Yearly Archives: 2019

Digital audio reboot

It costs a bit, but it takes a heck of a DAC to get me to like digital music.  I admired some of its attributes, but there were others I could never get around.  After using typical mass market CD players, I bought the Pioneer Elite DV-45A due to its ability to play back DVD-Audio and SACD; I bought a DV-578A as a backup.  “This is OK, but…” was one of those phrases that ran through my mind.  Higher resolution recordings did sound somewhat better, but again, digititis would set in and I’d be turning the volume way down or shutting the system off.

The Oppo BDP-105 was a big improvement–it has a wide and deep soundstage in comparison to the abysmal DV-45A, and lost some of that harsh digital edge and having a more full-bodied sound.  I lived with its sound for a few years with very few sonic issues, except that I was using it primarily as a network player to access FLAC files on my server.  It worked, but here were issues with the playback.  No gapless.  No way to stream DSD from the server over the network.  And JRiver was (and still does) crash often regardless of which computer it runs on, making the experience less than satisfactory.  Oppo’s own Media Control app was adequate.

I considered adding a streamer like the microRendu or ultraRendu, but I still wouldn’t be able to access the 300+ ripped SACDs on my server.

I upped my game and have a PS Audio DirectStream Jr. in house. I bought it shortly before they killed Junior, due to supply issues.  It uses the same basic concept as the big brother DirectStream, so it provides pretty much the same sound quality with that last little bit of refinement that I likely wouldn’t hear anyways.

I won’t go into the architecture, other than to say that unlike other DACs, the PS Audio DS DACs do not use a DAC chip.  They use FPGAs to process a digital signal which is upsampled to 20X DSD; for the output, this is downsampled to 2X DSD and output as analog, using only transformers to filter the output (so it avoids all the brickwall filtering of PCM-based DACs).  One additional advantage is that the unit is upgradeable.  If a new sampling rate or bit rate comes along, or another feature of some sort, it gets written into the next firmware.  I’m currently running Snowmass on the DSJ.

The improvement is noticeable.  The digital glare is pretty much gone. Anything remaining is probably on the original recording.  I hate to repeat marketing lingo, but there are times while listening to CDs that I did indeed here parts of the music that I have been missing.  It is a very nice, smooth sound that is also incredibly detailed.

One draw for me with the DSJ was the built-in Bridge II.  This is the network component which lets you use the DSJ as a player/streamer. (It also has the usual DAC inputs, and an I2S input for a compatible disc player.)  JRiver could use it, but I discovered the setup was still clumsy, and given JRiver’s multitudes of settings, there was no guarantee I was getting a bit perfect stream to the DAC.  Not only that, I was using the Qobuz beta program at the Studio level (which offers up to 24-bit/192kHz streaming), and I had no way to play it on the DSJ without using some kludge on a smartphone to make it happen.

The more I read about it, the more I started leaning towards Roon.  It has a reputation of getting a clean, lossless, bit-perfect audio to whatever device it streams to, up to its maximum resolution.  (So it can cast 24/96 to a Chromecast Audio, for instance.)  It also integrates Qobuz and Tidal, so you can not only browse your own library, you get access to millions of tracks that are only a few clicks or taps away.

There are two drawbacks to Roon.  First, it is expensive.  The cost is about $120/year or $500/lifetime.  But it is well supported, even if support of new features can be somewhat glacial.  It also requires some computing horsepower to make it work ideally.  They recommend a good quad core processor, plenty of memory and a SSD so that browsing the library is seamless.  In other words, unless you have a pricey NAS with a lot of computing power, you are going to have to run it on a separate computer.  This server component is called Roon Core.

To that end, I put together one of those Intel NUC mini-computers. These offer full-powered processors (i3, i5 or i7), two memory slots (load ’em up), and room for both a 3.5″ SSD and one of the smaller M.2 SSDs.  There are also numerous ports–plenty of USB ports, an HDMI port, you name it.  I went with an i3, processor, 8GB of memory (to start), and a Crucial M.2 SSD.  The size is no bigger than a stack of maybe four or five CD jewel cases.

It not only runs Roon Core, I decided to start using Mezzmo again, since now I had a silent, low-powered computer to run it on.  (Mezzmo runs only on Windows, but has features none of the other video servers have.)  I have a couple of other processes I can now offload from my NAS, which makes it run more efficiently.  I can tuck it in the basement on my network rack and run it headless, but decided to connect it to the 4K TV for now.

So far, this combination works mostly well enough, yet I am having a few glitches playing back DSD files (stored as .dsf files) over the network, as it is a bug in the Bridge II.  So I will wait another update or so before deciding if I might want to add a streamer of some sort.

