Hugo M Scaler by Chord Electronics - The Official Thread
Apr 15, 2024 at 8:31 PM Post #18,331 of 19,229
specific to his own DACs as he rarely comments on others, the DAC design is virtually immune to jitter, with the only limitation of Toslink being a guaranteed 24/96 bitrate and higher jitter than USB, BNC etc,
So using a Toslink input guarantees total source isolation from RF and any other possible noise, and with Chord DAC’s jitter isn’t an issue anyway
I use Toslink from CD transport and custom PC based server that uses a carefully chosen MB that included a native Toslink output …
Perfectly said.
 
Apr 15, 2024 at 10:02 PM Post #18,332 of 19,229
Is there a reference online where Rob said this? Curious the context for which he said it. I guess my amateur mind thinks that if toslink is easily better than all other connections, then it would have to mean one of the following:
1. The source going into the DAC has poor implementation in the other outputs
2. Chord products have poor implementation in the other inputs (doesn't seem logical as anal as Rob is on his designs)
3. Lastly, it is just a fact that toslink is THE best digital implementation on the market...which I have not personally heard.

So I would like to hear Rob's comments on this and in what context if there is a reference I could read/watch.

I covered this aspect in recent NYC and Singapore CanJam seminars - here are the relevant slides:

Slide16.JPG
Slide17.JPG


I think optical has a bad reputation because it sounds rich, warm and naturally rounded due to the reduction in noise floor modulation. This is a less impressive sound than the etched and forced sound quality, kidding you into thinking it's more transparent, that USB has due to increased noise floor modulation. The proof in that optical should be treated as the reference source is that USB and BNC can sound identical - when the source has no possibility of transmitting random RF noise currents into the DACs ground-plane such as with battery powered sources that are isolated from the mains or ground.
 
Apr 15, 2024 at 10:08 PM Post #18,333 of 19,229
I covered this aspect in recent NYC and Singapore CanJam seminars - here are the relevant slides:

Slide16.JPGSlide17.JPG

I think optical has a bad reputation because it sounds rich, warm and naturally rounded due to the reduction in noise floor modulation. This is a less impressive sound than the etched and forced sound quality, kidding you into thinking it's more transparent, that USB has due to increased noise floor modulation. The proof in that optical should be treated as the reference source is that USB and BNC can sound identical - when the source has no possibility of transmitting random RF noise currents into the DACs ground-plane such as with battery powered sources that are isolated from the mains or ground.
@Rob Watts : Thanks for sharing this and explaining. Per your explanation, sounds like I'm in a good place using my iPad as a source into (M Scaler/TT2) as I only run it on battery. 99.9% of my listening is streaming via my iPad. So to feel I HAVE to go to optical would completely upend how I run my set up and would be very undesirable.

Thanks,
Mike
 
Apr 15, 2024 at 10:15 PM Post #18,334 of 19,229
I covered this aspect in recent NYC and Singapore CanJam seminars - here are the relevant slides:



I think optical has a bad reputation because it sounds rich, warm and naturally rounded due to the reduction in noise floor modulation. This is a less impressive sound than the etched and forced sound quality, kidding you into thinking it's more transparent, that USB has due to increased noise floor modulation. The proof in that optical should be treated as the reference source is that USB and BNC can sound identical - when the source has no possibility of transmitting random RF noise currents into the DACs ground-plane such as with battery powered sources that are isolated from the mains or ground.
How’s about conversion from optical to electrical signal at DAC end? It can potentially introduce noise during the conversion.

Also the weakness of Toslink is only limited 4FS. It can’t transmit higher than 4FS.
 
Last edited:
Apr 16, 2024 at 7:56 AM Post #18,335 of 19,229
I covered this aspect in recent NYC and Singapore CanJam seminars - here are the relevant slides:

Slide16.JPGSlide17.JPG

I think optical has a bad reputation because it sounds rich, warm and naturally rounded due to the reduction in noise floor modulation. This is a less impressive sound than the etched and forced sound quality, kidding you into thinking it's more transparent, that USB has due to increased noise floor modulation. The proof in that optical should be treated as the reference source is that USB and BNC can sound identical - when the source has no possibility of transmitting random RF noise currents into the DACs ground-plane such as with battery powered sources that are isolated from the mains or ground.
Thank you Rob, every point you have here I heard when I switched to Toslink.
 
Apr 16, 2024 at 8:25 AM Post #18,336 of 19,229
How’s about conversion from optical to electrical signal at DAC end? It can potentially introduce noise during the conversion.
I think we have to assume it does. A friend with a Mojo 2 reports sound quality differences amongst three cables: KabelDirekt, Sys Concept and Supra ZAC (in ascending "sound quality").

