Review and Measurements of Benchmark DAC3

[Drakkar Noir]

about the headamp in the HGC... am I missing much by not upgrading to the HPA4?

I'm not in need of much power AFAIK, but the headphones I posses are somewhat different from one another. Apart from the sensitive JH Roxannes, I also posses Audeze's LCD-3 and might get a ZMF Atrium in the near future.

the HPA4's 256 step volume control is tempting, I admit. I have no experience trying to drive IEMs through a desktop amp, which are generally too noisy for these.

 

[John_Siau]

about the headamp in the HGC... am I missing much by not upgrading to the HPA4?

I'm not in need of much power AFAIK, but the headphones I posses are somewhat different from one another. Apart from the sensitive JH Roxannes, I also posses Audeze's LCD-3 and might get a ZMF Atrium in the near future.

the HPA4's 256 step volume control is tempting, I admit. I have no experience trying to drive IEMs through a desktop amp, which are generally too noisy for these.

The 256-step volume control and the 137 dB SNR will allow you to dial in the level that you need for you IEMs. There will be absolutely no audible noise from the HPA4 when driving IEMs.

 

[Barone Birra]

I use the HPA4 with a 64 Audio U18t, and the outcome is awesome. The background is pitch black.

 

[gabrielb]

Has anyone used the Benchmark DAC as a crossover between main/subwoofer and if so how?
I can install miniDSP miniDSP SHD-Studio, perform the XO in miniDSP and route the high frequencies to
DAC3->AHB2->Main Speakers leaving the low frequencies to the SHD internal DAC and from there to a self-powered subwoofer.

miniDSP ==>[internal DAC-> Analog out] ==> subwoofer
||
==> [digital out] ==> DAC3 ==> AHB2 ==> Main (passive) speakers

Did anyone tried this and can attest that there are no timing issues caused by two separate DAC ?

I wonder if Benchmark ever measured time deviation between miniDSP SHD analog output and Benchmark DAC3 when fed from miniDSP SHD digital out and if that can be corrected in DSP (by inserting a delay in one of the channels)?

It could be done by looping back the analog outputs of miniDSP SHD and DAC3 and testing time delays.
The next step is trying to calibrate the 2 streams by adding a delay on one of them inside miniDSP.

After calibrating the streams leave them running overnight and see if there is any deviation caused by clock skew

 

[ex audiophile]

Has anyone used the Benchmark DAC as a crossover between main/subwoofer and if so how?
I can install miniDSP miniDSP SHD-Studio, perform the XO in miniDSP and route the high frequencies to
DAC3->AHB2->Main Speakers leaving the low frequencies to the SHD internal DAC and from there to a self-powered subwoofer.

miniDSP ==>[internal DAC-> Analog out] ==> subwoofer
||
==> [digital out] ==> DAC3 ==> AHB2 ==> Main (passive) speakers

Did anyone tried this and can attest that there are no timing issues caused by two separate DAC ?

I wonder if Benchmark ever measured time deviation between miniDSP SHD analog output and Benchmark DAC3 when fed from miniDSP SHD digital out and if that can be corrected in DSP (by inserting a delay in one of the channels)?

It could be done by looping back the analog outputs of miniDSP SHD and DAC3 and testing time delays.
The next step is trying to calibrate the 2 streams by adding a delay on one of them inside miniDSP.

After calibrating the streams leave them running overnight and see if there is any deviation caused by clock skew

Interesting question. I've tried to figure out how to incorporate DSP into a fully Benchmark "stack" but it seemed to require two DAC3B units (one for mains, one for sub). That still left an issue with multiple subs, and Dirac Live doesn't really deal with the subs, DBC not available in minidsp units, the cost, not audible anyway, blah blah. Life is easier and cheaper with my STR pre, and if we are patient there'll be a Wiim super ultra with balanced outs and a decent room correction feature lol. Hopefully @John_Siau can weigh in on these issues.

