Daconomics and my DAC upgrade conundrum

opus111

Banned
Feb 10, 2012
1,286
3
0
Hangzhou, China
I haven't looked in detail at it but my comment was intended to spur you into thinking what kind of material you're wanting to play. If its DSD then I'm sure the EMM will do great, probably its amongst the very best. However DSD is at the opposite end of the spectrum from redbook - a DAC optimized for it will be poor at 44k1/16.

I don't understand your comment 'forget about Ed'. The designer of a DAC is for me the defining issue as to whether I'd recommend it or not when I've not actually heard it. You'll note I suggest AMR and Audial - primarily those suggestions are based on my respect for their designers, coz I've heard neither.
 

edorr

WBF Founding Member
May 10, 2010
3,139
14
36
Smyrna, GA
I haven't looked in detail at it but my comment was intended to spur you into thinking what kind of material you're wanting to play. If its DSD then I'm sure the EMM will do great, probably its amongst the very best. However DSD is at the opposite end of the spectrum from redbook - a DAC optimized for it will be poor at 44k1/16.

I don't understand your comment 'forget about Ed'. The designer of a DAC is for me the defining issue as to whether I'd recommend it or not when I've not actually heard it. You'll note I suggest AMR and Audial - primarily those suggestions are based on my respect for their designers, coz I've heard neither.

I see. I was not sure where you were comming from with a reference to Ed the man rather than EMM the DAC, but now I understand. So you're saying if you're all about 44/16 (like me) EMM may not be your best bet. Thanks for that input.

By the way, did you draw insipration for you forum name from the Beethoven sonata in C-minor? The Arietta from the Opus 111 is (along with the Adagio Hammerklavier and the Adagio from the string quartet Opus 131) some of the most achingly beatifull music ever written, either on DSD, PCM or live.
 

opus111

Banned
Feb 10, 2012
1,286
3
0
Hangzhou, China
I see. I was not sure where you were comming from with a reference to Ed the man rather than EMM the DAC, but now I understand.

Likewise I now understand your misunderstanding of what I wrote :) I don't have any personality issues about Ed the man, I just don't consider him experienced in the design of 'analog' DACs. Whereas if I wanted a 'digital' DAC then Ed would likely be my likely first port of call. I read at least one of his papers (AES I think) on 'Logic induced modulation' and found it enlightening at the time.

So you're saying if you're all about 44/16 (like me) EMM may not be your best bet.

Indeed a relatively poor bet given your requirements.

By the way, did you draw insipration for you forum name from the Beethoven sonata in C-minor? The Arietta from the Opus 111 is (along with the Adagio Hammerklavier and the Adagio from the string quartet Opus 131) some of the most achingly beatifull music ever written, either on DSD, PCM or live.

Yes, well spotted. I love that sonata and also its the name of a record label I've come across. If I were registering today though I'd probably choose Op101 as that sonata is where its at for me right now. Richter is pretty much unbeatable in his ravishing beauty : http://www.youtube.com/watch?v=HY0Ti_BURZA - the contrast between 1st and 2nd movements is stunning. The 2nd has pre-echos of my favourite variation on the Arietta of Op111 - sublime.
 

LL21

Well-Known Member
Dec 26, 2010
14,411
2,509
1,448
Likewise I now understand your misunderstanding of what I wrote :) I don't have any personality issues about Ed the man, I just don't consider him experienced in the design of 'analog' DACs. Whereas if I wanted a 'digital' DAC then Ed would likely be my likely first port of call. I read at least one of his papers (AES I think) on 'Logic induced modulation' and found it enlightening at the time.



Indeed a relatively poor bet given your requirements.



Yes, well spotted. I love that sonata and also its the name of a record label I've come across. If I were registering today though I'd probably choose Op101 as that sonata is where its at for me right now. Richter is pretty much unbeatable in his ravishing beauty : http://www.youtube.com/watch?v=HY0Ti_BURZA - the contrast between 1st and 2nd movements is stunning. The 2nd has pre-echos of my favourite variation on the Arietta of Op111 - sublime.

Hi Opus111 - i saw that too...i have a few Opus111 recordings from Vivaldi on my shelf. As for Richter and Beethoven Sonatas...i am aware that there have been some difficult to find recordings of all his Beethoven Sonata Works in Prague...for some reason, people say his stuff in other cities was not so well recorded and thus Prague is the best. Your advice is welcome!!! If i am looking for a good set of Beethoven Sonatas by Richter (on CD of course as you know ;) ) what do you recommend?
 

opus111

Banned
Feb 10, 2012
1,286
3
0
Hangzhou, China
Hey Lloyd you're already more up to speed than me on Richter and Beethoven sonatas! I don't have any complete set of Richter just odd sonatas on different CDs. I do have (acquired most recently) a box of Gulda (Decca, 1968) and also an excellent modern set by Paul Lewis (Harmonia Mundi). Bernard Roberts on Nimbus is also one I like.
 

