Hi Mark,
That's fair. You can only do your best. Your friend probably can measure some effects. Other than that, use logic accepting the jitter in the source material and know it presents some kind of limitation. That's all.
A better oscillator is great, and pursuing this is a good thing. Just be realistic in the application. Understanding limiting factors is what makes an engineer worth what they pay him/her. Otherwise you end up chasing things that don't matter - sometimes at great expense. That is really all I am saying.
-Chris
That's fair. You can only do your best. Your friend probably can measure some effects. Other than that, use logic accepting the jitter in the source material and know it presents some kind of limitation. That's all.
A better oscillator is great, and pursuing this is a good thing. Just be realistic in the application. Understanding limiting factors is what makes an engineer worth what they pay him/her. Otherwise you end up chasing things that don't matter - sometimes at great expense. That is really all I am saying.
-Chris
I have a nice picture of a 19" rack with about a dozen of timing units with the highest
grade 8607 in each of them. I took the photo when all of them displayed the same
funny date and time, cannot publish it because it discloses a customer's clean room.
My own attempt to buy a 8607 was a failure because it did not even draw current
for the oven. Luckily, the seller took it back without much ado.
Obviously, it is at least possible to develop a state of the art oscillator by measurement
and then strutting around about the continuous eargasm that it invokes.
Developing by eargasm and then demonstrating an outstanding phase noise
achievement has been shown to be a less than successful endeavour.
Gerhard
Hi Chris,
sorry but I don't think you can measure the phase noise or the jitter of SOTA oscillators with your HP 5372A, it has a jitter floor of around 150ps, much worse than the DUTs.
Andrea
As many others noted, any technical discussions are impossible to carry on in this thread. Lacking a dedicated Vendor Forum (where Andreas and his DIY partners would have moderation rights to clean up anything they dislike), I would think the Group Buys forum would be more appropriate. This would also clean some suspicions about the initiator scope. Myself, I would have no problem with anything it is posted here, once it would be posted in the right context.
Hi Gerhard,
as you know we have developed by measurements and it took very long time and several iterations.
About the audibility of the difference in digital audio sincerely I'm getting tired.
Everyone is free to use the oscillator who wants, while Me and Roberto we have our own approach to audio devices designing and we keep designing following our beliefs.
We are not looking for someone's approval, we are designing for ourselves.
Andrea
Oh, the ego-surfing of JC took a close to total hit when the thread was
renamed to "Black Hole" from something that contained HIS name.
And fame in the target group is a precondition for revenue. One must be
visible to one's sheeple.
Gerhard
Last edited:
Hi Andrea,
Quite often you set the tone.
I am using an external 10 MHz reference for the 5372A. But that doesn't matter if I am only trying to qualify your oscillator - which I am not. I have far more than enough resolution for proving what is required for digital reproduction. You're missing the entire point.
I am taking your claims for phase noise at face value because it doesn't matter for this discussion. If true, you are sitting on a fortune where industry is concerned. The fact they are not beating a path to your door, or having the military looking at your work gives me pause. Timing and frequency are critically important in GPS positioning. Also in advanced physics and other pursuits.
The requirements for digital audio reproduction are pretty low compared to what I have on my bench now. Gerhard has far better than I do and I will freely admit that. But what I have is far in excess of what is needed for audio work. In fact, the lab I have is so overqualified for audio work it is just plain silly. This means I have no doubt or uncertainty about my measurements at all. I do have some metrology training.
The way I see it, you have to determine realistic requirements for the job at hand. So far you have not. Have you had a real calibration lab characterize your oscillator yet? Probably need one with primary standards, and I believe you would want to be certified. Without such tests and certification you are merely talking out of your hat. At the levels you are talking about, no you cannot hear any improvements or impairments compared to what is on my bench, never mind a timing lab.
That's the real problem with listening to your own work, or listening all together. Your brain will fool you. Expectation bias is a real problem that you can't get around. When I make improvements, I send the equipment out to a different location to different people. I am never present. They do not know whether to expect improvement or not. A null response to the question of performance is good as well. So I have honest people assessing my work - because I trick them occasionally and they know that. The truth is more valuable to me than a pat on the back.
So, try to be honest with yourself. Know what the characteristics are, don't assume or guess. Anything less and you are living a lie and misleading others. Also, look at the oscillators used in the encoding equipment for starters. Once you pass that bar by a factor of at least 4, I sincerely doubt any further improvement will be audible. That is because the encoding instabilities are now the determining factor. You can`t go back in time and space to make it better, it is what it is. Trying to convince people that something better than this on the decoding end is misleading them.
