[Uncle Monty]

Well, I just went on a tear reading everything I could find on the Denafrips Terminator. There are a few users on the Terminator thread that compare it to the DAVE. As can be expected, they like the Terminator. One user borrowed a friend's DAVE for a week, and ended up keeping it in the closet after A/Bing it against the Denafrips.

Both are top tier DACs, but I'm having trouble getting a fix on the sonic differences. What is clear is that the Terminator has the more dense presentation. That makes sense to me, esp. as "thinness" can be a characteristic weakness of the DAVE, and the Terminator has those massive, back breaking, toroidal transformers.

What has me a bit confused, and frankly unnerved, is that I would then expect the DAVE to counter with greater transparency, clarity, and resolution. However, Jay Luong, over at Audiobacon believes that this is where the Terminator excels as well. Comparing the two specifically he writes,

"The Terminator does so more with spatial clarity, vividness, and air. You hear deeper into the trinkets, layers, and grooves of the sound. From the leading edges of a violin, drum skin flex, and finger movement on a trumpet. All are placed in plain view with uncanny control and speed. It just sounds more tonally “elegant” and agile.
The DAVE takes a more naturalistic approach. Its greatest strengths are timbre and coherence...."

At the end, he actually prefers the DAVE for its timbral accuracy. Yet, the way he describes the Terminator's superiority is what I'm accustomed to hearing as one of DAVE's signature strengths.

There are not a lot of posts comparing these units here. Has anyone had time with both, especially in the context of headphone listening?

There's an interview with Rob Watts where he's a bit dismissive of 'ladder DACs' , describing them as a 'dead end' I think - might be wrong - anyone?

 

[edwardsean]

Well I'm generally aligned to the nuances @llamaluv described a few posts back but here's a whimsical analogy in addition. Terminator is like putting on sunglasses outside... you're protected from the glare and you can see just fine, but your vision is also tinted, colored. DAVE is no sunglasses and 20/20 sight to boot. Look one way and wow what a beautiful glistening day. Look another way and you may wince, i.e. poor recordings.

That analogy makes a lot of sense to me and that was my general working assumption as well. What threw me off was the way that people related the hyper-detailed quality of the Terminator, even exceeding the DAVE. Micro details are described to pop out much more vividly than the DAVE. Negatively interpreted, this could be oversharpening that is perhaps artificial and fatiguing. Nevertheless, the consistent impression by proponents of Denafrips is that of greater engagement than the DAVE.

The very thing you lose with filtered vision is the detail and engagement with the image. Everything recedes and doesn't pop out.

There's an interview with Rob Watts where he's a bit dismissive of 'ladder DACs' , describing them as a 'dead end' I think - might be wrong - anyone?

I'm not overly concerned with this line of reasoning. it's like being dismissive of tube technology. Yes, there are inherent theoretical disadvantages compared to solid state. However, there are wonderful tube designs and a reason why they continue to exist at every level of audio. This isn't the first time R2R has been written off, yet smart designers keep finding a way to make it relevant. I'm team Chord FPGA, but I can't rule out Denafrips on spec, just because its R2R.

There's a rule among seasoned pro audio engineers. "If it sounds good, it is good." Doesn't matter how you got there.

 

[koven]

That analogy makes a lot of sense to me and that was my general working assumption as well. What threw me off was the way that people related the hyper-detailed quality of the Terminator, even exceeding the DAVE. Micro details are described to pop out much more vividly than the DAVE. Negatively interpreted, this could be oversharpening that is perhaps artificial and fatiguing. Nevertheless, the consistent impression by proponents of Denafrips is that of greater engagement than the DAVE.

The very thing you lose with filtered vision is the detail and engagement with the image. Everything recedes and doesn't pop out.



I'm not overly concerned with this line of reasoning. it's like being dismissive of tube technology. Yes, there are inherent theoretical disadvantages compared to solid state. However, there are wonderful tube designs and a reason why they continue to exist at every level of audio. This isn't the first time R2R has been written off, yet smart designers keep finding a way to make it relevant. I'm team Chord FPGA, but I can't rule out Denafrips on spec, just because its R2R.

There's a rule among seasoned pro audio engineers. "If it sounds good, it is good." Doesn't matter how you got there.