LL21

Well-Known Member
Dec 26, 2010
14,411
2,509
1,448
Hey Lloyd you're already more up to speed than me on Richter and Beethoven sonatas! I don't have any complete set of Richter just odd sonatas on different CDs. I do have (acquired most recently) a box of Gulda (Decca, 1968) and also an excellent modern set by Paul Lewis (Harmonia Mundi). Bernard Roberts on Nimbus is also one I like.

I have Paul Lewis and Emile Gilels for Beethoven Sonatas...i like Bernard Roberts on 2-set on Bach...a lot...Nimbus Recordings? I think late 50s, 60s is considered very good Richter by the reviews i have read. Some of those cd's are really expensive!
 

opus111

Banned
Feb 10, 2012
1,286
3
0
Hangzhou, China
I think I might have Gilels too. CDs being expensive is actually a two-edged sword - it makes for a good argument for spending more on your system to your other half. It goes something like this - 'We've so much money tied up in all these CDs, it doesn't make any kind of economic sense to have half-baked kit playing them all. We really need to maximize our return on investment by ensuring we extract every last drop of musical goodness from them with X DAC, Y amp and Z speakers don't you think?'.:p
 

LL21

Well-Known Member
Dec 26, 2010
14,411
2,509
1,448
I think I might have Gilels too. CDs being expensive is actually a two-edged sword - it makes for a good argument for spending more on your system to your other half. It goes something like this - 'We've so much money tied up in all these CDs, it doesn't make any kind of economic sense to have half-baked kit playing them all. We really need to maximize our return on investment by ensuring we extract every last drop of musical goodness from them with X DAC, Y amp and Z speakers don't you think?'.:p

Hah! Good one! I did tell her that a 15-CD collection of all of Richter's works in Prague (apparently considered some of his best in his career?) is asking over $5000!!! (Fortunately, i just ordered the 4-CD set of his Beethoven works in Prague for a 'normal used CD price'...i have been looking for several months, and it finally popped up used on Amazon.)
 

edorr

WBF Founding Member
May 10, 2010
3,139
14
36
Smyrna, GA
Indeed a relatively poor bet given your requirements.

There are a many reports of this being a phenomenal DAC on both 44/14 and DSD. I'm still a bit on the fence, but I may very well be placing this poor bet. :)

My only issue is the volume control. I currently control volume though dithering in my Trinnov processor that sits before the DAC. I was hoping to get a DAC with analog volume control (for two channel I could set the Trinnov to 0db and control volume on the DAC). The EMM has no volume control.

In my system, I'll would be attenuating in a range of between 50-20db. I know the volume control horse has been beaten to death, but can anyone confirm that 50db attenuation corresponds to 8bits, so with a DSP with 32 processor you loose no information on 44/16 material. What happens with native 24 bit sources is still not clear to me. Would this also have the 32 bits to work with so -50db (8bits) would not be a problem, or does a 24 bit source get chopped off at 24bit, so that at -50db you're down to 16bits, and have lost some information?
 

audioarcher

Well-Known Member
May 6, 2012
1,396
51
970
Seattle area
There are a many reports of this being a phenomenal DAC on both 44/14 and DSD. I'm still a bit on the fence, but I may very well be placing this poor bet. :)

My only issue is the volume control. I currently control volume though dithering in my Trinnov processor that sits before the DAC. I was hoping to get a DAC with analog volume control (for two channel I could set the Trinnov to 0db and control volume on the DAC). The EMM has no volume control.

In my system, I'll would be attenuating in a range of between 50-20db. I know the volume control horse has been beaten to death, but can anyone confirm that 50db attenuation corresponds to 8bits, so with a DSP with 32 processor you loose no information on 44/16 material. What happens with native 24 bit sources is still not clear to me. Would this also have the 32 bits to work with so -50db (8bits) would not be a problem, or does a 24 bit source get chopped off at 24bit, so that at -50db you're down to 16bits, and have lost some information?

Hi Edorr, I can't answer your question but have an additional one for those in the know. Maybe Bruce will know? Since DSD has only one bit how can digital volume control work with it without loosing info?

I'm guessing that it has to be converted to PCM first which will add noise?
 

edorr

WBF Founding Member
May 10, 2010
3,139
14
36
Smyrna, GA
Hi Edorr, I can't answer your question but have an additional one for those in the know. Maybe Bruce will know? Since DSD has only one bit how can digital volume control work with it without loosing info?

I'm guessing that it has to be converted to PCM first which will add noise?

Good question. However, to my knowledge the only DAC that processes DSD that has a digital volume control is DCS. Still interesting to learn how they do it though.
 

opus111

Banned
Feb 10, 2012
1,286
3
0
Hangzhou, China
The best recorded piano music I own is Opus 111 in MCH SACD on 2L nordic.

I have downloaded the first movement of Op111 as a sample from 2L. Also various others of their free samples - was listening to the Haydn string quartet (I think) movement only this evening and remarked to myself how it must be a DSD (low-bit) recording because the dynamics were just not there. So I must disagree that theirs constitutes the best recording - certainly its a clean sound, but dynamically compressed to my ears.
 

opus111

Banned
Feb 10, 2012
1,286
3
0
Hangzhou, China
In my system, I'll would be attenuating in a range of between 50-20db. I know the volume control horse has been beaten to death, but can anyone confirm that 50db attenuation corresponds to 8bits, so with a DSP with 32 processor you loose no information on 44/16 material. What happens with native 24 bit sources is still not clear to me. Would this also have the 32 bits to work with so -50db (8bits) would not be a problem, or does a 24 bit source get chopped off at 24bit, so that at -50db you're down to 16bits, and have lost some information?

