Summer Update! Huawei Matebook X Pro, Western Digital Gold HDD, and jitter simulation.

Well guys and gals, hope you're having a great summer (or winter as the case may be!). The weather here is certainly conducive to taking it easy and just enjoying the music instead of thinking much about the audiophile hobby apart from short comment/forum posts.

But tech things are happening in the background :-). Really exciting things as well which I'll discuss more in the days ahead. For today, I just wanted to discuss a couple changes / acquisitions this month...

First, I was rudely reminded the other week of the importance of making backups yet again... In early 2016, I mentioned that I upgraded my music library to Western Digital Red 6TB drives. They worked pretty well for the last 2-and-a-half years but sadly one of them decided to die on me the other day. It was literally a "hard" crash with sudden loud noises and no access to the data at all! Not much if any  SMART warning (I checked a month earlier). Thankfully it's still within the 3 year warranty so I mailed the device with an RMA to get a replacement. That's all well and good to get a replacement but if I did not have a backup, it would have meant a loss of 3TB of music library data including quite a number of vinyl rips which took some time to digitize and process!

Remember folks... Backup, BackUp, BACKUP! At least 2 copies which is what I usually do, but best to have 3 copies, one stored elsewhere just in case the home burned down for very important data.

Since I bought two WD Reds in 2016, the second device seems to be good so far - fingers crossed. Surprisingly the one that failed is the one that's kept in a cooler environment in the basement and accessed just as much as the other so I assume it must have just been defective. In the place of the WD Red, I decided to give the Western Digital Gold 6TB a try (~US$250). This is Western Digital's 7200rpm "Enterprise Class" hard drive which is supposed to be extremely reliable and meant for business servers and datacenters with 2.5M hours MTBF! Seriously, I'll believe it when I see it. I'd certainly be very impressed if it can run for the next 10 years without issue... Will keep you guys updated :-).

Just a note about the WD Gold. This drive runs at 7200rpm compared to the Red's 5400rpm. So it does take up a little more power, runs warmer but faster as well. It's also a tad noisier than the Red I had in there. The spinning drive doesn't seem any noisier, but head seek noise is more noticeable (ticking and clicking sounds) so best to not have this drive in the sound room while listening to music. I didn't run any benchmarks since the absolute speed isn't that important to me. However, I did notice that doing a full music library re-scan with Logitech Media Server was a few minutes quicker.

The other change recently is that I upgraded my "workhorse" laptop to the Huawei Matebook X Pro Signature Edition (quite a mouthful) as shown above. I went for the higher-end 16GB RAM, 512GB SSD, i7 model - you should be able to find a lower street price than the MSRP ~$1500. This is an upgrade to my Microsoft Surface Pro 3 from 2014 which has served me very well for the last 4 years but I was running out of space on the 256GB SSD and had a desire for better battery life and speed when on the road and wanting to do some more CPU-intensive tasks like image processing.

Check out this picture of the 14" Matebook X Pro (left) sitting beside the 12" 2015 MacBook (right):

As Charles Caleb Colton said a couple hundred years ago: "Imitation is the sincerest form of flattery". From the name ("Matebook" vs. "MacBook"), to the "space gray" color, to the general layout of the keyboard, about the same thinness (I think it's even thinner than the MacBook but at fractions of millimeters, I really don't care at this point!), to the size of the trackpad, this machine looks, feels, and acts in many ways like the MacBook and MacBook Pro. But better because it's not an Apple :-).

Okay, no need to get Apple fanboys on my back. As I've said before, due to work requirements, Windows is my preferred OS these days and I've become "all thumbs" when trying to do even the simplest things on Mac OS X. My wife is still an Apple faithful so we're an agnostic family...

Anyhow, as I was saying. The Matebook X Pro is like a MacBook Pro but better. More ports with 2 USB-C on the left (one of which Thunderbolt 3), a full sized USB-A 3.0 on the right. No doubt we're seeing the transition to USB-C as the universal power and data port for mobile computing and technology. It is rather impressive that such a small form factor is able to transmit at such speeds - Thunderbolt 3 over USB-C is capable of 40Gb/s! Good that the USB-C physical connector is symmetrical so no need to fiddle with orientation, but for power charging, it's not as easy as the light touch magnetic connector used in devices like the Microsoft Surface Pro.

