[Davesrose]

I have reread your first sentence of this post countless times and I am still having trouble understanding your point. The Bluesound post on the Node says this about “Identifying Signs of an Underpowered Drive”. Notice these are not an exhaustive list of the signs, but rather examples.

I'll try to multi quote here just to go point by point, and I might be the last member to then give up on this.

Identifying Signs of an Underpowered Drive​

Indications of an underpowered drive include, but are not restricted to:

• The PLAY/PAUSE button on the player does not turn WHITE, indicating that the storage device is not being indexed.
• The storage device emits repeated clicking sounds but fails to start.
• The storage device does not connect or initiate.

*Typically, underpowered drives have a size exceeding 500GB or 0.5TB.

Which point here is talking about diminished audio quality with underpowered drive, or USB drive types?? Your previous link indicated that if a device isn't indexed, the effect is that it loads 500 songs per minute vs 1,000. The device emitting repeated clicking and fails to start: that's a HDD that requires external power. It may have enough juice to spin and have heads unmount, but it stops and mounts because it has not communicated to the Node. That's what the USB protocols do. There has to be a handshake with the devices. If external drives that have a DC power input are plugged in without power...they may have enough power to begin a start process, but end before there's a handshake (so again, the drive isn't recognized....let alone is recognized and can play a file). Likewise storage device does not connect or initiate: SSD drives and thumb drives don't click because they don't have physical platters. If the Node recognizes FAT32, NTFS, or ext4 (Linux file system) then it will not see a drive that's exFAT. With computers, exFAT has become a popular format for SSD since they hold some better efficiency than NTFS for the format (or if you're strictly Mac, they have a APFS system that's optimized for SSD). Also note that when I say efficient, I mean best method for fast speeds and long longevity (not some issue with faults in file transfer). Finally I'm guessing the * 500GB is enforcing the point that compatibility issues may be more so with USB 2 drives (requiring more power to operate because that standard did not have as much power in the interface).

I did watch the video on jitter you shared. It is interesting, and I can see how jitter is no longer seen as an audible source of an error in modern digital front ends that utilize USB or Ethernet for asynchronous transfer of digital signals. From that and other posts here, I am assuming also that the signal from any of the external drives I used was transferred to the Node server asynchronously, so no timing errors were imparted in that process. Other noise perhaps, but not timing errors. But if the Node power supply is struggling to power an external drive, which is clearly a possibility given the caution above, then isn’t it conceivable it could affect the performance of the Node as a server, and potentially introduce jitter and/or other noise to the signal before it is transferred to the external DAC?

Nope, "it's got to be noise in my USB connection" is the same rabbit hole as jitter. Analog devices can be affected by noise. My house is an example. I live near a radio tower: so most my analog amplifier stages pick up the interference (and I can hear the radio station if I don't have a RF filter). I live in a townhouse that's now 20 years old. It's wired with I think cat5. Just recently I upgraded wifi to 6E. I tested my ethernet off the network switch I have plugged into my house's wiring. It read the max speed of my ISP (getting up to 1Gbps). So even though it would have had a long run to pick up the RF interference, there was no difference in digital bandwith. Conceptually, digital connections don't have to be as tolerant about noise, as they just need enough signal to read a 1 or a 0. There was a sponsor thread I engaged with about a LAN silencer: I wasn't outright banned or posts deleted because I was courteous with an audiophile who was adamant that any noise in a network is terrible for audio quality. The subject was how to test for audibility. He maintained the blind tests had to be conducted with the most expensive equipment and only "audio designers" from that brand. Seems kind of counter productive to the argument that computers (especially server rooms) are "too noisy" as sources, and that audio quality is easy to discern. If so, why not use some computer with internal sound card as source (because it's "too noisy") and sample as many people as you can (because there's not really a golden ear when it comes to any audio perception)?

