The Analog DAC

[edorr]

This is on RBCD material? If so it pretty much mirrors my own experience from comparing multibit DACs with S-D ones - the multibit win out on timbre and dynamics, the S-D ones have better bloom and soundstaging. My DAC developments over the past year or so have focussed on getting my multibit designs to the levels of air/bloom and SS depth to the levels I've heard on the best S-D ones.

Yes, this is on rbcd. I have a some dsd tracs and some high Rez, but I make my buying decisions based on good old 44/16 which is where most of the content is for me.

 

[Elberoth]

There is a substantial difference between the Signature and Diamond DACs, even after installing the Femto clock in both.

 

[edorr]

There is a substantial difference between the Signature and Diamond DACs, even after installing the Femto clock in both.

This is good to know next time I have $10K burning a hole in my pocket. Can you comment on the difference between Femto 140 and Galaxy clock?

Alexandre, I believe you went the Signature to Diamond route as well. What is your experience in terms of difference in SQ?

 

[asiufy]

It is significant, but in that "more of the same" way. Its an incremental improvement on MSB's take on digital. So, it's more resolving, more detailed, more punchy. But as you realised with the Signature, you're already ahead of most DACs in those aspects anyway, so the Diamond will only give you MORE of what you already have over the others.

Given that MSB is selling now a "plus" version, with the 140 clock built-in, maybe a lot of "non plus" Diamonds will come up on the used market.

Also, don't understimate the value of the power supplies, for the DAC and transport.

alexandre

 

[edorr]

It is significant, but in that "more of the same" way. Its an incremental improvement on MSB's take on digital. So, it's more resolving, more detailed, more punchy. But as you realised with the Signature, you're already ahead of most DACs in those aspects anyway, so the Diamond will only give you MORE of what you already have over the others.

Given that MSB is selling now a "plus" version, with the 140 clock built-in, maybe a lot of "non plus" Diamonds will come up on the used market.

Also, don't understimate the value of the power supplies, for the DAC and transport.

alexandre

Thanks. I am upgrading my powerbase to Diamond. I am not holding my breath about "old" Diamonds hitting the used market, because to get it to plus level, all you need to do is buy the Femto clock.... It appears most Diamond users already have a galaxy clock.

If I get the upgrade itch what I could do is wait for a fully loaded Diamond / Galaxy to show up used and sell my Signature with Femto 140. This type of transaction may work out a lot cheaper than buying the upgrade at retail. In any case, no rush. I'm looking forward to many blissful hours with the signature.

 

[Elberoth]

There is a guy on AudioAficionado selling his Diamond with Femto. You can check this ad out.

BTW - I believe the plus models come with the Femto 140 clock as standard, not the Galaxy one.

 

[rockitman]

the multibit win out on timbre and dynamics, the S-D ones have better bloom and soundstaging. My DAC developments over the past year or so have focussed on getting my multibit designs to the levels of air/bloom and SS depth to the levels I've heard on the best S-D ones.

So s-d chips yield better bloom and soundstaging ? To me that is what I like best about analog. Perhaps the emm labs sounds more like analog vs the Msb ?

 

[edorr]

So s-d chips yield better bloom and soundstaging ? To me that is what I like best about analog. Perhaps the emm labs sounds more like analog vs the Msb ?

I have limited reference of what analog "analog" sounds like, but I would say between the two dacs the MSB sounded more "analog" to me than the EMM Labs.

 

[LL21]

...my own experience from comparing multibit DACs with S-D ones - the multibit wins out on timbre and dynamics, the S-D ones have better bloom and soundstaging. My DAC developments over the past year or so have focussed on getting my multibit designs to the levels of air/bloom and SS depth to the levels I've heard on the best S-D ones.

So s-d chips yield better bloom and soundstaging ? To me that is what I like best about analog. Perhaps the emm labs sounds more like analog vs the Msb ?