I haven’t introduced the rest of the household to Roon yet, but I’m sure it’s simple enough to use. I may still use Pandora in many instances, but it is easy to direct music to any device, or group of devices, and each user can have their own profile.

So I’ve mostly got this nailed down.  I just need to refine things a bit, but in its current state, it’s certainly livable and working well.

Distributed bass array, Part 1

I read about a distributed bass array (DBA) a couple of years ago, and the concept was intriguing and made a lot of sense.  Bass in a listening room is affected by standing waves/room modes.  The distributed bass array alleviates this by strategically placing as many as four (!) subwoofers in a room to even out the bass response.  The volume of each subwoofer is set to a very low level, so it is not a matter of overpowering a room with too much bass.

Martin Logan 700W. ‘Tis wireless!

There is at least one complete DBA system available, comprised of a single 1,000 watt amplifier and four subwoofer modules.  Given my room layout, this would be somewhat inconvenient, primarily for all the wiring it would require.

Enter wireless subwoofers.

A Martin Logan Dynamo 700W sub came up on the local Craigslist dirt cheap and in excellent condition, so it followed me home. These subs are not very big at all–with the feet or spikes on, they are a hair less than the cubed size of an album jacket. Yet they can dip down into the 25Hz range. With class D amps on board, they offer up to 300w continuous and 900w peak. It can be changed from downward firing to front firing.

My Vandersteen 2ce pair dips down to about 28Hz for a -3dB point, thanks to the “bass coupler” which is a 10″ driver at the bottom rear of the speaker that pretty much acts as its own subwoofer. But that nasty bass suckout in my room bothered me. The DBA has been at the back of my mind as a means to minimize that suckout.

The 700W is wireless (the “W” in the model), and comes with the transmitter that hooks to a preamp output. I used Audioquest RCA cable splitters at the preamp, and ran the spare interconnects to the wireless transmitter. It paired up easily, and it works. (As an aside, my ML soundbar can also act as a transmitter for the 700W and 1000W subwoofers.)

I’ve tried the 700W in the corner by my desk in the back of the room, but found it a little too boomy. I moved it in front of the CD rack, by the opening to the kitchen at the rear of the room, and it’s a bit better. The key is to set the frequency low enough to cover the missing bass notes, and the volume low enough so the sub isn’t noticed. Setting the phase (0°, 90°, 180°) is another option.

OK, so far with dialing it in roughly, and having only one sub, the distributed bass array concept mostly works. I can sit in my preferred spot and the bass is now much more even, and I can hear just about all bass notes now. A clear improvement.

Here’s where it gets tricky. A DBA calls for as many as four subwoofers. And placement is somewhat critical, in that I would need to find the four spots in the room, along the walls, where the bass sounds the most even. (The fourth sub is run 180° out of phase.)

So, I’m three subwoofers short.

For the time being, I would live with the Vandersteens up front, and two rear subwoofers. But there is still a catch I hinted at above. I can still localize the bass in the rear, very slightly, even when I have the volume as low as I can logically set it. I believe this is partly due to time alignment. The bass from that rear sub is maybe a couple of feet closer to the listening position. Adding another 700W back near the desk is the plan, which should even it out just a little more–the bass feels slightly weighted towards one side of the room.

But what can I do with the time alignment? I am looking at a MiniDSP box with the right software plugin, where I could program in the distances for the rear subwoofers so they are delayed enough to match the perceived distance as the front speakers. In effect, pushing the virtual dimensions of the room so the subs appear to be further away.  I am not fond of digital processing (my amplification chain is, and shall remain, completely analog), but since this affects only the bass and is split out from the main signal path, it won’t make a difference in the grand scheme of things.

So for the time being, I will keep an eye out for another bargain 700W. They are plentiful.

Note that I still have the forlorn Sequel IIs to finish refurbishing (who has the time?), so if I put those into service, I will need to rethink the positioning since their response is not as low. Especially if I replaced the original woofers with the recommended aluminum woofers–the bass would not be nearly deep enough for me. Long term, I’m looking at the Martin Logan Summit, which has its own powered twin woofers that reach down as low as the 700W subwoofers, and that would be a natural for the two added rear subwoofers.

One subwoofer is only a small start on a DBA. Stay tuned as I work with this concept further.  Two subs may be enough if the front two speakers are responding well enough.  Or I may eventually need to take it up to four.  But my single 700W is more like a proof-of-concept trial and I can report that the basic premise is sound (no pun intended).

My “30 days of Roon” trial

The only apparently sane way I could get bit perfect audio into my DirectStream from both my server and from Qobuz was to use Roon.  I waited until my schedule was somewhat clear, bit the bullet and installed it.