The Node only has one USB port which can function as either an input or an output, and I am using it to accept the signal from the external drive. Given that, the only way to transfer the digital signal to an external DAC is via coax or optical toslink, which as I understand it provides a synchronous signal that includes timing information, good or bad, from the server. Anything that happens to the signal in the Node could affect the signal output via coax, which is what I use in my system.

During my life, I've had plenty of digital systems with toslink/coax....and I've never actually heard jitter. Practically, the cable run of them in a home system is short so there's not even a hiccup at the start. In real practice when we think of a clock with them, there might be a minuscule fraction of a second before the song starts. Then the connection is good enough that there's not timing issues while the file is streaming.

In the video you shared, the $8 DAC highlighted had passible jitter performance but nevertheless the narrator noted it was very noisy in other ways that could affect performance. Is it possible that the Node is functioning more like the $8 DAC than a $500 server/streamer/DAC when straining to power an external mechanical drive, and that could result in audible differences in performance as I have it configured at the speakers/headphones, all other things being equal? I am just trying to see if I am tracking your logic and that I am making my point clearly. I may be completely wrong here, but if so, I want to fully understand why.

kn

The video was just highlighting that the $8 DAC (which is not just the DAC chip) was the point in which there's bad enough engineering in the system. So with my systems, the only time I've had an issue with an actual storage medium was the micro-SD slot of a music streamer after a firmware update. After the update, all the songs had popping noises. It was software related effecting that input, and downgrading firmware fixed it. The only times I've heard differences with digital sources is a computer or smartphone going to my Benchmark DAC: the reason not being anything about jitter or noise. The reason being that they may not have a direct API (bit perfect source like ethernet/USB/toslink)....instead first going through their EQ settings.

 

[bigshot]

My only problem with underpowered hard drives is when there is a brownout from too much air conditioning being turned on at once in Los Angeles on a hot summer day. There is a brief drop in the power from the wall and the drive ejects. If the disk is writing at the time of the power drop, it can corrupt files. I have a battery backup on my disk arrays to prevent that.

Apple makes an $8 DAC that has fidelity to rival high end DACs. It isn't expensive to make a quality DAC. It isn't complicated either. Many DACs are made from off the shelf parts and it would take some major design bungling to make them sound bad.

Everyone always adds the caveat that the analog output can be noisy, but it is pretty darn simple to make a quiet line level output. I've got a $40 Walmart DVD player that has clean output. My brother has an old McIntosh amp with caps that have gone bad from being too old. That is to be expected after half a century.

Most of the time, noise in electronics is at the output to the transducers due to impedance issues... matching components with transducers that shouldn't be paired... in short, user error.

This case is just a plain old sloppy listening comparison. Applying the proper controls would remove all the mystery.

 

[knownothing2]

I'll try to multi quote here just to go point by point, and I might be the last member to then give up on this.

Which point here is talking about diminished audio quality with underpowered drive, or USB drive types?? Your previous link indicated that if a device isn't indexed, the effect is that it loads 500 songs per minute vs 1,000. The device emitting repeated clicking and fails to start: that's a HDD that requires external power. It may have enough juice to spin and have heads unmount, but it stops and mounts because it has not communicated to the Node. That's what the USB protocols do. There has to be a handshake with the devices. If external drives that have a DC power input are plugged in without power...they may have enough power to begin a start process, but end before there's a handshake (so again, the drive isn't recognized....let alone is recognized and can play a file). Likewise storage device does not connect or initiate: SSD drives and thumb drives don't click because they don't have physical platters. If the Node recognizes FAT32, NTFS, or ext4 (Linux file system) then it will not see a drive that's exFAT. With computers, exFAT has become a popular format for SSD since they hold some better efficiency than NTFS for the format (or if you're strictly Mac, they have a APFS system that's optimized for SSD). Also note that when I say efficient, I mean best method for fast speeds and long longevity (not some issue with faults in file transfer). Finally I'm guessing the * 500GB is enforcing the point that compatibility issues may be more so with USB 2 drives (requiring more power to operate because that standard did not have as much power in the interface).