I have limited reference of what analog "analog" sounds like, but I would say between the two dacs the MSB sounded more "analog" to me than the EMM Labs.

Let me take Opus' comments (which are always respected by me as nothing more than astudent of digital) and share my own non-techie observations.

I think the S-Ds that have impressed me do provide a more detailed soundstage, more clarity/delineation between instruments. I suppose one could technically include 'air/space' around instruments because of how precisely they seem to be placed within the soundstage. The Stahl-Tek is still the best i have ever heard for this, and the Scarlatti was probably right up there, thought i would prefer a detailed at home shoot-out to really know.

However, as for instrument "bloom" (which is not for me necessarily spacing, room around each instrument set back within the soundstage and separate from other instruments in the soundstage...but rather whether the instrument seems to 'pop out in 3D' into my living room...whether the singer almost seems like he/she is right in front of you and you could almost reach out and touch them with the breath coming from their mouth, as well as more stereotypical (pardon the pun) 'analog' traits of purity of tonality, etc...i have found i prefer the DACs which were not S-D technology. TDA1541A (ie, Zanden) still somehow capture that magic for me better, and i have prioritized those above the traits that Stahl-Tek do so well...so i have kept my 'old-school' Zanden for this reason.

Therefore, having heard Emm in the past, and also what i have read about MSB, i am not surprised to hear Edorr comment that it seems more analog to him than Emm.

 

[asiufy]

lloyd,

You must listen to the MSB gear then

You described exactly my impressions of my old gear (dCS), versus the new (MSB).
At first, you kinda miss all that lovely transparency, the pinpoint imaging... But you gain so much in presence, purity, palpability as you described, that you kinda leave that aside for a second
Overall, I'm happy with the trade off, but of course I'd love to be able to have the best of both worlds!


alexandre

 

[LL21]

lloyd,

You must listen to the MSB gear then

You described exactly my impressions of my old gear (dCS), versus the new (MSB).
At first, you kinda miss all that lovely transparency, the pinpoint imaging... But you gain so much in presence, purity, palpability as you described, that you kinda leave that aside for a second
Overall, I'm happy with the trade off, but of course I'd love to be able to have the best of both worlds!


alexandre

I just got off the phone with Andy Payor from Rockport...never spoken with him before. My impression of him has only gone from great to even greater. And i believe he uses MSB digital in his systems in Maine. Would love to hear it.

 

[NorthStar]

-- Andy Payor's not a member here at WBF?

 

[opus111]

So s-d chips yield better bloom and soundstaging ?

That has been my own experience - mainly from developing DACs based on AD1955 with external I/V. When I first heard multibit (NOS) after listening to my AD1955 implementations for several months its what I noticed as lacking - soundstage depth, ambience retrieval. The TDA1543 which was my first introduction to multibit was flat in comparison - the content was compressed (in space) towards the front of the soundstage and I couldn't hear the acoustic space where the performance was recorded.

To me that is what I like best about analog.

Yet to gain these qualities with S-D it turns out you have to sacrifice other desirable qualities of analog - timbral accuracy and dynamics. So which do you value most?

I've since got my multibit designs to have the soundstaging I was getting with S-D, and some incidentally. So its not an inherent feature of multibit to lose soundstaging and air, it depends on implementation.

 

[LL21]

That has been my own experience - mainly from developing DACs based on AD1955 with external I/V. When I first heard multibit (NOS) after listening to my AD1955 implementations for several months its what I noticed as lacking - soundstage depth, ambience retrieval. The TDA1543 which was my first introduction to multibit was flat in comparison - the content was compressed (in space) towards the front of the soundstage and I couldn't hear the acoustic space where the performance was recorded.



Yet to gain these qualities with S-D it turns out you have to sacrifice other desirable qualities of analog - timbral accuracy and dynamics. So which do you value most?

I've since got my multibit designs to have the soundstaging I was getting with S-D, and some incidentally. So its not an inherent feature of multibit to lose soundstaging and air, it depends on implementation.

