[EliasGwinn]

Quote:

Originally Posted by hh83917 /img/forum/go_quote.gif @Elias: Yes, I've just learned about the mute buttons. They are indeed very cool. It will remember the last volume position, which is nice. Also, I found the dim setting pretty useful. I can now press a button and it will go to the preset dim level I've set it as and answer a quick incoming call.

Great! Yeah, I love the way it works.

Quote:

Originally Posted by hh83917 /img/forum/go_quote.gif Another difference I found on the PRE and the HDR is that the HDR does not have auto standby after 15 seconds.

Thats true. We figured it was unnecessary with the remote.

Quote:

Originally Posted by hh83917 /img/forum/go_quote.gif Elias, can you explain a little about the "Phase-Accurate Multi-Track and 5.1" and how do one hooks up a 5.1 for listening (for future references)? Thanks.

Phase-accuracy is a very important consideration when it comes to multi-track audio. Many devices have phase differentials between the various channels of the audio. This will cause inaccuracy in the space perception of the sounds, and can cancel frequencies if the channels are ever summed.

To take advantage of this feature, you need a 5.1 source with 3 stereo digital outputs. These are very hard to find for consumer media playback systems. However, there is a gentleman who is modding Oppo DVD players to send 5.1 from three spdif outputs.

The people that take advantage of the phase accurate 5.1 are studio engineers who have digital interfaces with multiple stereo digital outputs.

All the best,
Elias

 

[lamikeith]

I traded in a DAC1 PRE that I had for a few weeks for a DAC1 HDR soon after it was announced. The units are functionally identical except for the remote control.

The remote volume performs better than expected. With this remote, there are three ways to "mute" the volume: dim, soft mute, and full mute.

dim: Mechanically lowers the volume POT to a lower "dim" level. The dim level is user adjustable and moves to stay within full volume level rather than remaining fixed. I like it.

soft mute: Mechanically lowers the volume POT to the dim level, then mutes.

mute/off: Pressing the remote "off" button once causes the unit to mute and indicate mute by lighting all LEDs. Pressing "off" again puts the unit in standby.

on: Pressing "on" returns the unit to an active, full volume state, no matter which mute/standby state the unit was in previously.

I've been playing with all of the remote buttons and mute modes, exercising every possible combination, and the unit consistently performs in a logical manner.

The sound is identical to the DAC1 PRE: excellent.

This is a nicely engineered, well thought out product. Thank you, Benchmark!

 

[HeadLover]

Is there any SQ different from the PRE to the HDR ???

BTW
EliasGwinn

I have heard of a new DAC chip that is capable of doing 32 bit

ESS 32 Bit Sabre DAC Chip


http://www.audiophilia.com/wp/?p=1268

Will there be a new DAC from Benchmark one day using this new chip? seem to be a better quality that what we have right now, not ?

 

[hh83917]

@Elias: Thank you for the explanations.

@HeadLover: The SQ should be the same as the DAC1 PRE. Main difference for me is that the volume controls on the HDR is more precise now (does not have the dials like the original volume knob) thanks to motorized knob.
The new 32bit chip looks interesting, I wonder when will manufacturers start to implement them.

@lamikeith: I felt the same way too. Other than that, I found myself pressing the remote and watch the volume goes up and down just for fun yesterday...

 

[emmodad]

.... except in the minds of ESS marketing people, that is.


remember that a DAC of this type, in very simplified terms, looks like:

input data interface > oversampling digital filter > DAC (modulator) > output conditioning

ESS adds claimed special jitter reduction stage before the modulator


before anyone's blood pressure rises, this chip is not a 32-bit DAC in terms of performance. this is a DAC chip which:

- accepts PCM input data in data word sizes up to 32 bits. "32" is simply the possible width of a data word (ie 16, 18, 20, 24 etc resolution data packed in a 32-bit word, which could be a typical data word size coming from ie a digital audio workstation during intermediate processing)

- contains internal "32-bit" digital filters. this "32" is unspecified, but from public info most probably refers to size of data word accepted at filter input; older public info for the recent/previous generation of this chip indicates that digital filters have internal 48-bit calculation precision, and that the digital filters contribute less than -170 dB of noise as they process signal before it is sent to the modulator. this is performance rather below 32-bit equivalent

