Chord Hugo TT 2 - A DAC, preamplifier & headphone amplifier

[Summit]

6 hours ago, mozes said:

Will leave the answer to someone techy.

What I do know is that all DSD tracks sound richer with more layering under the DSD plus mode (no decimation)

 

For PCM music its best to set the Dave to PCM plus, because it will when use 164,000 taps. If the DAC is set to DSD plus while playing PCM music it will only use 88,000 taps. For DSD records it’s best to set the DAC to DSD plus. So it’s all about how many taps (and filter) that are used 164,000 taps for PCM plus and 82,000 for DSD plus.  

 

If am not mistaking Dave has an auto detector for this.

[Summit]

Rob Watts are saying that as a pure DAC the SQ from Dave is better becuse of 10 more element, but if you need to drive hard to drive HPs or speakers the TT2 has the advantage of more power. How much better is the question?

[ElviaCaprice]

3 hours ago, Summit said:

 

For PCM music its best to set the Dave to PCM plus, because it will when use 164,000 taps. If the DAC is set to DSD plus while playing PCM music it will only use 88,000 taps. For DSD records it’s best to set the DAC to DSD plus. So it’s all about how many taps (and filter) that are used 164,000 taps for PCM plus and 82,000 for DSD plus.  

 

If am not mistaking Dave has an auto detector for this.

You are mistaken, there is no auto detect.  It's not just about how many taps and filter are used.  

[auricgoldfinger]

4 hours ago, One and a half said:

The supercaps in the MarkII is a better idea, but only worthwhile if there's isolation from the charging system as in the LPS1x

 

+1

[ElviaCaprice]

8 hours ago, austinpop said:

 

Thanks Mark.

 

I followed the reference and watched the video, but nowhere have I found what decimation is.

 

will keep digging.

Okay, Rajiv.

Keep us informed if you find more information.  Would have been nice if we could find exact details from Rob, but I can't seem to find further details.

[Summit]

1 hour ago, ElviaCaprice said:

You are mistaken, there is no auto detect.  It's not just about how many taps and filter are used.  

 

Rob Watts @

“It's simply because the DSD+ filter is too big to fit with the PCM+ 164,000 tap filter. It was under 70% capacity but the design was not routable. But the low latency 82,000 tap is much smaller...”

 

https://www.head-fi.org/threads/chord-electronics-dave.766517/page-30

 

ecwl @ explanation about decimation.

 

“From what I read over at Head-Fi (pages 29 & 30), I think there are actually two possible filters to choose from on the Chord DAVE. One is the WTA 164k tap filter, which Rob Watts called PCM+. And the other filter is called DSD+. He said that if the WTA 164k tap filter causes too much audio delay for video, people can use the DSD+ filter which has lower latency. He said he created the DSD+ filter so that it is non-decimating for DSD playback and obviously it must remove HF distortion and noise. But from the sounds of it, you can use either filters for DSD or PCM playback. If you use the WTA 164k tap filter (PCM+) to playback DSD, the DSD would probably be decimated to 176kHz/24-bit first and then go through the PCM+ filter as PCM but I'm speculating. This is supposedly how Hugo currently plays DSD it seems. I don't know what would happen when you feed PCM into the DSD+ filter but since Rob Watts seem to imply that you can watch video with that filter, it has to accept PCM (because how else are you going to get DSD audio stream for video). And this is where it gets a bit confusing. It sounds like if you are in DSD+ filter mode, and if you're sending PCM signals, the DAC will use a WTA 82k tap filter to lower latency. And the reason for the two modes is because he couldn't fit the WTA 164k tap filter with the DSD+ filter. So one mode has the WTA 164k tap filter and decimates DSD to some sort of PCM first. The other mode has the non-decimating DSD filter and a WTA 82k tap filter? Either way, I'm really banking on the fact that there won't be significant audio delay that my Oppo blu-ray player cannot handle, since I already asked my dealer to pre-order my Chord DAVE.”

 

 

[ecwl]

Just to update on what I said a long time ago...

 

DAVE DSD+ mode takes 2.8MHz DSD signal and probably converts to 2.8MHz 24-bit signal and then is upsampled to the 104MHz 5-bit signal that is sent to the pulse array DAC in DAVE. Whereas DAVE PCM+ mode probably takes the 2.8MHz DSD signal and decimates it to ?176kHz 24-bit or maybe even 705.6kHz 24-bit, and then upsampled to 104MHz 5-bit signal for the pulse array DAC.