Hey Opus, thanks for continuing to post here...enjoying all the feedback! My own far more limited experience with good S-D and good NOS on soundstage mirrors your description, in that the Stahl-Tek depth and delineation of soundstage was like nothing i have come across before. And it was an incredibly low noise floor (the entire system to be fair) which allowed minute details to surface effortlessly (and therefore it seemed without distortion or mud). I still felt like when a female vocalist was front and center...her face started to peer through space at me more with the NOS dac, whereas the the Stahl-Tek placed her further back properly proportioned and delineated in space but not coming thru the barrier between the speakers and me. TDA 1541 double crown DAC (Zanden) still that better for me. And with 'loads' of damping around the Zanden, the DAC did start to set the soundstage back quite a bit though still not in same league as the Stahl-Tek.

 

[opus111]

Hi Lloyd, no worries, its fun comparing notes on how things sound and trying to correlate that with circuit details

The 'delineation in space' thing is something that, to my ears at least, is rather artificial. I did find that my S-D DAC imaged well but this is unlike reality - when I've been to concerts I consciously compare what I hear there with what I've been hearing on my system, and I don't find 'pin point imaging' at concerts at all. Rather I tend to rely on my eyes to get the location of the instrument/soloist rather than on my ears - the actual sound of them is more diffuse than I used to get with my S-D DAC. I now put that down to what I call 'MSG' - added HF 'detail' which isn't actually on the recording. Its distortion contributed by HF overload of opamps in my experience because it can be eliminated by RF filtering prior to the opamp. This takes the 'edge' off the perceived sound, but makes it much more 'organic' in exchange, giving that 'coming through the barrier' effect - more 'presence' somehow, more breathiness on vocals as their tonal balance is shifted lower in frequency.

 

[LL21]

Hi Lloyd, no worries, its fun comparing notes on how things sound and trying to correlate that with circuit details

The 'delineation in space' thing is something that, to my ears at least, is rather artificial. I did find that my S-D DAC imaged well but this is unlike reality - when I've been to concerts I consciously compare what I hear there with what I've been hearing on my system, and I don't find 'pin point imaging' at concerts at all. Rather I tend to rely on my eyes to get the location of the instrument/soloist rather than on my ears - the actual sound of them is more diffuse than I used to get with my S-D DAC. I now put that down to what I call 'MSG' - added HF 'detail' which isn't actually on the recording. Its distortion contributed by HF overload of opamps in my experience because it can be eliminated by RF filtering prior to the opamp. This takes the 'edge' off the perceived sound, but makes it much more 'organic' in exchange, giving that 'coming through the barrier' effect - more 'presence' somehow, more breathiness on vocals as their tonal balance is shifted lower in frequency.

Now THAT's interesting...that is the first time i have read (or maybe i missed it elsewhere) that the detail is artificial. I agree that live venues are always more diffuse, but i somehow imagine that the microphones were picking something up due to their multi-mike locations or something that was getting that detail.

You are suggesting that this is NOT the case, and that in fact this is artificial. Can you explain further to a layman? Thanks.

 

[opus111]

Well I can just talk in a bit more detail about my discoveries with DACs and hopefully you'll be able to relate that to your own experience with what's commonly called 'detail' and make up your mind accordingly.

At first listening to my TDA1543 DAC I really missed the detail which was so evident on the S-D (AD1955) I'd had as my reference. Missing the detail meant no 'spotlighting' of the front of stage interest any more, less 'focus' on vocals. I must admit though that there was always something alluring about the timbre of orchestral instruments with NOS, but for me the trade off wasn't sufficient for me to pursue NOS development seriously because of its lack of soundstage. Then one day I experimented with the output filter on my TDA1543 and made the soundstage come back - not to the same extent as with the AD1955, but perhaps 80% as good. Only then did I start to think that maybe NOS DACs had a future.