- has a modulator (trademarked name: "HyperStream") with architecture somewhat different that typical sigma-delta. published info on the previous-generation version of this chip specifies a modulator noise floor (ie the noise contribution from the modulator) which is "well below -160 dB" ie again rather short of 32-bit equivalent performance

- achieves a claimed THD+N figure of -120 dB which is certainly good, but which is similar to several other high-end DAC integrated circuit products from ie Wolfson, TI/Burr-Brown. it is about 19-20 equivalent bits from a performance perspective.

- achieves a claimed 129 dB DNR performance measurement which again is similar to several other high-end DAC integrated circuit products from ie Wolfson, TI/Burr-Brown. this is ~20-21 equivalent bits from a performance perspective. The ESS 9018 contains many (16?) independent channels of this DAC architecture; by connecting several in parallel, it is possible to increase the effective DNR by 6 dB. this is where the other published performance claim of 135 dB comes from.

use the search button and google to find numerous scientific and academic articles explaining why reasonable real-world limitations constrain achieveable performance to this range of 21-23 bit-equivalent.


bottom line: although initial (very uncontrolled) listening tests are indeed reporting positive impressions (which I have no reason to doubt and on which I have no opinion), the ESS 9018 accepts input data words in format up to 32-bit wide. calling this a "32-bit DAC" is simply disingenuous marketing nonsense from ESS. they're not the only ones, AKM is doing something similar in current marketing.

 

[Quaddy]

/\ sure you have got the right thread here?

 

[The Monkey]

Quote:

Originally Posted by hh83917 /img/forum/go_quote.gif 0000 The new 32bit chip looks interesting, I wonder when will manufacturers start to implement them.

Here's one.

 

[The Monkey]

Quote:

Originally Posted by EliasGwinn /img/forum/go_quote.gif The 'garage-door' optical ports were susceptible to breaking. The hinged-door broke on too many units, so we are currently using the original style on all units. All the best, Elias

A very smart move. I just had a "garage door" port break on me (on a different brand unit) and it pisses me off to no end.

 

[emmodad]

well, info responding to posts 2532 & 2533.....

 

[hh83917]

@The Monkey: I think the new Headroom one uses the 9008 chip instead of the 9018 that claims 32 bit.

@emmodad: Thanks for the long explanation. Unfortunately, that's what marketing people do most of the time. Anyways, I'm happy to have my DAC1 now and probably will stick with it for years as I gradually upgrade other components, such as speakers, in the future.

 

[EliasGwinn]

Quote:

Originally Posted by emmodad /img/forum/go_quote.gif before anyone's blood pressure rises, this chip is not a 32-bit DAC in terms of performance.

Exactly. This chip does have some interesting potential characteristics, but it does not acheive 32-bit performance. Bit-depth only affects the dynamic range (aka signal-to-noise ratio). In other words, it won't affect the tone of the DAC, merely the noise floor.

The aspect of DAC chips that could use a real breakthrough is filter performance. Currently, there is not enought horsepower (DSP) in a DAC chip to effectively eliminate all frequencies above Nyquist while remaining linear for all frequencies below Nyquist.

All the best,
Elias

 

[emmodad]

Quote:

Originally Posted by EliasGwinn /img/forum/go_quote.gif Exactly. This chip does have some interesting potential characteristics, but it does not acheive 32-bit performance. Bit-depth only affects the dynamic range (aka signal-to-noise ratio). In other words, it won't affect the tone of the DAC, merely the noise floor.

yes, although this chip provides no better dynamic range ability than several other high-end devices as mentioned (TI, Wolfson,). It provides only an input data format ability to accept 32-bit wide data words, and (apparently) to pass those 32-bit wide data words to input of the digital filters. Some other high-quality DAC chips have data input formats only up to and including 24-bit wide words.


Quote:

Originally Posted by EliasGwinn /img/forum/go_quote.gif The aspect of DAC chips that could use a real breakthrough is filter performance. Currently, there is not enought horsepower (DSP) in a DAC chip to effectively eliminate all frequencies above Nyquist while remaining linear for all frequencies below Nyquist.