It may be that Mojo 2 is more susceptible than other Chord DACs (or HMS).

Also the weakness of Toslink is only limited 4FS. It can’t transmit higher than 4FS.
The well known streaming services don't offer higher than 4FS, which is lucky :)

@Rob Watts does the count of elements in the pulse array correlate with the ability to reject jitter?
 
Apr 16, 2024 at 8:34 AM Post #18,337 of 19,229
I think we have to assume it does. A friend with a Mojo 2 reports sound quality differences amongst three cables: KabelDirekt, Sys Concept and Supra ZAC (in ascending "sound quality").

It may be that Mojo 2 is more susceptible than other Chord DACs (or HMS).


The well known streaming services don't offer higher than 4FS, which is lucky :)

@Rob Watts does the count of elements in the pulse array correlate with the ability to reject jitter?
My point is Toslink is dated technology. It uses a very long wavelength and limits the bitrate. If Chord is serious about optical technology, they should develop their own optical connection from Scaler to DAC rather keep using BNC. Many DAC manufacturers develop their own optical connection.
 
Apr 16, 2024 at 9:33 AM Post #18,339 of 19,229
I told you many times, with Chord stuff don't bother with expensive streamers.
Just use decent toslink source and you are done. Unfortunately price and tales fool people...
Yup I just learned that a few days ago. With the Bluesound On optical it sounds better than streamers 10 times the price 😂 😂 😂. You're correct.
 
Apr 16, 2024 at 11:42 AM Post #18,340 of 19,229
Yup I just learned that a few days ago. With the Bluesound On optical it sounds better than streamers 10 times the price 😂 😂 😂. You're correct.
I can go lower: Wiim Pro😁
 
Apr 16, 2024 at 12:06 PM Post #18,341 of 19,229
I agree that Rob should be adding an optical connection from the new scaler to future dacs that could have that optical input, to completely solve the issue of rf noise coming from the scaler. As others said, quite a few other manufacturers have done this. I will be pretty disappointed if it doesn't have such a thing.
 
Apr 16, 2024 at 11:56 PM Post #18,342 of 19,229
How’s about conversion from optical to electrical signal at DAC end? It can potentially introduce noise during the conversion.

Also the weakness of Toslink is only limited 4FS. It can’t transmit higher than 4FS.

Any digital circuit can introduce RF noise. The issue here is how much, and this depends upon power - Toslink optical receivers have the lowest power consumption. By far the worst is USB, as this requires a processor to buffer and decode, and that creates two orders of magnitudes more random RF noise. Indeed, how the USB decoding is powered makes substantial differences to sound quality. But with optical, I treat it as a RF noisy source, with it's own dedicated RF filters; not that I have any evidence for RF noise from optical. The RF filters I employ covers ground noise as well. So optical certainly wins out on the RF noise implementation0 issue too.
@Rob Watts does the count of elements in the pulse array correlate with the ability to reject jitter?

Not source jitter. But more elements gives lower noise, distortion and lower overall impedance; the lower impedance makes for better immunity from audio frequency ground currents.
 
Apr 17, 2024 at 3:05 PM Post #18,345 of 19,229
Indeed, how the USB decoding is powered makes substantial differences to sound quality.

Note: A bit off-topic as I'm applying this to other Chord DACs not the HMS.

Thank you for sharing this snippet. I'm assuming this applies across the board to the DACs and not just the HMS. I can't recall the manufacturer # for the USB chipset ATM for my Chord DAC to compare against the HMS USB chipset manufacturer #.

I H8ed USB for Audio with a passion, but now I quite enjoy because it sounds like Toslink after powering USB decoding with Off Mains Super Capacitors (Native 5V, no regulators, 30 times lower impedance than the TT₂ Super Caps, almost zero impedance and ~6000 Farads).

Initially, I pointed the SQ improvements at powering the length of the USB cable (10M/33 Feet). By giving it Nuclear Power Plant Clean Power (Super Capacitors), I thought it smoothed out the power over a long distance. But now with your comment, I believe I'm actually directly powering the USB chipset with sufficient power so that the USB chipset can do it's job optimally with a substantial increase in SQ. It's been in the back of my head of what is going on since beginning of 2024 so it makes sense now it's about directly powering USB decoding that made substantial difference in SQ and not smoothing out power over a long distance which was my previous theory.