 

However, since the Hugo 2, Blu Mk 2 and Hugo TT 2 all take DSD signals and decimates them to 705.6kHz 24-bit and then upsampled to 104MHz 5-bit signal for the pulse array DAC.

 

What Rob Watts said was that when he developed DAVE, he found DSD sounds better in DSD+ mode than PCM+ mode. But with a different decimation filter, he now finds DSD sounds better using the new Hugo 2/Blu2/TT2 decimation filter to convert the DSD signals to 705.6kHz 24-bit first and then upsampled to 104MHz 5-bit signal for the pulse array DAC. This is why these new devices do not have an additional option for a specific DSD mode...

[ElviaCaprice]

26 minutes ago, ecwl said:

Just to update on what I said a long time ago...

 

DAVE DSD+ mode takes 2.8MHz DSD signal and probably converts to 2.8MHz 24-bit signal and then is upsampled to the 104MHz 5-bit signal that is sent to the pulse array DAC in DAVE. Whereas DAVE PCM+ mode probably takes the 2.8MHz DSD signal and decimates it to ?176kHz 24-bit or maybe even 705.6kHz 24-bit, and then upsampled to 104MHz 5-bit signal for the pulse array DAC.

 

However, since the Hugo 2, Blu Mk 2 and Hugo TT 2 all take DSD signals and decimates them to 705.6kHz 24-bit and then upsampled to 104MHz 5-bit signal for the pulse array DAC.

 

What Rob Watts said was that when he developed DAVE, he found DSD sounds better in DSD+ mode than PCM+ mode. But with a different decimation filter, he now finds DSD sounds better using the new Hugo 2/Blu2/TT2 decimation filter to convert the DSD signals to 705.6kHz 24-bit first and then upsampled to 104MHz 5-bit signal for the pulse array DAC. This is why these new devices do not have an additional option for a specific DSD mode...

Thanks @ecwl for that update.  What have you found with DSD in your listening?  Is it better decimated thru the Blu2 followed by the DAVE, which I take is PCM+?  Or does it sound better direct into DAVE with DSD+ non decimated?  

 

Nothing against Robs stated preference, but he is the designer, thus possible biased.

[ecwl]

6 hours ago, One and a half said:

My impression with Chord in that DACs Mark II versions arrive too frequently... Frequent model changes means the design has flaws, whether in the initial stage or the manufacturing process or there's an irrational push from the crowd in the football field.

Oh my goodness, all our electronic gadgets must be completely flawed. Time to abandon computer audio? Time to go back to vinyl?

 

Kidding aside, I suspect/think Hugo 2 & Hugo TT 2 are primarily a result of improved computing power from FPGAs which allows for more taps. It is also an improved design from the Chord DAVE experience that 256fs WTA upsampling sounds better than 16fs WTA upsampling (if you have the computing resources to do it). I consider the latter to be a new design discovery, not a flaw. Because that would be saying that plasma or LCD TVs are flawed because we now have OLED screens. (I was actually going to use the analogies of LED backlighting vs fluorescent backlighting or local dimming LED for LCD TVs but that may be too esoteric, even if it is more appropriate as an analogy).

 

But I also think part of the issue is cost and volume of sales. DAVE is a flagship product so it can have a 20-element pulse array DAC and still make money. I suspect the cost to have more elements adds up and you need slightly more FPGA processing power for more elements. As a result, Hugo 1 and Mojo only have 4-elements for the pulse array DAC because the product needs to be profitable and Chord didn't know how many units they'll sell. Now that they know Hugo can sell tremendously well at its price point, they can afford to put in a 10-element pulse array DAC because the sales volume would justify the increase cost of production. 

 

That said, I think One and a Half is right in saying that there are certain "design flaws" or I'd rather call them "suboptimal design decisions" that Chord might have made with Hugo TT that they did not fully recognize at the time. But that is true for any product. Until a product is out in the field, in use by lots of users, and you hear user feedback, it is difficult to know what optimization users actually want and how the products are being used and how the products are not optimized for real-life usage scenarios. Even with that said, it is still unclear to me that at the time when Hugo TT was developed, whether these alternative design solutions should have been thought of, or whether these design decisions are made possible now because of lower cost of components or better technological innovations in component parts.