Nope, "it's got to be noise in my USB connection" is the same rabbit hole as jitter. Analog devices can be affected by noise. My house is an example. I live near a radio tower: so most my analog amplifier stages pick up the interference (and I can hear the radio station if I don't have a RF filter). I live in a townhouse that's now 20 years old. It's wired with I think cat5. Just recently I upgraded wifi to 6E. I tested my ethernet off the network switch I have plugged into my house's wiring. It read the max speed of my ISP (getting up to 1Gbps). So even though it would have had a long run to pick up the RF interference, there was no difference in digital bandwith. Conceptually, digital connections don't have to be as tolerant about noise, as they just need enough signal to read a 1 or a 0. There was a sponsor thread I engaged with about a LAN silencer: I wasn't outright banned or posts deleted because I was courteous with an audiophile who was adamant that any noise in a network is terrible for audio quality. The subject was how to test for audibility. He maintained the blind tests had to be conducted with the most expensive equipment and only "audio designers" from that brand. Seems kind of counter productive to the argument that computers (especially server rooms) are "too noisy" as sources, and that audio quality is easy to discern. If so, why not use some computer with internal sound card as source (because it's "too noisy") and sample as many people as you can (because there's not really a golden ear when it comes to any audio perception)?

During my life, I've had plenty of digital systems with toslink/coax....and I've never actually heard jitter. Practically, the cable run of them in a home system is short so there's not even a hiccup at the start. In real practice when we think of a clock with them, there might be a minuscule fraction of a second before the song starts. Then the connection is good enough that there's not timing issues while the file is streaming.

The video was just highlighting that the $8 DAC (which is not just the DAC chip) was the point in which there's bad enough engineering in the system. So with my systems, the only time I've had an issue with an actual storage medium was the micro-SD slot of a music streamer after a firmware update. After the update, all the songs had popping noises. It was software related effecting that input, and downgrading firmware fixed it. The only times I've heard differences with digital sources is a computer or smartphone going to my Benchmark DAC: the reason not being anything about jitter or noise. The reason being that they may not have a direct API (bit perfect source like ethernet/USB/toslink)....instead first going through their EQ settings.

Thanks for taking the time to respond. I appreciate it. Just a couple of quick questions to clarify my understanding further. The point I think I was trying to make is that the switching power supply in the Node is likely working pretty hard to power the external drive. Assuming all the data is getting off the drive accurately via USB to the server, is it possible that noise from the power supply inside the server could be non constant, ie increases with current draw? If that is the case, is it possible that noise could affect the clocking of the data in the server before sending it to the DAC via coax? Do I understand correctly that the coax connection provides a clocked data stream from the server to the DAC, ie it’s not asynchronous? In your reply, I did not see an example of a set up exactly as I am describing it here. Perhaps the examples you provided are different in detail, but functionally the same. Note that I am no longer invoking any noise or jitter arising within the USB or coax cables or connections, just the possibility of excess noise generated in the overtaxed server that is conserved via the coax connection to the DAC.

Thanks again. As has been recommended by @bigshot and others here, I am attempting to remain open minded and learn. My lack of expertise in all things digital requires some spoon feeding.

kn

 

[Davesrose]

My only problem with hard drives is when there is a brownout from too much air conditioning being turned on at once in Los Angeles on a hot summer day. There is a brief drop in the power from the wall and the drive ejects. If it is writing at the time of the power drop, it can corrupt files. I have a battery backup on my disk arrays to prevent that.

Hey, lets have a conversation here now that we're devoting pages to circular arguments. I now have two NAS systems. One as a Plex server (WD PR4100) and another for file server (Synology DS1812+). Hotlanta can get hotter than LA: but we don't get brownouts. Every once in awhile I'll get really bad power fluctuations. It happened to me last week. Power kept switching on and off--the Synology just powered down. The WD kept trying to recover. Overall it didn't seem to do anything with my RAID configuration as far as integrity. While I had thought I'd only maybe need UPS for workstations, I'm now thinking of maybe getting a cheaper one for the two NASs (but it still only is for these once in a year fluctuations).

