|
|
Post by rbk123 on Nov 15, 2018 9:29:12 GMT -5
He really needs to actually listen to some options. +1 Paralysis by analysis at its finest. |
|
|
|
Post by pedrocols on Nov 15, 2018 10:45:41 GMT -5
Page 24 interesting 🤔
|
|
KeithL
Administrator 
Posts: 10,273
|
Post by KeithL on Nov 15, 2018 12:14:47 GMT -5
The definition of "a perfect USB input circuit" is actually pretty simple. (Let's be specific and say "a perfect asynchronous USB input circuit".)
It should accept whatever quality of signal it is presented with, including signals of the lowest possible quality that still fulfill "the USB standard".
It should extract the digital audio from that signal, and pass it on to the next stage, with nothing added, removed, or altered... It should eliminate or ignore any flaws present in the original USB input data - and should introduce no new flaws or imperfections. (So, by definition, the signal exiting our "perfect USB input circuit" is perfect - so there is no possible room for improving it.) Since the specification doesn't call for any noise to be present at the output there should be none. It should simply be "perfect data, a perfect clock, and nothing else".
In practice, it's relatively simple to achieve bit-perfect data (where every single bit is correct)... so that is a reasonable requirement... And, in practice, the asynch USB input creates a new clock for that data, so the clock is totally under our control...
Because noise is so pervasive, and impossible to absolutely eliminate, we're going to end up defining what noise level we consider "acceptable". In most cases, there will also be a clear definition of what the input of the next stage considered to be "acceptable noise that can be totally ignored".
(This would usually be defined as "how much noise can we tolerate without risking degrading the performance of the next stage in a significant way".)
This is where I take exception to many claims about "audiophile Ethernet cables". Up to a certain, and rather high, level, neither noise nor jitter have ANY EFFECT WHATSOEVER on Ethernet packets. Past a certain level, noise or jitter may cause correctable errors; if so, the network speed will be degraded, but the data will remain perfect. And, past some much higher level, jitter or noise may result in uncorrectable errors (the network "stops working".).
One of the major benefits of most digital systems is that they have a "hard fail level". They either work perfectly... or they stop working... and the "shades of grey in between" have been eliminated - by design.
An Ethernet cable is either "good enough to meet spec" or not. Yes, you may find "borderline situations" where a piece of equipment that exceeds spec will still work with another that fails to meet spec. But, as long as every piece meets spec, the system will work as intended... and there will be no benefit to "improving any of the pieces".
(That's the whole point; as long as all the pieces meet spec, you don't need to worry about how well they will work together.)
I think it's just a matter of semantics..... People like to overgeneralize.... and this is where it can get confusing.
For example, jitter can be a problem at certain places in the signal chain; however, in others, it is totally harmless, or even irrelevant.
For another example, 0.1V of analog noise on your Ethernet cable, or on a USB data line, is TOTALLY HARMLESS; it only becomes a problem if it finds its way into the analog circuitry. (And you can prevent it from becoming a problem by eliminating it to begin with - or by simply by keeping the proverbial door shut - or by removing it after it sneaks in... although one of those options may be the easiest.)
A "perfectly performing" asynchronous USB input would be able to deliver a perfect audio signal, and a perfect clock, so there would be no room for improvement. What the Eitr does is to replace the current USB input on your device with a new one - which is, we hope, able to perform more perfectly. Can you define, more perfectly, please?☺ |
|
|
|
Post by Casey Leedom on Nov 15, 2018 12:22:55 GMT -5
KeithL, when you refer to "Noise" in this context, what do you mean? I'm guessing "Electrical Noise" on the USB cable into the "perfect" USB Input receiver? Casey
|
|
klinemj
Emo VIPs 
Official Emofest Scribe
Posts: 15,095
|
Post by klinemj on Nov 15, 2018 19:58:38 GMT -5
The definition of "a perfect USB input circuit" is actually pretty simple. (Let's be specific and say "a perfect asynchronous USB input circuit".)
It should accept whatever quality of signal it is presented with, including signals of the lowest possible quality that still fulfill "the USB standard".
It should extract the digital audio from that signal, and pass it on to the next stage, with nothing added, removed, or altered... It should eliminate or ignore any flaws present in the original USB input data - and should introduce no new flaws or imperfections. (So, by definition, the signal exiting our "perfect USB input circuit" is perfect - so there is no possible room for improving it.) Since the specification doesn't call for any noise to be present at the output there should be none. It should simply be "perfect data, a perfect clock, and nothing else".
