 conductors plus a foil shield. In essence they are built of conductors:

• Positive conductor
• Negative conductor
• A super thin and delicate foil shield
• The drain wire which is used to attach the foil to a ground conductor on the RCA or XLR jack

On RCA cables the drain wire should be attached to the negative conductor at the source. The destination end does not use it but instead uses the positive and negative wires. With an XLR cable all 3 wires are attached at both ends, unless the destination does not have a ground lift pin in which case the ground may go unattached at the destination.

My favorite brands for non-esoteric cables:

• DH Labs
• Connex
• Mogami
• Belden

I personally use Connex solid silver cables for everything, but they are delicate. Get the more expensive DH Labs varieties if you need rugged.

Cable TV & Antenna Problems

This problem can also cause Internet access issues. I use a dedicated Cable TV isolator like this one. You put it immediately inside the RF plug, unless you have a satellite dish.  More on that, below.

Satellite Dishes

Make sure the satellite dish cable is grounded before entering your home. Even though it is required by code, installers often fail to do this, and ends up in having your receiver or antenna getting fried by heavy wind causing static as it blows across the dish.

Unlike Cable TV and overhead local antennas, satellite dishes require DC power to operate the RF amps built into the little head. For this reason isolating them is a little trickier. Normal isolators block DC in all forms.

The trick is to buy a separate DC power supply for your antenna. Place the ground loop eliminator closest to your receiver, and your antenna power supply closer to the antenna.

HDMI

Your best / cheapest way to eliminate issues from your television over HDMI are to fix any connections going to it such as cable tv, satellite, a PC, etc.

USB/DAC

Early in the history of external DACs ground loops could occur through coaxial cables. Most good DAC's today provide what is called "galvanic isolation" meaning that there is no DC or ground loop path between the input plugs and the rest of the circuits in a DAC. This can be done by purpose built transformers or modern monolythic IC's. Unfortunately no magazine or agency I know of tests for this so there is no way to 100% guarantee a DAC's isolation. Of course, the way to test this is to disconnect your USB input and see if your hum goes away. There are also cases where you have very little ground loop noise. To fix either use a USB isolator like this inexpensive model designed for medical professionals, but works just as well with USB 2.0 DACs.

Another fix is to use a purely optical cable between your source and DAC such as the Audioquest Forest or Monoprice both of which come in a variety of lengths.  Check the size of the plugs, some may interfere with other sockets on your equipment. Optical cables are also a very good choice for going from PC to a DAC or receiver if the PC supports it since PC's are such noisy environments.

Of course, since the entire point is to avoid a current path, gold plated optical cables are kind of silly.

Audio Signal Isolators

Audiophiles hate putting anything in the signal path, and some fussy recording engineers may also, but pro's also know that it's far better to put in a transformer than delaying a show, so here are a couple of isolation products from Ebtech that are reasonably inexpensive.

I'm currently working on Cable Mittens, a new concept to warm up the sound of an amplifier while breaking ground loops and reducing noise.  Until then, the choices below are the best available!

The Hum Eliminator takes 1/4" jacks, but adapters are easily found. For a little more you can get the XLR version shown here.

The EbTech models, especially at their prices, are very good, but audiophiles who only want the very best turn to Jensen Transformers for the gold standard in high quality audio isolation. If that's what you need, I present the RCA Jensen Iso-Max for your approval. It's usually the best solution for PC audio problems.

The XLR Iso-Max, below, is also available for around $250.

Last Ditch Efforts

If your problem is your electronics and the single power strip idea doesn't work, the only remaining almost safe way to prevent the problem I know of is the Ebtech Hum X. The Hum X is only rated for 6A which limits it to line  level electronics. You cannot use it on power amps, which is not really a problem because we can prevent the ground loop at either end. You will find it just as effective by putting this on a preamp, TV or source as on the amps. It should also be effective on PCs, but I'm not sure if it could cause other problems.

I say it's "almost safe" because it has not yet been UL approved. The 6A rating is probably why, as there's no way to guarantee users will only plug-in 6A devices.

Iffy Solutions

If you are an audiophile you might have gotten to the end of this article wondering why the real "power conditioners" weren't mentioned. The truth is that the solutions provided above are the most effective in solving ground-loops than almost any high-end power conditioner.   Balanced power conditioners, which are often touted as the best solution for this, may or may not fix actually fix a ground loop problem.  Bill Whitlock and Jamie Fox of Jensen wrote a great paper for the Audio Engineering Society on the matter. Balanced conditioners ARE completely effective at removing DC from an AC line however, and very effective at reducing other types of incoming AC line noise.