Don`t worry Andreas. Many CD player clocks are not very good. You can still sell your upgrades with a positive effect, an improvement. It`s just that pushing the claims to extreme ends does not help your customers. I should check one day, but the jitter introduced by some DSP chips may far exceed some stock clocks, and certainly undo the best clock in the world. Certainly possible skew between the various output clock signals may also greatly limit improvement too. In other words, what you feed the DSP and what the DAC gets may be two completely different things. You have to look at the entire system, not just the clock source.
-Chris
Quite often you set the tone.
I am using an external 10 MHz reference for the 5372A. But that doesn't matter if I am only trying to qualify your oscillator - which I am not. I have far more than enough resolution for proving what is required for digital reproduction. You're missing the entire point.
I am taking your claims for phase noise at face value because it doesn't matter for this discussion. If true, you are sitting on a fortune where industry is concerned. The fact they are not beating a path to your door, or having the military looking at your work gives me pause. Timing and frequency are critically important in GPS positioning. Also in advanced physics and other pursuits.
The requirements for digital audio reproduction are pretty low compared to what I have on my bench now. Gerhard has far better than I do and I will freely admit that. But what I have is far in excess of what is needed for audio work. In fact, the lab I have is so overqualified for audio work it is just plain silly. This means I have no doubt or uncertainty about my measurements at all. I do have some metrology training.
The way I see it, you have to determine realistic requirements for the job at hand. So far you have not. Have you had a real calibration lab characterize your oscillator yet? Probably need one with primary standards, and I believe you would want to be certified. Without such tests and certification you are merely talking out of your hat. At the levels you are talking about, no you cannot hear any improvements or impairments compared to what is on my bench, never mind a timing lab.
That's the real problem with listening to your own work, or listening all together. Your brain will fool you. Expectation bias is a real problem that you can't get around. When I make improvements, I send the equipment out to a different location to different people. I am never present. They do not know whether to expect improvement or not. A null response to the question of performance is good as well. So I have honest people assessing my work - because I trick them occasionally and they know that. The truth is more valuable to me than a pat on the back.
So, try to be honest with yourself. Know what the characteristics are, don't assume or guess. Anything less and you are living a lie and misleading others. Also, look at the oscillators used in the encoding equipment for starters. Once you pass that bar by a factor of at least 4, I sincerely doubt any further improvement will be audible. That is because the encoding instabilities are now the determining factor. You can`t go back in time and space to make it better, it is what it is. Trying to convince people that something better than this on the decoding end is misleading them.
Don`t worry Andreas. Many CD player clocks are not very good. You can still sell your upgrades with a positive effect, an improvement. It`s just that pushing the claims to extreme ends does not help your customers. I should check one day, but the jitter introduced by some DSP chips may far exceed some stock clocks, and certainly undo the best clock in the world. Certainly possible skew between the various output clock signals may also greatly limit improvement too. In other words, what you feed the DSP and what the DAC gets may be two completely different things. You have to look at the entire system, not just the clock source.
-Chris
Chris, all the good point you made above (including the jitter in the on silicon chip clock processing and distribution) were previously made by myself, chris719, Gerhard and many others over the past months. All technical arguments were rejected, so I have no reason to believe yours will be heard.
But that didn't stop someone from judging someone's oscillator solely based on measurements.
Their anecdotes are irrelevant outside of the room it happened on that day. That is unless it is an objective test, proctored and documented.
Hi syn08,
Yes, you`re right. I have posted nothing new. I was hoping the context makes it easier for others to follow. I am appealing to the objective nature in people to do some thinking before simply accepting arguments at face value. As long as people think first, whatever they decide is fine.
-Chris
Yes, you`re right. I have posted nothing new. I was hoping the context makes it easier for others to follow. I am appealing to the objective nature in people to do some thinking before simply accepting arguments at face value. As long as people think first, whatever they decide is fine.
-Chris
But then those with business agenda will (& have been) construct roadblocks to steer the traffic towards their interest. "sheeple" was a mentioned earlier and yes, that is the target demographics by these business posters. The difference between real sheep and sheeple is that sheep will go where they are redirected by the right shepherd to good path while sheeple will sometimes fight the right shepherd.
Hi Chris,
you can argue about the audibility of the difference using a better clock, but you can't argue about the reliability of my instrument.
So, we have real calibration lab to characterize our oscillator.
We have used our tools to develop the oscillators, we have measured the final result, I have posted the phase noise plots several times, and as you know Joseph K has measured the same oscillator getting the same results, maybe better.
So, please stop arguing about the reliability of our measurements, unless someone demontrates the opposite.
To be short on the other point we are just looking at the entire system, not just the clock source, since we have already designed the front end (the FIFO Lite buffer) and the DAC).