In practice, it's relatively simple to achieve bit-perfect data (where every single bit is correct)... so that is a reasonable requirement... And, in practice, the asynch USB input creates a new clock for that data, so the clock is totally under our control...
Because noise is so pervasive, and impossible to absolutely eliminate, we're going to end up defining what noise level we consider "acceptable". In most cases, there will also be a clear definition of what the input of the next stage considered to be "acceptable noise that can be totally ignored".
(This would usually be defined as "how much noise can we tolerate without risking degrading the performance of the next stage in a significant way".) This is where I take exception to many claims about "audiophile Ethernet cables". Up to a certain, and rather high, level, neither noise nor jitter have ANY EFFECT WHATSOEVER on Ethernet packets. Past a certain level, noise or jitter may cause correctable errors; if so, the network speed will be degraded, but the data will remain perfect. And, past some much higher level, jitter or noise may result in uncorrectable errors (the network "stops working".).
One of the major benefits of most digital systems is that they have a "hard fail level". They either work perfectly... or they stop working... and the "shades of grey in between" have been eliminated - by design. An Ethernet cable is either "good enough to meet spec" or not. Yes, you may find "borderline situations" where a piece of equipment that exceeds spec will still work with another that fails to meet spec. But, as long as every piece meets spec, the system will work as intended... and there will be no benefit to "improving any of the pieces".
(That's the whole point; as long as all the pieces meet spec, you don't need to worry about how well they will work together.)
Can you define, more perfectly, please?☺ I've got a simpler answer. A perfect one sounds better to the ears than a non-perfect one. Mark |
|
|
|
Post by novisnick on Nov 15, 2018 20:09:36 GMT -5
The definition of "a perfect USB input circuit" is actually pretty simple. (Let's be specific and say "a perfect asynchronous USB input circuit".)
It should accept whatever quality of signal it is presented with, including signals of the lowest possible quality that still fulfill "the USB standard".
It should extract the digital audio from that signal, and pass it on to the next stage, with nothing added, removed, or altered... It should eliminate or ignore any flaws present in the original USB input data - and should introduce no new flaws or imperfections. (So, by definition, the signal exiting our "perfect USB input circuit" is perfect - so there is no possible room for improving it.) Since the specification doesn't call for any noise to be present at the output there should be none. It should simply be "perfect data, a perfect clock, and nothing else".
In practice, it's relatively simple to achieve bit-perfect data (where every single bit is correct)... so that is a reasonable requirement... And, in practice, the asynch USB input creates a new clock for that data, so the clock is totally under our control...
Because noise is so pervasive, and impossible to absolutely eliminate, we're going to end up defining what noise level we consider "acceptable". In most cases, there will also be a clear definition of what the input of the next stage considered to be "acceptable noise that can be totally ignored".
(This would usually be defined as "how much noise can we tolerate without risking degrading the performance of the next stage in a significant way".) This is where I take exception to many claims about "audiophile Ethernet cables". Up to a certain, and rather high, level, neither noise nor jitter have ANY EFFECT WHATSOEVER on Ethernet packets. Past a certain level, noise or jitter may cause correctable errors; if so, the network speed will be degraded, but the data will remain perfect. And, past some much higher level, jitter or noise may result in uncorrectable errors (the network "stops working".).
One of the major benefits of most digital systems is that they have a "hard fail level". They either work perfectly... or they stop working... and the "shades of grey in between" have been eliminated - by design. An Ethernet cable is either "good enough to meet spec" or not. Yes, you may find "borderline situations" where a piece of equipment that exceeds spec will still work with another that fails to meet spec. But, as long as every piece meets spec, the system will work as intended... and there will be no benefit to "improving any of the pieces".
(That's the whole point; as long as all the pieces meet spec, you don't need to worry about how well they will work together.)