How much information you're left with depends on the DAC's noisefloor, assuming that more bits are used in the processing than are provided by the DAC. I have yet to hear of a 32bit (engineering, not marketing bits) DAC. I think the Sabre (in Medea+ incarnation) gets a noise floor around -136dB so approx 23bits. With this DAC you could in theory apply 7 bits (42dB) attenuation to 44k1/16 before losing information - i.e. exceeding the noise floor on the original recording. In practice since noise floors on redbook are generally above the theoretical minimum this 7bits is conservative.
 

edorr

WBF Founding Member
May 10, 2010
3,139
14
36
Smyrna, GA
How much information you're left with depends on the DAC's noisefloor, assuming that more bits are used in the processing than are provided by the DAC. I have yet to hear of a 32bit (engineering, not marketing bits) DAC. I think the Sabre (in Medea+ incarnation) gets a noise floor around -136dB so approx 23bits. With this DAC you could in theory apply 7 bits (42dB) attenuation to 44k1/16 before losing information - i.e. exceeding the noise floor on the original recording. In practice since noise floors on redbook are generally above the theoretical minimum this 7bits is conservative.

Keep in mind, I do the dithering not in the DAC, but in the Trinnov processor that sets before the DAC. This is a 64 bit floating point DSP engine. What is your bottom line in this scenario. Will I or will I not loose information?
 

edorr

WBF Founding Member
May 10, 2010
3,139
14
36
Smyrna, GA
I have downloaded the first movement of Op111 as a sample from 2L. Also various others of their free samples - was listening to the Haydn string quartet (I think) movement only this evening and remarked to myself how it must be a DSD (low-bit) recording because the dynamics were just not there. So I must disagree that theirs constitutes the best recording - certainly its a clean sound, but dynamically compressed to my ears.

Nordic sounds records in DXD (352.8kHz/24bit). This gets transfered to SACD (DSD) and BR (192/24). They sell packages with both the BR and SACD. I completely don't understand the dynamics are not there comment. May be an issue with the flac you downloaded. The SACD is superb.
 

opus111

Banned
Feb 10, 2012
1,286
3
0
Hangzhou, China
Keep in mind, I do the dithering not in the DAC, but in the Trinnov processor that sets before the DAC. This is a 64 bit floating point DSP engine. What is your bottom line in this scenario. Will I or will I not loose information?

Its easy to work out, irrespective of where the dithering takes place - if you're in need of dither that would only be because you lost information. Dither's only required (to randomize the quantization distortion) where there's word-length truncation, and this reduction in the word length means information is being lost. The number of bits in the DSP isn't relevant.
 

opus111

Banned
Feb 10, 2012
1,286
3
0
Hangzhou, China
Nordic sounds records in DXD (352.8kHz/24bit). This gets transfered to SACD (DSD) and BR (192/24). They sell packages with both the BR and SACD. I completely don't understand the dynamics are not there comment. May be an issue with the flac you downloaded. The SACD is superb.

Perhaps you've not heard redbook played back over an optimised multibit converter. Only once I was exposed to what was possible then did I notice what was missing on low-bit recordings (and low-bit DACs). I'm betting that Nordic's ADC uses a low-bit modulator prior to creating the DXD datastream. Can't think how there could be any issues with FLAC. I get the same loss of dynamics on 88k2 material which was originally obtained from a (Chandos) DSD stream so I take it that the dynamics loss is inherent with DSD. Stanley Lipshitz's technical treatment of DSD is entirely in accord with this perceived loss of dynamics.
 

edorr

WBF Founding Member
May 10, 2010
3,139
14
36
Smyrna, GA
Its easy to work out, irrespective of where the dithering takes place.

It is not easy enough for me. What is your take? These are the parameters: 50db attenuation, 96/24 source, volume control in DSP with 64 bit floating point processor. Do I od do I not loose information?
 

edorr

WBF Founding Member
May 10, 2010
3,139
14
36
Smyrna, GA
Perhaps you've not heard redbook played back over an optimised multibit converter. Only once I was exposed to what was possible then did I notice what was missing on low-bit recordings (and low-bit DACs). I'm betting that Nordic's ADC uses a low-bit modulator prior to creating the DXD datastream. Can't think how there could be any issues with FLAC. I get the same loss of dynamics on 88k2 material which was originally obtained from a (Chandos) DSD stream so I take it that the dynamics loss is inherent with DSD. Stanley Lipshitz's technical treatment of DSD is entirely in accord with this perceived loss of dynamics.

All I can say is that of all the recordings I have, the 2L nordic sounds the closest to an actual piano in my room. It is reassuring to know things can get better still. Just out of curiosity, if you had $10K to spend (used or new), what "optimised multibit converter" would you get?
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing