GUSTARD DAC-R26 Balanced Decoder R2R+1Bit Dual Native Decoding Music Bridge

[Johnjc1]

If you in the UK come over and visit
I be open to anyone doing this
I find the whole placebo argument a bit weak
I have reported on another forum the on the Audio gd Di20He I heard no difference when i expected it to make a difference

 

[camrector]

@Johnjc1 It’s not electrical noise you will hear. You will hear phase noise/shift and added odd number harmonics.
It’s non linear.
Nelson Pass makes a 2nd order harmonic generator. This adds a pleasing distortion.
Ask him why he doesn’t make a 3rd order one.

 

[BShaw]

Right, I have ordered the R26, but haven't received it yet.

If there is a real difference with the LB clock connected to the R26, it may be because some people prefer a sound with more jitter and noise, which most likely would be the case here. If so, that would be completely okay with me.

Regarding placebo, the point is that, if you expect to hear an improvement, you will hear it. If a dozen guys all hear it, it just confirms the point. I remember reading a former high-end audio salesman recounting a true story: He had a customer who wanted to listen to some very expensive speaker cables, and the salesman was changing the cables behind the amplifier. The customer heard a very clear difference between the cables he was already using at home and the more expensive ones he was testing. He even said that he couldn't understand how he could have lived with the cheaper ones, because the expensive ones were so much better. However, the salesman fooled him by not changing the cables as the customer thought. In short, the customer was subjectively hearing a clear and obvious difference, but in reality, there was none. The placebo effect is that strong (but subtle in its workings).

Well, fortunately it’s really easy to A-B test an external clock with the R26 by simply toggling on or off. Surely someone has tried this with the LB clock?

 

[Johnjc1]

Maybe best to lets go back to what jitter and digital noise do.
In my opinion any digital noise takes away from the noise floor of the music it makes long term listening for me hard and music has less flow and detail.
This is the opposite of what I am experiencing. i have more depth detail clout timbre etc; not less.
So unless we have measurements coming from the DAC its just opinion to what being heard In my case expensive in other people's cases the belief that the phase noise will be past to the DAC (which makes sense).
In the end I think its healthy to have an enquiring mind and be open to experience.
I personally can not go back to listening to R26 without the clock. To my ears it's just a backwards step.

 

[Johnjc1]

@Johnjc1 It’s not electrical noise you will hear. You will hear phase noise/shift and added odd number harmonics.
It’s non linear.
Nelson Pass makes a 2nd order harmonic generator. This adds a pleasing distortion.
Ask him why he doesn’t make a 3rd order one.

Interesting point.

 

[Johnjc1]

I am actually okay with that of it means me finding this more enjoyable
If second harmonics give more bass depth space clout and texture I am all for it. However it's okay if others are not

 

[LarsHP]

-70 dBc @1Hz phase noise is absolutely atrocious. No thank you.

I just checked the info. It says:

"Measured phase noise of GPS clock at 10MHz output is equal or better than:
-70 dBc/Hz at 1 Hz offset from the carrier
-100 dBc/Hz at 10 Hz
-125 dBc/Hz at 100 Hz
-143 dBc/Hz at 1 kHz
-150 dBc/Hz at 10 kHz
-152 dBc/Hz at 100 kHz
-155 dBc/Hz at 1 MHz"

Note that it says "offset from the carrier". In other words, the phase noise is below 70dBC around +/- 1Hz at 10MHz.
At 990kHz (10MHz - 10Hz) it's down 100dBC and so on, which means the noise we are looking at is far, far, far above the audible range. Even 1MHz is far beyond 20kHz. What we are looking at here is timing, not in the audible frequency range, but timing errors as such are definitely audible. Whether they are in the picosecond range is another story.

(Sorry if this is obvious to you.)

 

[LarsHP]

Well, fortunately it’s really easy to A-B test an external clock with the R26 by simply toggling on or off. Surely someone has tried this with the LB clock?

Yes, people have done that. However, testing it un-blinded opens the gate to the notorious placebo effect.

