[adrianm]

Rob,
In what way is optical behind USB on Dave?
Is not All jitter discarded by buffering?

Sorry i reply on a 2016 post.. im catching up from the start of this thread ..oof

He actually changed his mind later. And I agreed until I got the Meze Elite. Now I think optical sounds considerably softer and more diffuse, but with better depth and width of the soundstage. Still looking for a way to get the best of both worlds. Time to try Taiko? LOL.

 

[zen87192]

What’s best out of DAVE in to a Headphone Amp? RCA or XLR?

 

[wasupdog]

RCA since Chord dac topology is single ended, but would also depend on the amp pairing.

 

[number1sixerfan]

I guess you are correct regarding little "compelling business case and profit model" with Dave. Too late and frankly the really demanding headphone I mentioned did not even exist when Dave was designed,but in his defence, Rob did make TT2 notably more capable in that respect, although he compromised a bit in some other respects compared to Dave.
Unfortunately it has a lot to do with market segmentation as I see it, possibly also with max profit margins carefully calculated for each segment in his and the rest of Chord´s product line.
Actual parts and building costs are probably only a mere fraction of it all.

With Rob and Chord we very much pay for the many hours he has spent on his products designing and writing the code,both optimised and suitably compromised for each segment in the product line.

Absolutely. R&D is definitely a major cost factor. I appreciate the design and thought that goes into Chord products, which is why I'm generally ok with their price points.

"Now for something completely diffferent"

Stax SRX900 on order? Did you audition it?

Or are you buying it unheard knowing how transparent Stax electrostats generally tend to be ,expecting this one to be their best ever??

I have not auditioned any Stax headphones since the 009S,but I am also a bit tempted to get back into electrostats again with headphones ,not only via speakers.
What about Dan Clark´s closed planars "Stealth" which Rob Watts and many but not all reviewers tend to laud ,have you heard it?
I want to audition it asap, but so far no closed headphone I have ever heard has been transparent and open sounding enough for me. The Raal SR1A on the other hand, which I see that you also own sound about as open as a headphone can get imho.
But unlike the Susvara which only needs lots of clean power and current, the RAAL will not even play directly out of Dave it needs a step up box or dedicated amp like an electrostat.

Cheers CC

I don't want to clog the Dave thread, but in short I have the x9000 now. It's definitely one of the best headphones Stax ever made (if not the best, haven't heard Omegas), but is also quite different in presentation than the 009 and others--warmer, more laid back. You can read more in-depth impressions here.

Also, haven't had the chance to hear the Stealth.

 

[Reactcore]

He actually changed his mind later. And I agreed until I got the Meze Elite. Now I think optical sounds considerably softer and more diffuse, but with better depth and width of the soundstage. Still looking for a way to get the best of both worlds. Time to try Taiko? LOL.

Oh im enjoying my Wyse D90D7 thinclient PC which i placed a 4TB HDD running stripped simple embedded Win7, F2k and Mediacenter controlling it with my phone using remote control app.

Its as noise quiet a pc can get.. passive cooling and external PSU.
I recently built in a CM6631A USB - optical converter.

Optical to HMS with no other inputs connected to it.

 

[zen87192]

RCA since Chord dac topology is single ended, but would also depend on the amp pairing.

Currently giving a Singxer SA-1 a go which is pretty impressive but may aim at a Benchmark HPA4 or equivalent when I get used to what’s out there and their value for money.

 

[chesebert]

"could not keep up with high instantaneous power requirement during high dynamic and complex passages"
It could be the same reason why switched mode power supplies suck? Those now days have lower noise figures and measurements dont show their shortcomings...
Most of the good dacs and amps have massive linear power supplies with massive amount of high end capacitors...
Bakoon, Sparkos, Headamps, Niimbus and Ferrums amps show worse measurements than Topping amps, but unlike Chinese amps they have Beefy power supplies.
I can bet that those power supplies that are bought for Dave wont show improvements on Audio Analyser, but people are happy with that upgrade right?

I would not generalize all SMPs as inferior to linear. All my Linn dacs and my Emm Labs have SMPs in there. No issue with dynamic, bass slam, body, weight and warmth. I believe dCS also uses SMP for their dacs (I think Bartok uses both SMP and linear, with linear used for headphone amp). However, all of them roll their own SMPs. For example, Emm Lab's SMP is a fanless, zero-feedback design with two custom-wound transformers, active power-factor correction, and is synchronized to the dac’s high-precision audio clock to further reduce jitter.

 

[801evan]

Mscaler may do transients a bit better than PGGB ,but PGGB and Qutest on battery deliver a slightly lower noise floor than Mscaler,audible via headphones and with large scale symphonic music with a huge dynamic range.
And when I say huge dynamic range I do NOT mean the loud parts.