Yes I think you have the right grasp on it. In a DAVE thread, it should be unsurprising that DAVE is defended w/ prejudice but holistically there isn't and never will be a consensus on best. DAVE is simply one of many high end discrete options among Bricasti, dCS, CH, MSB, Nagra, Playback, Rockna, Lampi, Mola Mola, the list goes on. Each has their own take on DAC design with their own fanbase. As a consumer you can peruse hundreds of rose-colored impressions only to end up more confused than when you started. There's really no bad choices at this tier, merely different flavors. At some point you just have to bite the bullet and try some first-hand.

Enzo Ferrari smugly told tractor maker Ferruccio Lamborghini that his criticisms of Ferrari was unfounded nonsense. Lo and behold 50+ years later Lamborghini remains one of Ferrari's top competitors.

 

[DaveRedRef-III]

Just checked the architecture specs for PCM on the Denafrips and it is 26bit! I wouldn’t bother even sound testing such a product against something like a Dave DAC. Bit depth architecture has a huge affect on image focus. Listen to any DAW setup for processing in 32bit and then compare to the 64bit version. Any DAW you like to try. You would never give 32bit system time of day from then on.
In this day an age I amazed there are still DACs trying to claim the high ground with such low bit depth tbh.

 

[edwardsean]

@koven, First of all, the Ferrari/Lamborghini story was choice.

And, I know we're talking about this in the Headfi house of DAVE. I am one of the family. I've paid my DAVE entrance fee and learned the secret handshake.

Yet, I'm not crazy about judging audio "prejudicially." What is it again? Digital to Audio Veritas Extremus? "Truth to the extreme" right? We chose DAVE because we want biases overturned in favor of the true sound. We have to let the actual audio speak its facts for itself. Specs aren't meaningless. @DaveRedRef-III, I get what you're saying, but they're not everything either.

The Terminator can also do DSD1024 and PCM1536. Does that now make it better than the DAVE, which is stuck at 512/768? That's not even theoretical for me. I use HQP and I actually want 1024/1536. Srajan Ebaen over at 6Moons tested out the 1536KHz upscaling quite favorably on the Denafrips. But, this such a small piece of all of this right?

In the DAW 32bit/64bit comparison–you mean all other things being equal right? It's not just 64bit processing no matter how bad it is. In the case of these DACs all other things aren't equal. That's what we're discussing. Consider....

Lamborghini to Ferrari, "I beat you, V12, baby, V12! How many cylinders you got, 6, pfff, 8? I got 12. Count em, 12 cylinders. I win. I win!"
Ferrari: "Mmm... I think there might be a little bit more to an engine than the cylinder count. It's just a little bit more complicated."
Porsche: "Hey fellas. I really don't give a rip about any of this. What say we take it to LeMans and we'll see who wins. Veritas extremus."

 

[audio_1]

Yes I think you have the right grasp on it. In a DAVE thread, it should be unsurprising that DAVE is defended w/ prejudice but holistically there isn't and never will be a consensus on best. DAVE is simply one of many high end discrete options among Bricasti, dCS, CH, MSB, Nagra, Playback, Rockna, Lampi, Mola Mola, the list goes on. Each has their own take on DAC design with their own fanbase. As a consumer you can peruse hundreds of rose-colored impressions only to end up more confused than when you started. There's really no bad choices at this tier, merely different flavors. At some point you just have to bite the bullet and try some first-hand.

Enzo Ferrari smugly told tractor maker Ferruccio Lamborghini that his criticisms of Ferrari was unfounded nonsense. Lo and behold 50+ years later Lamborghini remains one of Ferrari's top competitors.

The difference with the Dave and the mScaler is that Rob is the only DAC designer using long tap reconstruction filters. It it the only DAC that can recover transient information correctly from 44.1 kHz as far as I know. In another automotive analogy, it is really the only car with round wheels, all the other ones have square ones, irrespective of cost! It obviously could be improved, better power supply and of course dual optical inputs, with perhaps LC or ST connectors for attaching to a 2m tap mScaler.