Apple makes an $8 DAC that has fidelity to rival high end DACs. It isn't expensive to make a quality DAC. It isn't complicated either. Many DACs are made from off the shelf parts and it would take some major design bungling to make them sound bad.

I'm assuming you mean their lightening to earphone dongle. That's probably the only reasonably priced item they introduced: especially considering what they're providing now. Their lightening to HDMI was overpriced: a similar USB to HDMI having better specs. Really can't get over new Macs charging so much for memory upgrades (that you're tied into because it's all soldered on).

 

[Davesrose]

Thanks for taking the time to respond. I appreciate it. Just a couple of quick questions to clarify my understanding further. The point I think I was trying to make is that the switching power supply in the Node is likely working pretty hard to power the external drive. Assuming all the data is getting off the drive accurately via USB to the server, is it possible that noise from the power supply inside the server could be non constant, ie increases with current draw? If that is the case, is it possible that noise could affect the clocking of the data in the server before sending it to the DAC via coax? Do I understand correctly that the coax connection provides a clocked data stream from the server to the DAC, ie it’s not asynchronous? In your reply, I did not see an example of a set up exactly as I am describing it here. Perhaps the examples you provided are different in detail, but functionally the same. Note that I am no longer invoking any noise or jitter arising within the USB or coax cables or connections, just the possibility of excess noise generated in the overtaxed server that is conserved via the coax connection to the DAC.

Thanks again. As has been recommended by @bigshot and others here, I am attempting to remain open minded and learn. My lack of expertise in all things digital requires some spoon feeding.

kn

I try to not be as condescending as some others....when they wonder what kind of "magical" thinking you're pursing, maybe it boils down to a misunderstanding of what electric current is. A USB storage device (even USB 3) has a relatively low power consumption. As the Bluesound's links indicate: if a USB 2 device is used (which usually assumed a separate DC power input), it simply won't be recognized (because it doesn't have enough power to function). Your references are what I've been saying: USB devices are either recognized or not: audio reproduction is bit perfect when it comes to streaming off them if recognized. When it came to real world examples of synchronous streaming with coax: I was merely interjecting that there might be a hundredths of a millisecond off of it starting (something that's imperceptible in the real world). You say "overtaxed server that is conserved via the coax connection".....you've also interchangeably said "underpowered" drives tax the system. But that's not how things work. A DAC is independent of a player outputting coax signal or USB input. I don't know how many times I can continue that an underpowered USB device just won't be recognized by the DAC. With a coax device, it has to function to send a digital stream...and in practical applications we don't see jitter or noise distortion.

When it comes to current, a power supply only needs to provide enough current to which the system draws. These USB systems are fairly minuscule in the scheme of things. Now with laptops, the push is having USB-C interfaces that can accept over 100W for higher performance systems. With these complicated systems, if they don't get enough power, then it means the performance is hampered (which the OS is under powering the GPU and CPU).

 

[FunkyBassMan]

My lack of expertise in all things digital requires some spoon feeding.

kn

The problem is, you're being spoon fed by monks of the order, not scientists.

 

[Davesrose]

The problem is, you're being spoon fed by monks of the order, not scientists.

Monks of the order? Folks of various disciplines who all have varying knowledge of computer systems? I think most of us have been patient supplying answers about how storage mediums work with computers (and how power, digital interface, capacity size, or file system are bit perfect).

 

[FunkyBassMan]

Monks of the order?

Correct. You constantly state as fact things you can't back with any science.

"I don't know how many times I can continue that an underpowered USB device just won't be recognized by the DAC. "

And things that are demonstrably contradicted in the real world.

"a power supply only needs to provide enough current to which the system draws"

Folks of various disciplines who all have varying knowledge of computer systems?

Thank you for evidence for my point precisely.

I think most of us have been patient supplying answers about how storage mediums work with computers (and how power, digital interface, capacity size, or file system are bit perfect).