I've got a simpler answer. A perfect one sounds better to the ears than a non-perfect one. Mark Id like to add that if a motor meets spec and it runs then thats all there is to it? I call BS! Lets make it run smoother, more efficient, maybe even a little bit less noisy! According to KeithL, all this matters not! 🤯 we are dealing with the pleasure sound brings us, not weather the switch meets spec and is either on or off. Thats a rather narrow view. |
|
|
|
Post by Loop 7 on Nov 15, 2018 20:36:03 GMT -5
A timely release from Darko on the potential the Raspberry Pi has to provide good quality sound as a streamer (with attachments, but still at a lower cost than competitors): Click here for link to Youtube videoReally, watch it all the way through and there are some really interesting results/conclusions towards the end (subjective, I know)! Glad you posted the video. I was hesitant to do so as it may have ignited more fires. I find Darko's writing and videos to be full of substance and this video has me wanting to at least try Raspberry Pi and a hat or two. |
|
|
|
Post by Casey Leedom on Nov 15, 2018 21:32:14 GMT -5
Personally, I'm hoping that one of the crew who believe that USB Sources matter gets the new RMC-1 and report back. Because, iff you do hear a distinct difference, it would be interesting to have the Emotiva folks — *cough* *cough* KeithL, et.al. — grab the same two sources and see what's happening ... (And please do remember, I'm also in KeithL's camp, but would love to see a case of two different USB Sources resulting in discernibly different sound which could open up the possibility of this investigation for the actual cause.) Casey |
|
|
|
Post by garbulky on Nov 15, 2018 22:27:01 GMT -5
A timely release from Darko on the potential the Raspberry Pi has to provide good quality sound as a streamer (with attachments, but still at a lower cost than competitors): Click here for link to Youtube videoReally, watch it all the way through and there are some really interesting results/conclusions towards the end (subjective, I know)! Glad you posted the video. I was hesitant to do so as it may have ignited more fires. I find Darko's writing and videos to be full of substance and this video has me wanting to at least try Raspberry Pi and a hat or two. Full of food subtance? His reviews make me hungry with all his food analogies! I remember when he compared the Gungnir to the difference between a lime or lemon and another dac to that of some sort of chocolate cake. Yum! f you like his style of writings you might like those of the part time audiophile. |
|
|
|
Post by liv2teach on Nov 16, 2018 2:11:07 GMT -5
So apparently I somehow subscribed to Tidal through Amazon...which charges $5 more a month for the service... It's not the $5, but the principle of the thing. Anyway, looks like I have to delete Tidal and start anew with a different email, etc. Anyone else run into this issue? I hope this doesn't screw me when it comes to resetting Roon that currently is feeding through my Amazon Tidal account.
|
|
|
|
Post by bluemeanies on Nov 16, 2018 5:01:30 GMT -5
So apparently I somehow subscribed to Tidal through Amazon...which charges $5 more a month for the service... It's not the $5, but the principle of the thing. Anyway, looks like I have to delete Tidal and start anew with a different email, etc. Anyone else run into this issue? I hope this doesn't screw me when it comes to resetting Roon that currently is feeding through my Amazon Tidal account. I think Boom had a similar or I will say bad experience with BEST BUY in regards to TIDAL. BUMMER! That said, why not go directly to the source...TIDAL in this case. I pay $20.00 per month and if YOU are a VETERN there is a substantial discount. Never buy anything from a third party when you have the opportunity to get to the main source. |
|
klinemj
Emo VIPs
Official Emofest Scribe
Posts: 15,095
|
Post by klinemj on Nov 16, 2018 9:13:56 GMT -5
So apparently I somehow subscribed to Tidal through Amazon...which charges $5 more a month for the service... It's not the $5, but the principle of the thing. Anyway, looks like I have to delete Tidal and start anew with a different email, etc. Anyone else run into this issue? I hope this doesn't screw me when it comes to resetting Roon that currently is feeding through my Amazon Tidal account. You should not have any issue with Roon. It just logs into Tidal account you tell it to log into. If you end up needing to change the credentials, you just change them in Roon's settings. Mark |
|
|
|
Post by mr on Nov 16, 2018 9:58:39 GMT -5
So apparently I somehow subscribed to Tidal through Amazon...which charges $5 more a month for the service... It's not the $5, but the principle of the thing. Anyway, looks like I have to delete Tidal and start anew with a different email, etc. Anyone else run into this issue? I hope this doesn't screw me when it comes to resetting Roon that currently is feeding through my Amazon Tidal account. I think Boom had a similar or I will say bad experience with BEST BUY in regards to TIDAL. BUMMER! That said, why not go directly to the source...TIDAL in this case. I pay $20.00 per month and if YOU are a VETERN there is a substantial discount.Never buy anything from a third party when you have the opportunity to get to the main source. The Veteran's discount is 40%, well worth filling out a short form. |
|
|
|
Post by liv2teach on Nov 16, 2018 11:05:16 GMT -5
I think Boom had a similar or I will say bad experience with BEST BUY in regards to TIDAL. BUMMER! That said, why not go directly to the source...TIDAL in this case. I pay $20.00 per month and if YOU are a VETERN there is a substantial discount.Never buy anything from a third party when you have the opportunity to get to the main source. The Veteran's discount is 40%, well worth filling out a short form. I am a vet...an old one now.... Thanks for the heads up, I'll get on it.... |
|