 

[edwardsean]

The difference with the Dave and the mScaler is that Rob is the only DAC designer using long tap reconstruction filters. It it the only DAC that can recover transient information correctly from 44.1 kHz as far as I know. In another automotive analogy, it is really the only car with round wheels, all the other ones have square ones, irrespective of cost! It obviously could be improved, better power supply and of course dual optical inputs, with perhaps LC or ST connectors for attaching to a 2m tap mScaler.

I'm sure you're well aware that there is also a software approach in HQP. The technical differences between the two are continuously debated, and there is no need to rehash it here. Both Rob Watts and Jussi Lasko are building filters with the computationally complex algorithms necessary to properly reconstruct the original waveforms. Jussi's long tap filters are different than Rob Watts, but also sound wonderful. Being software based they can be readily applied to any DAC.

HQP filters also go up to double the sample rate of Mscaler filters, which is why it is nice to see that DACs like the Terminator can take advantage of them and scale to 1536KHz.

Again, same side. I'm team DAVE. Though I'm sure that is highly suspect over the course of this discussion. I'm just fascinated by the long tap length technology that Rob Watts pioneered and enjoy exploring it in different forms.

 

[GreenBow]

The difference with the Dave and the mScaler is that Rob is the only DAC designer using long tap reconstruction filters. It it the only DAC that can recover transient information correctly from 44.1 kHz as far as I know. In another automotive analogy, it is really the only car with round wheels, all the other ones have square ones, irrespective of cost! It obviously could be improved, better power supply and of course dual optical inputs, with perhaps LC or ST connectors for attaching to a 2m tap mScaler.

I like how you pointed out your vision for stuff for a next gen system.

Only today I was thinking similarly. Going with the TT2 +M-Scaler model, but the same with the DAVE and the M-Scaler. I think it would be great if Chord built another module the same dimensions as the TT and M-Scaler. The purpose of the module in my mind, would be to supply power to the TT2 and M-Scaler. Also to filter the signal between the M-Scaler and the TT2, in the same way that the Opto-DX does. (Converting to optical and then back to coaxial.)

All of this would of course be isolated and noise immune, so that noise could not travel from M-Scaler to TT2. Same with the powering of the TT2 and M-Scaler from the unit. Meaning the unit would need one plug in, and the internals separate and clean power to TT2 and M-Scaler. Therefore eliminating any chance of noise skipping from TT2 to M-Scaler via mains. TT2 and M-Scaler would not even plug into mains. Just a power cable from clean power from the unit, to TT2 and one to M-Scaler.

It could be just a power supply, and then use Triode's WAVE cables. However it would be really nice if the unit could clean the signal too, like the Opto-DX.

It's exciting to think about. Might become a large unit though.

 

[audio_1]

I like how you pointed out your vision for stuff for a next gen system.

Only today I was thinking similarly. Going with the TT2 +M-Scaler model, but the same with the DAVE and the M-Scaler. I think it would be great if Chord built another module the same dimensions as the TT and M-Scaler. The purpose of the module in my mind, would be to supply power to the TT2 and M-Scaler. Also to filter the signal between the M-Scaler and the TT2, in the same way that the Opto-DX does. (Converting to optical and then back to coaxial.)

All of this would of course be isolated and noise immune, so that noise could not travel from M-Scaler to TT2. Same with the powering of the TT2 and M-Scaler from the unit. Meaning the unit would need one plug in, and the internals separate and clean power to TT2 and M-Scaler. Therefore eliminating any chance of noise skipping from TT2 to M-Scaler via mains. TT2 and M-Scaler would not even plug into mains. Just a power cable from clean power from the unit, to TT2 and one to M-Scaler.

It could be just a power supply, and then use Triode's WAVE cables. However it would be really nice if the unit could clean the signal too, like the Opto-DX.

It's exciting to think about. Might become a large unit though.

The big attraction for me, as well as the sound quality benefits is that it would allow using long optical cables. All digital source components can be placed away from between the speakers to the side of the listening room without the need for long analogue interconnect cables. The mScaler would be used to select sources. My system is currently configured like this, but the opto-DX is required. Physical distance also ensures that RF noise from digital components is well away from sensitive analogue components.

