Merging Hapi - A few words about one of the best DACs I ever listened to, and that money can buy to make happy an Audiophile like me.

[matthias]

22 minutes ago, Em2016 said:

Unless it's power over ethernet? I hope not... for selfish reasons.

 

I hope it will have at least a DC input like Hapi.

 

Matt

[matthias]

1 hour ago, Kal Rubinson said:

"Two channels come out through XLR connectors, two through line level phone jacks, and the remaining two through the headphone jacks."

 

I see discussions coming which of the three pairs sounds best:-)

 

Matt

[tailspn]

31 minutes ago, matthias said:

BTW, does anyone know if Anubis has the power supply internal or external?

Thanks

 

Matt

External. The prototype I was shown at AES was not powered, and I don't remember the DC In connector or voltage. Functionally, Anubis is the same as a Hapi with DA8P board, restricted to 6 channels of playout. It's strength primarily lies in its utility as a 4 channel recorder of all formats and bit/sample rates that Merging supports, its unrestricted monitoring capability (compared to Hapi/Hours without a DAC board), its Zman onboard processor for networked components control as well as monitoring processing, and its extremely small size for that kind of utility. It's designed as a recording systems tool. to flesh out the Hapi/Horus line, as well as a small stand alone 2-4 channel recorder.

 

Hope this helps,

 

Tom

[Kal Rubinson]

7 hours ago, Em2016 said:

In your reference system, is your multi-channel D-to-A done by Merging? A NADAC or Hapi/Horus?

Or something else at the moment?

exaSound e38

7 hours ago, Em2016 said:

If Merging - playing of multi-channel files is from a Mac or Windows PC via Ravenna?

When I had the NADAC+, I used it from a Windows PC via Ravenna.

 

7 hours ago, Em2016 said:

Can you share your entire playback system/chain, include PC/Mac, software etc.

WinPC-based Server (Baetis Prodigy) running Jriver.  USB output via SoTM to e38 to AR MP-1 preamp (or bypassed directly) to:

--3xBenchmark AHB2 amps to 3 B&W 802D3 (LCR). 

--McCormack DNA-1 Rev. A to 2 B&W 804D3 (SL/SR).

--JLAudio Fathom f113 subwoofer.

Other amps, DACs around.

Occasional use of ROON, HQPlayer.  General use of DiracLive.

Big changes coming soon, though.

 

[Kal Rubinson]

6 hours ago, matthias said:

I see discussions coming which of the three pairs sounds best:-)

And the number of possible combinations goes up if you use it for multichannel. 😐

[matthias]

On 3/15/2019 at 8:39 PM, improvedsound said:

For my needs, in my  two-channel Hi-Fi system, I hade configuration the DAC,  assigning for the Left and the Right channel a group of four converters that work in parallel.

 

Please can you share with us your complete hardware and software set-up?

I am particularly interested in your equipment upstream to the Hapi.

Thanks

 

Matt

[tailspn]

Several years ago, an engineer and DAC company owner who everyone here would recognize immediately, said that paralleling DAC channels was one of the stupidest ideas he'd ever heard. It does have the benefit of reducing the noise floor by 3dB  per channel pairing theoretically  (actually less because noise is not completely random in a bounded frequency response system), at the expense of smearing the sound due to timing differences in each of the parallel channels. DAC's today, which approach a  -130dB noise floor, hardly need another 6+dB of noise reduction.

 

I know it's a very salable spec feature today, but just sayin 

[asdf1000]

5 hours ago, tailspn said:

Several years ago, an engineer and DAC company owner who everyone here would recognize immediately, said that paralleling DAC channels was one of the stupidest ideas he'd ever heard. It does have the benefit of reducing the noise floor by 3dB  per channel pairing theoretically  (actually less because noise is not completely random in a bounded frequency response system), at the expense of smearing the sound due to timing differences in each of the parallel channels. DAC's today, which approach a  -130dB noise floor, hardly need another 6+dB of noise reduction.

 

I know it's a very salable spec feature today, but just sayin 

 

Hi Tom

 

Various people over the years have mentioned (or guesstimated) that ESS Sabre DAC chips convert DSD (or 1 bit SDM) without direct conversion to from 1 bit SDM to analogue - there may be a conversion to 5 or 6 bit SDM before conversion to analogue. 

 

At the same time there are DAC chips on the market that are believed to convert 1 bit SDM direct to analogue, like certain TI/Burr Brown chips and AKM , without some intermediate stage.

 

I was wondering why companies like Mytek and Ayre and Merging which have a number of A-to-DSD converts used in recording to DSD, all use ESS Sabre chips for their D-to-A's, instead of chips that convert DSD directly to analogue?

 

I know you're not working for a DAC manufacturer but from your discussions with DAC designers over the years, what's the reason they stick with ESS for their D-to-A's , specifically for DSD conversion to analogue? Instead of chips that convert DSD directly to analogue?