This is true with my experience. Mscaler slightly lifts the noise floor but the reason is perimeter and in-line RF. People have mentioned keeping it 2 feet away from devices and using wavestorms.

After many tests, end of the day, PGGB and HQplayer makes the sound worse and is worse than the mscaler. Best to spend on a better transport. MQA and native 44.1 sounds better than PGGB and HQplayer on a great transport. Vice versa on a weak one.

"The Achilles Heel" of Dave could actually be a case of not having enough power or current reserves to drive some the most demanding headphones like Susvara to their full capability with dynamically non compressed large scale symphonic material.

Dave is great on the Susvara, and Stealth. And Susvara is more demanding that the 1266 TC. Dave is very revealing so you need to give it a good transport, a good line conditioner is required and use spdif in. You get good sub bass and good tone even on the stock cable.

Of course a battery capable of handling instantaneous power and current requirements optimally while avoiding the drawbacks and limitations of fluctuating and possibly dirty? grid power and the RF issues involved with grid power as well.

I've done a full dc setup. Dedicated powerbank on every device with silver occ dc cables. While better than ac mains with a Weiduka line conditioner, an ipower x with a better line conditioner can beat the powerbank. Powerbank is grainy, very slow, small soundstage (13 inch tablet size) and you can hear the battery churn.

It could be the same reason why switched mode power supplies suck? Those now days have lower noise figures and measurements dont show their shortcomings...

Ifi ipower can beat practically beat all the LPS I've tried. LPS can giving a ringing sound at best and can still have a poor noise floor. Ifi elite can even beat the Hypsos easily. Even better than the lps 1.2 and the farad 3.

What about Dan Clark´s closed planars "Stealth" which Rob Watts and many but not all reviewers tend to laud ,have you heard it?

Best HP I've heard. Like the Dave, it'll just reveal the quality of your chain upstream. It sounds more open than my open cans. A brilliant device that makes chain diagnosing easier too. All criticisms on the Stealth is just them actually criticizing their chain, not the HP.

 

[atya35mm]

This is true with my experience. Mscaler slightly lifts the noise floor but the reason is perimeter and in-line RF. People have mentioned keeping it 2 feet away from devices and using wavestorms.

After many tests, end of the day, PGGB and HQplayer makes the sound worse and is worse than the mscaler. Best to spend on a better transport. MQA and native 44.1 sounds better than PGGB and HQplayer on a great transport. Vice versa on a weak one.



Dave is great on the Susvara, and Stealth. And Susvara is more demanding that the 1266 TC. Dave is very revealing so you need to give it a good transport, a good line conditioner is required and use spdif in. You get good sub bass and good tone even on the stock cable.



I've done a full dc setup. Dedicated powerbank on every device with silver occ dc cables. While better than ac mains with a Weiduka line conditioner, an ipower x with a better line conditioner can beat the powerbank. Powerbank is grainy, very slow, small soundstage (13 inch tablet size) and you can hear the battery churn.



Ifi ipower can beat practically beat all the LPS I've tried. LPS can giving a ringing sound at best and can still have a poor noise floor. Ifi elite can even beat the Hypsos easily. Even better than the lps 1.2 and the farad 3.



Best HP I've heard. Like the Dave, it'll just reveal the quality of your chain upstream. It sounds more open than my open cans. A brilliant device that makes chain diagnosing easier too. All criticisms on the Stealth is just them actually criticizing their chain, not the HP.

Can I check what is your typical volume listening well when plugging Susvara direct to Dave? I'm at -3db, and it's about right for me. Would you go past to positive level?

 

[801evan]

Can I check what is your typical volume listening well when plugging Susvara direct to Dave? I'm at -3db, and it's about right for me. Would you go past to positive level?

That's the same for me. I only went to max volume because the src dx was half volume when I got it brand new and had to fix it. After that, it was around -3dB on the Dave. but you are on HMS which is -3dB. I was mostly not using HMS as I preferred straight CDT. I do range around -6dB to -3dB depending on the loudness of the album.

 

[Rob Watts]

I would recommend against battery power for best performance because this has been done to death 20 years ago and the consensus at the time was that you get dynamic limited when you are on battery - even for headphone use. I suspect battery-based power supply simply could not keep up with high instantaneous power requirement during high dynamic and complex passages. You can try again but thought you may find this data point useful.

Just because some do not like the sound of batteries does not mean they are right. Using batteries with conventional amps gives a completely different sound quality, with things sounding hugely warmer and softer - and this is solely down to reduced noise floor modulation. But because it sounds warmer, softer with much better instrument separation and focus, people mistakenly think it lacks dynamics - due to the apparent softness.