Roon’s basic setup is such that one Core runs on your network, and any other device acts in essence as a remote to control the Core, such as, which device to send music to, volume, DSP, and of course the mechanisms in place to seek, search, discover and play the music from multiple sources.

My first plan was to get a QNAP NAS to add to the Synology NAS I’m already using on the network, yet I found the QNAPs to be pricey.  The reason was due to the QNAP having more powerful processors and expandable memory.  Once I thought about it, that didn’t seem like a good idea.  A NAS really should only be storing files and performing a few minor functions.  What if my processing needs change in a few years?  Then I’m having to purchase an even more powerful NAS.

I initially balked at the idea, but I settled on buying one of the Intel NUC boxes (basically, it’s a full-fledged computer in a very small package, and is expandable in terms of memory and storage), and loaded up Windows 10. This way, I can run Roon Core, as well as the Mezzmo video server, which runs only on Windows, but is far superior to anything else I have tried in terms of organizing and cataloging a video collection.  I’ve used a 500GB M.2 SSD in the NUC, and am starting with 8GB of memory.  Everything is running well on the NUC in this configuration so far.

So far, Roon seems to be easy to use, and performs nearly flawless.  One thing I want to try is to have others in the household try to use it, without any assistance from myself–I want to see if usability is intuitive to others who aren’t tech-oriented.

There are some quirks, however, but I will cover them separately.  There is so much to cover that I will split things up into separate posts.

For the time being, Roon seems to be about the best way to get the audio to the DirectStream without compromising it in any way, and is easy enough to use that I’m not worrying about not being able to find my music.  JRiver still gets me access to my music faster, but over the years it is still an unstable piece of software, freezing up often or crashing at least once daily.  I’ve had no such pauses with Roon.

Soundbar: Martin Logan Vision

With a 65 inch 4K screen, I knew going in that the built-in speakers would be lacking. They are in reality not too bad, but still can’t touch a dedicated soundbar.

I searched. I tried researching all of my favorite brands as well as some I don’t care for, and without fail, the reviews on most of those I looked at pointed to some fundamental flaws that I could not overlook. In a lot of cases, it was the electronics that caused problems, especially if wireless technology was used.

Out of the blue, one of my saved Craigslist searches returned a hit on a Martin Logan Vision soundbar.  It was a three hour drive away, but given the price, there was no way I could turn it down.

As I picked it up, I noticed how long the box was–this is not some tiny soundbar. For now, it fits on my low IKEA TV stand with about 6 inches on either side to spare.  It is also quite heavy.  The contour of the enclosure is curved and sleek. It comes with a wall mounting plate, which I plan to put to use once I hang the LED panel on the wall.

The rear hosts the inputs–two S/PDIF optical inputs, one coaxial, and one analog input.  There is an analog subwoofer output, and the Vision is also capable of wirelessly transmitting to the Dynamo 700W and 1000W subwoofers (which also have wireless capabilities built in).

The drivers include Martin Logan’s “folded motion” tweeters, which act more like a folded accordion to push the sound out.

The Folded Motion tweeter

As the TV has a digital optical output, I utilized one input for that on the Vision, and used the second input for the digital output of a Chromecast Audio.  That way, anyone can turn on the soundbar, select the source, and cast to it.

The Vision sounds really good so far–it has a much larger sound than its cabinet would have you believe.  The surround mode is also a nice touch–it can decode Dolby surround and DTS, and also reaches back to Dolby Pro Logic II if needed.  Bass is adjustable, and there is also a compressed Night mode for late night use.

I pressed an old subwoofer into action, but unfortunately the subwoofer output is like others I have used–it is only activated when fed 5.1 surround sound, with bass present on the .1 channel.  And that is sad, since I would love to use this soundbar with the video games, which don’t utilize the .1 channel.  I have looked a little for a bass management system that could work inline via the optical connection, but I doubt I will find one easily, if at all. (I just need to route the bass from the main channels over to the .1 output.)

Aside from that, the sound is quite good.  The highs are not piercing or grating at all–the folded motion tweeters are smooth and revealing.  Bass is tubby at the moment, due to having it tucked into the shelf space of the TV stand, but that will change once I can wall mount the soundbar.  The subwoofer output is a nice touch but again, I’ve rarely heard my subwoofer due to the 5.1 digital surround signal often not having any “.1” content. (I would need some sort of external bass management box that would inject all bass into the .1 channel, one that had both optical S/PDIF inputs and outputs. Nothing exists, that I can find.)  It does a good surround effect with the right input signal, extending the sound far beyond the confines of the soundbar without having any weird phasey effects (as many of these do).