 

[John_Siau]

Has anyone used the Benchmark DAC as a crossover between main/subwoofer and if so how?
I can install miniDSP miniDSP SHD-Studio, perform the XO in miniDSP and route the high frequencies to
DAC3->AHB2->Main Speakers leaving the low frequencies to the SHD internal DAC and from there to a self-powered subwoofer.

miniDSP ==>[internal DAC-> Analog out] ==> subwoofer
||
==> [digital out] ==> DAC3 ==> AHB2 ==> Main (passive) speakers

Did anyone tried this and can attest that there are no timing issues caused by two separate DAC ?

I wonder if Benchmark ever measured time deviation between miniDSP SHD analog output and Benchmark DAC3 when fed from miniDSP SHD digital out and if that can be corrected in DSP (by inserting a delay in one of the channels)?

It could be done by looping back the analog outputs of miniDSP SHD and DAC3 and testing time delays.
The next step is trying to calibrate the 2 streams by adding a delay on one of them inside miniDSP.

After calibrating the streams leave them running overnight and see if there is any deviation caused by clock skew

The four output channels on the miniDSP have individual delay adjustments. You should be able to match the delays without a problem.

We have miniDSP SHD devices in all of our listening rooms. Some are being used as crossovers and these drive a pair of DAC3 B converters using the SPDIF outputs on the mini. The SHD version does not have analog outputs. MiniDSP does make a version with built-in DACs (in addition to having the digital outputs). The built-in DACs will have a different delay than the DAC3, but it should be possible to match the delays using the settings in the mini.

 

[ex audiophile]

The SHD does have analog outs. From their website:

"The SHD has extensive connectivity options. Three digital inputs, two analog inputs and USB Audio enable the SHD to fit right into any modern audio system. The four-channel DSP-controlled outputs are available as single-ended (RCA) and balanced (XLR) analog as well as SPDIF digital."

 

[RichB]

The SHD does have analog outs. From their website:

"The SHD has extensive connectivity options. Three digital inputs, two analog inputs and USB Audio enable the SHD to fit right into any modern audio system. The four-channel DSP-controlled outputs are available as single-ended (RCA) and balanced (XLR) analog as well as SPDIF digital."

There are a range of SHD products, this one is digital:

SHD Studio

All-digital Roon Ready network player with Dirac Live® room correction, miniDSP audio toolbox for subwoofer integration and multiway speakers, Volumio music player, headphone output, 450MHz Sharc DSP processor

www.minidsp.com

- Rich

 

[John_Siau]

There are a range of SHD products, this one is digital:

SHD Studio

All-digital Roon Ready network player with Dirac Live® room correction, miniDSP audio toolbox for subwoofer integration and multiway speakers, Volumio music player, headphone output, 450MHz Sharc DSP processor

www.minidsp.com

- Rich

We are using the miniDSP SHD Studio. Sorry for the confusion on the product name.

 

[gabrielb]

The four output channels on the miniDSP have individual delay adjustments. You should be able to match the delays without a problem.

We have miniDSP SHD devices in all of our listening rooms. Some are being used as crossovers and these drive a pair of DAC3 B converters using the SPDIF outputs on the mini. The SHD version does not have analog outputs. MiniDSP does make a version with built-in DACs (in addition to having the digital outputs). The built-in DACs will have a different delay than the DAC3, but it should be possible to match the delays using the settings in the mini.

Thanks for replying!
I spent some time this weekend reading the SPDIF spec (extremely simple and short) and my understanding is that there are multiple sources for delay:

1) Syncing to the preamble and then to the beginning of an audio block (which at 96KHZ means up-to 2msc) after each stop/start action, but after further investigation I learned that miniDSP output continuous data stream (with NULL-DATA when nothing plays) keeping both sides in-sync at all times.