This situation reminds me of the time I was designing the Qutest DAC. This was a replacement to the Qute DAC - and some people had replaced the PSU with a linear and stating it sounded better. So I investigated this, and did indeed find that the Qute DAC was sensitive to the PSU. Replacing the 12v PSU with a 12v car battery (300A, huge dynamic current capacity, zero in-band noise, zero RF noise) gave a big improvement in SQ. So I redesigned the internal PSU regulation on Qutest, and found that there was no difference in sound quality going from a car battery to the supplied 12v PSU - this was down to the improved PSU regulation and RF filtering. But of course once Qutest was launched people still preferred the degradations of a linear supply (which are just RF noise and audio noise sources) compared to the supplied PSU.

Rob,
In what way is optical behind USB on Dave?
Is not All jitter discarded by buffering?
And is this the same case on HMS?

Sorry i reply on a 2016 post.. im catching up from the start of this thread ..oof

You have discovered a post where I made a mistake. In 2016 with Dave I did indeed prefer the brightness of USB, and I thought it was down to the technical benefits of Dave supplying the clock with USB. But later I bought an MSI gaming laptop that had USB and an optical out. On this laptop I could hear zero difference between USB and optical in sound quality - so the "benefits" of clocking from Dave was an illusion, and the extra brightness was down to USB injecting more RF noise into Dave's earth plane and that creating (unmeasurable) increase in noise floor modulation. This illustrates how easy it is to get it wrong from a listening test - it's almost impossible to tell the difference from a genuine improvement in transparency (brighter sound better detail resolution) to a very small increase in noise floor modulation (brighter sound better detail resolution). Listening tests need to be done very carefully, and one needs to re-evaluate and re-test your assumptions - otherwise true progress will not get made.

Optical should always be considered the reference (best) input. Depending upon the source and your system, you may find USB is as good as optical, or it's not as good when it will sound brighter with degraded depth performance.

Source jitter is removed by my DPLL, as this acts like a buffer, where the data is taken out from the buffer via an FPGA low jitter derived clock that is frequency locked (not phase locked) to the incoming data. Thus source jitter is eliminated, as the data is processed by the local 104.25 MHz low jitter clock.

 

[adrianm]

Just because some do not like the sound of batteries does not mean they are right. Using batteries with conventional amps gives a completely different sound quality, with things sounding hugely warmer and softer - and this is solely down to reduced noise floor modulation. But because it sounds warmer, softer with much better instrument separation and focus, people mistakenly think it lacks dynamics - due to the apparent softness.

This situation reminds me of the time I was designing the Qutest DAC. This was a replacement to the Qute DAC - and some people had replaced the PSU with a linear and stating it sounded better. So I investigated this, and did indeed find that the Qute DAC was sensitive to the PSU. Replacing the 12v PSU with a 12v car battery (300A, huge dynamic current capacity, zero in-band noise, zero RF noise) gave a big improvement in SQ. So I redesigned the internal PSU regulation on Qutest, and found that there was no difference in sound quality going from a car battery to the supplied 12v PSU - this was down to the improved PSU regulation and RF filtering. But of course once Qutest was launched people still preferred the degradations of a linear supply (which are just RF noise and audio noise sources) compared to the supplied PSU.



You have discovered a post where I made a mistake. In 2016 with Dave I did indeed prefer the brightness of USB, and I thought it was down to the technical benefits of Dave supplying the clock with USB. But later I bought an MSI gaming laptop that had USB and an optical out. On this laptop I could hear zero difference between USB and optical in sound quality - so the "benefits" of clocking from Dave was an illusion, and the extra brightness was down to USB injecting more RF noise into Dave's earth plane and that creating (unmeasurable) increase in noise floor modulation. This illustrates how easy it is to get it wrong from a listening test - it's almost impossible to tell the difference from a genuine improvement in transparency (brighter sound better detail resolution) to a very small increase in noise floor modulation (brighter sound better detail resolution). Listening tests need to be done very carefully, and one needs to re-evaluate and re-test your assumptions - otherwise true progress will not get made.

Optical should always be considered the reference (best) input. Depending upon the source and your system, you may find USB is as good as optical, or it's not as good when it will sound brighter with degraded depth performance.

Source jitter is removed by my DPLL, as this acts like a buffer, where the data is taken out from the buffer via an FPGA low jitter derived clock that is frequency locked (not phase locked) to the incoming data. Thus source jitter is eliminated, as the data is processed by the local 104.25 MHz low jitter clock.