Other types of power conditioners will have no effect at all on ground loops but may reduce other types of noise or provide surge protection and, as mentioned, connecting all your electronics to a single strip or conditioner may also eliminate the problem but don't go spending big bucks on them trying to fix ground loop issues, or ignore the solutions above because they don't seem high-tech or expensive enough.

The LM-1 vs. Home Theater

As mentioned The LM-1 is most at home on a bookcase, with the back panel wall within 6" to 12" of the rear wall. For the all-analog audiophile that wants a purist system this is perfect.

However, if you are more progressive, and have access to room-correction then you can use the LM-1 pretty much anywhere.  Of course, this usually means a home theater.

If you are going to use them with a sub you may wish to plug the ports though, as that will give you the most dynamic range and make for the easiest crossover matching.

Good listening!

The LM-1 vs. Stock

Unfortunately it seems my designs are so popular that in the US the LM-1 and its sibling the LM-1C have caused Madisound and Parts-Express to run out of stock. Fear not, Parts-Express expects more in February of 2017.

If you absolutely cannot wait, the Peerless 830860 will substitute. Same motor with a polypropelyne cone. Probably won't be as transparent, but I've not heard or measured it. Specs say it's going to be very close.

Lepai 2020A Micro Amplifier

I purchased a Lepai 2020A 20 watt micro amplifier to power the LM-1 speakers I recently put on my desktop.I used the micro plug for the PC sound and the RCA inputs are being fed from my Logitech Squeezebox Touch.

The Lepai 2020A is part of a family of amps that visually look identically but are based on different chipsets or have higher power ratings.  The "A" variant is based on a Yamaha chipset. Yamaha has been making professional amplifiers for decades, including very high efficiency models for studio and touring use.  The 2020A uses a 12V/3A supply.

Based on a variety of reading material, I doubt the 2020A does 20 watts/channel. It's probably closer to 7 when you limit power rating to 1% THD. Still, sounds very good.

For the price, it's really kind of ridiculously good.

The next model up, the 2020Ti takes up to a 24V/3A supply and gets closer to 14w/channel at 1%. Sort of.

The sound is smooth, very quiet, and not fatiguing at all. It has plenty of power for watching Hulu while in bed or playing games, as well as listening to Jazz FM 91 via the Internet.

The tone controls are subtle! They may not do it for you if you are looking for big changes, but if you like to gently adjust bass and treble they'll be great for you.

Some users report even better performance with beefier power supplies which I did not try. In my case small and hidden is key! Plus I never really felt the amp was too small. One thing though, the LM-1 speakers are VERY easy to drive. Your success with speakers that have lower impedance may not be as good.

The Lepai 2020A does not have any input switching. Both the RCA and mini jack inputs are live at all times. Sadly they are mixed through a resistor, not buffered. and then mixed together. This means that if two devices are operating, you'll get half the volume of either. Turn one off and the volume doubles.

If you don't intend to head bang, but are looking instead for a solid mini-amp to put under your desk, or in a boat or RV, I highly recommend the Lepai 2020A. 

The amp runs ice cold at all times. The one negative thing is this desire for Asian manufacturers to put blue LED's on every damn thing. In this case it's around the volume control. It's big, bright and blue, so you can forget about sleeping with it on nearby.

Stereophile Hatchets the Crystal Cable Arabesque Minissimo Diamond

This is kind of a follow up on my original article Stereophile - The Data Doesn't Lie and, to be honest, this is a tempest in a teacup as well. It's just one sentence that lit a fire under me.

First, I've never heard the Arabesque Minissimo, nor can I really morally justify the purchase of small home speakers that cost $20,000, whether from Crystal or Magico.

Even then I am nothing if not a hero for truth and justice so it was inevitable that the latest online review from Stereophile would completely piss me off. For the most part, it's a too crappy and short for a pair of speakers with such pedigree.

Stereophile Listening Tests 

The listening section is only one page, and very little devoted to actual music listening. Also, John Atkinson seems to write that he relies on test tones and his ears to tell "tonality" instead of oh, I don't know, music? I mean, your ears are for music, not test tones. If you have the gear, you don't need to be using test tones and aural guessing.

The biggest problem is that by JA's own admission, the speakers could not be placed optimally in his room. Instead of realizing there was significant "operator error" JA goes on to blame the speaker's and claims "treble tailoring" will only play nice with some music.

So what happened? Speaker/reviewer mismatch.