You've supplied answers, yes. The bible supplies answers too.

Neither contains science.

 

[Davesrose]

Correct. You constantly state as fact things you can't back with any science.

"I don't know how many times I can continue that an underpowered USB device just won't be recognized by the DAC. "

And things that are demonstrably contradicted in the real world.

I'll just full stop here.....My assertions are backed by not just science but the manufacturer's website and anyone's observance. Do I have to interject Einstein's quote about insanity being repeating the same thing expecting something different? As in the real world, and manufacturer's site....a drive that's underpowered won't be recognized. It will not show up. How then, since it's not showing up as a source....then can magically be a source and then have noticeable audio degradation?

 

[bigshot]

An underpowered hard disk drive won't mount. If it won't mount, it can't be accessed by the DAC. Davesrose is right. This isn't a matter of science. Just RTFM.

I'm nursing an old Drobo 8D that is maxed out with 14TB drives. Most of the time it performs fine, but I had a brownout where the power dropped and the disk ejected. I suspect it was running some sort of automatic repair or housekeeping function when it dropped. In any case, when it came back it lost a drive and I had to add a new one. With a Drobo this big, it takes over a week to do that. With a battery backup, it has been working fine for a few years with no problem. Drobo is out of business, so when this is ready to retire, I'll have to get something newer and hopefully more dependable. But for now, the Drobo is working and everything on it is safely double backed up.

 

[Davesrose]

I'm nursing an old Drobo 8D that is maxed out with 14TB drives. Most of the time it performs fine, but I had a brownout where the power dropped and the disk ejected. I suspect it was running some sort of automatic repair or housekeeping function when it dropped. In any case, when it came back it lost a drive and I had to add a new one. With a Drobo this big, it takes over a week to do that. With a battery backup, it has been working fine for a few years with no problem. Drobo is out of business, so when this is ready to retire, I'll have to get something newer and hopefully more dependable. But for now, the Drobo is working and everything on it is safely double backed up.

Oh, I see it's also different than the NAS systems I have. So I first got the WD PR4100 thinking it would hold me for a Plex server. Even loading 4K (compressed) movies, it's still going to take awhile before I fill it. I've typically been going 18TB per drive. But being a computer nerd, I really hate how proprietary the WD is: I can't access it the way I can with the open source Synology I have. The Synology can be my main NAS for quite awhile: it being 8 bay with expansion 2x 5 bays. At least when both go offline, it doesn't take much time to recover. The only main time there is when you add more drives to a current RAID: it can take 2 or 3 days considering how it goes cluster to cluster indexing everything.

 

[bigshot]

The more and bigger drives, the longer indexing takes. I’m up to about a week and a half. The blinking amber lights scare me, so I’m nervous the whole time.

 

[Davesrose]

The more and bigger drives, the longer indexing takes. I’m up to about a week and a half. The blinking amber lights scare me, so I’m nervous the whole time.

Well after the power fluctuations I had last week: I also think it depends on system. My Synology, which I've filled 6 of 8 bays 18TB it apparently shut off. So after I waited after the fluctuations finished, I saw the NAS had shut down. It was completely off, so I turned it on and didn't take too much time for it to beep to indicate it was online. I think the WD PR4100 was a bit more complicated about how it was always on and me trying to get it online.

 

[bigshot]

It’s taken me a decade to figure out the weird little things about Dronos. So much is undocumented and customer service was terrible.

 

[gregorio]

No, seriously, I hadn’t given it any thought at all until I experienced a degredation in sound quality when switching from a flash drive to an HDD. I literally had no expectations at all, but was disappointed when the quality of sound was clearly degraded. To have an “expectation bias”, one has to have “expectations”.

