About what one can hear and relevance of measurements (XMOS+Crystals and PS)

For sure measurements are are relevant and important but those don't always match with the hearing of the average person.

I am not talking about the personal preference in sound and timbre of persons. For sure that is the case too and nothing against that. We don't want to fall all for the same girl or boy either don't we.

Need to know:
All DAC's have a digital input clocked by the audio source directly regardless what the other DAC circuitry uses as the internal clock source.
Better crystals only got effective after the introduction of intermediate digital buffers like provided by XMOS or Amanero processors. The audio source feeds the XMOS input using its own clock and the XMOS output runs on the ultra stable clock. XMOS is not a simple buffer but is processing the data stream.
If the data path does not run via one of these intermediate processors there is no difference in SQ improving crystals.
You can also buy e.g. an Innuos Phoenix USB unit which does the intermediate re-clocking as a separate unit. This way your favorite very good DAC can get a second live!

The difference between a Standard Edition A6 and the Master Edition A6 is a swap of the crystals clocking most important the XMOS processor. The DAC gets the SQ benefit indirectly!

Ran into a Zidoo UHD3000 (now replaced by UHD5000) which strangely sounded better when wired externally from player USB Output to DAC USB input on the same box compared with set to Direct DAC input from the media player section. Diagnosing with Zidoo engineers it turned out they sent the data directly from the media player processor to the DAC Input while the USB path ran via the Intermediate XMOS processor. Only changing the FW to deploy the XMOS path also for the direct internal data path corrected this.

Talked with the engineers of McIntosh and they told me after discussing intermediate processing that they use a custom chip doing so also for their HDMI inputs feeding the DAC's. The DOH chips used by Neo Alpha, Eversolo and UHD8000 do clock stabilization. When paired with a quality crystal on the motherboard HDMI output is optimal.

Something which is confirmed by many is the effect of the quality of crystals regarding preciseness, jitter and stability. The best technology can produce these days are crystal in the femto range. These crystals can be measured as stand alone products but actual effect on audio reproduction remains mostly in the dark.

Could myself testdrive a fair number of DAC's and audio setups. The quality range of the Crystal deployed often could be predicted without going to the specifications. The big thing here is that only in recent years special attention was given to this component which often turns out to have become the critical component in good DAC's designs. Audio equipment often specifies the DAC brand and type but hardly ever does so for the crystal used. Even most of todays (very) expensive Amplifiers still don't specify it. Even the most expensive CD and UHD BD players of the past never did and still never do.

Now for my own experience doing A/B testing in all relevant cases joined by other persons:
- The relevance of the Crystal used by the DAC on the Sound Quality was always clear and could be picked reliable doing A/B switching.
- Also SOC (Sytem On a Chip) integrated DAC's using a general crystal on the motherboard could be ranked reliable doing A/B tests. This applies using HDMI audio output directed via the integrated DAC's of the AMP. An entry level Dune media player e.g. sounds better than an equivalent Zidoo media player sharing the same AMP for that exact reason. The Zidoo Neo Class X and Alpha media players sound similar to an OPPO 203/5 via HDMI which again is somewhat better than those Dune's. Only asking the Zidoo engineers directly could confirm these actually use a better crystal on their respective dedicated motherboards.

Expensive DAC/AMP brands for sure paid a lot of attention to the PS often resulting in the (dedicated external) PS being (far) more expensive than the total unit being powered by it. Lots of demos have been organized with A/B switching just the PS from Internal to External.

My conclusion:
No I have no clue what so ever how so tiny differences in crystal accuracy can have such huge effects on the actual listing experience. I only can conclude it is real and not fake; it is perceived by most.
The exact effect of the Power Supply on the crystal/DAC accuracy remains in the dark for me but using a good PS for sure helps. If using dual transformers and using super oversized capacitors really does improve SQ? It for sure does not harm but the value/money ratio remains questionable for me.

Other A6 inputs than USB on most current External DAC's don't benefit from better crystals as these mostly bypass the XMOS circuitry. The same happens using the A6.