[ecwl]

4 minutes ago, ElviaCaprice said:

What have you found with DSD in your listening?  Is it better decimated thru the Blu2 followed by the DAVE, which I take as PCM+?  Or does it sound better direct into DAVE with DSD+ non decimated?

 

I have to admit I have very few DSD tracks and I rarely listen to them. And the vast majority of those tracks are actually hi-res PCM converted to DSD. And I get so lazy that I sometimes don't even bother with DSD+ mode.

I like the sound of DSD tracks through Blu2 better than the DAVE's DSD+ mode. However, since I have so few DSD tracks, most are not direct DSD recordings, I didn't and can't do ABX comparisons, and I'm totally biased by whatever Rob Watts said, I think you should really take my opinion with a grain of salt.

[barrows]

8 hours ago, One and a half said:

My impression with Chord in that DACs Mark II versions arrive too frequently. Lampizator is worse though, in the end, the longevity and appreciation of the product wanders downhill. Also makes a buying decision difficult, when a newer model is around the corner or with so many choices and options, choice is paralysed. Frequent model changes means the design has flaws, whether in the initial stage or the manufacturing process or there's an irrational push from the crowd in the football field.

 

For the Hugo TT, the internal battery was a novelty, with one day, the battery WILL fall over, and try find a battery that fits and meets the spec in ten years time. The supercaps in the MarkII is a better idea, but only worthwhile if there's isolation from the charging system as in the LPS1x. Forget the fancy case, rainbow LEDs and shove in a transformer with large caps that last forever. My 1.5 cents.

 

 

I agree, but of course what you describe is DAVE!  Chord does it this way because they can reach their price point this way.  If they made the chassis big enough to accommodate a nice linear supply, and enough space to isolate that supply from the critical circuits, they would have a full size component, and the cost of the chassis would go way up.  Add that to the price of the power supply parts (vs. the SMPS they provide) and we ar talking the TT@ at $10K again, and no one whats that.

Remember the chassis cost is the single biggest cost of most audio products on the BOM.

Also, Chord does not do linear supplies, they are completely committed to SMPS in their products.

[barrows]

There are no "design flaws" in the TT2, or DAVE!  OMG!  There are decisions based on what makes sense within the product line,what gets the customer the most "bang for the buck", and what features Chord wants to provide in their products.  Chord now has a coherent, sensible, line in DACs for home use at 3 price points:  good (Qutest), better (TT2), best (DAVE).  It is as simple as that.  As a customer, one can now choose which product best suits their needs and budget, or one can choose a DAC from another company.

Personally, I expect that the TT2 will be the best DAC under $10K USD available, especially for users who do not listen to a ton of DSD, as it is clear DSD is better served by DAVE which does not decimate the sample rate of DSD files.

 

No single product is going to please all users, or be perfect for all users, luckily we have many DACs to choose from.

 

It kind of makes me laugh, as everyone seems to want DAVE level performance at a Qutest price...  Well with the TT2 you get quite close to that (except with DSD), but not all the way there.  If you want the "best" get DAVE...  

[austinpop]

So I guess "decimation" is nothing more than the "process of reducing the sample rate of a signal" - https://en.wikipedia.org/wiki/Decimation_(signal_processing)

 

In English, the word decimation has connotations of doom and destruction - armies being decimated in wars, or the ancient Roman practice of decimation:

 

...was a form of military discipline used by senior commanders in the Roman Army to punish units or large groups guilty of capital offences, such as mutiny or desertion. The word decimation is derived from Latin meaning "removal of a tenth."

 

 Thankfully, in terms of sample rate conversions, DSD to PCM decimation is relatively benign, as DSD is 2.8224 MHz, which is an integer multiple (16x) of 176.4 KHz, or 4x of 705.6 KHz.

 

Beyond this, every DAC does voodoo to the signal presented to its inputs, so best not to worry too much about it. Either it sounds good or not!

 

 

[barrows]

7 minutes ago, austinpop said:

So I guess "decimation" is nothing more than the "process of reducing the sample rate of a signal" - https://en.wikipedia.org/wiki/Decimation_(signal_processing)

 

In English, the word decimation has connotations of doom and destruction - armies being decimated in wars, or the ancient Roman practice of decimation:

 

...was a form of military discipline used by senior commanders in the Roman Army to punish units or large groups guilty of capital offences, such as mutiny or desertion. The word decimation is derived from Latin meaning "removal of a tenth."

