[ultrabike]

It seems that the chord pulse array is a 2048FS 5th order multi bit pulse density modulation with dither and noise shaping,

Something like that it seems. Would be interesting to know roughly what the quantization Noise Transfer Function (NTF) is (or the equivalent to whatevs his deal is) as that determines to some extent the level of out of band quantizaiton noise filtering needed. That's quite a large oversampling rate also. So is he interpolating the imput signal to the (maybe) delta sigma? If so, using what type of filter? WTA?

Maybe that's why he needs a lot of taps. Because running the delta sigma that fast might require a fairly large interpolation filter before actually hitting the delta sigma (or PDM, or PA, or WMD).

 

[evillamer]

System diagram of chord Hugo

 

[negura]

You're asking a bit much, not a pro-studio engineer. Saw some of those though (nothing of say abbey road size & prestige) and my general impression was: very detailed and clear sound but not a relaxing listening space. Some studio speaker samples: k+h o300-400, focal (biggest 3way ones), barefoot, pmc (dont remember models, pretty big & shinny new ones). I did not measure anything but the sound was quite linear, no obviously accentuated freqs and all had some sort of room treats.
Their components did sound quite linear/neutral but also kinda shrill for my ears. Same for many ultra highend (and suposedly very neutral) hifi setups in various shops.

As about the slewrate, I think most modern amps (at least the ss kind) are linear to 50khz or more .. even "lousy" receivers are. Doubt that is a big issue nowadays. A poor slew rate would prolly also defeat the linear to 20khz rule .. same about the peaks you mentioned.

So, I think I still have a Q. Other than odd order harmonics why would a linear-to-20k component sound shrill?

 
Where imo some of the studio monitors (in my experience mostly PMCs - have had two different models of these) are really good compared to some "nice" consumer speakers, is the former don't have any design goals in exaggerating or colouring to please. That would not impress their core market. It could be an error of commission, but likely not intended. They try to stay accurate and transparent to the source, which, while may sometimes not instantly impress, in the long term can actually be more pleasant and enjoyable imho, by sounding well a bit more realistic. That's not to say other non-monitor speakers don't achieve the same goals of course.
 
Harshness/Shrillness: Usually source material and upstream rig would be the cause of this. The PMCs I have are surprisingly well behaved, for how resolved and transparent they are. One of the reasons is that say comparatively to my HD800s, the PMCs don't have audibly nasty treble peaks and (in my room, which isn't that special) reflections like the unmodded former. That said if I were to hook the PMCs to a bright/harsh sounding DAC running bright records, I would always hear that. There are other factors too. Hardly surprising - most good transparent speakers would react the same. 
 
I would encourage anyone to try and hear a pair of studio PMCs or ATCs on a good R2R DAC or vinyl and a good amp. With my Theta Gen V the PMCs sound noticeably that notch or two warmer/softer compared to the Yggdrasil, or the PWD2. But then I realise preferences are everything and some people would be enamoured with a different type of sound.

 

[bmichels]

Has someone heard the new BAKOON DAC-21 ?  It is battery operated, with SATRI connection to connect to the BAKOON HPA-21 amp.
 

 

[skeptic]

Could very well be wrong, I was just thinkin out loud. But there are quite a few empiricals that seem to fit my "linear is quite sharp" hypothesis. Most studio speakers will fit it too and those are known for being very neutral/linear. (and of course nwavdude is not the only builder of neutral components, it was just a sample.)

But I still wonder, what exactly makes a well built and neutral/linear component sound harsh !? The only theoretical explanation I know is about the odd harmonics sounding harsh (as opposed to the pleasing even ones) ... and apparently a lot of negative-feedback may produce lots of those. That could very well be the O2's case. But modern DACs do not produce much of any harmonics ... at least not in the audible range (I'd put that at -100, -110dB and most DACs are below that already).
Any other ideas?

