HDSPe AES and Word Clock, jitter, etc.

If its Nonsence why you could measure better Jitta after putting an Mutec MC3+ in between with reclocking!!!

Thats Physical measured much cleaner afterwards on DAC performance!!!

How that could be than if its nonsence less THD+N also

Nonsense is to put Copperbowls around the speaker and say its sounds now better

Why the RME Card has than Stadyclock thats makes the Jitter better if its nonsense to have clean Clock.

Also what does matter more Jitta with low Value 0.5pS or

Clocksatbility over a Year 0.5ppm or 0.01ppm if i talk abot Musik playing not Recording!!!

Read this but its german language it's Farytale for you its youre Dessision.

https://www.lowbeats.de/test-neuer-takt … -mc-3-usb/

But for me the Measurments telling me a different story!

There still Beliver in GOD out without any Prove of the Existence so anyway

The Boock (Bible) was written before India or America was Discovered and they has belived at this time the Earth was an plane surface LOL If God exists and Jesus belong to God why they never got the information that Adam and Eva wasn't the First People on that Eart and Erth was an Ball ??? Jesus was an Woodworkes that i can believe but thats all of the Truth... at the Time this Boock was writen the most people Couldn't read or Write only realy rich person or People from Chirch could do so so no posebility to prove the written word at this Time from other Side of Few.

I believ in Thor's Story than i like it more LOOOL

Nothing writen in Books could change my Life in any direction but the Stupidy off People related to me could do it or desccisions of stupid Politicions

So no i am not an Believer at all

Here is an Articel about Masterclocks:
https://www.soundonsound.com/techniques … ster-clock

But means it the same for WORDclocks ? Some of them has Masterclock inputs but not all.

I talk about WORDClocking like aviable on RME Card

Please stop talking if you can't provide prove off what you talking about.

My questions are to RME Developer not to anyone thinks he know something an do blabla all the ttime or i need to go elswhere !!

I use an RME Product and want to know how i could use the WORDCLOCK in and Out the right way and when it will help!

So if you didn't has an Audio Deveöloper Backround stop talking!!! If you are not belive that Jitta has something to do with sound quality stop talking!

I am not an Believer and i am only belive in Prove off concepts not in BLABLA BLA BLA

So well i can measure it by my selfe after i moved in to my new Home, makes no sense to speak with Forum Trolls

I am Out!

On that mentioned german Articel above, they used an special Testsignal called:

"The J-Testsignal 2448 (commonly referred to as J-Test 24/48) is an industry-standard digital audio test file developed by Julian Dunn in the 1990s to evaluate the jitter susceptibility and clock recovery of DACs (Digital-to-Analog Converters)"

to feed it into an DA Converter and test his JITTA Supression capabillitys and what an MUTEC MC3+ or MC3+ USB can do in between as an reclocker.

And the Measurements show that an MUTEC MC3+ can damp JItta at cartain ammount and an MUTEC MC3+ USB is even better removed the Jitta spiks completly.

So this tested DAC coudn't supress this JITTA at all it seems by his own (AMI DDH-1).

And on my Setup i use an DYI HDMI to 4x AES Extractor to feed the Source into my RME Card and HDMI is known for let say not that good digital connection and could introduce some JITTA, but sure i need to measure it by my own and look how bad it is.
I could  compare it to SPDIF Out on the same source. Bluerayplayer has SPDIF and HDMI Out so i can test it easely with the same DAC only. But for full surroundsupport i need to use HDMI.

If i could use an VanityPRO for HDMI to AES it would reclock the signal on his own also SRC included but for the Moment not aviable in Germany till end off August and i like to use an Hyperion DPR-16 Surroundprocessor for ATMOS when it is released.
Not sure if it has an SRC integrated that i could set to 96kHz Out like on an MUTEC MC-4. CamillaDSP (used RME HDSPe AES Card) needs fixed 96kHz input. My Crossover runs FIR Filter at 96kHz thrue CamillaDSP for all Channels (16 Out).

And DAC's get his clock from RME Card thrue AES in. I measured already 3 Stereo DAC's at once and they don't Drift (Analogue Outs) at all so this worked already very well this way without external Clock Sync. This Consumer DAC's don't have an seperate Clock input so anyway.

@Frunse:

Regarding your core question - if I understood it corrrectly -
Black Lion clock on the Word Clock input of MUTEC MC-4 ahead of the HDSPe AES.

The HDSPe AES uses RME's SteadyClock technology, specified at >50 dB jitter suppression at 2.4 kHz (see section 26.3 of the technical specs). That's the same published figure as RME's reference converter, e.g. the ADI-2 Pro FS R BE.

