The last couple of weeks has been a bit crazy - I had a couple of projects to be completed before CanJam Chicago, then back to the UK to pack for a family holiday in Florida. Now I am on holiday I now have time to answer!
What is supposedly wrong with pcm DACs per se? And what exactly do people mean by pcm dac? Pcm encoding adapted for a truncated word length d to a core ?
I'd have thought a full r2r converter would have the most straightforward characteristics if it works properly.
So the distinction is between PCM or DSD DACs. But pure DSD DACs are pretty useless, as you can't do EQ or volume control digitally without converting to PCM. And anyway most DSD chips convert to PCM anyway. To get the best SQ from DSD you need to remove the huge amount of distortion and noise from the DSD, and that means conversion to PCM...
R2R can never work properly, the R2R technology is fundamentally limited.
So if one is using an Mscaler.. Dave's WTA1 is bypassed and this would then only eat up FPGA space ..and use less energy..
It might be said before.. but could WTA2 noise shaper be expanded further than 17th order and achieve even better small signal accuracy? Or was FPGA space not the limiting factor?
Offcourse HMS must always be used then or Dave would become a NOS DAC
Whats Dave's WTA1's usage of space in % actually?
It could be programmed as a selectable 'menu item' same as the choice of DSD+ and PCM+.
If im not mistaken this also loads some pre-stored code ..from a ROM ..im not sure here..
The final noise shaper used in Dave was not limited to 17th order - going higher actually gave worse SQ performance. The situation is much more complicated than I have talked about. The limitations aren't with the FPGA but the number of elements.
@Rob Watts , do you think that the TT2 can handle both single driver speakers (on XLR, sensitivity of 94.5db) and a passive subwoofer (on RCA) at the same time?
Sure. But don't use the XLR OP's for driving the speaker.
@Rob Watts apologies if this seems off-topic, but in your FPGA implementation in the Qutest, are implementations of each different type of filter e.g. WTA2 filter 'shared', or do you have 2 separate physical copies of e.g. the WTA2 filter implementation, one for white filter & one for green? I'm assuming the latter, for performance reasons, but I don't know much about this subject. Thanks in anticipation!
For TT2, Hugo2, and Qutest the HF filter is done in the filters after WTA2, that is the filter that goes from 256FS to 2048FS. This is a multi stage IIR type filter, and the time constants are adjusted to give the HF filter. So WTA2 always stays the same.
Interesting points you raise here and I would even add that to be able to "sort the wheat from the chaff" in a really meaningful and reasonably objective way it is best to use test recordings you have either made yourself as is the case with Rob´s own recent recording or one where you have been present at the sessions.
I am aware most people do not have such direct paths to decide or judge what sounds really good or better than some other link in a reproduction chain.
But imho in the real sense and definition of the term, HIFI can only really be judged with unamplified acoustic instruments and the human voice. All else lacks a direct reference point and can only be of relative and subjective value in any comparison between different products.
I am aware this is very controversial to some here. And I am also aware that "at the end of the day" we all listen to whatever genre or music that appeals to us.
Absolutely nothing wrong with that!
Each to his or her own taste.
But HIFI to me has a VERY clear and defined reference point which also fits my own tastes in music.
Personally I am lucky to have been working as a photographer with some world class labels and orchestras and I always use some of my direct masterfiles to judge SQ. And with the introduction of Mscaler a few years ago Rob´s tech brings me closer to the real sound of acoustic instruments as I am used to hearing them both live and from sessions masters in the studio than most of the competition did.
There could be other equally accurate dacs and upsampling methods out there that I have not yet heard. But to me ,one thing is sure Rob knows what is he is doing and his recent test recording is also proof of that.
My very limited maths knowledge also leads me to suspect that he is right 2+2=4? Or am I wrong?
Cheers CC
2+2 = 4 except for people that believe their own realities and live in a fantasy world.
Absolutely I agree with unamplified music being the reality we should judge by. But - most modern concerts, and all opera I have been to, use sound reinforcement - discretely done - which IMO destroys the performance. I always go for seats as close as possible to the performers, so that the PA doesn't ruin the sound.
The Covent Garden Royal Opera House do not use the 200 odd Chord power amps to heat the concert hall...
I have a general question for Rob (and our concert-going members) which touches on M-Scaler and also Rob's experiments with recording technology:-
- why is it that (for me, anyway) classic EMI and Decca recordings from the 1960's and 70's "sound" so much better than modern recordings?
Specifically, I'm referring to the overall "ambience": the sense of being in a physical venue, rather than an anechoic chamber (i.e., the typical Deutsche Grammophon sound - which, sadly, characterises most later Karajan recordings).
For example, the other evening I was listening to some modern Chandos recordings of Richard Hickox and Sir Andrew Davis conducting Delius. Fine performances, from a respected label - but the recordings themselves gave little in the way of "depth" perception. I then switched to a classic Sir John Barbirolli recording, made in 1969 by EMI - and at once, the sound - through my newly-acquired Chord M-Scaler/DAVE combo (thank you Rob!) - was almost three-dimensional in nature (albeit with thin and screeching strings).
My assumption is that this may boil down to differing microphone techniques. We know that the EMI and Decca engineers pioneered stereo recording - using a minimum of microphones: e.g., "The Decca Tree" and occasional "outriggers". By contrast, their counterparts at Deutsche Grammophon (DGG) seemed obsessed with multi-miking, which - to my ears, at any rate -destroys the ambience, providing only discombobulated point sounds. Looking at session photos these days, it seems that multi-miking is the norm - presumably because companies are increasingly risk-averse and would rather "fix post production" using their mixing desks, rather than experiment with better microphone placement during sessions.
I didn't think that Chandos used multi-miking, and maybe the differences were more a reflection of the venue's inherently dry acoustics - but such a stark difference. Any thoughts?