The source does not matter if we had developed a true FIFO buffer which well isolates the source from the DAC making them operating in different time domains.
Andrea
you can argue about the audibility of the difference using a better clock, but you can't argue about the reliability of my instrument.
So, we have real calibration lab to characterize our oscillator.
We have used our tools to develop the oscillators, we have measured the final result, I have posted the phase noise plots several times, and as you know Joseph K has measured the same oscillator getting the same results, maybe better.
So, please stop arguing about the reliability of our measurements, unless someone demontrates the opposite.
To be short on the other point we are just looking at the entire system, not just the clock source, since we have already designed the front end (the FIFO Lite buffer) and the DAC).
The source does not matter if we had developed a true FIFO buffer which well isolates the source from the DAC making them operating in different time domains.
Andrea
As usual, you are missing the point entirely.
Can you guarantee that each and every oscillator board meets the advertised performance benchmark (that already famous noise plot you are posted multiple times)? More precisely, are you measuring each and every SC cut crystal you are delivering and every assembled and finished board? Or are you simply extrapolating the performance of a "best case" prototype build some time ago?
You were asked by myself, 1audio, and others, several times, to send random oscillator sample boards for an independent evaluation (Joseph and anybody else of your fan club do not qualify), and you refused to do so. I could do it as well, for anything over 10MHz, as you know I have a HP5052A, but then I decided I also do not qualify as an "independent", since I am obviously a biased skeptic.
P.S. I can only hope this thread won't be closed, but only moved to the appropriate board.
Last edited:
yeah this turned sour quickly in lots of directions, I hope the "good" people stay around :/
I guess it's an internet problem, people feel like they can say what they want without regarding other people. Seems like constructive critique is an art form nowadays ^^
That's not meant as an attack to anybody I just hope the guys who make the effort to post their measurements and listening experiences are not discouraged as this is the diyaudio forum so it's kind of the heart of the thing
I guess it's an internet problem, people feel like they can say what they want without regarding other people. Seems like constructive critique is an art form nowadays ^^
That's not meant as an attack to anybody I just hope the guys who make the effort to post their measurements and listening experiences are not discouraged as this is the diyaudio forum so it's kind of the heart of the thing
My personal observation from this topic the last few days:
-All the people here that criticize Andrea’s clock achievements have not tested them with their expensive instruments and self claimed knowledge.
-Nor have they listened to them with their own ears (oh, how dumb of me: ears are not to be trusted
)
-They also do not contribute to this diy community with constructive advise
-Nor do they offer a better product or design for this community.
-They ridicule Andrea and users of his products with their posts.
For the average audio diy-er there are some clock choices in the market, not many, and the designs from Andrea certainly stand out from them. That is not only my personal observation, there are a lot of people on this forum with similar experience.
So if the critics cannot deliver anything better (or even similar) to Andrea’s designs can they please stop this endless discussion?
It has no purpose other than exposing your own ego.
-All the people here that criticize Andrea’s clock achievements have not tested them with their expensive instruments and self claimed knowledge.
-Nor have they listened to them with their own ears (oh, how dumb of me: ears are not to be trusted
)-They also do not contribute to this diy community with constructive advise
-Nor do they offer a better product or design for this community.
-They ridicule Andrea and users of his products with their posts.
For the average audio diy-er there are some clock choices in the market, not many, and the designs from Andrea certainly stand out from them. That is not only my personal observation, there are a lot of people on this forum with similar experience.
So if the critics cannot deliver anything better (or even similar) to Andrea’s designs can they please stop this endless discussion?
It has no purpose other than exposing your own ego.
You can trust your ears if the test is objectively set up so that only your ears are being used for the evaluation of audible traits of audio gear. It's not be trusted when used along with sight and bias (we all have it and cannot control it on our own), a.k.a. subjective listening evaluation.-Nor have they listened to them with their own ears (oh, how dumb of me: ears are not to be trusted
You see, post like that is what prompts the kind of response quoted below, no matter the behavior of Andrea.
No problem at all.
I have to ask the members for whom I have done the measurements if they agree if I publish the phase noise plot of their oscillators.
Audio curiosity, dirkwunderl, Jooseph K, NicMac, pmeij, sligolad, tbrowne
may I publish the measurements of your oscillators?
What you measured for others can't be considered as "independent", in particular because that was a paid service you and your DIY partners are offering.
Friendly warning: you are digging it in an ever deeper hole; a quick search shows some (not all) people out of your list with a handful (5-10) of posts, mostly related to guess what? The group sales of your boards. Happy customers, I would guess. Also interesting, Joseph K has the tools (or at least he claimed so, he posted his measurement results), now you are claiming you did measurements for him (allegedly not for free).
- Status
- Not open for further replies.