 

Since these companies mentioned design both A-to-DSD converters and DSD-to-A converters, they have the unique opportunity to know what the analogue output of their DACs is supposed to be, to improve the designs of their DACs.

 

I assume these designers of these companies are more than happy with the performance of conversion of DSD to analogue with the ESS chips? The performance is close enough to direct conversion of DSD to analogue?

 

Obviously you are also in a similar unique position as DAC designers, as you know what the output of DSD-to-analogue converts is supposed to be. 

 

I assume you're also happy with the performance of the ESS chips for DSD-to-analogue conversion, assuming the implementation of the DAC chip is done well of course (analogue section design and power supply designs etc) like Merging and Ayre and Mytek have a long reputation of doing of course.

 

[matthias]

34 minutes ago, Em2016 said:

 

Hi Tom

 

Various people over the years have mentioned (or guesstimated) that ESS Sabre DAC chips convert DSD (or 1 bit SDM) without direct conversion to from 1 bit SDM to analogue - there may be a conversion to 5 or 6 bit SDM before conversion to analogue. 

 

At the same time there are DAC chips on the market that are believed to convert 1 bit SDM direct to analogue, like certain TI/Burr Brown chips and AKM , without some intermediate stage.

 

I was wondering why companies like Mytek and Ayre and Merging which have a number of A-to-DSD converts used in recording to DSD, all use ESS Sabre chips for their D-to-A's, instead of chips that convert DSD directly to analogue?

 

I know you're not working for a DAC manufacturer but from your discussions with DAC designers over the years, what's the reason they stick with ESS for their D-to-A's , specifically for DSD conversion to analogue? Instead of chips that convert DSD directly to analogue?

 

Since these companies mentioned design both A-to-DSD converters and DSD-to-A converters, they have the unique opportunity to know what the analogue output of their DACs is supposed to be, to improve the designs of their DACs.

 

I assume these designers of these companies are more than happy with the performance of conversion of DSD to analogue with the ESS chips? The performance is close enough to direct conversion of DSD to analogue?

 

Obviously you are also in a similar unique position as DAC designers, as you know what the output of DSD-to-analogue converts is supposed to be. 

 

I assume you're also happy with the performance of the ESS chips for DSD-to-analogue conversion, assuming the implementation of the DAC chip is done well of course (analogue section design and power supply designs etc) like Merging and Ayre and Mytek have a long reputation of doing of course.

 

 

Good post,

particularly interesting because Merging use for the AD section AKM chips and for DA section ESS chips.

 

Matt

[asdf1000]

13 minutes ago, matthias said:

 

Good post,

particularly interesting because Merging use for the AD section AKM chips and for DA section ESS chips.

 

Matt

 

That makes it even more interesting indeed.

 

I can only guess that the 'direct' DSD conversion to analogue of suitable TI/BB/AKM chips offer no performance advantage over the way ESS do DSD conversion to analogue... to the DAC designers anyway.

 

If there was a technical advantage I would assume TI/BB/AKM would be used in Merging's DSD-to-A converters.

 

Would love to know what Tom has heard from these DAC engineers, and experienced in his professional job too of course, since like the DAC designers, he is also in the unique position to know what the analogue out of a DAC is supposed to sound/measure like.

 

Of course we know there's more to DAC performance than just chip, like analogue section design and power supply sections design. But still the choice of chip is interesting.

 

I'll tag another DSD guru also @Miska

[matthias]

9 minutes ago, Em2016 said:

 

That makes it even more interesting indeed.

 

I can only guess that the 'direct' DSD conversion to analogue of suitable TI/BB/AKM chips offer no performance advantage over the way ESS do DSD conversion to analogue... to the DAC designers anyway.

 

If there was a technical advantage I would assume TI/BB/AKM would be used in Merging's DSD-to-A converters.

 

Would love to know what Tom has heard from these DAC engineers, and experienced in his profession job too of course, since he is also in the unique position to know what the analogue out of a DAC is supposed to sound/measure like.

 

Linn, another top manufacturer compared several top DAC chips for their Klimax range of streamers/DACs and came to the conclusion that AK4497 is the best sounding one, so why not AKM for both ADC and DAC at Merging?

And when ESS why not the top chip ES9038PRO?

 

Matt

[asdf1000]

30 minutes ago, matthias said:

 

Linn, another top manufacturer compared several top DAC chips for their Klimax range of streamers/DACs and came to the conclusion that AK4497 is the best sounding one, so why not AKM for both ADC and DAC at Merging?

And when ESS why not the top chip ES9038PRO?

 

Matt

 

I know about Linn but did want to keep it to discussions about the companies that make both A-to-DSD converters AND DSD-to-A converters, specifically. 

 

Like Merging, Ayre, Mytek. And these 3 use ESS in their D-to-A converters.