Only the A8 has an IIS Output running via XMOS and all A8 external inputs can be directed to run via IIS Output. This is the setup to upgrade the sound of a top quality external DAC without XMOS. If you want to give your TV (or CD player) a real SQ boost then SPDIF => A8 IIS Output => External IIS DAC input is the way to achieve it.

Example: A (top) quality CD-Player connected via Coax/Toslink to an A6/A8 will still run on its Internal CD-Player clock with additional jitter added by the SPDIF connection. If the CD-Player has a similar quality DAC chip build-in it should in fact sound better that way (also claimed on this forum). Connecting a dirt cheap USB CD/DVD-player to the A6/A8 doing just DAE (Digital Audio Extraction) will benefit from the integrated DAC as designed and sounds amazing well. Tried this myself beating my OPPO 203 regarding SQ playing the same CD's. Have all my physical CD's ripped to FLAC and play them always that way using my A6.

Interface choice connecting an external DAC
- Only for the A8: Connecting an external DAC via IIS is the best option as all external inputs will then take advantage of the Accusilicon crystals as this path runs via XMOS. The External DAC does not need to include XMOS and/or USB ports itself; so relative old but further top quality DAC's may do an excellent job here.
- The second best choice is USB as also that input runs via XMOS with A6 and A8 (but now not for SPDIF digital inputs).
- Using SPDIF coaxial is the third choice as it does not run via XMOS and has less bandwidth than USB.
- SPDIF optical is the worst choice as it has a lot more jitter and the absolute lowest bandwidth.

 

No need to remain in the dark just look at the measurements of a well engineered component such as the A-6 , which cannot be improved by a clock or linear power supply.
Keith

 

Sorry Keith to contradict: Did a lot of A/B compares with others and the outcome was very consistent with a very mixed audience.

 

Uh huh and by any chance was this comparison sighted?
Keith

 

Pure A/B blind tests. Did a lot of those including the person doing the actual switching being swapped. Always using the same AMP, Speakers and room.

I have a degree in electronics myself and do value measurements. But still a lot is unknown about the human capabilities.

A personal story:
I have a home which had 2 owners before. Learned from an old time neighbor that a son of the first owner hanged himself in my attic. Don't like that fact obviously but it does not affect me in any way when going there.
But had twice that somebody coming along with me upstairs got cold with a nasty feeling. They told me so without knowing the actual history. With which instrument can we measure this? It still is real I am afraid!

 

There seem to be two camps of music enthusiasts. Those for whom signal-to-noise ratio is everything, and those for whom other qualities in the signal are also possible. In the ASR forum, supporters of group 1 predominate, and in some cases there are even missionary tendencies. You believe DACs sound different, how dare you!
Very interesting thread: Serious Question: How can DAC's have a SOUND SIGNATURE if they measure as transparent? Are that many confused?

https://www.audiosciencereview.com/...re-as-transparent-are-that-many-confused.9245

I find it contradictory that this is being discussed in a forum environment where a rating of DACs according to SINAD is celebrated. The better the signal-to-noise ratio, the higher the rating. Quite inexpensive DACs lead the list of the best here. Topping D50 for about 230 EURO or the SMSL SU1, which costs maybe 80 EURO, so the whole DAC costs as much as the clock generator in the A6 ME.

I think what many do not think of noise reduction in such devices. Due to digital filter or by external filters in electronics. Such filters will remove noise, but also noise which is part of the original music. I believe it is such noise, which transports some of the emotions we hear in music.

 

Dacs are digital to analogue converters, they are entirely characterised by their measurements, ‘noise, which transports some of the emotions we hear in music’!
Really!
Keith

 

Hey Keith, instead of playing the same old broken record, it would be interesting to hear what you got to say about Nice Monkey being able to consistently tell the standard A6 from the ME in an unsighted AB comparison….just curious

 

I don’t believe he can, if the comparisons were conducted properly.
Without wishing to disparage Monkey he states he has a degree in electronics then unbelievably states that noise contains ‘emotion’.
Keith

 

Wrong Monkey, I am not Nice and do not have a degree in electronics either.
If you do not understand what I mean then I am sorry - you seem to not hear the blowing sound of the flute, bowing sound of the violin or rustling of leaves. Thats all kind of random "noise" - and, at least for me, central to enjoy sound.
Guess what happens to such sound when you treat it with a noise filter.