I started with TDA1543s but from the measurements was convinced that TDA1545 was the better DAC, technically. It has 'continuous calibration' to ensure the bit weights are accurately proportioned, but when I tried it, it didn't sound as good. It hadn't got the 'organic' quality of the TDA1543 - what was actually happening though was it was adding 'detail' - and by this time I'd listened to TDA1543 long enough that I was totally weaned off detail in favour of 'organic' and accurate timbres. So when it came back, I wasn't a fan and wondered if I could get rid of it and make the TDA1545 sound more 'organic'.

It turned out the answer to this question was 'yes' - the solution was ferrite beads, lots and lots of them (30 or so). A very long string of ferrite beads between the TDA1545 and the amplifier (using passive I/V in between) got rid of all that detail and made the TDA1545 sound considerably better (as measurements predicted it should) than the TDA1543. The ferrite beads were absorbing all the excess RF energy (glitches) produced by the TDA1545 which weren't coming out of the TDA1543 because this latter chip is a bipolar one (not CMOS). With this discovery, I was satisfied that this is the reason DIYers have shied away from the TDA1545 in favour of the SQ of the 1543 - they were looking for the 'organic' sound and hadn't cottoned on to the reason the 1545 sounded poorer in this department - glitches.

Does any of this help?

 

[LL21]

Well I can just talk in a bit more detail about my discoveries with DACs and hopefully you'll be able to relate that to your own experience with what's commonly called 'detail' and make up your mind accordingly.

At first listening to my TDA1543 DAC I really missed the detail which was so evident on the S-D (AD1955) I'd had as my reference. Missing the detail meant no 'spotlighting' of the front of stage interest any more, less 'focus' on vocals. I must admit though that there was always something alluring about the timbre of orchestral instruments with NOS, but for me the trade off wasn't sufficient for me to pursue NOS development seriously because of its lack of soundstage. Then one day I experimented with the output filter on my TDA1543 and made the soundstage come back - not to the same extent as with the AD1955, but perhaps 80% as good. Only then did I start to think that maybe NOS DACs had a future.

I started with TDA1543s but from the measurements was convinced that TDA1545 was the better DAC, technically. It has 'continuous calibration' to ensure the bit weights are accurately proportioned, but when I tried it, it didn't sound as good. It hadn't got the 'organic' quality of the TDA1543 - what was actually happening though was it was adding 'detail' - and by this time I'd listened to TDA1543 long enough that I was totally weaned off detail in favour of 'organic' and accurate timbres. So when it came back, I wasn't a fan and wondered if I could get rid of it and make the TDA1545 sound more 'organic'.

It turned out the answer to this question was 'yes' - the solution was ferrite beads, lots and lots of them (30 or so). A very long string of ferrite beads between the TDA1545 and the amplifier (using passive I/V in between) got rid of all that detail and made the TDA1545 sound considerably better (as measurements predicted it should) than the TDA1543. The ferrite beads were absorbing all the excess RF energy (glitches) produced by the TDA1545 which weren't coming out of the TDA1543 because this latter chip is a bipolar one (not CMOS). With this discovery, I was satisfied that this is the reason DIYers have shied away from the TDA1545 in favour of the SQ of the 1543 - they were looking for the 'organic' sound and hadn't cottoned on to the reason the 1545 sounded poorer in this department - glitches.

Does any of this help?

Yes! Very interesting read...thanks, Opus111

 

[LL21]

Opus - these ferrite beads...can you buy them and mod existing products, or is this too difficult?

 

[opus111]

Certainly you can, they're cheap things - about 3 cents each in a sizeable volume - I buy full reels. As you can imagine, I use quite a few

http://www.mouser.com/ProductDetail/TDK/MPZ2012S102A/?qs=%2fha2pyFaduhOuNZScbHe%2fvQQza9t9FbZXD82NWnA%2fbI%3d