Come on, none of this makes any sense and isn’t even possible.
1. How did you switch from a flash drive to a HDD without “any thought at all”? Did you decide to make the switch, buy the drive and physically do the switch all while you were in a coma or when you were sleep walking?
2. Again, a flash drive and a HDD look different right? I presume you also know that they operate differently, the HDD uses spinning disks and the flash drive just uses solid state memory chips? You also stated the HDD required a cable and the flash drive just plugged straight in. So how on Earth is it possible that you did not have any “expectations” that they were in fact significantly different? Indeed, if you had no expectations of them being different in anyway at all, what would be the point of switching from one to the other in the first place?

So, clearly there must have been both “thought” and “expectations” and in addition, we obviously spend our entire lives in the real world, constantly and consistently experiencing the fact that things which have different sizes, that look different, that are different and that operate on completely different principles ALWAYS sound at least somewhat different. Now maybe you didn’t actively/consciously think about that at the time but your brain “knows” this to be the case!

It is curious to me that the issue manifested in a slight but noticeable difference in the sound quality of the music reproduced when using the different drives …

In which case, the rational approach to satisfying that curiosity is a standard, logical trouble shooting procedure, a series of steps to progressively narrow down the possibilities to a single actual/truthful answer. For example, first ascertain if there is actually a difference in the sound or the analogue signal coming out of the DAC. If there isn’t, then it must be a perceptual issue. If there is, then perform further steps to isolate where that difference is occurring, for instance is there a difference in the data arriving at the input to the DAC or is the difference occurring in the DAC?

Your problem, which is typical in the audiophile world, is that you do NOT take a rational approach, you do not apply a logical procedure and in fact you don’t even take the very first step to ascertain if there is actually a difference to start with. What you actually do is irrational and illogical, you just assume your perception is infallible (even though it can’t be), that there really is a “difference in sound quality” which is audible and then spend all your time inventing or regurgitating nonsense explanations created by marketers, that were debunked years/decades ago, in order to explain that audible difference (which doesn’t exist!). It’s ridiculous, with just a fraction of the time you’ve spent posting falsehoods and regurgitating tired old marketing fallacies, you could have easily ascertained the actual facts!

Assuming all the data is getting off the drive accurately via USB to the server, is it possible that noise from the power supply inside the server could be non constant, ie increases with current draw? If that is the case, is it possible that noise could affect the clocking of the data in the server before sending it to the DAC via coax? Do I understand correctly that the coax connection provides a clocked data stream from the server to the DAC, ie it’s not asynchronous?

What a great example of my previous paragraph! You raise “a possibility” invented by audiophile marketers 2-3 decades ago, when that’s debunked as impossible or nonsensical you just raise another, “throwing spaghetti against the wall” ad infinitum hoping something will stick, all the while ignoring/dismissing the obvious! EG. Is noise a possibility? Yes, if you ignore the whole point of why digital audio was invented in the first place or if the power supply is faulty. Or, is it possible that noise could affect the clocking of the data? Yes, but so what, unless the DAC is faulty then it cannot affect the output. Or now, can a noisy synchronous coax signal cause some audible effect? Yes, if the DAC or server is faulty/broken, otherwise the effect is too low in level for it even it to be reproducible as sound, let alone be audible.

The whole clocking noise/jitter thing was debunked as audiophile marketing BS decades ago. The audibility of jitter has been known for half a century, the BBC published a paper in 1974 indicating around 200ns - 500ns (nanosecs) as being the typical threshold with music. There has been considerable research since, effectively confirming that. Probably the most comprehensive is the Benjamin and Gannon paper in 1998 (link here). From memory: 50 cheap consumer devices (mostly CD and DVD drives and players) were tested for jitter, obviously all on the market prior to 1998, along with the effects of coax cable with SPDIF (synchronous clocking) and comprehensive blind testing of trained subjects. The absolute threshold of jitter for the most sensitive subject with test signals was 3ns (average was about 10ns), the lowest with a specific music recording was 30ns, but with typical music recordings the BBC’s conclusion of around 200ns - 500ns was confirmed. But the actual measured jitter of these cheap audio devices from ~30 years ago was on average around 0.15ns!

G