 
I'm inclined to think that linear is a misnomer here, and the standard array of published audio measurements just isn't telling us the full story.  On the feedback front, I don't think it is so much about odd or even harmonics but rather disproportionate high order distortions.  According to Pass ( https://passlabs.com/articles/audio-distortion-and-feedback), feedback designs with superlative conventional measurements manifest more complex non-linear distortion products than higher THD nonfeedback designs.  Depending on their analog output stages, I would think dacs could share the same traits since opamps require negative feedback to operate in a linear mode (i.e. as an audio amplifier), and many dacs employ opamps in the output.  
 
That said, although I tend to prefer nonfeedback amps, some opamp based gear really does sound good to me, and not at all harsh, with the right headphones.  So I think even the above must be oversimplifying what is really going on.  

 

[ultrabike]

System diagram of chord Hugo

To some extent it makes sense. To some other, not so much. That's a lot of interpolation filters and one obscure "noise shapper". Who knows. If it's a multi-bit 5th order delta sigma, I guess it's not much different from some consumer audio DACs ICs out there. The sampling rate is quite high though. Maybe that's what they are trying to market as their differentiator.

Having a high sampling rate may result in lower quantization noise, in paper. Depends on the actual implementation.

Wish things were more up front with those guys though. Given impressions by some it may be that they are over designing for one set of requirements and neglecting others. I dunno.

 

[maverickronin]

To some extent it makes sense. To some other, not so much. That's a lot of interpolation filters and one obscure "noise shapper". Who knows. If it's a multi-bit 5th order delta sigma, I guess it's not much different from some consumer audio DACs ICs out there. The sampling rate is quite high though. Maybe that's what they are trying to market as their differentiator.

 
Doing the crossfeed in the digital domain on a portable device is pretty damn cool though.
 
It's a shame that it looks like toy for toddlers though.  It might as well be Fisher-Price My First DAC or something...

 

[evillamer]

 

System diagram of chord Hugo

To some extent it makes sense. To some other, not so much. That's a lot of interpolation filters and one obscure "noise shapper". Who knows. If it's a multi-bit 5th order delta sigma, I guess it's not much different from some consumer audio DACs ICs out there. The sampling rate is quite high though. Maybe that's what they are trying to market as their differentiator.

Having a high sampling rate may result in lower quantization noise, in paper. Depends on the actual implementation.

Wish things were more up front with those guys though. Given impressions by some it may be that they are over designing for one set of requirements and neglecting others. I dunno.

 
The advantage of Hugo over other sigma delta dac is that it has a dedicated FPGA which has more processing horsepower which means more complex digital filters can be implemented. If you look at higher end dacs that feature ES9018s/ES9018K2M(e.g. Audiolab MDAC, Grace design M920) they usually have their own custom filters and also may bypass the ASRC/PLL on the Sabre.

 

[ultrabike]

Doing the crossfeed in the digital domain on a portable device is pretty damn cool though.

It's a shame that it looks like toy for toddlers though.  It might as well be Fisher-Price My First DAC or something...

LOL! Yup. Looking at it brings to my mind "brick", which is not a good thing.

Crossfeed in the digital domain is not too bad IMO, but a good feature to have. You know the Sansa Clip+ Rockboxed does some of that. It works to some extent. But it lacks head tracking which I found to be very important after giving the Smith Realiser a shot.

The advantage of Hugo over other sigma delta dac is that it has a dedicated FPGA which has more processing horsepower which means more complex digital filters can be implemented. If you look at higher end dacs that feature ES9018s/ES9018K2M(e.g. Audiolab MDAC, Grace design M920) they usually have their own custom filters and also may bypass the ASRC/PLL on the Sabre.

The FPGA is more flexible, but not necessarily higher performance. Depends on implementation details I guess.

 

[DreamKing]

Just that live, non-amplified music can sound "rolled off" compared to the harsh, glaring treble you were speaking about in your posts. Real music usually doesn't have a hot treble. Some recordings of course do have that artifact, but a great DAC, amp, speaker, headphone should convey the tone, dimensionality, naturalness of un-amplified music, no?