The screen is a 13.9" 3000x2000 LTPS display with ~100% sRGB color gamut and up to 450 nits brightness - great for productivity and photography but maybe not so much for 16:9 movies. Screen to body area is a stunning 91%. It is a touchscreen as well. Easily "Retina" resolution at 257ppi for a laptop (the upcoming 2018 15" MacBook Pro remains at 2880x1800 - still great of course). The screen is as bright as the MacBook Pro and crisper. I typically don't go beyond 60% brightness. There's a biometric finger scanner built into the power button which is fast and accurate, providing one-touch turn on and secure login; excellent. The round-edged aluminum feels solid in hand, and the machine runs quietly with a soft fan. The 8th Generation i7-8550U 4-core/8-thread processor at 1.8GHz is fast. I'm seeing easily 9+ hours of battery life. Since I don't play games on this thing, I usually turn off the nVidia MX150 discrete graphics with 2GB video RAM and even set my max CPU speed to 80% on battery in my Windows power settings which will keep it even quieter and with better battery life.

By the way, one interesting "feature" (or hindrance depending on your needs) in order to keep the screen bezel very thin is that they moved the camera to the top row of function keys on the keyboard:

Notice the camera popped up in the middle of the function key row. You can also see the USB-A connector near the hinge.

Interesting actually. For work, I rarely use the camera so this is not an issue. In order to capture the face properly, one has to sit a little further back and keep the hands off the keyboard. This could be a little problematic for those who video chat a lot although on the plus side, it supposedly has 4 microphones for high quality voice pickup. The camera only needs to pop up when in use and it also has a little white LED to indicate when on, so that's also good for privacy. I actually like this design as a professional user with little need for the webcam.

Speaking of the keyboard, it's certainly better than the low-travel (<1mm) recent MacBook and MacBook Pro "butterfly" switch keyboards. The Matebook X Pro has a key travel of 1.2mm which is of course a far cry from the typing pleasure and accuracy of my HyperX Alloy Elite mechanical keyboard but still has a good feel coming from the slightly mushy Microsoft Surface Pro Type Cover.

It is unfortunate that the machine comes only with Windows 10 Home rather than Windows 10 Pro. At least Huawei doesn't install all kinds of bloatware on the OS like McAfee. They also provided a USB-C dongle for HDMI, VGA, USB-A, and USB-C:

Huawei accessories: USB-C dongle for HDMI and VGA output as well as USB-C and USB-A (behind), power supply for charging on the right.

I remember having to purchase a USB-C hub separately for the MacBook.

As for the audio of the Matebook X Pro, you can't of course expect too much from laptop speakers. Having said this it does sound fine though not the loudest built-in speakers I have heard. There's also a 3.5mm headphone jack on the left side near the hinge - notice the 2 USB-C ports:

Curiously, there's a "Dolby Atmos Sound System" moniker attached to the speaker system. Apparently Dolby was involved in the speakers and placement. There's also a sound control scheme which looks like some EQ changes and such:

As you can see on the left are the "Music profiles" with a few EQ images you can choose from. And on the right are the actual custom settings you can adjust for yourself based on a 10-band EQ with a few DSP settings like "Surround enhancement".

Overall it does sound pretty good. There is a good amount of perceived stereo separation and in some videos I watched, the illusion of depth in the sound field seemed very nicely enhanced. For example, this YouTube video of binaural audio envelops you nicely when seated in front of the computer with the illusion of sound around and behind! No... This is not the same as Dolby Atmos as witnessed in movie theaters with speakers overhead and around. As expected, with 4 speakers located in the thin strips on either side of the keyboard (like a MacBook Pro), you're not going to hear loud, thundering, accurate bass. At high volumes, distortion is audible. But as far as small speakers on a laptop go, this is good.

As usual, at some point I'll probably measure the audio output from the 3.5mm jack to examine the sound resolution and weather a modern "ultrabook" laptop like this can handle hi-res audio.

