@Markw4 You must argue your points on technical grounds and without making it personal.
Post #3477 breaks no rules and there are no direct quotes of anything you said. Quote what you feel is in error and challenge it from a technical perspective.
That applies to everyone, don't make arguments personal.
We draw a line under this now. Finished.
Technically speaking, that's not necessarily true. I already explained why in #3,470. If the explanation is not clear then please ask.
Technically speaking, I made no such suggestion about white noise. Again, if what I wrote was not understood or not clear then better to ask.
Technically speaking, standard measurements of phase noise may or may not show a difference. I have attempted to explain how that can happen based on how phase noise measurement works, don't know if the explanation was clear/understood. That said, I do believe all physical phenomena are in principle measurable. I do not believe all possible test instruments for physical phenomena have been developed as of this point in human history.
Also, I believe Marcel understood the points I was trying to make. Can't know about other folks if they don't ask; I can only assume they understand the explanations if they don't ask.
For people who apparently don't understand the points yet argue they are wrong, or people who summarily dismiss the points without ever trying to understand, well, its not very clear to do about that.
Last edited:
You suggested that it is akin to the differences in different types of white noise without any evidence pointing to that. So it is speculation at this stage.
Another speculation at this stage as you have not shown any phase noise measurements with/without ferrite beads.
Please don't make it personal or this is likely to escalate again.
If the question were rephased into a technical one rather than personal, then I might have an technical response.
Again, this appears to be going in a personal direction. If there is a purely non-personal technical question then I might have a technical response.
What I said is an analogy which is based on mathematical truth. It has little or nothing to do with speculation.
If it is not clear why that is, then please ask. I can't know if I was not clear if nobody asks.
It is speculation unless you can show that such "noise" exists in this case.
Furthermore you have not given any evidence about the audibility. Your claim alone does not merit further analysis.
It is a well known part of physics that magnetic materials have hysteresis effects. It is also well known that hysteresis effects can produce noise/distortion if a current flowing in a wire interacts with the magnetic material. In order for a ferrite bead to suppress noise there must be such interaction. All of that is well known basic physics. It is not speculation.
Some time out from this thread for some of you.
Markw4, do not change the meaning of the content of quotations by underling and/or making text in the quote bold or omitting key words as you have done. This totally changes the original meaning and intent.
Some recent measurements (a bit down in the weeds unfortunately) of the RTZ DSD dac:
Took a scope shot of "clocksup_F" across C85. Passive probe with spring ground. Here is the pic:
Measurement location:
...Looked again a bit later with an active probe. There is a sort of hump artifact that I focused in on. Turned out was some aliasing with AC line ripple only when the active probe ground was connected to the dac ground. Same thing with two different active probes. Probes are fine when not connected to dac ground. Thus, I kind of discount the hump and look at the RF spikes which are probably not AC line related. Also, variations in spike shapes may be due to multiple sources of noise, possibly including some measurement noise due to my location on top of a hill overlooking a large valley with many RF noise sources (including multiple airports, etc.). However, the particular spike time intervals suggest some relation to the dac operation (probably should have aligned the cursors a little better though).
EDIT: Thing is I don't see the spikes when I ground the scope probe spring ground to the dac ground, then put the probe tip at the same ground point. There is some noise, but not the spikes that look clock frequency related. Also, the above image is from a single sweep so no noise averaging.
EDIT 2: Only reason for looking at this was to see if anything to account for a little bit of something still slightly off about the sound, still a little bit "raw." Just wondering where its coming from.
Took a scope shot of "clocksup_F" across C85. Passive probe with spring ground. Here is the pic:
Measurement location:
...Looked again a bit later with an active probe. There is a sort of hump artifact that I focused in on. Turned out was some aliasing with AC line ripple only when the active probe ground was connected to the dac ground. Same thing with two different active probes. Probes are fine when not connected to dac ground. Thus, I kind of discount the hump and look at the RF spikes which are probably not AC line related. Also, variations in spike shapes may be due to multiple sources of noise, possibly including some measurement noise due to my location on top of a hill overlooking a large valley with many RF noise sources (including multiple airports, etc.). However, the particular spike time intervals suggest some relation to the dac operation (probably should have aligned the cursors a little better though).
EDIT: Thing is I don't see the spikes when I ground the scope probe spring ground to the dac ground, then put the probe tip at the same ground point. There is some noise, but not the spikes that look clock frequency related. Also, the above image is from a single sweep so no noise averaging.
EDIT 2: Only reason for looking at this was to see if anything to account for a little bit of something still slightly off about the sound, still a little bit "raw." Just wondering where its coming from.