This soundbar is a keeper so far.

MQA Is Dead

Maybe I’m the first to say it. But let’s face it:  That lossy digital audio codec known as MQA is dead.

Sure, the pundits will point to streaming MQA titles on Tidal.

And that’s all fine and dandy, until you realize that Qobuz is coming out of beta on February 14th and offers true lossless high-res streaming up to 24-bit, 192kHz. Qobuz has already been the “official” streaming service at a few audio shows, and is doing the same at AXPONA this year.  Not only that, Qobuz fired a direct shot across Tidal’s bow by having Roon support right out of the gate.

Will audiophiles leave Tidal in droves?  Maybe.  Maybe not.  But an appreciable number of them will.  And given Tidal’s past financial troubles, and the perception of being an also-ran among the Spotify users of the world (“what’s a Tidal?”), it may not take much to push them over the precipice of failure. (Or ripe for acquisition, by a corporation looking more at dollar signs than “those crazy audiophiles.”)

Many will hang onto MQA for only one single reason–they spent a small fortune on compatible hardware just to play it back.  They can’t justify their purchase of “just another DAC” any other way.

Let’s also consider MQA CDs. A lame concept right from the beginning. The discs have stiffed!  Many were purchased by curiosity seekers, and many ended up for sale rather quickly.  MQA obviously did not take a lesson from HDCD’s failure. Another format the music buying public was apathetic about.  Essentially the sound quality of CD-resolution is compromised in order to take advantage of MQA’s lossy encoding.

The public did not ask for HDCDs.  Nor MQA CDs.  Which are essentially HDCD 2.0.

MQA already has a lot going against it.  Apparently not even its creators can give us a clear explanation of what it does.  Rather, we get “white-paper-speak.” In essence, a lot of technobabble that makes sense only to its creators, intended to baffle everyone else who asks.  Plus, we keep hearing of so many things that MQA does, or promises, as we go along–it is at the point where I don’t think even its creators know what it does anymore.

MQA. Master Quality Authenticated.  It turns out that there is nothing “authenticated” about its “masters.”  “Quality” is debatable–it’s a better lossy format than MP3, but it’s still a lossy format.  Call it some ridiculous name like unfolding but…it is still a lossy format. Only now, depending on what you play it on, you get different levels of lossy.  Don’t forget it also does some magical “deblurring” or whatever they call it–essentially, using DSP to tamper further with the original sonics of the recording.

Nice.  Why not add a smiley-face EQ and maximize it while you’re at it?

What was laughable a short time ago was how the high-end press were gaslighting us. Of course it was us, the consumers who didn’t understand MQA. “We” didn’t get it. “We” couldn’t hear its many benefits.  But, we mistrusted their motives for publishing what amounts to propaganda. Why were they so eager to shove this down our throats? That remains unanswered.

What they also failed to point out was that MQA was a licensing scheme, on many levels (equipment manufacturers, labels, etc.).  Not only that, it offered a back-door scheme to enable DRM (digital rights management).  This appeases the record companies to make them embrace MQA, but even the hint of DRM turns off consumers who learn it may affect the music the purchase.

So OK, right.  What about us, the consumers?  Most audiophiles that I know personally distrust MQA.  Few ever hear any improvement; most hear no improvement at all.  Others have found that MQA really is just a lossy codec which barely resolves to a level above CD, not to mention having DSP (digital signal processings) to perform some vague “deblurring” that is questionable at best.  And some report the MQA-processed signal to be very slightly louder.  This itself can be deceptive–it has long been proven that given a comparison between two sound samples, listeners will usually choose the louder of the two samples as sounding “better.”

Others feel it’s a money grab by a company that failed to make its last compression technology a household item. (That was MLP–Meridian Lossless Packing, which was used on DVD-Audio discs, which of course have now died a slow and painful death themselves.  MQA is essentially a spinoff of Meridian; if they did one thing right, they’ve planned for the eventual failure of MQA by separating it from Meridian.) What a great way to sell more hardware!  Oh, and all that licensing income can’t be hurting them either.

So, where does that leave us?

Qobuz is blazing the path of true high-res music streaming.  Many will leave Tidal for Qobuz.  The discs have flopped.  Why buy lossy MQA CDs when you can download true high-res files from several places online now?  There will be audiophiles who hang onto MQA, most likely because they purchased expensive equipment to play it back on.  But that will not sustain a dead format.

The bottom line? Nobody asked for MQA. Nobody needs it or particularly wants it.  It was an answer to a question nobody asked.  While it arguably wasn’t stillborn, it had lived a short, miserable existence.

Time to order the headstone.  RIP, MQA.