To have the maximum possible sensitivity and simplest crossover it seems that the Minissimo foregoes "baffle-step compensation" or BSC, and requires close wall placement. I have highlighted this region under the line in the chart at the left. This rising bass effect happens as the woofer's output goes from omni-directional to half space, and is related to the woofer and the baffle width. Notice it does not show up in his near field graphs...that's a whole other story of mess too long for this post. 

You might look at that and say "what a crappy speaker!" but you'd be wrong.  There's nothing wrong with designing a speaker for a specific environment, and we need more speakers that don't try to dominate your living room. This kind of design approach can yield exceptional results. Unlike almost all "high end" speakers, the Minissimo's were never designed to be a "one size fits all" model. They are ideal for a salon where discretely beautiful speakers could play wonderfully while leaving space for everything else. Listening to the Minissimo's in the middle of the room will sound like a mosquito orchestra, which seems to be just what happened.

JA rigidly reviewed them as middle-of-the-room stand mounts and blames the speaker when he should be blaming his closed mind and lack of understanding. With these measurements in hand, JA should have stopped or found a better place to listen to the Minissimos, or held of on publishing the review. Instead he attributed the weak response to the "tailored treble." His excuse that his furniture did not allow him to place the speakers correctly is the height of professional whinery. Either move them to the right place, or let some one who can and wants to them properly so. Don't compromise by writing a half-ass review with the speakers in a poor location.

The Bottom Line

The Crystal Cable Minissimo designers took truly refined components with a minimalist design aesthetic and made a high quality, beautiful looking pair that will stay out of your living space. Comparing the Minissimo to the slightly less expensive, but equally insensitive, Magico S1 Mk. II, the S1 requires much more floor space to sound it's best. To me, a small speaker that actually needs so much space is kind of a weird situation. On the other hand, the Minissimo is small (for a speaker with a 6" woofer) and perfectly happy to get cozy to the art on the wall.

With this in mind, if you are a music lover who wants top tier, beautiful speakers made from rare materials that don't take over your room I strongly encourage you to listen to them for yourself and pay no attention to the critic behind the curtain.

Ubuntu Mytek Brooklyn DSD Setup Guide

Update August, 2018:

These instructions are still valid, but I've gotten direct DSD working as well.  Please visit my updated post here:

http://pqltd.blogspot.com/2018/08/mytek-brooklyn-and-manhattan-dsd-with.html

I'm happy to report that the Mytek Brooklyn DAC is not only great sounding but it's a real treat to set up with Ubuntu. Why would you care if you are an OSX or Windows person? Because it's really inexpensive to set up a music server with Ubuntu.  From under $100 Raspberri Pi 3 to my own mini box which was under $500. In all cases, an Ubuntu server competes very well with $2,000 or more dedicated streamers. 

One major feature that is missing from a lot of "audiophile" solutions is Android app support.  This is something we will overcome quite easily.

Also, since this is really a driverless installation, these instructions will probably work for any other modern USB 2.0 DAC that claims to be "driverless" for OSX and Linux. The catch will be in setting the right device (SL_SOUNDCARD, below) but that part is easy.

There is a known limitation. Currently Squeezelite only supports DoP, which is horribly bandwidth inefficient. DoP is DSD over PCM. It is not the same thing as direct or native DSD play.

It's not my code, but I'm working on it.

Introducton to the Logitech Media Server

There are a lot of Lunix music players out there, I'm going to go with the easiest one that gives me bit-perfect playback: Logitech Media Server 7.9 and Squeezelite.  You need at least 7.9 to get DSD player support as it's a fairly recent feature. I think it was introduced in 7.8, but I never tried it.

Originally developed for the highly regarded streamers created by Squeezebox and bought by Logitech, LMS has pretty much transcended Logitech AND Squeezebox and now has a robust and active Open Source development community. It's easy to use, convenient and free, and has Android and iPhone apps available. On Android I use Squeezer but on iPhone you can try the original controller app or SlimLibrary.

The one thing that is missing from LMS is a music management feature. It doesn't tag, organize, or do much else but play music. On the other hand, it does this VERY well and without complications. For this you'll have to find another application of which there are legion.

Also, LMS has fantastic and simple Internet radio discovery and playback. It's a real marvel in this manner for which on Linux I have no better alternative, nor do I want one.

Preparation

First, you must have Xubuntu/Ubuntu 14 or greater set up. I am on 16.04 LTS, but 14.04 LTS was identical.  If you have another Linux distro, see the link to the details, below. I will say in the past I used to run Fedora and LMS was one of the main reasons I quit it, so if that's your distro and this is difficult, you have my sympathies. May the gods have mercy on your heathen soul.