So Many Thanks. I was stumped why I got USB to sound so good similar to Toslink that I'm seriously considering moving away from Toslink. I like to follow Strict Sound Fundamentals so sometimes it inadvertently pays off like in this case. I didn't plan on powering USB decoding. I didn't plan on having the purest Galvanic Isolation possible. I didn't plan on unlocking the "Soundstage Depth" feature in my transportable Chord DAC. It just inadvertently happened by staying Sound Fundamentally.

I tried this recently too as I'm scaling Super Caps ATM so I had to discharge all my Super Caps to 0V. It takes a good week to discharge. The USB was not bearable in this time. Once I scaled (double the time, 1/2 the already super, super low impedance), I was back in USB heaven. My production Super Caps are usually 24/7/365 so it was a painful downtime to discharge. Progress.

I've been very Pro-Toslink my entire Chord Life, but now I'm leaning Pro-USB given I'm properly powering the USB decoding. My Chord Philosophy is less is more. Give the Chord DAC a clean low latency signal integrity with clean Off Mains power and let the Chord DAC do it's magic. Get out of it's way. I see so many that add "more" such as software upsampling that would skew the graphs below. Less is More. Keep it Clean, get out of the Chord DACs way. The wrong algorithm with software upsampling and LPS will skew the Graphs below. One by adding small errors (software upsampling) and one by noise floor modulation (LPS). It's not even a Rob Watt's DAC anymore after software upsampling, it's a hybrid with tons of small errors by not using a Chord algorithm.

Slide26.JPG


That's how I was able to unlock Soundstage Depth. By keeping the graphs above super tight and super clean so no small non-linearity signal errors. When you add "More", your just tricking your own Brain. My chain is completely 100% free of the tiniest of small non-linearity signal errors. The brain adapts. It took 6 weeks to unlock Soundstage Depth, so at that time I knew small non-linearity signal errors were no longer a factor. That's why I want to stick with USB and move away from Toslink. Toslink is just a complication.

Slide29.JPG


Software upsampling, LPS, etc., your just tricking your brain with a plethora of small signal errors. The brain is a quantum computer, your just confusing it so it won't compute. The trick I learned over the years is to stop confusing the Brain by eliminating small signal errors by downsizing and letting the DAC do it's job.

-- RW comments --

smallsig.png

brain.png

--

Toslink for now is just more. If I can clean power the whole chain from end-2-end and keep it simple with USB, I think that is the optimal route. At least in my case, that's Plan A. With the facts and powering USB decoder is now an official thing, I think I'll stick to Plan A. Toslink will likely be a relic by the EOY. I'm the only one running Modern Toslink, not legacy Toslink too so it will be difficult to let go and move on but to stay fundamentally "Less is More" sound it should be the right thing to do.

Focus on 3/3. Once you add "More", your lucky to to get 1/3 or 2/3:
3areas.png


In conclusion, I get all the benefits of Toslink while maintaining all the benefits of USB. All Pros, No Cons. I also get the benefit of "Less Is More" and to keep things clean and tight. I decouple from the complication which is Toslink. A Zero variance Off Mains Chord chain. I wouldn't be able to unlock Soundstage Depth if there was a leak in the chain or if I induced small signal errors.

I'm glad too that the HMS₂ will apply across the board to other production Chord DACs and not just to the modern flagship.

EDIT: Oh, don't worry about the RF Thing-a-ma-jig. RF doesn't apply in my case because it is a certified USB Fibre Optic Cable which 100% decouples. My source is also powered by Off Mains Super Capacitors which 100% decouples so Off Mains Source doesn't apply either. Unfortunately, the Super Capacitor Supply Chain has dried up for at least the Summer, so I have to delay further projects for Summer 2025. Last time, the Supply Chain dried up for five years. Those EU Public Transportation Projects seem to be taking off as the demand is ridiculously high for Super Caps. Clean power tho is good for the environment. As a consumer, I can't compete against Industry for procuring Super Caps. But my Super Cap foundation is already solid, I just want to scale going forward. I'm looking to scale from 4 Large Super Caps to 8 Large Super Caps in 2025. Mission already accomplished by scaling 2 Large Super Caps to 4 Large Super Caps this Summer 2024. BTW, Music is super buttery smooth with this USB setup. Enough that it will melt. It's so smooth, I can loop a single track all day without flinching. What is surprising is I can play 384k over USB Fibre Optic while Toslink Fibre Optic maxes out at 192k with a Lifatec. I have yet to try 768k as it is a low priority.

4x4.jpg
 
Last edited:

Users who are viewing this thread

Back
Top