 

I'm not so concerned about the choice of which ESS chip - Merging's Dominique Brulhart already engaged in that discussion in another thread, regarding the choice of ESS chip for NADAC and gave the Engineering decisions as to why they chose that chip, instead of the flagship ESS chip at the time.

 

I'm moreso interested in the choice of ESS overall, not which particular ESS chip - as it's believed that ESS DAC chips don't convert DSD direct to analogue, the way some TI/BB/AKM chips do.

 

[matthias]

Recalling posts of Tom here on AS he always found recordings with Hapi/Horus in DSD256 superior to those in DXD.

So I may assume that there is some difference between handling DSD and PCM inside the ESS chips. 

However a more pure DSD path may yield to even greater differences between DSD and PCM.

 

Matt

[matthias]

1 hour ago, Em2016 said:

 

I know about Linn but did want to keep it to discussions about the companies that make both A-to-DSD converters AND DSD-to-A converters, specifically. 

 

Linn do ADC as well inside their streamers but do not sell stand alone ADC converters. 

AFAIK, all ADC converters in production are A-to-DSD.

 

Matt

[asdf1000]

32 minutes ago, matthias said:

Linn do ADC as well inside their streamers but do not sell stand alone ADC converters. 

 

Noted but wanted to keep it to Merging/Ayre/Mytek since that's Pro Audio and Tom has direct experience there.

 

If we start talking about why HiFi DAC manufacturers choose other chips over ESS, Tom may lose interest in the HiFi/consumer world conversation 🙂

 

The point about Merging A-to-DSD using AKM chips is interesting, for sure.

[tailspn]

6 hours ago, matthias said:

 

Linn, another top manufacturer compared several top DAC chips for their Klimax range of streamers/DACs and came to the conclusion that AK4497 is the best sounding one, so why not AKM for both ADC and DAC at Merging?

And when ESS why not the top chip ES9038PRO?

 

Matt

The "why not AKM for both functions" is a good question for  Merging. But please bear in mind that Merging uses only the SDM output conversion part of the 9028 chip, which is the smallest percentage of ESS's design. The format conversions, and other magic, are performed by their own software processes in the neighboring  processor. By the time digital content gets to the 9028, it's all Pulse Density Modulation (DSD). That's also one of reasons they do no use the more expensive 9038, whose richer front end conversion  processes they don't use.

 

Jussi Laako may want to chime in here, with his wealth of knowledge of these chip architectures.

[asdf1000]

4 hours ago, tailspn said:

By the time digital content gets to the 9028, it's all Pulse Density Modulation (DSD).

 

Is that processor before the 9028 converting to 1 bit PDM? 

 

Or is it 5  or 6 bits?

[Miska]

51 minutes ago, Em2016 said:

Is that processor before the 9028 converting to 1 bit PDM? 

 

Or is it 5  or 6 bits?

 

ESS takes only PCM or DSD input. There's not that much choice on that front...

[asdf1000]

10 minutes ago, Miska said:

 

ESS takes only PCM or DSD input. There's not that much choice on that front...

 

Noted, thanks. So it's indeed single bit.

 

But then how the 9028 converts that DSD to analogue is still that mystery of being "DSD like" but not really direct conversion

 

 

[asdf1000]

23 hours ago, Kal Rubinson said:

Big changes coming soon, though.

 

Okto DAC8 8Ch DAC?

 

https://www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-okto-dac8-8ch-dac-amp.7064/

[Miska]

14 minutes ago, Em2016 said:

Okto DAC8 8Ch DAC?

 

https://www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-okto-dac8-8ch-dac-amp.7064/

 

Looks quite nice, but no DSD support, limited to 192k PCM?

 

If one is fine being limited to 192k PCM for multichannel, there are plenty of pro-audio options. Almost too many to count.

 

[Kal Rubinson]

2 hours ago, Em2016 said:

Okto DAC8 8Ch DAC?

 

Among other things.

[Ralf11]

only $1,100 (!!)

 

 

their PS module is likely to be what they are using in the DAC...

 

https://www.oktoresearch.com/modules.htm

 

 

[asdf1000]

2 hours ago, Miska said:

Looks quite nice, but no DSD support, limited to 192k PCM?

 

Hmmm there is mention of DSD256 input

 

 

 

https://www.oktoresearch.com/modules.htm

 

[asdf1000]

11 hours ago, tailspn said:

Merging uses only the SDM output conversion part of the 9028 chip, which is the smallest percentage of ESS's design. The format conversions, and other magic, are performed by their own software processes in the neighboring  processor. By the time digital content gets to the 9028, it's all Pulse Density Modulation (DSD).

 

Hi Tom

 

What sample rate (at 1 bit) do Merging have this neighbouring processor converting to, before the 9028 DAC chip?

 

DSD256?