After running the laptop for about a month now, I'm not surprised that The Verge is calling this machine the "best laptop right now". Looks like Apple is coming out with their new generation of 2018 MacBook Pros soon - some very impressive benchmark numbers (for a price of course!). Like many other areas of consumer technology, we're basically at the point I think where most of us are satisfied with the speed, storage space, size, lower fan noise, and functionality of today's laptops. I guess one could never have enough battery life. There was a time when I would be updating the laptop every couple of years due to one thing or another. Thankfully, those days are gone and I suspect that other than the machine breaking down, I'd be quite happy with this for a number of years.

Apple - take note of what your competitor is doing at a significantly lower price point. I'm thinking after some negative press against Huawei and communications security, they're being very aggressive in pricing. From my perspective, essentially everything high-tech these days is made in China anyway - including Apple products of course. Here in Canada, currently I would not be able to buy a bare-bones basic MacBook with half the amount of RAM, half the SSD storage, no discrete GPU, previous generation CPU for the price of this Huawei! Also, Apple, please stop nickle-and-diming the consumer for IMO essential accessories like dongles for HDMI or VGA through the USB-C connector. Finally, don't expect that there will always be fanboys to be there to support decisions made based on "form over function" and a "locked in" ecosystem. Innovation is the hallmark of technological progress and I suspect Apple's innovation curve has very markedly flattened over the last few years.

-------------------------------------

Before ending, let's talk a little about audiophile stuff...

I recently came across an interesting and powerful piece of software. Check out Yamamoto2002's WWAudioFilter. Among other features, one of the cool things you can do with it is to simulate a form of periodic jitter with an input audio file. For example, here's the addition of 1ns of 225Hz sinusoidal jitter. Click on "Use This Filter" in the "Add Jitter" box to add to the list of processing you want to do:

Here's what happens to the 16-bit Dunn J-Test when you add 1ns of simulated sinusoidal 225Hz jitter played through WavSpectra (131072 bins examining 5kHz to 18kHz) which I typically use to analyze the equipment I test:

Notice the addition of the -100dB sidebands.

From what I've seen, 1 nanosecond of periodic jitter at a specific frequency creates more sideband anomaly than what is a typical, good digital interfaces these days; especially decent asynchronous ones like USB and ethernet. In fact, here's the old Chromecast Audio's digital output through S/PDIF optical using a generic TosLink cable tested a few years back:

Far from a terrible jitter result for a <US$50 low-priced consumer device that does not advertise any special temporal accuracy, right?

So, you might be thinking... "Hmmm, that 1ns amount seems pretty bad! I bet it's audible!"

Not so fast :-).

By all means, run some music through WWAudioFilter using the settings above and listen for yourself. Even better, try using foobar's ABX plugin to compare the original audio track and one with jitter anomaly added.

Remember, you can also stack a number of the jitter simulation settings together. For example, here's one where we add 2ns of simulated 225Hz sinusoidal jitter + 1ns at 1125Hz + 0.5ns at 2475Hz to create sidebands that look like this with the J-Test signal:

Over the years you might have heard/seen other jitter "simulations" - for example, listen to Stereophile Test CD 2 track 26:

Notice how unrealistic this "simulation" is even for 1992! This is a huge amount of periodic jitter creating these -80dB sidebands with the primary signal down at -20dB. Remember that typical estimates for jitter with good CD players and DACs would be in the tens or maybe hundreds of picoseconds. Notice that the strongest and most proximal set of sidebands in the image above is +/-4kHz from 11kHz such that the FFT was expanded to cover the full 0 to 22kHz! I've yet to see/hear such a thing even with very cheap equipment, perhaps Stereophile chose to do this to accentuate the importance of jitter despite the obvious exaggeration. With any half decent audio system, it would not be hard to hear the overlay of tones in that "simulation" track. In fact, over the years, I have only come across either very cheap or very expensive >US$100,000 systems where it became difficult to hear the effect of those sidebands suggesting very poor fidelity of these systems (the most memorable was a system consisting of highly regarded Japanese "hand made" preamp and tube monoblocks with Avantgarde horns)! IMO, this says something about the actual fidelity of some of the very expensive stuff out there created by small boutique companies. Of course, you'd never hear subjective-only reviewers speak of such a thing when wowed by components and sound systems with an asking price in the 6-figures, right?