Decide where your music and playlists are going to be. I prefer to put public media in /opt, so:

cd /opt
sudo mkdir Music Playlist log
sudo chown bob:bob *

Replace "bob" with your user account name.  Notice we are creating a log directory as well.  This is where any messages from the player (squeezelite) will be put.

Server Installation

First, install LMS from the public repo.  There are only nightly builds but seem pretty stable. You'll find detailed instructions here.

I used this script:

url="http://www.mysqueezebox.com/update/?version=7.9.0&revision=1&geturl=1&os=deb"
latest_lms=$(wget -q -O - "$url")
mkdir -p ~/sources
cd ~/sources
wget $latest_lms
lms_deb=${latest_lms##*/}
sudo dpkg -i $lms_deb

Mine is just a little different because I don't like cluttering up the root directory. This will look for the latest build, download to ~/source and install it.

After it's installed LMS should be running and available on port 9000. So open up firefox and click here:

http://localhost:9000

Follow instructions on the screen.  You'll need to setup your directories. Click on the lower right of the page, "Settings." and setup your music and playlist folders. The only tricky part left in LMS now is to enable the DSD player.  Click on the "Plugins" tab and enable "DSDPlayer."

Very cool part of this is that DSD files will be automatically transcoded if your player doesn't support DSD directly. I have a Squeezebox Touch that does not, so that gets transcoded to 88.2kHz PCM automatically for me.

Make sure your music and playlist folders are universally readable.

Player Installation

OK, that's it for the server.  Now we need to get your PC to talk to the DAC. This is even easier since the Ubuntu distribution includes squeezelite. Let's install it:

sudo apt-get install squeezelite

I've tried the distro version with Ubuntu 14 and 16, and both work great with the Mytek Brooklyn. DSD and HiRes PCM played without a hitch on both so don't feel put out if you don't have the latest version. It's not worth the trouble to get the nightly or the source and build.

Make sure your DAC is now plugged in and turned on.

You'll find the configuration file for Squeezelite in /etc/default/squeezelite.

Here is how I configured it:

# The name for the squeezelite player:

SL_NAME="Brooklyn_DAC"

# ALSA output device:
SL_SOUNDCARD=""hw:CARD=DAC,DEV=0"

# Squeezebox server (Logitech Media Server):
# Uncomment the next line if you want to point squeezelite at the IP address of
# your squeezebox server. This is usually unnecessary as the server is
# automatically discovered.
#SB_SERVER_IP="192.168.x.y"

# Additional options to pass to squeezelite:
SB_EXTRA_ARGS="-D -f /opt/log/squeezelite.log -d all=info"

# As above, but without the log file

# SB_EXTRA_ARGS="-D"

The key part is the SL_SOUNDCARD setting. If you aren't using the Brooklyn, I'm not sure this will be right.  I found it by using "squeezelite -l" and looking for the entry that said :

USB Audio - Direct hardware device without any conversions

There were other entries that worked, but I found they would convert whatever I played to 48/32, no matter the original.

Another important switch is -D which enables playing of DSD without transcoding to PCM.  Without this switch your DAC will always get PCM.

You'll see two versions of SB_EXTRA_ARGS. The first adds a log file, but really once you've got it all working it's not very useful. Once you are happy it's working comment it, and uncomment the last line, then restart.

One last command and you should be running:

sudo service squeezelite restart

Assuming all went well, you should see "Brooklyn_DAC" (set by SL_NAME, above) appear as an available player in LMS on the top right.  Select it and play some music! If you have music but it's not showing up you need to scan your library. Click on the bottom right "Settings" and in the "Basic" tab check the media folder is correct, then click "Rescan" and your files should get imported quickly.

If you need to free up your DAC for movies, or games use the player OFF switch in the LMS screen on the top right.

Now of course it is time to download some good sounding DSD music, for which I highly recommend Blue Coast Records.


Happy Listening!

Meridian Quality Authenticated - It's Snake Oil, until it's not

MQA 

Update: After a lot of reading, MQA is NOT lossless. What a shame.
A full treatment is available from this post at Benchmark Media

In fact, it's such a good and thorough post that I'm going to delete the rest of my original posting and let it speak for itself. 

I will also add that you can listen to MQA files and compare them to the originals via the 2L test site here. 

After listening for a while with a Mytek Brooklyn DAC I can honestly say I hear no difference between an MQA file which claims 384k to the "merely" PCM 96k/24 bit files. 