 

[koven]

The difference with the Dave and the mScaler is that Rob is the only DAC designer using long tap reconstruction filters. It it the only DAC that can recover transient information correctly from 44.1 kHz as far as I know. In another automotive analogy, it is really the only car with round wheels, all the other ones have square ones, irrespective of cost! It obviously could be improved, better power supply and of course dual optical inputs, with perhaps LC or ST connectors for attaching to a 2m tap mScaler.

HQP can turn those square wheels into round wheels...

 

[PhenixS1970]

Hi All,

I'm thinking about having a listen to a Purifi Eigentakt digital amplifier using DAVE as pre.

Below some specs of the amp:

Output power (max)

8Ω @1%THD : 227W 4Ω @1%THD : 425W 2Ω @1%THD : 450W

• Gain of + 7.2dB, for a total of 20dB (default) - Maximum power reached with a signal of ~ 4.2V RMS
• Gain of + 12.5dB, for a total of 25.3dB - Maximum power reached with a signal of ~ 2.2V RMS
• Bypass: in this case the signal passes directly from the inputs to the modules, allowing the lowest possible gain of 12.8dB. This mode can be interesting when using a powerful preamplifier, capable of driving a load of 2 to 4kOhm and reaching a voltage of ~ 10V RMS

Any recommendations on which gain with DAVE?

Thank you in advance.

 

[ecwl]

Hi All,

I'm thinking about having a listen to a Purifi Eigentakt digital amplifier using DAVE as pre.

Below some specs of the amp:

Output power (max)

8Ω @1%THD : 227W 4Ω @1%THD : 425W 2Ω @1%THD : 450W

• Gain of + 7.2dB, for a total of 20dB (default) - Maximum power reached with a signal of ~ 4.2V RMS
• Gain of + 12.5dB, for a total of 25.3dB - Maximum power reached with a signal of ~ 2.2V RMS
• Bypass: in this case the signal passes directly from the inputs to the modules, allowing the lowest possible gain of 12.8dB. This mode can be interesting when using a powerful preamplifier, capable of driving a load of 2 to 4kOhm and reaching a voltage of ~ 10V RMS

Any recommendations on which gain with DAVE?

Thank you in advance.

It’s whatever gain you actually need. There are several ways to figure this out...
1) if the Chord Etude is giving you 150W into 4ohms maximum and loud enough for you and the Etude has somewhere between 28dB-30dB gain, you might have enough power going to the Purifi with default of a total 20dB gain.
2) DAVE digitally clips at +4dB which is about 6-7V. If you add the M-Scaler, DAVE would clip at +7dB But the max music amplitude would still be 6-7V. So in theory, you can try Bypass mode with Purifi to see if that’s enough power
3) If you normally set your DAVE say at -30dB to-10dB with the Etude, and the gain of Etude is 28-30dB, you should at least be able to set your DAVE volume to -23dB to -3dB for the Purifi in default more for a gain of 20dB.

That said, the reason why pairing Chord DACs with class D amplifier is not recommended is because class D amplifiers switch at a very low frequency, e.g. Purifi @ 500kHz. By comparison, your Chord DACs are first upsampling to 705/768kHz and then to 104MHz to create an analog waveform and preserve timing accuracy. If the amplifier switches at 500kHz, you’re going to worsen the timing accuracy of Chord DACs. Sonically, what you would hear is that transients are less precise, e.g. finger snaps, drum strikes, guitar plucks. Moreover, the worsening transients can also alter the timbre of instruments and vocals. Moreover, many class D amplifiers have high noise floor modulation (although maybe not Purifi). Increased noise floor modulation can make the music sound bright although sometimes it can make the music sound exciting. The end result is less accurate music reproduction.

However, I know many people in person and on head-fi/audiophile style forums that pair their Chord DACs with class D amplifiers. There is no doubt class D amplifiers frequently have significantly higher signal-to-noise ratio so you’re getting much better soundstage depths from your Chord DACs. Moreover, it just depends on what you’re listening for. Because for some people, the blurred transients can sound smoother (like DSD DACs) and the noise floor modulation can make the music sound exciting so even though the music is less accurate, people sometimes just like the sound of their class D amplifiers more. I personally owned DAVE and Mojo for over 1-1/2 years and cannot tell the difference in transient accuracy until I went to a Boston Symphony Orchestra Concert and reminded myself what live instruments sound like. But I’ve also read people online who frequent live orchestral performance who believe that their R2R DACs with class D amps sound more realistic than Chord DAVE with class A amps.