 

 Thankfully, in terms of sample rate conversions, DSD to PCM decimation is relatively benign, as DSD is 2.8224 MHz, which is an integer multiple (16x) of 176.4 KHz, or 4x of 705.6 KHz.

 

Beyond this, every DAC does voodoo to the signal presented to its inputs, so best not to worry too much about it. Either it sounds good or not!

 

 

Well, I would not be so quick to suggest that decimation of DSD signals is not a problem.  Chord does not think so either, hence the approach used in the DAVE to preserve the original sample rate of DSD files.  tis is also an advantage of DACs which use the ESS Sabre chips, as they do not decorate DSD signals either.

Of course with DAVE, one can listen to both methods and decide for themselves, any DAVE owners do this comparison?

[monteverdi]

I listen today to the Hugo TT2 and was very impressed. I had a short conversation with Rob Watts and he has no firm US price. He said that it does not approach the sound quality of Dave (and that is not only about computational abilities but also about part quality and output stage) but clearly above Hugo. It has enough output to drive sensitive speakers and that what he did during the development phase.

[mozes]

4 hours ago, barrows said:

 

Of course with DAVE, one can listen to both methods and decide for themselves, any DAVE owners do this comparison?

DSD does sound better with DSD plus mode with Dave. 

[jamesg11]

Where does the Qutest fit into this spectrum of the handling of dsd files?

[barrows]

3 minutes ago, jamesg11 said:

Where does the Qutest fit into this spectrum of the handling of dsd files?

Only DAVE offers the separate DSD mode in the Chord DACs.

[seeteeyou]

Holy Smoke, this beast actually got (part of?) the upcoming DX amp integrated

 

https://www.head-fi.org/threads/hugo-tt-2-by-chord-electronics-the-official-thread.879425/page-9#post-14234876

Quote

Yes the last 9 months has been very busy, with the TT2 and power pulse array - the DX amp. Initially TT2 had a beefed up Hugo 2 OP stage. And the DX amp had a much more complex OP stage - but this turned out to be not suitable. So this got redesigned, and I got the performance I needed - but then I realised I could update TT2 with this new OP stage, so it got added into the next TT 2 prototype. This increased the short circuit current delivery by 10 times, without sacrificing speed or propagation delay. The speed is crucial, as the 2nd order analogue noise shaper OP stage absolutely requires very fast switching speed, (of order of 30 nS) for phase margin and stability reasons.

 

 

[Superdad]

On 5/11/2018 at 1:11 AM, One and a half said:

The supercaps in the MarkII is a better idea, but only worthwhile if there's isolation from the charging system as in the LPS1x.

 

Can pretty much guarantee that is not the case.  You have seen the circuitry and board real estate it takes to do an isolated, bank-alternating supercap supply, yes?  

[ecwl]

https://www.head-fi.org/threads/hugo-tt-2-by-chord-electronics-the-official-thread.879425/page-10#post-14236842

 

Rob Watts just posted his presentation on Hugo TT 2. It sort of details the design changes/improvements. 

[rickca]

Why does Chord have to make their stuff look like Fisher Price toys?  I can't stand the DAVE industrial design.

[johndoe21ro]

4 hours ago, rickca said:

Why does Chord have to make their stuff look like Fisher Price toys?  I can't stand the DAVE industrial design.

That's what I keep asking myself every time I see Chord gear...

[rodrigaj]

13 hours ago, rickca said:

Why does Chord have to make their stuff look like Fisher Price toys?  I can't stand the DAVE industrial design.

I'm pretty sure that it is "Steampunk" in conceptual origin.

 

https://www.google.com/search?q=steampunk+furniture&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwi5gq6-x57bAhVi5YMKHT9UAGEQsAQIhwI&biw=1280&bih=625

 

Not very original, but I'm sure they think it is.

[Hauser]

On 5/24/2018 at 12:37 PM, rickca said:

Why does Chord have to make their stuff look like Fisher Price toys?  I can't stand the DAVE industrial design.

 

If you don't like the look of Dave, see what you think of Blu2.

 

Martin.