Wyred4Sound Remedy - First Listening Impressions

I have hooked up the Remedy between my Logitech Squeezebox and my Audio Research DAC 8.  The only true jitter measurements I could find were for it's slightly newer cousin, the DSPre.  It seems that the DSPre is wildly sensitive to jitter, at least the measurements are, so I'm just going to assume the DAC 8 is the same or worse.

Most of my listening these days is to Internet radio, including Toronto Jazz FM 91, which in addition to having great programming also streams at 24/96.   The other station was KDFC 90.1, Bay Area Classical.

In the middle of this my power regulator has started to hum, so I can't do the remedy justice until I move it to a quieter location.

Still, here's what I think so far.

The remedy works much more noticeably with low-resolution stations, but it IS better. The sense of space inside the sound stage and the treble decay. Sometimes it feels worse though. During mass string crescendos the sound gets too complicated.  With Jazz 91 the improvements seem much less pronounced, but still there.

However this is really hard to gauge with radio.  I'll give this a better listen soon, when my biggest noise sources have been fixed.


Update August 28, 2016
So I had thought my impressions might greatly change, but they did not. I think this is a good tool but only for more source sensitive DAC's like I had (just sold it), the ARC DAC 8. That DAC played brilliantly when driven by an Ayre CD player, but when I brought it home it was much fussier.

With high resolution music (96k and above) I could honestly not tell if it was working or not, which I guess is a good thing. So, overall I would recommend this to clean up the sound of a mid-Fi CD player or Internet radio or inexpensive streamer like my Squeezebox Touch. 

I have however switched over to a Mytek Brooklyn DAC which sounds as good as the Wyred4Sound + ARC DAC 8 without the remedy in place, regardless of how it was driven.

I'm now driving the Brooklyn with a 2TB Linux streamer I built myself ($650) and it's very happy to play PCM, MQA and DSD from it.

Digital Audio - Upsampling and Oversampling Explained

Many types of digital sources, accessories and Digital to Analog Converters (DACs) provide some sort of sample data magic called oversampling or upsampling.  Put simply it means you end up with more digital data than you started with.

There are some benefits, but none of these methods truly gets you closer to the original music. They are all just ways of trying to make the experience more pleasant. Think of it as looking out your window with a screen. You may take a picture and find that you can see the screen itself in the image, or you can do some editing with Gimp or PhotoShop and remove it. The new image can't possibly contain more true to life data than you started with, but the picture should be much more pleasant to look at.

Many audiophiles have been led to believe that this kind of digital math can do things like you might see on the TV shows CSI or NCIS. Somehow four pixels on a grainy satellite image can be processed over and over again until the criminal's face is clearly visible. It's just not true.

Looking at it another way, the frequency response of up and oversampling does not change. A 44.1 kHz file is not going to have 30kHz created after 4x upsampling.  The frequency range and content density is unchanged. What may happen is that digital filtering becomes smoother and easier on the ears, or that jitter is improved somewhat by the use of higher data rates.

Differences Explained

Let's take original data.  Since digital music is always integer, I'll imagine two consecutive samples with convenient values of 24 and 28. Now lets see what happens at 4x up or oversampling. If the original data was 44kHz/16 bits the DAC will now see a sample rate of 176.4 kHz but the bit depth may or may not. So, just to be thorough, here is our original data:

• 24
• 28

at a sample rate of 44.1 kHz these two samples represent:

• 2 samples / 44,100  = 45 microseconds of music. 

 Remember that we are adding samples in between the time slots, so we don't want to stretch out our time, that would result in pitch changes. Instead we increase the rate (samples / second) at which we feed the DAC, keeping the pitch constant.
So, instead of 2 samples, we have 8, but with a new sample rate.  Lets redo the math:

• 8 samples / 176.4kHz = 45 microseconds of music.

Thhat's great, because if that didn't work the sound would be 4 times slower. :)

Oversampling

This is the oldest trick in the book. Almost immediately after CD players became commercially available oversampling became a buzz-word. I am no longer sure, but this may have only worked with so-called Delta-Sigma or 1-bit DAC's.

It's so simple you don't think it should work. Take a sample, and repeat it several times. It's that simple. It does not attempt to provide any more data but may shift some noise far above the Nyquist frequency.  No math is involved, just counting.  With 4x oversampling the DAC our orignal two samples become:

• 24
• 24
• 24
• 24
• 28
• 28
• 28
• 28

It's weird it helps, but it does. In fact, with oversampling, only 1 sample really matters at a time.