I’m personally super curious about the new Purifi Eigentakt amplifier myself and would love to have an opportunity to hear it. Ultimately, if you think you enjoy the pairing of an amp to your Chord DAC more, you should go for it. It’s just good to know the technical side of things.

 

[HeeBroG]

Hi @ecwl,

"DAVE digitally clips at +4dB which is about 6-7V. If you add the M-Scaler, DAVE would clip at +7dB But the max music amplitude would still be 6-7V."

I have some DSD recordings that are particularly soft eg. Hope by Hugh Masekela

I use DAVE direct to 105dB quasi full range drivers.

I sometimes turn DAVE/Blu2 up beyond +7dB with a commensurate increase in volume and it still sounds clean without any obvious clipping.

Can you explain how that happen if "max music amplitude" of 6-7V occurs at +7dB?

Thanks

 

[JaZZ]

+7 dB doesn't necessarily mean clipping. Clipping just occurs with signal peaks reaching -0 dB, = full level in the digital domain. That means a recording that never reaches higher than -3 db won't begin to clip before +10 dB.

 

[PhenixS1970]

It’s whatever gain you actually need. There are several ways to figure this out...
1) if the Chord Etude is giving you 150W into 4ohms maximum and loud enough for you and the Etude has somewhere between 28dB-30dB gain, you might have enough power going to the Purifi with default of a total 20dB gain.
2) DAVE digitally clips at +4dB which is about 6-7V. If you add the M-Scaler, DAVE would clip at +7dB But the max music amplitude would still be 6-7V. So in theory, you can try Bypass mode with Purifi to see if that’s enough power
3) If you normally set your DAVE say at -30dB to-10dB with the Etude, and the gain of Etude is 28-30dB, you should at least be able to set your DAVE volume to -23dB to -3dB for the Purifi in default more for a gain of 20dB.

That said, the reason why pairing Chord DACs with class D amplifier is not recommended is because class D amplifiers switch at a very low frequency, e.g. Purifi @ 500kHz. By comparison, your Chord DACs are first upsampling to 705/768kHz and then to 104MHz to create an analog waveform and preserve timing accuracy. If the amplifier switches at 500kHz, you’re going to worsen the timing accuracy of Chord DACs. Sonically, what you would hear is that transients are less precise, e.g. finger snaps, drum strikes, guitar plucks. Moreover, the worsening transients can also alter the timbre of instruments and vocals. Moreover, many class D amplifiers have high noise floor modulation (although maybe not Purifi). Increased noise floor modulation can make the music sound bright although sometimes it can make the music sound exciting. The end result is less accurate music reproduction.

However, I know many people in person and on head-fi/audiophile style forums that pair their Chord DACs with class D amplifiers. There is no doubt class D amplifiers frequently have significantly higher signal-to-noise ratio so you’re getting much better soundstage depths from your Chord DACs. Moreover, it just depends on what you’re listening for. Because for some people, the blurred transients can sound smoother (like DSD DACs) and the noise floor modulation can make the music sound exciting so even though the music is less accurate, people sometimes just like the sound of their class D amplifiers more. I personally owned DAVE and Mojo for over 1-1/2 years and cannot tell the difference in transient accuracy until I went to a Boston Symphony Orchestra Concert and reminded myself what live instruments sound like. But I’ve also read people online who frequent live orchestral performance who believe that their R2R DACs with class D amps sound more realistic than Chord DAVE with class A amps.

I’m personally super curious about the new Purifi Eigentakt amplifier myself and would love to have an opportunity to hear it. Ultimately, if you think you enjoy the pairing of an amp to your Chord DAC more, you should go for it. It’s just good to know the technical side of things.

Thank you very much for your very detailed feedback, ecwl. I'll have to read it a couple of times to let it sink but happy to hijack your knowledge this way I'll certainly report back after I have received the amp (end of Jun).

Edit: According to specs Etude has a gain of 30db. Listening volume is between -40 (relax) and -30 (loud) on DAVE using Vivid Audio Kaya 25 speakers.