 
If the recording is at fault or the headphone is harsh (which is demonstrably more plausible) , why should the dac and amp roll off harshness in them? I mean they could but does that mean that they're superior because they roll off everything and should every electronic strive for this? Where this harshness is coming from should be evaluated before coming to the conclusion that the dac or amp is at fault. And then you're in the realm of perception with all of these terms, so it's all relative.

 

[maverickronin]

Crossfeed in the digital domain is not too bad IMO, but a good feature to have. You know the Sansa Clip+ Rockboxed does some of that. It works to some extent. But it lacks head tracking which I found to be very important after giving the Smith Realiser a shot.

 
I'm just glad that another 'high end" company is taking the idea of crossfeed seriously.  A full on convolution DSP like the Realiser uses plus the head tracking can help quite a bit, but if you're listening to stereo recordings any kind of crossfeed is better than nothing.
 
It's the biggest problem in the fidelity of headphones that almost no one is addressing.  IMO not using analog crossfeed or a more advanced DSP is the headphone world's equivalent of just throwing a pair of speakers wherever they'll fit and not bothering with toe-in, stands, room treatment, etc.

 

[hans030390]

 
I guess the reason is the same for desktop DACs as well: more components, larger the board => harder to make it transparent.     All DACs from $100 to $1xxx I have tried, are not as clear and detailed as the $2 simple sound system.   Small, less components => easy to make it transparent.   Only after > $2000 level, desktop DAC start to be able to compete with this $2 system.

 
I'm curious where you got this rule that more components and larger boards = harder to make transparent. I'm also wondering if your idea of "transparent" is different than how others describe transparent (and it probably is).

 

[wahsmoh]

   
I'm just glad that another 'high end" company is taking the idea of crossfeed seriously.  A full on convolution DSP like the Realiser uses plus the head tracking can help quite a bit, but if you're listening to stereo recordings any kind of crossfeed is better than nothing.
 
It's the biggest problem in the fidelity of headphones that almost no one is addressing.  IMO not using analog crossfeed or a more advanced DSP is the headphone world's equivalent of just throwing a pair of speakers wherever they'll fit and not bothering with toe-in, stands, room treatment, etc.

Seriously. Check out Legacy audio.. they have some serious $$$$ speakers I have heard once in a music room at a high end stereo shop. 
 
http://legacyaudio.com/products/dsp-solutions/

 

[Sapientiam]

But I still wonder, what exactly makes a well built and neutral/linear component sound harsh !? The only theoretical explanation I know is about the odd harmonics sounding harsh (as opposed to the pleasing even ones) ... and apparently a lot of negative-feedback may produce lots of those. That could very well be the O2's case. But modern DACs do not produce much of any harmonics ... at least not in the audible range (I'd put that at -100, -110dB and most DACs are below that already).
Any other ideas?

 
If you're looking for technical reasons - its poor PSRR, coupled with power supplies which haven't low enough impedance.
 
Simple experiment to try on the O2 - add lots of low ESR capacitors across the supply rails for the NJM4556s which are doing all the heavy lifting. I'd suggest a dozen 3,330uF/10V, six each side. The reason for the power supply noise is that the opamps are operating in classAB, meaning nasty non-linear currents are drawn.
 
These are the caps - they're bulky so the whole thing will no longer fit in the case. But its a $10 upgrade.
 
http://www.digikey.com/product-detail/en/EKZE100ELL332MK25S/565-1651-ND/756167

 

[aqsw]

Although I found out the $2 MB chip inside Dell XPS 9100 is extremely clear and detailed,   but it is just an exception. 
All OTHER PC's sound card I have tried are not good.    Either noisy, or blurred.     Like the Creative chips, it is very blurred.  They charged a premium for loaded features, but loaded leads to less transparency.

Only talking about clearness and transparency. The $2 MB chip inside Dell XPS 9100 is nothing musical or enjoyable.

I guess the reason is the same for desktop DACs as well: more components, larger the board => harder to make it transparent.     All DACs from $100 to $1xxx I have tried, are not as clear and detailed as the $2 simple sound system.   Small, less components => easy to make it transparent.   Only after > $2000 level, desktop DAC start to be able to compete with this $2 system.

Please dont feed the troll.