Remember also, with real music, you will need to add much more jitter than the equivalent of a few nanoseconds sinusoidal jitter to significantly hear audio degradation (as suggested above, go test for yourself)! Actual recordings are noisier than test signals of course and the ebb, flow, and dynamics of the music will make it difficult to "lock on" to evidence of the distortion from typically miniscule jitter effects such that even if present, it's not likely going to ultimately affect the emotional impact.

If you guys want, I can put together some sample files with varying amounts of simulated jitter for you to listen to. Just let me know in the comments.

If you missed it the first time, here's how an engineer, golden ear, and magician describes the sound of the dreaded jitter:



Apparently magician-engineers have no need to show evidence with objective demonstrations - words and statements of opinion are good enough. After all, "we all know that we don't like it", and apparently toes don't tap as much when jitter is present. As for hearing jitter from "early computer systems", I can only imagine he must be talking about computers from before Y2K and GHz clockspeeds. Anyone disturbed by jitter with even US$50 Raspberry Pi's these days? He uses even more words and talks about rubidium clocks in another video apparently fearing that he will get flak for saying one doesn't need rubidium (vis-à-vis Antelope Audio perhaps)!

Thank goodness, in the last few years, I seem to be noticing less anxiety around jitter performance among the forums and with reasonable audiophiles (the video above was from 2014). Could it be that audiophiles are recognizing that in digital audio, the temporal accuracy of our gear has already way outstripped the ear/brain's ability to detect temporal errors? Could it be that the reason why most people don't know what "jitter sounds like" is because listening for jitter is akin to looking for giant pandas in the wild? Sure, they're out there but obviously very rare and you need to know where to look! (For more chatter about jitter, see here and here among other articles over the years in this blog.)

Maybe the "high-end" Industry will continue in their never-ending claims of ever lower jitter as relevant to the listening experience. And audio magazine writers will continue to blame jitter for all kinds of faults they believe they hear. At some point, perhaps the audiophile community will realize that certain concerns like jitter have actually become obsolete with almost any normal-priced consumer device over the last decade.

Okay... Back to work and trying to catch some R&R over July and August.

Enjoy the music! :-)

Addendum - July 22, 2018:
Like many things in whatever field we might explore, I was reminded in the comments and also through E-mail of the complexity that one could delve into with jitter and all the variables involved! Jitter is timing anomaly and there are various types. Remember that what I'm showing above with WWAudioFilter is periodic sinusoidal jitter that the program can simulate to create the frequency sideband distortions shown on the FFT.

While this is but one of the types (another type would be random jitter which can create the "skirting" at the base of a primary signal on the J-Test), it is one we often run into. To have an easy way to create the effect is nice for the sake of running music through and hearing the distortion ourselves and deciding whether we're "worried" about audible degradation when found in testing. Great work Yamamoto2002!

One of the problems with jitter is that audibility depends very much on which frequencies are affected (since human hearing acuity is not the same through the audioband), the level of the distortion of course, and as well the shape of the distortion (Is it a specific or broad frequency effect? Does it appear as side bands or spurious noise?). As such while we can plug a number like "1ns" into the DSP program to create the distortion sidebands at a specific frequency, obviously the same "1ns" equivalent will have different levels of audibility depending on the frequency / amount / shape of the distortion. Therefore, there is no single "standard 1ns jitter sound". It all depends on amount and ultimate effect which could be complex especially on devices which may show idiosyncratic "data dependent jitter".

Yes... One can get very complex about jitter and all the manifestations. But the point remains that in modern devices, the likelihood that we can hear jitter anomalies these days with good DAC hardware would be rather unlikely although I was told that R-2R DACs (which some audiophiles seem to be very interested in) are more prone to jitter anomalies. Hmmm, maybe I'll have a look at this in the future.