Last edited:
Using the same passive probe technique on one of the output flip-flop film bypass caps on my reclocker board in-use with Marcel's dac, noise looks like this (at 20mv/div for this one):
What I am thinking is that the RTZ DSD dac might make good use of some kind of a isolator/reclocker circuit between the signal processing part of the circuitry and the dac shift registers. Of course, only speculation at this point. Wouldn't be surprised if some benefit though.
What I am thinking is that the RTZ DSD dac might make good use of some kind of a isolator/reclocker circuit between the signal processing part of the circuitry and the dac shift registers. Of course, only speculation at this point. Wouldn't be surprised if some benefit though.
Last edited:
Thank you! I for one am learning a lot from many of the participants in this thread. At some point I would really like to do some prototyping of my own, but still collecting the appropriate equipment.
For purposes of listening tests for this design, can good quality headphones (I mean really good stuff) make up for a lack of a good room and excellent speakers? Headphone amp seems to be taken care of, via alexcp's Omicron, but I'm way down on the totem pole when it comes to experience.
For purposes of listening tests for this design, can good quality headphones (I mean really good stuff) make up for a lack of a good room and excellent speakers? Headphone amp seems to be taken care of, via alexcp's Omicron, but I'm way down on the totem pole when it comes to experience.
Last edited:
I am having a similar experience with an I/V converter in a CD player sounding the same... a little bit "raw", otherwise it doesn't sound thin or lacking in body, weight, harmonic structure or overall dynamics. My question is it necessarily wrong... perhaps at the limits of the DAC perhaps? Don't know.Some recent measurements (a bit down in the weeds unfortunately) of the RTZ DSD dac:
Took a scope shot of "clocksup_F" across C85. Passive probe with spring ground. Here is the pic:
Measurement location:View attachment 1337579
...Looked again a bit later with an active probe. There is a sort of hump artifact that I focused in on. Turned out was some aliasing with AC line ripple only when the active probe ground was connected to the dac ground. Same thing with two different active probes. Probes are fine when not connected to dac ground. Thus, I kind of discount the hump and look at the RF spikes which are probably not AC line related. Also, variations in spike shapes may be due to multiple sources of noise, possibly including some measurement noise due to my location on top of a hill overlooking a large valley with many RF noise sources (including multiple airports, etc.). However, the particular spike time intervals suggest some relation to the dac operation (probably should have aligned the cursors a little better though).
EDIT: Thing is I don't see the spikes when I ground the scope probe spring ground to the dac ground, then put the probe tip at the same ground point. There is some noise, but not the spikes that look clock frequency related. Also, the above image is from a single sweep so no noise averaging.
EDIT 2: Only reason for looking at this was to see if anything to account for a little bit of something still slightly off about the sound, still a little bit "raw." Just wondering where its coming from.
Interestingly the noise levels of my I/V are also not low, though I wouldn't want to trade off anything else sonically to make it sound less raw or noisy. On the other hand I would like to trade my RIGOL scope for yours (5 G Sa/s)... and for your active probes...
For Marcel's dac when I said a little "raw," I think its probably so small most people would have trouble hearing it unless quick A/B and on a very good system. Its very slight at this point, given the changes to the external circuitry made so far. Other than that, the sound stage of this dac can be exceptional, but the output stage problem will still need to be solved for most people (common mode reduction theory may have to be set aside too, no matter how difficult it is to believe; IMHO one has to trust one's ears on this issue). The transformers in use here are proprietary prototypes. With that problem solved and the other external stuff in place I think it can already compete well with some of the best dacs available in the several thousand dollar price range. I just want to get it up to where it beats all those dacs, not just competes. Its mostly nitpicky stuff at this point, however IMHO its the small details that mostly tend to differentiate the best dacs.
On the subject of the scope, it was painful to acquire, trust me. If you are willing to accept the burden then you could probably have one too.
On the subject of the scope, it was painful to acquire, trust me. If you are willing to accept the burden then you could probably have one too.
Last edited:
Yes. My wife got me thinking about that scope. I suggested switching out our cats FancyFeast for servings less fanciful. She said it would be alright if I ate it too. I'm thinking dog milk bones weren't all that bad. Some time ago I dared my little nephew to eat one with me. He didn't think I would... being very resistant in doing so after I did. He finally relented after repeated declarations stating we had a deal and that I could never trust him again. Later on my wife commented about having to clean up our bathroom... too which I was silent.
My comments referencing that my digital CD source sounded a little "raw" was more to highlight this as the prominent sonic attribute of nature most negative to me, though to others the presentation seems likely to be more alien in its dynamic whole, seemingly would bring that more into question than its rawness. Bear in mind I have difficulty in describing most any sonic character, but it seems to me it captures the power envelope exceptionally well, meaning it holds together the full bodied character of individual presentations within a piece dynamically, without leaning out, much smearing, veiling or disintegrating. I don't know the character of Marcel's FIRDAC or others like it, but expect it to have a similar character, for better or worse.
