The same way as in all the listening tests I've run. Line output direct into PC soundcard and captured with Audacity. It may not be perfect but any characteristics of the recording chain are present for both samples.
The use of large capacity caps have to be done however carefully, and not just soldering such caps on the power pins of whatsoever chip.
Indeed such decoupling should not be used on digital devices, as it foolish the whole system.
These caps and the approach is useful (my opinion) for analogue stages, which are not supposing to be a part of a logic sequence in that system.
Large capacities may not be only a issue for power up sequences, but an possible equal issue for power down. Devices decoupled so, it can still be powered even hours after the main power it disappear from the rest of the system.
The circuits and even the system these are part from, have to be carefully analysed before use of such caps.
I use large decoupling capacities on ES9018, but only on its analogue power rails. It work very well and I never ever had any issue. The digital part of the chip it start quite fast, the clock it do its function inside the chip digital stage, and everything is works fine. I allays use discharging circuits for power down, or I ensure the caps normal discharging rate is in the safe zone for the rest of the system.
But again, I appreciate the use of large capacities for experiments with this cap filter as not a must in itself. It can well be an improvement of the approach applied in a second phase.
Last edited:
Things added are not critical, but critical small differences are. It is those small differences that can be the difference between good & great. They can remove differences if they exist because, given the descriptions of the changes, they are "small" details that can make a big difference.
How good is your PC sound card?
dave
I listened with an Asus Xonar and Sennheiser 650's and could not tell any difference of note, certainly not at the extreme level as some comments on the mod. No preference on either.
Perchance a run through foobar (or other software's) ABX plugin to put any number on hearing the difference in Mooly's files?
I'll try at home on my HD280's.
In absolute terms I wouldn't like to say (Dell Vostro PC and inbuilt sound card) but it has proved capable in past listening tests.
Thanks Scott.
I didn't use Foobar on this occasion. I just listened via my normal amp and speakers... many times it must be admitted.
Panacea? I don't think that has been suggested. But you do have similar concerns that I had. My idea for using them is what I believe they need to do, whether that is what they are doing is up to interpretations, what does it fix?
But your concerns:
Indeed, that was of concern when using the first time. Some tests were set up before feeling comfortable using them. The Oppo 105D carries all those potentials for problems, discharge affecting logic, the potential delay on turn-on etc. But so far no sign whatsoever - I know Coris is using even higher values than I recommended at the beginning of the thread.
So even before using them, with a lab supply it is fairly easy to set up tests and study their behaviour or misbehaviour. Using a scope across the supercap itself and another floating scope across a series 1R resistor (fraction of ESR), with 3.3V and 5V, and they behave surprisingly well and are also well damped - and they do have a lot of dielectric absorption and that seems to be a plus in this instance. The working voltage does drop and the logic does not play up when turning back on.
I got an engineer friend of mine also conducting his own tests - and he ended up using them in a USB DAC. His summing up agreed with mine (he reads this thread). I also have a friend who actually works for a company that develops supercap technology. And he was extremely surprised when I got his DAC and applied them - unorthodox use, but he was not complaining about the result.
This was my number one concern - I didn't want to see things blowing up. But this is where the ESR defines the peak current when the cap is in a depleted state. The two-scopes revealed what I had hoped, with 75R the peak currrent is limited to <44mA @ 3.3V and <66mA @ 5V rails. This is not going to be a problem and besides, these regulators have self-protection too, but I don't think they are triggered.
So says we all... I hope.
As Hilary, a scientist friend of mine for nearly thirty years said, "People like you discover these things and us scientists have to figure it out."
Sometimes you just have to try the unorthodox - especially when new components like this becomes so accessible. It's then a no-brainer to try. But I did have a specific idea in mind before I tried them - and that Paul Miller measurement is at the heart of it and also the concern that LF jitter is the worst kind.
Exactly. The Paul Miller test shows this is a real phenomenon - And look at the amplitude as it clings up the carrier like wine in a wine glass. That amplitude is far greater than regular jitter and largely masked by the carrier. Paul Miller also made the observation that the two measurements overlaid as shown, had virtually identical jitter in 'conventional' measurements (I have a friend who hates the word jitter - but it's what we use) and that means the carrier is largely masking jitter, and of the highest amplitude because of the nature of 1/F - no wonder then that this more than anything is what makes digital sound digital (not excluding other things). Certainly Miller said that it was very audible.
He even gave it a name: "Low Rate Uncorrelated or Noise-Like Jitter"
Last edited:
Jitter is not special to audio, and there is no need to make up a name for something that has been well known in the comms business for decades. 1/f jitter exists on all clocks at some level it's basic physics of uncertainty. The sidebands on the 1kHz test tone on an LP are orders of magnitude worse.
BTW I came across a very interesting article by Dave Greisinger where he tried to quantify the effect of turbulence and non-uniformity of density and temperature in a listening environment. This would translate into, as you say, "Low Rate Uncorrelated or Noise-Like Jitter". Your listening room is far from stationary on a sub-Hz time scale.
I'll pass on the jitter stuff. Any test like Mooly's will inject his sound card jitter and the users sound card jitter so there is nothing but a face to face with two units left to make any determination. Like faeries you can't take it's picture.
Last edited:
Here are the files for the 'final' modification outlined in post #743. You still don't know which of the first set were original and which were of the first modification tried.
See what you think to this one by comparing it to the others. Joe... you need to listen to all these and write your impressions down. I have done that as I have gone along and I'll post what I wrote and which file is which when we are all done.
Also, something quite surprising has just been brought to my attention with these files... so listen closely to them.
Modification 2
(with hindsight I should have made more files of the un-modified player, perhaps some different tracks and then we could have shuffled things round a bit more. As it is it relies on honest audible impressions, in other words as you hear it)
See what you think to this one by comparing it to the others. Joe... you need to listen to all these and write your impressions down. I have done that as I have gone along and I'll post what I wrote and which file is which when we are all done.
Also, something quite surprising has just been brought to my attention with these files... so listen closely to them.
Modification 2
(with hindsight I should have made more files of the un-modified player, perhaps some different tracks and then we could have shuffled things round a bit more. As it is it relies on honest audible impressions, in other words as you hear it)
Which album was the genesis track from? Not familiar with that one.
Initial listen on headphones inconclusive. Everytime I thought I heard something different going back showed it was there. Will try again on the speakers at home.
Initial listen on headphones inconclusive. Everytime I thought I heard something different going back showed it was there. Will try again on the speakers at home.
- Status
- Not open for further replies.