2) Resync after FS change (up-to 6 msec on DAC3), but that is not a problem with miniDSP as it is locked to 96/24

3) Processing time on DAC3 (set to 0.82 ms at 96 kHz) which is a fixed value and can easily be solved using miniDSP delay adjustments

4) Clock skew/drift between the source (miniDSP) and target (DAC3) as the S/PDIF spec allows time difference of 1000 ppm between sides translating to almost 1 second delay after 15 minutes of music (what King Crimson refers to as single track )
I assume that well built modern products like miniDSP and DAC3 aim much higher than the 1000 PPM allowed by the standard, but even at 1 PPM drift we accumulate almost 1 msec delay every 15 minutes of play time.
The S/PDIF protocol doesn't define any feedback channel from the destination (DAC) to the source (miniDSP) so there is no way to slow down or speedup the source.

This seems to be the real problem.
I'm still trying to understand if and how this issue is solved -
A. A simple approach will buffer the incoming stream reclocking it for perfect timing.
This will create unavoidable drift between the channels after the XO (assuming 2 standalone DAC).
B. An adaptive approach might sample the incoming stream at power up and if the drift seen after a 10-15 second of sample time is proven to be very low (say 1PPM) will humbly accept the source clock and disable the reclocking.
This might be possible with miniDSP being a high quality source.
C. Use ASMR to keep both sides in-sync stretching time by removing or adding samples (after extrapolation) to keep source and destination in-sync.
This will eliminate the drift, but might cause *some* artifact.
Maybe if the clocks difference is very low and the extrapolation done wisely this won't be detectable by human ear

 

[John_Siau]

Thanks for replying!
I spent some time this weekend reading the SPDIF spec (extremely simple and short) and my understanding is that there are multiple sources for delay:

1) Syncing to the preamble and then to the beginning of an audio block (which at 96KHZ means up-to 2msc) after each stop/start action, but after further investigation I learned that miniDSP output continuous data stream (with NULL-DATA when nothing plays) keeping both sides in-sync at all times.

2) Resync after FS change (up-to 6 msec on DAC3), but that is not a problem with miniDSP as it is locked to 96/24

3) Processing time on DAC3 (set to 0.82 ms at 96 kHz) which is a fixed value and can easily be solved using miniDSP delay adjustments

4) Clock skew/drift between the source (miniDSP) and target (DAC3) as the S/PDIF spec allows time difference of 1000 ppm between sides translating to almost 1 second delay after 15 minutes of music (what King Crimson refers to as single track )
I assume that well built modern products like miniDSP and DAC3 aim much higher than the 1000 PPM allowed by the standard, but even at 1 PPM drift we accumulate almost 1 msec delay every 15 minutes of play time.
The S/PDIF protocol doesn't define any feedback channel from the destination (DAC) to the source (miniDSP) so there is no way to slow down or speedup the source.

This seems to be the real problem.
I'm still trying to understand if and how this issue is solved -
A. A simple approach will buffer the incoming stream reclocking it for perfect timing.
This will create unavoidable drift between the channels after the XO (assuming 2 standalone DAC).
B. An adaptive approach might sample the incoming stream at power up and if the drift seen after a 10-15 second of sample time is proven to be very low (say 1PPM) will humbly accept the source clock and disable the reclocking.
This might be possible with miniDSP being a high quality source.
C. Use ASMR to keep both sides in-sync stretching time by removing or adding samples (after extrapolation) to keep source and destination in-sync.
This will eliminate the drift, but might cause *some* artifact.
Maybe if the clocks difference is very low and the extrapolation done wisely this won't be detectable by human ear

The DAC3 has a fixed time delay. This delay is sample rate dependent, but does not change if the input sample rate is constant. The PLL in the upsampler locks to a fixed delay. 1.36 ms at 44.1 kHz, 0.82 ms at 96 kHz, and 0.47 ms at 192 kHz. See pages 48, 58, and 59 of the DAC3 B manual for more information.