RME's own measurements show this isn't just a theoretical figure: they tested it with input signals carrying deliberately injected jitter of up to 500 ns, far beyond anything a real-world clock source would ever produce, and demonstrated the resulting jitter compression.
See this RME video on SteadyClock FS:
https://www.youtube.com/watch?v=Ti0aHW-zYcs

We don't actually know the self-jitter figures of the HDSPe AES's own circuitry or of the Black Lion clock in any comparable, apples-to-apples way; isolated picosecond specs from different manufacturers, measured under different conditions, are not directly comparable.

What we do have is RME's own jitter suppression and THD+N data, which gives good reason to trust that this part of the chain is handled very well already.
Given that the jump from the original ADI-2 Pro to the FS version, roughly a 20 dB improvement in jitter suppression, is widely reported by users as not producing an audible difference, it's hard to make the case that an external clock (here Black Lion) would move the needle further upstream of that.

So feeding a very low-jitter external clock into the HDSPe AES's Word Clock input is unlikely to make any measurable or audible difference at the end of the chain, since the card already handles that task to a standard that's apparently already beyond the threshold of audibility.

For the MUTEC MC-4, it depends on what you actually need it for:

• As a reclocker ahead of the HDSPe AES: same reasoning applies, the card downstream (HDSPe AES) already does that job.

• As a format converter, if your sources use different digital formats (PC, TV, streaming box) and sample rates. That's a genuinely useful role. Its internal clock is sufficient for that purpose.

• As a sample rate converter, if you want to unify everything to one sample rate: that's a fair reason to use it. But worth noting that upsampling is purely digital number-crunching, interpolating new sample values from existing ones. There's no waveform yet at this stage, so clock quality doesn't affect the calculated result. The MC-4's internal clock is entirely sufficient; an external clock (-> Black Lion) adds nothing here. And whatever comes out of that step gets reclocked again by the HDSPe AES afterward anyway.

Whilst jitter suppression in the digital domain is certainly important if any AD or DA converters are connected to these devices, it is most crucial at the point where AD and DA conversion takes place. Here, the HDSPe AES in front of the DAC stage in your setup delivers excellent performance in terms of jitter suppression, and the use of RME Reference converters with SteadyClock FS/EX would go one step further by incorporating the internal clock.

That said, the jump from the standard ADI-2 Pro to the FS version, despite the measurable jitter improvement, was not reported as audible by users, so even that edge appears to be more of a measurable refinement than something you'd actually hear.

Bottom line: Comparing picosecond self-jitter specs across manufacturers is not meaningful without a common, comparable measurement basis. RME's SteadyClock FS technology delivers jitter suppression that is measurable but, by available reports, not audible. My recommendation: trust that and avoid overcomplicating the setup. The MC-4 makes sense if you want to consolidate everything to AES and a single sample rate, that's its real value here. The Black Lion Word Clock adds IMHO no audible value in this chain.

The resulting setup

different audio sources
   |
MC-4 (roles: SRC, format converter, clock master)
   |
HDSPe AES (SteadyClock, jitter suppression, ideal clock slave due to SteadyClock technology)
   |
Connected D/A converter (clock slave).
Side note: ideal would be an RME reference converter with SteadyClock FS/EX, utilizing the internal clock for D/A conversion additionally.

Thank you thats an answer i can count on!
So thats what i want to know.
The MC-4 is still usefull for me at the Momennt, to provide 96kHz in to RME Card so it don't need to bother about different Sources like 48 kHz or 192kHz to feed in.

I read here that the newest Windows Driver support different Samplerates on Driver side already but i use Linux and on Linux i didn't have official supported Driver.

I am also not sure how CamillaDSP on windows 10 (ASIO) could handle it, need to test it by my self, no problem.

CamillaDSP was already running on that RME Card with Windows 10 and the official Driver, not the newes here mentiond.

But at that point i used an PC with HDMI Out for testing and it was set to 96kHz fixed Output smaplerate, i used only Wav files for testing. That worked well already.

But for my Bluerayplayer , TV i need to use HDMI Out with different Samplerates comming from Source.

So now i know that i don't need to buy any Clocks on that Setup but need to figure it out how to handle the different Samplerates in between the best way.

CamillaDSP offers something for that in Software but only on Linux aviable, but thats also makes sometimes Problems i read if you updates Softwareversions and so on, so an external option i would prefer, switch on and forget style.

If i get an Surroundprocessor (Hyperion DPR-16 or Nuprime HX16-AES) in the future may be it handle it already by his own not sure yet.