To be clear I never perform AB(x) testing and don't plan to (for readers to weight such declaration accordingly), hence what is revealed is only opinion. For me to conduct AB(x) testing is a don't care case of outcome stemming from the belief, being realistic or otherwise, that complex differences exist that are still not necessarily quantifiable or extractable of implications from measurements. In my intention to advance the technology (hopelessly or otherwise) there seems most always the requirement to choose one of two paths. If in the evidence performing AB(x) testing the differences can't be resolved (or can't be resolved of one being necessarily better or worse than the other), this doesn't change the requirement to still choose a path to the best of ones ability. It doesn't matter in the delusion, knowing or guessing.
For reasons perhaps in part understood the CD player was IMO markedly better sonically than the streaming vis a Wii Mini into an SMSL SU-1 and also better than the SU-9. Streaming using the SU-1 was like listening through a thickish veil with dynamics being lost, perhaps in part because of actual low distortion preventing a kind of punch through excitement that higher distortion devices seem often to provide. This coupled with a veiled lossy background is nothing to keep. In contrast the SU-9 was considerable clearer in the background, hence had a much better dynamic presentation, though the artists seemed more "going through the motions" in comparison to the much more intimate and dynamic liveness of the CD source. Perhaps more likeable distortion from the I/V in the CD player, this is not clear yet though it is doubtful the distortion is that high.
As a guess there seems hasn't been enough study into the mechanisms of impedance of sonic implications, specifically in lowering network impedances. Bear in mind that impedance and noise can be closely aligned, with current emphasis being to some degree perhaps misdirected on noise alone. In being completely off the wall, perhaps zero impedance has some implication in the correlation to a black hole.
My comments referencing that my digital CD source sounded a little "raw" was more to highlight this as the prominent sonic attribute of nature most negative to me, though to others the presentation seems likely to be more alien in its dynamic whole, seemingly would bring that more into question than its rawness. Bear in mind I have difficulty in describing most any sonic character, but it seems to me it captures the power envelope exceptionally well, meaning it holds together the full bodied character of individual presentations within a piece dynamically, without leaning out, much smearing, veiling or disintegrating. I don't know the character of Marcel's FIRDAC or others like it, but expect it to have a similar character, for better or worse.
To be clear I never perform AB(x) testing and don't plan to (for readers to weight such declaration accordingly), hence what is revealed is only opinion. For me to conduct AB(x) testing is a don't care case of outcome stemming from the belief, being realistic or otherwise, that complex differences exist that are still not necessarily quantifiable or extractable of implications from measurements. In my intention to advance the technology (hopelessly or otherwise) there seems most always the requirement to choose one of two paths. If in the evidence performing AB(x) testing the differences can't be resolved (or can't be resolved of one being necessarily better or worse than the other), this doesn't change the requirement to still choose a path to the best of ones ability. It doesn't matter in the delusion, knowing or guessing.
For reasons perhaps in part understood the CD player was IMO markedly better sonically than the streaming vis a Wii Mini into an SMSL SU-1 and also better than the SU-9. Streaming using the SU-1 was like listening through a thickish veil with dynamics being lost, perhaps in part because of actual low distortion preventing a kind of punch through excitement that higher distortion devices seem often to provide. This coupled with a veiled lossy background is nothing to keep. In contrast the SU-9 was considerable clearer in the background, hence had a much better dynamic presentation, though the artists seemed more "going through the motions" in comparison to the much more intimate and dynamic liveness of the CD source. Perhaps more likeable distortion from the I/V in the CD player, this is not clear yet though it is doubtful the distortion is that high.
As a guess there seems hasn't been enough study into the mechanisms of impedance of sonic implications, specifically in lowering network impedances. Bear in mind that impedance and noise can be closely aligned, with current emphasis being to some degree perhaps misdirected on noise alone. In being completely off the wall, perhaps zero impedance has some implication in the correlation to a black hole.
A couple more measurements to share today. Measurement point is C48, using passive probe with spring ground:
This is with my SE output hooked up:
Thought maybe the unbalanced load could be causing some extra noise, so here is a pic with no loading on the dac outputs:
This is with my SE output hooked up:
Thought maybe the unbalanced load could be causing some extra noise, so here is a pic with no loading on the dac outputs:
If the bit clock frequency was 11.575 MHz, both waveforms consist almost completely of bit clock harmonics.
True. However given this is in an analog part of the circuitry, is the noise magnitude and ringing of negligible significance?
To look at it from another perspective, would something like this be "okay" on, say, op amp power pins so long as it is clearly related to the signal being processed?
To look at it from another perspective, would something like this be "okay" on, say, op amp power pins so long as it is clearly related to the signal being processed?