Upsampling

Bit Perfection

One of the objections to upsampling, is that the signal is no longer bit-perfect.  The DAC no longer gets the original facts, but the original facts, plus a lot more. That "lots more" is pure mathematical conjecture. However, there are some real benefits to be had.

Things get even more muddled when upsampling is used for ASRC, Asynchronous Sample Rate Conversion, but also more beneficial, as it's one of the best ways to reduce jitter.  More on that in a future post.

Technically and mathematically more challenging, there are two general approaches. To take the best advantage of this it's better if the bit depth increases beyond the original. So if the original was 16 bits, 24 or 32 bits will provide better resolution.  However remember that this doesn't really make it more true to life. It just makes some things easier to do and helps us keep more of our results. There are some VERY nice 32 bit DAC chips out there though, so taking full advantage of them may also get us much closer to true 24 bit resolution. That's a topic for someone else.

Linear Interpolation

Imagine two points on a chart. Draw a straight line between them. That's simple interpolation. It's no more complicated than simple algebra. Calculate the rise, divide it by the number of intervening samples, and add that much for each "new" sample. For linear interpolation, the sample rate converter needs to know two samples at a time in order to figure out the rate at which the intermediate samples should change.


Again, consider our original two samples, 24 and 28. The rate of change is 4/sample.  4/4 = 1. Now the DAC gets:

• 24 +1 =
• 25 +1 =
• 26 +1 =
• 27 +1 =
• 28

We'll just assume there's no bit-depth changes, or that in this case no extra resolution was required. Of course, I chose 24 and 28 to make the math here easy.

Spline

A much more advanced way to create more samples is by using what are called splines. Remember the "French Curve" tools you may have used in drawing school?

Technically you only need 2 samples for a spline, but the result is the same as linear interpolation, so we'll ignore that case. With spline math we take a number of samples, usually under 20,  to draw a much softer curve. Wadia was the first company I know of who introduced this concept. In this case it really helps to have more bits, as the extra bits help with more fine grained results. As you might imagine, the math and CPU power required is greatest for this example.

If this was floating point math our working data set would be:

• (nine samples before)
• 24.000
• 25.185
• 26.355
• 27.888
• 28.000
• (nine  samples after)

Remember that what's really going on is that the algorithm is taking more samples into account than our original two in order to fit the curve properly.  So why the third sample is 27.888 instead of 27.978 or 26.500 has to do with the nine samples in the original file before the first (24) and after the last (28) shown here. It is believed, without a lot of proof, that this method may provide the most natural resulting sound.

Are Splines Really Better?

Splines are very cool, but it may be argued, convincingly, that we are not doing much more than you could achieve with a capacitor and resistor with the proper time constants. In other words, it's a lot of math and hardware for what could be done with $2 or less in parts. The real potential benefit of this advanced though is in custom algorithms. You can be as creative as you want to in your algorithms.

What About Sound Quality? 

Personally I have come to believe that the analog output stages matter much more than interpolating algorithms and sample rates or bit depth but the devil is in the implementation details. As always, buy what you like, and what is most pleasant to your ears. Don't buy algorithms or chips. Buy results, and spend no money that isn't pleasing to you.

Wyred4Sound Remedy - Snake Oil or a True Panacea?

I've just ordered a Wyred4Sound Remedy.  I had been drinking and needed a pick-me-up and based on Digital Audio Review's positive impressions I ordered it.

What I have just realized however is that the Remedy is not the product I thought it was in a couple of ways.

Mind you, it's clear that for some kinds of low-grade digital audio sources the Remedy is probably a very good solution. I'm thinking of Sonos, Apple TV and Chromecast specifically. However, it is a terribly over-hyped product which smells of snake oil.

What Kind of Product is the Remedy? 

Of course, marketing people, being devoid of souls at birth, are free to call a tomato a vacuum cleaner and there's rarely any legal consequences. In my world however the Remedy should properly be called a sample rate converter (SRC) or Asynchronous Sample Rate Converter, a feature built into many of the top DAC chips today.  SRCs always includes re-clocking, so calling it an SRC with reclocker is redundant.

Using an ASRC is a very good way (and a little lazy by itself) to ensure minimal jitter with possibly very jittery sources, such as Internet radio.  The reason I'm kind of on the fence about this is that an ASRC is no longer bit perfect, but time-perfect. To ensure that every x picoseconds a new sample is processed, regardless of how the input signal may vary in the short and long term an ASRC resorts to a mathematical brute-force method, the details of which are beyond the scope of this posting.  Suffice it to say you can kiss bit-perfection goodbye, and not just for the interpolated samples either.