And yes i would ike to use RME DAC's but the most multichannel Devices have not an Multichannel AES input or you have to pay for also for a lot of Inputs that i don't need

I was looking into RME M-32 DA Pro II-D but it has only DANTE or ASE67 and Madi inputs such a Device with AES in and i am on youre Side
And for me i didn't want to add an AES to DANTE converter i would than go for an HDSPe AoX D but i didn't

HAPI mk3 cost near double that and has only 16 Channel out but it has AES inputs.

And i will stay with AES it has 192kHz cability and didn't shrink the Channel Counts on high Samplerates used vor Ages and i can thrust on.

I need to feed Wires to the Speaker and it makes no differences for me if its an Ethernet Cable or an AES or XLR Cable

The upcomming Surroundprocessor has also AES and DANTE capabillitys for 16 Channels Out.

The way to deal with comments you don't like is to ignore them. You don't have to read them, much less respond with personal attacks.

I prefere AMT'S over Ribbon Tweeter, i am obsessed with Harwood AMT 25, it isn't to big and goes up to 40kHz -6dB.

I use it with an Scanspeak 18W/4434G00 in a 2 Way Design down to 60Hz Bassreflex (5 Speaker build from my Design thrue an Woodworker).

Also Designed an Waveguide (3D printed) for it and tested it already that gives me the possebility to use it from 1800Hz 96dB/octave FIR filtered on CamillaDSP with low Distotion. The Waveguide i will CNC cut than in to the Baffel and place the Chassi from behind every where Round eges for smooth response. At the Moment it works at 3kHz Crossover point without Waveguide.

For Bass i use right now 2 10" Bassreflex Subwoofer and buoght already 4x 12" Chassis for 2 Ripole on an Axel Ridthaler aproved design But that work in Progress.

So i will design 2 new Frontspeaker with it and 2 18W/8434G00 Side by Side for even bigger Waveguide, I use HYPEX OEM Amp's NC122 2x125Watt 4Ohm and NC250 1x 250Watt 4 Ohm for Subs
So i have than an 7.2 Setup for my new Apartment

My actual DSP Project runs here:

https://www.diyaudio.com/community/thre … up.440807/

English is my third language so i am not every time able to understand every Joke people made with native English capabillity.
And also not able to descripe all well like on Motherlanguage.

I would like to test Planar Chassis like this:

https://www.soundimports.eu/de/grs-pt50 … gKZzfD_BwE

But never did right now.

Piega has also something like Planar chassis not directly Ribbon like this:

https://www.hifisound.de/de/Lautspreche … oener.html

My Waveguide design is descriped here:

https://www.diyaudio.com/community/thre … st-8023326

And here you can find some Picture from my Speaker and Work in Progress with Waveguide:

https://www.diyaudio.com/community/thre … 25.428467/

Network switches don't have a word clock input in the first place, that concept doesn't even apply here. The TCXO mentioned is just the switch's own internal timing reference for its Ethernet PHY, completely unrelated to any audio clock domain.

Sorry, but this is nothing special. There are mechanisms in the underlying internet transport protocols that prevent data loss.
Either the protocol handles it itself (TCP), or when using UDP, this is the task of the higher protocol layers.

According to the product description of the network switch:
"The SW-9X is a high-end 8-port gigabit Ethernet Network Switch with built-in low noise power supply, AC filter, and ground shielding. It improves the network performance by reducing signal noise. By reducing circuit and power supply noise, and using a customized TCXO clock, SW-9X delivers accurate network signal that ensures stable music data transmission."

Well, a network switch with audiophile-quality claims.
It sounds like they adopted the vocabulary of audiophile customers and tried to translate this to the network business
to entice audiophiles who need LAN switches into buying these products, convincing them it's the best they can do.

An 8-gigabit switch is nowadays absolutely no rocket science anymore. The best that you can do is to buy a standard, manageable 8-port Gigabit switch with the feature set that you need and to save some money for other things.

Sorry. but with €1500 this device is overpriced and the marketing claims don't hold up technically.

AVB and Dante have their own mechanisms to ensure proper transfer of audio streams. AVB focuses on Layer 2, Dante supports Layer 3 (routing). Dante's clock synchronization is handled via PTP (IEEE 1588) directly between the audio endpoints, not by the switches in between. A switch's internal oscillator quality is irrelevant to that process, regardless of marketing claims to the contrary.

Sorry, but the irony was not recognizable for me because it sounded as if you were using this device yourself.

"I have transferred a lot of data between computers through networks..."