I'm curious, is this also valid when taking the M-scaler into consideration? I've been using optical for 2 years, but I recently got the Meze Elite, that are much more resolving than my old headphones, and going optical from the streamer into the M-scaler does result in better depth and width, but also a very "soft" and "diffuse" image, with slightly less clarity in vocals for example. USB out of my gaming pc clearly results in more noise( even with a Jitterbug) but Coax out of the streamer gives deeper bass than all above (except maybe optical), a much more focused image (on par with usb), even if slightly more compressed than optical.
I'm starting to believe the wider image out of Optical was just distortion, since the image is so fuzzy. This has happened before with an optical cable that turned out to be half working, so I went back to the stock optical cable provided, which was fully lit.

For reference, this is the cable that sounded blurry before, a 150 euro "Qed quartz reference optical cable " before replacement:

And this is the stock cable which sounded clearer back then (at least with less resolving headphones) :

Granted ,I have been moving them around for testing ,but i don't think i have damaged it in the process. Will check again when i get back home. Travelling with Dave atm

 

[801evan]

I'm curious, is this also valid when taking the M-scaler into consideration? I've been using optical for 2 years, but I recently got the Meze Elite, that are much more resolving than my old headphones, and going optical from the streamer into the M-scaler does result in better depth and width, but also a very "soft" and "diffuse" image, with slightly less clarity in vocals for example. USB out of my gaming pc clearly results in more noise( even with a Jitterbug) but Coax out of the streamer gives deeper bass than all above (except maybe optical), a much more focused image (on par with usb), even if slightly more compressed than optical.
I'm starting to believe the wider image out of Optical was just distortion, since the image is so fuzzy. This has happened before with an optical cable that turned out to be half working, so I went back to the stock optical cable provided, which was fully lit.

For reference, this is the cable that sounded blurry before, a 150 euro "Qed quartz reference optical cable " before replacement:


And this is the stock cable which sounded clearer back then (at least with less resolving headphones) :

Granted ,I have been moving them around for testing ,but i don't think i have damaged it in the process. Will check again when i get back home. Travelling with Dave atm

The lights on the multi core optical is dependent on the bend of the cable.

Optical is still the best for me,if it sounds fuzzy, gotta upgrade the upstream device as optical is very sensitive to psu quality and vibration. Place sorbothanes.

 

[Reactcore]

and the extra brightness was down to USB injecting more RF noise into Dave's earth plane and that creating (unmeasurable) increase in noise floor modulation.

Optical should always be considered the reference (best) input. Depending upon the source and your system, you may find USB is as good as optical, or it's not as good when it will sound brighter with degraded depth performance.

So it is possible to get a tiny piece of the RF devil through USB ground with some sources..

Softer/easier to listen to sound is what i found comparing optical to USB out of my thinclient PC to HMS > Dave.. and i use Dave's HP out and nothing else connected but HMS. And i like how it sounds now..

In my USB case it must come from my ethernet connection (or less likely the PC's SMPS) i know of users that use ethernet over fiber to work this issue for PC's

Have you heard the DX's optical DBNC solution between HMS-DAC once? I have read your opinion in the beginning of HMS thread stating the lossses if normal ferrites used and benefits of tight certain freq. ferrites (after which Wave jumped in with their products)
Or even the cable length.

Im just wondering if you came to new knowledge concerning the DBNC link and RF.

I use short 25cm standard 75ohm CCTV RG59 cables as opposed to long ones.

Can it be they function as antennas picking up radio interference? (i know the air is polluted with radio signals) So thats why certain lengths work better/worse

 

[Reactcore]

I'm curious, is this also valid when taking the M-scaler into consideration? I've been using optical for 2 years, but I recently got the Meze Elite, that are much more resolving than my old headphones, and going optical from the streamer into the M-scaler does result in better depth and width, but also a very "soft" and "diffuse" image, with slightly less clarity in vocals for example. USB out of my gaming pc clearly results in more noise( even with a Jitterbug) but Coax out of the streamer gives deeper bass than all above (except maybe optical), a much more focused image (on par with usb), even if slightly more compressed than optical.
I'm starting to believe the wider image out of Optical was just distortion, since the image is so fuzzy. This has happened before with an optical cable that turned out to be half working, so I went back to the stock optical cable provided, which was fully lit.

For reference, this is the cable that sounded blurry before, a 150 euro "Qed quartz reference optical cable " before replacement:


And this is the stock cable which sounded clearer back then (at least with less resolving headphones) :

Granted ,I have been moving them around for testing ,but i don't think i have damaged it in the process. Will check again when i get back home. Travelling with Dave atm

I have my doubts if multi stranded optical is better than single ones.

Ok thinner strands shorten the inner reflected travel distance of the signal forcing a more straight line.. but on the other side the signal arrives a multitude of times slightly different in time with each strand that may confuse the receiver side.

Longer cables give bigger timing differences than short ones. And bending amplifies this further

I played alot with optical cables with my AK120 DAP that didnt accepted 192khz with some cables .. and landed with a €25 0.5m Audioquest forest single stranded cable.