One major annoyance, that relegates the Remedy to mid-fi sources is that the input signal is ALWAYS recreated. What's worse is that if you have music with a higher sample rate, such as  24/192kHz, the Remedy will actually DOWN-sample it to 24/96kHz. This, plus having no input switching makes his solution seem kind of dopey.

For about three times more a more robust option is the Mytek Stereo 192 SRC. It gives you much better control over what you want to do with the signal, up, down or no change with equivalent or better jitter reduction, as well as being able to convert up to 24/192kHz

Of course, this is all spec-manship. Listening is the true arbiter of what you should buy.

Is it really a femto-clock?

Having examined one, it seems W4S has used one of the best commonly available Crystek oscillators, the CCHD-957 series, which does in fact have very good phase noise characteristics, among those affordable to mere mortals.  I'm still confused though. Based on W4S's own measurements this does not appear to be a femto-clock grade solution but a pico clock. What's the difference? About 1,000 times worse performance. It is possible that the internal clock device inside the case is a femto-clock class part, but that the other circuitry used can't take full advantage of it, or that it can only do so much in one pass. It would be very interesting to see measured comparisons using a standard Mac Mini or Apple TV to see how it measures to Mytek, M2Tech or Auralic with and without.

Consider this. The Auralic Vega with a true femto clock (and 10x more expensive) has jitter around 80 femto seconds. The image on W4S's own Remedy page shows jitter around 87 pico seconds. That's about 1,000 times worse performance. Of course, many would argue that you can't hear 80 pico seconds of jitter, but the point is the marketing hype. I don't like being lied to or misled.

Another similar device with a price point kind of in between is the M2Tech HiFace Evo 2. It is intended as a USB to SPDIF interface, but it will also take a coaxial SPDIF as an iput and allow you to select sample rate conversion.  Price is around $700 USD.

Why does this matter? 

My point to all of this is that the Remedy is doing more than just jitter reduction, and I would really have liked to know this before I ordered. Hiding major behavior is not a sign of a trustworthy vendor. Remedy is playing with the bits and I should have known that first. For more on why this is different, see my post on Upsampling and Oversampling.

It's a little odd as many DAC's make upsampling a key feature. They charge more for it and often tout their proprietary algorithms as being better one way or another.

In the end though I'll have to listen to it to evaluate the Remedy as having any sort of meaningful benefit. More on that in the next several weeks. One of my sources however will be Toronto Jazz 91 which streams at 24/96kHz already, this will let me do direct comparison to see if the jitter reduction is worthwhile, in addition to a small selection of 24/96kHz FLAC recordings I have.

Many listeners are easily swayed by "different." Even John Coltrane suffered this, always thinking his next performance was better. If jitter or SRC at these levels is audible it's quite possible many will be swayed by a different sound, but not necessarily a better sound.  I can imagine many will get worse jitter than they started with, and then proclaim how audible and beneficent the differences are!

So your CPAP treatment isn't working?

I've gone through seven or eight sleep studies. I've lost count of exactly how many, but the last three at the same sleep lab, without my symptoms being fully improved. I would estimate that at the beginning they were about 20-30% better, but years later seemed to fail completely.

What I eventually discovered is that exercise actually made my sleep quality much worse. If this sounds like  you, it might, then it is possible we share an odd, and previously unreported condition. The only other two conditions which I've read that can cause this is Cushing's Syndrome and Chronic Fatigue.  However, sleep apnea that is not fully treated can mimic both of these, including elevated cortisol levels. So if you have noticed, or are able to determine if exercise is bad for you, then we may share a condition, which is treatable!

The problem in my case only was that I have two modes of sleeping. The lazy day, fully rested mode, and the "I've worked out" mode.

Why didn't the previous sleep studies help?

The previous sleep studies did find a partial solution.

I didn't notice the problem at the time, but looking back it seems that since I slept poorly after exercising I learned not to exercise, even a little. I stopped taking the stairs and always chose an elvator, had groceries and food delivered, anything I could do to avoid physical labor. That is, my symptoms conditioned me not to exercise, so I would go get a study done after a couple of days of being fully rested, and as a result my full symptoms did not occur. I only know this after-the-fact because of the use of a Jawbone UP (lasted 3 months, but did the job) which allowed me to discover exactly when I had good and bad days of sleep.


After using the Jawbone for several months looking for clues to my problem I was able to completely correlate exercise with poor sleep the next several nights. This is the exact opposite of what should happen for most people but for me was the wall I could not overcome.