If you feed multiple DAC3 converters from the same SPDIF signal, they will all be perfectly in phase with each other. On page 58 of the manual you will see that the interchannel phase accuracy between channels is preserved across multiple DAC3 units.

The miniDSP operates at a fixed 96 kHz output frequency. Just connect a DAC3 to each output and they will be in phase. If you are using a DAC3 and the miniDSP analog outputs, you will just need to adjust the delay on one of the channel pairs.

 

[gabrielb]

The DAC3 has a fixed time delay. This delay is sample rate dependent, but does not change if the input sample rate is constant. The PLL in the upsampler locks to a fixed delay. 1.36 ms at 44.1 kHz, 0.82 ms at 96 kHz, and 0.47 ms at 192 kHz. See pages 48, 58, and 59 of the DAC3 B manual for more information.

If you feed multiple DAC3 converters from the same SPDIF signal, they will all be perfectly in phase with each other. On page 58 of the manual you will see that the interchannel phase accuracy between channels is preserved across multiple DAC3 units.

The miniDSP operates at a fixed 96 kHz output frequency. Just connect a DAC3 to each output and they will be in phase. If you are using a DAC3 and the miniDSP analog outputs, you will just need to adjust the delay on one of the channel pairs.

Thanks you so much for clarifying this!

 

[F_M]

I read the manual for DAC3 HGC and I'm not sure what happens with headphone volume when the HT mode is activated. Is it also maxed out same as analog outputs on the back panel? I would appreciate some input from those who have experience with this device. Thank you!

 

[John_Siau]

I read the manual for DAC3 HGC and I'm not sure what happens with headphone volume when the HT mode is activated. Is it also maxed out same as analog outputs on the back panel? I would appreciate some input from those who have experience with this device. Thank you!

On the DAC3 HGC the headphone volume and line volume are both controlled by a single volume knob.

The DAC is set to maximum volume (line and headphone) when the HT mode (home theater bypass mode) is active. This volume control bypass is intended for applications where the volume is controlled by an upstream device. This mode is generally not recommended for headphone use because the upstream device is in control of the headphone level.

In our HPA4, we provide completely separate control levels for the headphone and line outputs.

 

[F_M]

Thank you for clarifying, much appreciated!

 

[Sal1950]

On the DAC3 HGC the headphone volume and line volume are both controlled by a single volume knob.

John, Thanks again for your strong continued support of our members.
Very few others lend this type of direct customer interaction.
Awesome, Sal1950

 

[AudioSceptic]

John, Thanks again for your strong continued support of our members.
Very few others lend this type of direct customer interaction.
Awesome, Sal1950

Indeed. I wish I had an excuse to buy a Benchmark product!

 

[Rischa]

Hello, all. I'm receiving a DAC3 b today as the first component in what will eventually be an all Benchmark system. In the meantime I'm wondering if it will work to use the DAC3's unbalanced outs with my Denon PMA-1700NE integrated amplifier. According to Amir's tests here, the DAC3 outputs 3.1 volts unbalanced, while the Denon's input sensitivity is rated at .83 volts. Does this mean they're incompatible?

Thanks in advance for the help. I'm new to the world of specs and measurements but eager to learn.

 

[NTK]

Hello, all. I'm receiving a DAC3 b today as the first component in what will eventually be an all Benchmark system. In the meantime I'm wondering if it will work to use the DAC3's unbalanced outs with my Denon PMA-1700NE integrated amplifier. According to Amir's tests here, the DAC3 outputs 3.1 volts unbalanced, while the Denon's input sensitivity is rated at .83 volts. Does this mean they're incompatible?

Thanks in advance for the help. I'm new to the world of specs and measurements but eager to learn.

Welcome to ASR!

No. They will work just fine together. It only means you may (more likely may not) need to turn your volume knob a little bit lower.

 

[Rischa]

Welcome to ASR!

No. They will work just fine together. It only means you may (more likely may not) need to turn your volume knob a little bit lower.

Wonderful, thank you!