Another technical question that probably only will find it's answer in a pro forum!
Is there any way to measure digital output from a source, by comparing it to all the bits from the original file? For example the spdif out in a media extender.
Does digital-out-clock upgrades on for example a squeezebox be worth if you put the signal into a linn dac, for example classik music 08? uppgrade costs around 200€
I've heard Ikemi has got one of the best spdif outs in the world, wouldn't an upgrade on squeeze do the same for the music?
will an ikemi-level-digital signal in to a classik music 08 still be a lot worse than the bulit in CD in the music 08? (i know the ikemi is a better source then the cd in music 08, but it's just an example of a good transporter)
Jitter measuring?
Moderator: Staff
You can measure the jitter of a digital stream, sure. No need to compare it to the bits in the original file, as jitter is timing errors. So, all the bits will arrive intact, but with slight variations in the timing of the clock. These variations can be either random (as in noise), periodical or dependent on something else (like the musical signal).
When the receiver converts the digital stream, the timing errors will show up in the analogue signal as distortion. Different types of jitter will cause different types of distortion.
Correlating the distortion created by jitter with sound quality is very difficult! This is the same as with other sources of distortion; over the years, various types have been claimed to be especially bad, only later to be replaced by another type. In the experiments I've made, using the Tune Method for measuring quality, I've reached these general conclusions:
1. An optimal balance of different types of distortion is very important.
2. The less overall distortion, the better.
3. Number 2 above is less important than number one.
Sorry to use so many words, but what I'm trying to say is that even if you measure the jitter, you won't know much about the sound. You really have to listen to the options you have in mind.
When the receiver converts the digital stream, the timing errors will show up in the analogue signal as distortion. Different types of jitter will cause different types of distortion.
Correlating the distortion created by jitter with sound quality is very difficult! This is the same as with other sources of distortion; over the years, various types have been claimed to be especially bad, only later to be replaced by another type. In the experiments I've made, using the Tune Method for measuring quality, I've reached these general conclusions:
1. An optimal balance of different types of distortion is very important.
2. The less overall distortion, the better.
3. Number 2 above is less important than number one.
Sorry to use so many words, but what I'm trying to say is that even if you measure the jitter, you won't know much about the sound. You really have to listen to the options you have in mind.
But if you have a receiver that makes the spdif signal back into a wave file again. Then the transporter giving the best file (most similar to the original file) on the other side of the receiving clock, shouldn't that be the best transporter?lejonklou wrote:You can measure the jitter of a digital stream, sure. No need to compare it to the bits in the original file, as jitter is timing errors. So, all the bits will arrive intact, but with slight variations in the timing of the clock. These variations can be either random (as in noise), periodical or dependent on something else (like the musical signal).
When the receiver converts the digital stream, the timing errors will show up in the analogue signal as distortion. Different types of jitter will cause different types of distortion.
Correlating the distortion created by jitter with sound quality is very difficult! This is the same as with other sources of distortion; over the years, various types have been claimed to be especially bad, only later to be replaced by another type. In the experiments I've made, using the Tune Method for measuring quality, I've reached these general conclusions:
1. An optimal balance of different types of distortion is very important.
2. The less overall distortion, the better.
3. Number 2 above is less important than number one.
Sorry to use so many words, but what I'm trying to say is that even if you measure the jitter, you won't know much about the sound. You really have to listen to the options you have in mind.
The most similar recreation would be the one with the lowest amount of distortion, yes. But will it sound better than the one with slightly more distortion? Perhaps not.
The same applies to analogue amplifiers. Is one with 0.001% distortion better than one with 0.002%? If the distortion is of the exact same type, I would say yes. But if the distortion is of different types, it could be the other way around. The 0.001% could be particularly nasty while the 0.002% could be rather unobtrusive.
I find this field interesting, as a couple of years ago, I knew very little about it. But one day I worked on a circuit, which actually gained sound quality when I traded a tiny bit of one type of distortion for more of another type. When listening, it was possible to detect that the altered circuit probably had a little more distortion, but it sounded like it was in balance. And that balance let the musical message pass through much better than with the original version. This result proved to be repeatable in many other circuits. The best approach that I've found so far is to start with a circuit with as little distortion as possible, and then tune it until it's in balance. Before the tuning, it can sound really disappointing. Even if in theory it's the most "clean" solution, with the least total amount of distortion.
The jitter problems usually create distortion levels that are very, very low. So those who rely on distortion measurments can sometimes claim their effect is irrelevant (and since loudspeakers create lots of distortion in comparison, they will claim those are the weak links in the chain). We who use the Tune Method often find that products that measure extremely well don't sound any good. That's psychoacoustics; we may not notice some big flaws in the reproduction, but can get really annoyed by tiny ones.
The same applies to analogue amplifiers. Is one with 0.001% distortion better than one with 0.002%? If the distortion is of the exact same type, I would say yes. But if the distortion is of different types, it could be the other way around. The 0.001% could be particularly nasty while the 0.002% could be rather unobtrusive.