 

uh huh
Keith

 

To summarize, you bust our balls saying that comparisons must be conducted unsighted, with 2 devices that can be swapped within seconds, level matched and so on…..but when someone does all that and proves you wrong, you don’t believe it. Interesting…..

Your total lacking of coherence would suggest your credibility is equal to nihil.

BTW you are mixing up two Monkeys here, perhaps a strong coffee would help you . The one with technical background is not the one of the noise…..btw, besides selling overpriced Hypex amps in a nice enclosure, what is YOUR background?

 

Has nice monkey actually conducted a controlled comparison?
Perhaps he could detail exactly how and what he compared?
I made a controlled unsighted comparison between the A-6 and A-6 ‘special edition’, and wrote a small piece on it here.
https://www.puriteaudio.co.uk/post/eversolo-dmp-a-6-versus-eversolo-dmp-a-6-gold


Keith

 

The sound differences perceived by me and my audiences is not differences in frequencies or noise. Those are mostly influenced by the AMP and mainly speakers plus room parameters. Measuring those is fairly easy to do with a special microphone. Playing with DIRAC will for sure follow to address that part of my audio chain.

Some aspects correlated with the DAC build:
- The sound is more natural compared with attending the live event. It sounds more like being there in person.
- The sound is "crispier" for some instruments or music fragments.
- Your can better hear the direction of the sound and point where each instrument actually is.
- One can pinpoint the sound of individual instruments better/more precise.
Don't ask me the why of this but it can be reproduced for sure. Good source material is needed but any quality standard CD recording may already do. I have the Dutch Evergreen 2000 collection in 44.1-16 format and many tracks are usable for the test.

Yes I needed to point some of my listeners to paying special attention to these aspects. The first impression may be that it all sounds the same.

 

Your introduction appears to me to be meant sarcastically, or am I wrong:
I made this comparison between the two Eversolo DMP A-6 units before Christmas but it was so dull, and not wishing to soil my reputation as a witty raconteur:- But while I wait for my expensive new cables to undergo their 600 hour burn-in. If it was, that already shows your bias.

I do not understand how "Level matching was easy peasy because they both have the same output voltages" - because the voltage is not constant, is it?
I would assume to do this really scientifically, you would need a 2 channel oscilloscope to exactly match the output curve to adjust to same voltage.

Why do you use "Roon" to switch between 2 different devices when you can just connect devices to the same amplifier and switch the input with a physical switch.

And what was the amplifier and the speakers? If the speaker limit the resolution of the output, I think we do not have to expect that there will be a difference.

 

Unsighted level matched comparisons are purely used to determine whether there is a discernible difference between two components.
Keith

 

Roon allows instant switching of the same programme at the exact same point between the two components and can be operated out of sight of the subject.
Both the A-6 and the A-6 special edition have the same outputs 5.2v over XLR
https://www.eversolo.com/Product/sp...+Edition/target/SGlS2hlXYyJKKmVViAFMcQ==.html
https://www.eversolo.com/Product/spec/model/DMP-A6/target/7abWHw++oHhKKmVViAFMcQ==.html
Keith

 

I did a comparision with the DMP A6 ME vs a CD Player. With the attached headphone amp I was able to switch between sources with just a little switch. I noticed a more "open space" with the CD-Player. Later I looked it up, and found out that it (Onkyo C 755) has a linear power supply.

I am doing that right now, playing : Sahra Jaffe, Pretender PT.1 That is an accustic recording with significant microphone noise - and for me that noise sounds totally different on the A6, dampend, more compressed.

People find my "noise" thoughts funny, but I think if one wants to compare two devices it would be best to have as input pure white noise. If the noise sounds the same on both devices, they will likely also sound the same with music.

 

It is perfectly possible for two components to sound different.
But for them to sound different and be reliably ‘ picked’ in an unsighted comparison one of them must be adding audible distortion.
This distortion would be clearly evident in the components measurements.
One has yo remove the sighted component because cognitive bias is so very powerful and affects every human being.
Keith

 

There are other things (OCD), that are very powerful and affect especially one human being.