Agreed. I don't know how much is down to simple techniques, or early ADCs that used good analogue anti-aliasing filters rather than the rubbish half band decimation filters that are used today. My pulse array ADC and simple recording techniques will answer that question. What you get from these pre Dolby recordings is a sense of life, speed and power - and that is certainly indicative of aliasing issues. But it's also the sound you get from lack of processing and an electronics chain that is simple and direct.
I agree with your observations ,but as I have mentioned before here, I was a bit suprised when Rob recently mentioned not having any? pure Blumlein "takes" in his collection.
Maybe because, at least on his laptop, which I had the opportunity to partly "scan through" and listen to via DAVE/BLU 2 in Singapore some years ago, he seemed to have only DECCA recordings?
But even those were enough to hear the obvious benefits of mscaling particuarly with 16/44.1 material.
Although unless I am wrong some of the late 50s and early 60s EMI recordings were Blumlein recordings?
And even Karajan whom you mention as an example of later 50-55 dB compressed, DGG recordings are indeed very multimic´d and not as open and cohererent as some of the Karajan/ Philharmonia Orchestra recordings made by EMI in the Royal Albert Hall in the late 50s and the 60s before EMI, like DGG began multimiking.
But as far as Karajan is concerned luckily for mainly "Classical Music HIFI nerds" like me, Karajan and the BPO also recorded for both DECCA and EMI again, during the 70s.
And although those are not pure Blumlein recordings his recording engineer, Wolfgang Gülich at those 70s Karajan /BPO recordings for EMI took some pride in the fact that he used only HALF the number of mics that DGG engineers ,did in those days: And the more coherent,more open sounding recordings by him are clearly audible in my system via my electrostatic speakers.
Some of those recordings like Otello and Tristan und Isolde and the late Sibelius symphonies ,are still good examples of what could actually be achieved with "no Dolby" analogue tape at the end of the 70s.
DECCA was imo, a bit hampered by DOLBY and "outriggers" in the 70s .
To me the simply mic´d "Crystal Clear" or "Sheffield Labs" Direct to Disc LPs still sound about as good as analogue recordings ever got and without the obvious dynamic range compression of early anaogue tape .
But even those early EMI and DGG LPs with 50-55db limited dynamic range is still wider DR than most popular music genres ever reach, even today!.
Cheers CC
Vinyl only ever sounded half decent with Sheffield Lab direct to disc.
No, I am not advertising for Topping here, it is more of a question.
This device is advertised as a USB galvanic isolator. It removes ground loop issues, by creating an isolated new ground, USB data lines and power are also isolated.
It is an active device (true isolator not just a filter) running on power from host USB, drawing about 20mA. The side USB-C socket, is to plug in a separate power source for those DACs that require power (Not Chord, I believe). It is full USB 2 compliant.
Here is the question:
Will it eliminate ground RF pollution that might occur using Hugo2 or Mojo2 in desktop mode? Assuming it does what it says on the tin!
It seems is does just that.
Here is a lab measurement done on the device.
If it is up to the task, at its low price, it would be useful.
This looks great for Hugo 2 and Mojo 2 for desktop applications. I would order one if it was available in the UK. Also, there are measurement problems with the APx555 on USB with ground loops from the analogue oscillator creating issues that would not be present in real situations.
@Rob Watts
Did or have you ever considered using opto couplers in the coaxial signal paths of HMS and your DAC's? And if yes and not already implemented.. is there a reason to avoid it?
Could creating a separate (2 battery operated) optocoupler yield any benefits?
Thanks in advance for your response,
DIY guy Rick
Sure I used to use 100Mbs opto-isolators. But today I use isolators that use a RF carrier - these work faster with lower skew, and similar or lower coupling capacitance on all my isolations.
Which powerbank are you using? Does it totally lack a switching circuit?
Im designing a battery supply with integrated charger for Dave and HMS which also makes 5vdc using 8v lead acid accu's and a linear regulator. This has none switching activity.
The capacity can be much more than regular powerbanks.
If u want i can share a schematic.
Car batteries are the ultimate PSU - huge dynamic currents, unlimited current, zero impedance, zero noise. And for 12v or 24v operation, no regulators are needed, which would degrade everything.
Perhaps you are right on digital cables... improving the power supply of source components could help to nullify the audible differences/performance delta between different 'Audiophile' USB cables... on the other hand, the delta might actually improve. I'll try it.
Power supply changes to qutest have minimal effect in comparison. You need to spend some serious money on a preamp to better Chord's inbuilt pre in Hugo2 etc..., if you run something like an Audio-gd preamp (e.g. Master 9 like used to own) it's just not good enough. Really, you need a top tier preamp (something like a REF 6 or perhaps Holo Serene) to be able to really hear these differences.
I don't really care if a USB cable follows standard spec or not, it could be voiced using different shielding or conductor arrangements - because really, power supplies, whilst potentially lowering the noise floor, are just another way of voicing a component (e.g. Innuos Phoenix uses Mundorf caps, changing them would change the sound). Who cares so long as it sounds good and is still technically high performing as well? Just different approaches. Of course I'm talking ultra high-end here, were any change in the system can be heard... its both a blessing and a curse.
Anyhow, I mainly dropped back in to ask if
@Rob Watts had seen the new TI ISOUSB211 chip. Very good feedback emerging on this and it could be a very nice addition to Chord DACs. I'm going to purchase the EVAL board.
Thanks for that. Just looked at the data sheet - but the power dissipation is a problem at nearly 1.2W, but I guess that is at 480Mbs, something audio would never use.
But my USB isolation solution has another benefit in that the processor noise (which makes a significant SQ degradation) is eliminated by my isolation, as this noise is on the source side not the DAC side.