Integrated Circuit architectures are always subject to tradeoffs of cost / performance / etc. DAC chips (such as those used in the DAC1) designed in the early 1990s were subject to constraints of manufacturing technology and costs of that era, hence the amount of silicon area dedicated to the digital filtering engine was limited (and generally hardwired, ie not a "software code on DSP" implementation)

this is why, ie, the architecture of the Berkeley Audio Design alpha DAC is an interesting example (digital filtering is performed in an ADI SHARC DSP external to and prior to the actual DAC chip); and one of the interesting points of the new flagship DAC chips from Wolfson (a product designer has the choice of using Wolfson's on-chip digital filtering engine, for very good performance at lowest parts cost, or putting a DSP (as digital filtering engine) in front of the DAC chip, and use the Wolfson with its on-chip digital filtering engine bypassed).

so, I'm waiting to see what John does with a block of dsp horsepower in front of a DAC chip with excellent performing modulator (hint Wolfson or ADI). Oh: and remind him to do HDCD decoding in that dsp; price the product at a premium to DAC1 but well below alpha; and the resulting "DAC2HDCD" will sell like wildfire.

wishful thinking ps: give it a native FireWire interface...... 24/192 through a modern John Siau design...... mmmmmmmmmmm.

 

[gregeas]

Agreed with emmodad we are entering an era of new potential for high-end DACs. For the last five years or so it has struck me how little evolution there was in DAC design. But now we have new, interesting products from Ayre, Weiss, Berkeley, PS Audio, etc. The ESS Sabre chip is going into some interesting products as well. The acceptance of PC audio by audiophiles is opening up the market.

In the future DAC designers should consider adding a HDMI input. My laptop has an HDMI port, and I'm experimenting with this connection to my Cary pre-pro. I was surprised that everything worked, and the quality was quite good. Blu-ray players like the new Oppo are universal and can output hi-res PCM and DSD via HDMI.

BTW, I'm back to using a Benchmark (Pre) in my main rig. I have a PS3, cable box, and laptop connected. Works great connected to my speaker amp!

 

[HeadLover]

I hope some really new stuff be out soon, something really "revolutionary" kind of.

And who knows, maybe there will be a DAC2 some day

 

[EliasGwinn]

Quote:

Originally Posted by emmodad /img/forum/go_quote.gif Integrated Circuit architectures are always subject to tradeoffs of cost / performance / etc. DAC chips (such as those used in the DAC1) designed in the early 1990s were subject to constraints of manufacturing technology and costs of that era, hence the amount of silicon area dedicated to the digital filtering engine was limited (and generally hardwired, ie not a "software code on DSP" implementation) this is why, ie, the architecture of the Berkeley Audio Design alpha DAC is an interesting example (digital filtering is performed in an ADI SHARC DSP external to and prior to the actual DAC chip); and one of the interesting points of the new flagship DAC chips from Wolfson (a product designer has the choice of using Wolfson's on-chip digital filtering engine, for very good performance at lowest parts cost, or putting a DSP (as digital filtering engine) in front of the DAC chip, and use the Wolfson with its on-chip digital filtering engine bypassed).

Exactly. This is what we do with the ASRC, as well. The bulk of our filtering is done in an Analog Devices AD1896 ASRC chip. The AD1896 is a digital-only chip, so it doesn't have to be 'room-mates' with analog circuitry. This free's the chip designer to build a DSP work-horse that can execute thoroughly.

Quote:

Originally Posted by emmodad /img/forum/go_quote.gif so, I'm waiting to see what John does with a block of dsp horsepower in front of a DAC chip with excellent performing modulator (hint Wolfson or ADI). Oh: and remind him to do HDCD decoding in that dsp; price the product at a premium to DAC1 but well below alpha; and the resulting "DAC2HDCD" will sell like wildfire. wishful thinking ps: give it a native FireWire interface...... 24/192 through a modern John Siau design...... mmmmmmmmmmm.

Great feedback. I will forward this to the man, himself.

All the best,
Elias