I find this field interesting, as a couple of years ago, I knew very little about it. But one day I worked on a circuit, which actually gained sound quality when I traded a tiny bit of one type of distortion for more of another type. When listening, it was possible to detect that the altered circuit probably had a little more distortion, but it sounded like it was in balance. And that balance let the musical message pass through much better than with the original version. This result proved to be repeatable in many other circuits. The best approach that I've found so far is to start with a circuit with as little distortion as possible, and then tune it until it's in balance. Before the tuning, it can sound really disappointing. Even if in theory it's the most "clean" solution, with the least total amount of distortion.
The jitter problems usually create distortion levels that are very, very low. So those who rely on distortion measurments can sometimes claim their effect is irrelevant (and since loudspeakers create lots of distortion in comparison, they will claim those are the weak links in the chain). We who use the Tune Method often find that products that measure extremely well don't sound any good. That's psychoacoustics; we may not notice some big flaws in the reproduction, but can get really annoyed by tiny ones.
Does this conclude that a clock uppgrade to a Superclock 4 on a squeezebox classik is a 100% gambling?lejonklou wrote:The most similar recreation would be the one with the lowest amount of distortion, yes. But will it sound better than the one with slightly more distortion? Perhaps not.
The same applies to analogue amplifiers. Is one with 0.001% distortion better than one with 0.002%? If the distortion is of the exact same type, I would say yes. But if the distortion is of different types, it could be the other way around. The 0.001% could be particularly nasty while the 0.002% could be rather unobtrusive.
I find this field interesting, as a couple of years ago, I knew very little about it. But one day I worked on a circuit, which actually gained sound quality when I traded a tiny bit of one type of distortion for more of another type. When listening, it was possible to detect that the altered circuit probably had a little more distortion, but it sounded like it was in balance. And that balance let the musical message pass through much better than with the original version. This result proved to be repeatable in many other circuits. The best approach that I've found so far is to start with a circuit with as little distortion as possible, and then tune it until it's in balance. Before the tuning, it can sound really disappointing. Even if in theory it's the most "clean" solution, with the least total amount of distortion.
The jitter problems usually create distortion levels that are very, very low. So those who rely on distortion measurments can sometimes claim their effect is irrelevant (and since loudspeakers create lots of distortion in comparison, they will claim those are the weak links in the chain). We who use the Tune Method often find that products that measure extremely well don't sound any good. That's psychoacoustics; we may not notice some big flaws in the reproduction, but can get really annoyed by tiny ones.
It's looks like a pure technical upgrade.lejonklou wrote:Sorry, I don't know anything about Superclock 4 or the company behind it.
If anyone else does, I hope they will chime in.
I think this text reveals how much they know about tune dem...
"Once upgraded the Squeezebox Classic the overall presentation is cleaner, more open, with a marked reduction in digital artefacts such as glare and hardening. Detailing and dynamics are improved, the music gains substance and depth with low frequencies benefiting from increased weight with the highs sounding sweeter, more natural.
Squeezebox Classic (SB3) products can now be upgraded with the Audiocom Superclock 4. This modification can be installed as a stand-alone upgrade or in conjunction with our level-1 or level-2 modifications.
The factory clock fitted in the Squeezebox players is a ‘Pierce’ type using a 74HCU04 inverter chip. This type of oscillator is simple and easy to implement but also very noisy, jittery. With jitter minimized, low to mid frequencies gain substance, high frequencies are cleaner, image focus and dynamics are significantly improved."
...that is zero
but i was thinking that maybe a less jittery clock, would, mean less timing errors and a more musical presentation. But that depends whether they've succeded to make a balance in reducing all types of jitter.?
since i conclude from what you wrote above, that there's no correlation between an expensive digital part and musicality.
A few years ago I upgraded a Nakamichi CD 4 with an LClock XO (bought a used one cheap on the net) and it was a significant upgrade. The CD player got much more rythmic and easy to listen to.
But I do not know if it is a good idea to buy a Superclock 4($295!) to the cheaper SB3. Maybe better to buy a used Linn Sneaky DS?
But I do not know if it is a good idea to buy a Superclock 4($295!) to the cheaper SB3. Maybe better to buy a used Linn Sneaky DS?
Thanks for the input!TMV wrote:A few years ago I upgraded a Nakamichi CD 4 with an LClock XO (bought a used one cheap on the net) and it was a significant upgrade. The CD player got much more rythmic and easy to listen to.
But I do not know if it is a good idea to buy a Superclock 4($295!) to the cheaper SB3. Maybe better to buy a used Linn Sneaky DS?
A SB + clock upgrade is still half price, though, if one prefers the SB user interface.
But you're right a sneakt will probably have a lot more worth for the money, but maybe one could buy a used SB , and it is really interesting if there is a some sort of correlationship between an expensive clock and more musical involving presentation.
Since there's no correlation between an expensive analog cd-player and musical presentation.