 

[Stellabagpuss]

Yes, people have done that. However, testing it un-blinded opens the gate to the notorious placebo effect.

l understand your suspicion, and l too scoffed at the specs, but there is no way this is the placebo effect, give it a go ,like l did ,and be ready to be surprised like l was.
You can always return, in point of reference, it replaced my AD Emperor Giesemann, and AD psu.

 

[camrector]

@Johnjc1 I’m not meaning to come off here as a measurements means it sounds better ASR type. I’m actually the opposite (because have you seen the R26 measurements lol)
R2R DACs are a linear circuit. So introducing a non linear 3rd order harmonic will cause distortions to the the output waveform. We want a wave form as close to the fundamental as possible ideally.
Unlike positive and negative sequence harmonics, 3rd order (triplen harmonics) do not cancel out and continue to add up.
Maybe this device is also adding 2nd order harmonics too?
My comment as -70db phase noise being very bad, is directed towards a performance metric for all external clocks. This is measureable worse than the internal R26 clock.

 

[Clivem2]

This is absolutely not true and not how a clock works.
Phase noise is very relevant and determential to SQ. This clock in particular introduces 3rd order harmonics which are no bueno.
There is no such thing as “clock assist” Even if works with a small phase-locked loop bandwidth, the onboard R26 clock is better than -70db@1hz

The term assist was my was very much my word for it. My point is that the internal clock still operates when an external clock is connected. We don’t know any more than this. If I force EXT ERR then is no interruption in playback, not even a click or tick.

Btw I had been sceptical about attaching external clocks. Logically it makes sense when several digital devices are being run from a single clock. A single device…yes you’d expect the internal clock would be best, especially with the external connections that are involved.

 

[Johnjc1]

I am open to doing a blind test
But the difference is not subtle

 

[LarsHP]

Regarding the R26 with the LB clock: Is it actually better, or is it just placebo effect?

I should perhaps have emphasized that my comments regarding placebo effect in this thread are not an expression of a general suspicion towards subjective evaluations of sound, but are specifically related to the fact that we are talking about the purely digital domain (before conversion to analog) and that the built-in clock is a femtosecond clock while the LB clock is only a "sub-picosecond" clock. That's almost 1000x worse jitter than the built-in clock.

If the facts were the other way around, I would simply have ordered the LB clock because it would objectively be a much better clock. However, when the facts are as they are, and people claim better sound from a technically much, much lesser clock, then I do question the validity of those subjective experiences and suspect the placebo effect.

This is just my honest viewpoint.

 

[Johnjc1]

Hear for yourself if you don't like it fair enough just report your findings and return

 

[Jake2]

Regarding the R26 with the LB clock: Is it actually better, or is it just placebo effect?

I should perhaps have emphasized that my comments regarding placebo effect in this thread are not an expression of a general suspicion towards subjective evaluations of sound, but are specifically related to the fact that we are talking about the purely digital domain (before conversion to analog) and that the built-in clock is a femtosecond clock while the LB clock is only a "sub-picosecond" clock. That's almost 1000x worse jitter than the built-in clock.

If the facts were the other way around, I would simply have ordered the LB clock because it would objectively be a much better clock. However, when the facts are as they are, and people claim better sound from a technically much, much lesser clock, then I do question the validity of those subjective experiences and suspect the placebo effect.

This is just my honest viewpoint.

With respect, I wouldn't call these general statements using the terms picosecond and femtosecond as establishing any useful or specific specs differences between the two. Especially with use of the term 'sub' which generally means less than or below, what's less than/below pico? - that'd be femto!

In terms of establishing an R26 baseline of phase noise specs (for what they're worth) for comparison I've looked hard online without success on several occasions to find either the exact oscillator used in the R26 (e g Crystek model X) and/or it's claimed phase noise specs at 1hz and 10hz offsets from carrier. There was one suggestion I came across that the oscillator used in the R26 is not the same or as good as the low phase noise Accusilicon AS338 as used in the U18 (-108dBc/10hz offset, no claimed 1hz spec but this is normally around 30db more/worse).