Confounds

Moderate exercise (5,000 steps / day) would not cause this problem to occur immediately.  It would sneak up on me.  I needed it to be vigorous, or weight bearing to cause the symptoms to show up faster.

Another thing that seems to trigger this effect is calcium channel blocking medications, or medications which are related. These can be prescribed for a range of issues from depression to high blood pressure.  The good news is that if this is you then your sleep studies should show up without having to go through the exercise portion of this experiment. 

Further, the post-exercise crashes would last for days, sometimes even a week.

How can I tell if this fits me?

If you have a Jawbone, FitBit or ResMed S+ you may see your deep sleep very disturbed, or lots of wakings, despite the CPAP measuring otherwise low AHI numbers and tolerable leaks.

In the picture on the left I share an image from the Jawbone UP application showing my very worst night of sleep.  See how many different periods I have of deep and light sleep?  There should be about 8 of them.  Instead I have about 17.  Also notice the time I slept. Over 11 hours! That's because the sleep quality was so poor.

You should have big chunks, around 6-9 of deep sleep/light sleep cycles with no awakenings. 

Be Your Own Guinea Pig

If you want to investigate on your own first, try to correlate it over a couple of weekends.  One weekend take it completely easy. Do no physical labor for two days, see how well you sleep Sunday night. If weekends are your housework days, hire a cleaning service this once. If you are already in a disturbed sleep cycle though you may need to do this for several days until your symptoms clear.  Only after you are in a "good" cycle should you attempt to disturb your sleep with exercise.

The next weekend, attempt to exercise normally both days. I don't mean do your normal routine, I mean work out 30-45 minutes each day, see how much of a difference this makes to your sleep Sunday night. If you find that your first Monday was great, and your second Monday terrible, you fit this profile.  However, there are many unanswered questions that remain, and I've noticed in addition to exercise a secondary, cyclical pattern which I have yet to pin down.

Notice that the ResMed S+ was less sensitive for me than the Jawbone for this particular issue.

How can I get better?

Schedule another sleep study, but this time make sure you exercise normally the day of the study, if not two days before the study. In my case this was really a challenge, since my symptoms had been getting worse, my energy levels and moods had been suffering, causing me to exercise less and less.

It really helps to hire house keepers and have friends to support you.  Let them know what you are up to and that you'll need extra help while you try this. Take time off from work if needed. If you are like me, you'll crash for 3-5 days and be completely unproductive after the exercise portions.

It's also important to stress that immediately after exercising I felt great. My lungs felt clear, I felt light on my feet, it was such a great thing. It was only after I slept and failed to recover that symptoms appear.

Tips:

Remember that exercise IS recommended as part of almost every sleep improvement regimen, so you aren't breaking any rules by following the advice here. In all cases follow the advice of your doctors and personal trainers, but the guidelines here should be in compliance with them.

Right now as far as I know, there is no guidance on exercise before a sleep study/titration, however it's generally recommended not to exercise in the late afternoon evening as it can wake you up and prevent you from sleeping fully. 

The day before the sleep study:

• Have no alcohol.  Zero.
• Stop caffeine and chocolate after noon. 
• Have a couple of bananas.  

The bananas and stretching do nothing for sleep, they are to help prevent leg cramps, skewing your study, and possibly disconnecting your wires. If you aren't used to exercising, a little extra potassium will help.  Any other potassium-rich food will do of course, if you are allergic, or just don't like them.

The day of the sleep study:

• Have no alcohol, Zero, nada, zilch. 
• Exercise as much as you should be able to do if you had no sleep issues.
• Exercise vigorously 30-45 minutes in the morning if possible, but no later than early afternoon. Get into your aerobic zone for most of it. Again, be an adult and follow the advice of your doctors and qualified exercise instructors. For me, it seems that weight bearing causes the symptoms faster than aerobic exercise, so a few squats to tire your legs out may be all you need. Climb stairs with a couple of gallons of water in each hand should do the trick as well if a gym is out of your reach.
• Make sure to cool down and stretch your legs.
• Have a couple of bananas after your workout.

You may be desperate for an answer, but if you are like me, this doesn't take THAT much effort to tip your body over, so don't give yourself a heart attack before your sleep lab!

In my case, it was like magic, symptoms that had never shown up before became clearly and consistently visible during the lab work. The CPAP machine that was fine for the first 6 studies was completely inadequate now.


By challenging your body, and your CPAP machine you'll have the best study possible.

If you are a sleep specialist, or researcher, and would like detailed information I would be willing to make that available privately.  Please leave me a comment with information about your publicly visible contact information and I'll reach out to you.