MSB Diamond IV Select dac

[Priaptor]

The folks at MSB were pretty firm in their belief that their new Quad i2s USB input sonics were on a par with any alternative source, which was a first for them with their new implementation.

My IV should come back as an upgraded V with a Quad USB later this week. I'll test the new quad i2s USB against the Ethernet-based chain (which I used previously), which in prior iterations did beat their best USB based alternative

I will be upgrading as well just want to do it when I am out of out of town so my downtime is mitigated. I look forward to your impressions. One benefit I will get is I will be able to see how much of a difference the Quad USB makes vs the V upgrade.

I think you will be in for a big surprise with Quad solution.

 

[Elberoth]

Lastly "Ethernet" is no panacea of eliminated noise. However again the buffering by the ProI2s board "should" alleviate most if not all issues no matter which solution one likes.

Buffering can't elimanate noise, which travels from the computer via the ground connection. Once it enters the DAC (via USB input) it contaminates the whole DAC. Buffering can only eliminate jitter.

 

[wizard]

 

[Priaptor]

Buffering can't elimanate noise, which travels from the computer via the ground connection. Once it enters the DAC (via USB input) it contaminates the whole DAC. Buffering can only eliminate jitter.

Adam

I get it. However Ethernet can also pick up noise and is no panacea. In fact probably the electronically "dirtiest" items in ones computer audio are hubs/switches. I have read as you have the benefits and negatives of both, the proponents and antagonists of both, etc.

MSB's approach is to electrically and fundementally isolate a bit perfect data stream to their PROI2S board. "Theoretically" either way should yield similar results but I am all in favor of that which sounds best. If the renderer is better great. My luck with renderer interfaces hasn't been the best but I too wouldn't mind a simple Ethernet connection to my DAC and if it functioned as well and sounded equal to or better I am game!

 

[Elberoth]

However Ethernet can also pick up noise and is no panacea. In fact probably the electronically "dirtiest" items in ones computer audio are hubs/switches. I have read as you have the benefits and negatives of both, the proponents and antagonists of both, etc.

Noise pickup was never a problem in computer audio, as far more noise is transfered from the computer as beeing picked up by the cable from the air. Besides, if you got for CAT 7 cables, those are heavily shielded.

Ethernet may not be a panacea for noise (afterall, galvanic isolation never works 100% - otherwise signal would not get through) but as a standard is hugely better at isolating the noise comming from the computer / hub etc than USB which has no galvanic isolation at all.

Add to that the superior data transfer protocol (TCP/IP), and it becomes clear that once more we had chosen the inferior standard (like we did in the past with VHS over BetaMax).

BTW - wher did you get that from that a hub is more noise than a computer ? Never heard that before.

 

[Priaptor]

Noise pickup was never a problem in computer audio, as far more noise is transfered from the computer as beeing picked up by the cable from the air. Besides, if you got for CAT 7 cables, those are heavily shielded.

Ethernet may not be a panacea for noise (afterall, galvanic isolation never works 100% - otherwise signal would not get through) but as a standard is hugely better at isolating the noise comming from the computer / hub etc than USB which has no galvanic isolation at all.

Add to that the superior data transfer protocol (TCP/IP), and it becomes clear that once more we had chosen the inferior standard (like we did in the past with VHS over BetaMax).

BTW - wher did you get that from that a hub is more noise than a computer ? Never heard that before.

My history goes far beyond musical networks. I have designed and was one of the original architects of functioning imaging DICOM networks in the medical industry, dependent on clean data pathways, often involving developing complex routing tables utilizing multiple smart routers, usually from CISCO with guaranteed "galvanic isolation". These were huge breakthroughs which I helped pioneer leaving ancient networks like SINET behind, adapting TCP/IP protocols with what is known as AE Titles to develop routing tables. I ran into more problems than you can imagine, usually the simplest things such as a crappy cable, a malfunctioning plug in module in a high speed copper module that was "galvanically shielded" that would manifest its problems to a network TCP/IP transfer of 1/1000 of expected, to shorting out a module, to shorting out a server or taking down the network and at times, the problems could be as simple as a crappy termination on a wall plug. Adam, there are many many many variables in a network. I am not disputing that in the theoretical world a properly setup network, with great shielded cables, properly isolated and terminated, terminating in a high end switch that is set up for "galvanic isolation" as is supposedly the default in CAT 7 can't or shouldn't eliminate ALL electrical noise. From my experience dating back to early 90s when DICOM rolled on the scenes I have learned my lessons. The vendors would love to use mini networks so that if one modality's network went down it didn't take down the rest. My experience was the more "complex" and more numerous the talking among isolated networks the higher the rate of error and I now ONLY homerun all my cables to high end router, ensure my cables are truly isolated and check and check again that my cables are properly terminated. Now I am working in the "perfect" world where I am designing these things in new buildings or hospitals designed for such things. AND even then I still run into issues with noise, which I can track down and fix. The real world, even in a simple home environment is quite different and you can take that to the bank. Again, I am all on board and would love nothing more than just plugging an ethernet cable into a renderer built-in to my MSB if it works properly ( few do in my experience at THIS TIME) and the sound is equal to or better.

My original peeve was with the poster saying a renderer will ALWAYS be sonically superior, which is nonsense. When I hear these absolutes I always criticize. You use a Trinity DAC which is a whole different ballgame than say a Vega in terms of USB interface. To compare one to the other and make an absolute statement about some commonality between the two DACs is absurd.

That is the thrust of my commentary on this thread. Time will tell which input, renderer or QUAD USB is better. My "bias" is leaning, given what I have read about and spoken to the people at MSB, the methods used for both should be pretty similar sonically, with some recommending one over the other, and if I had to project, I would think sonically they both will probably perform similarly and implementation and ease of use with be what determines ones choice. BUT I keep an open mind as I am married to no audio gear. If it does gapless without a hiccup, it will be a huge step from their UMT+ regardless.

 

[seriouslyhifi]

Thank you for your detailed explanation in the Medical world.
For the record I have spoken to a few friends and have also personally heard the MSB Analog DAC with both the Renderer and the Quad Rate USB as a comparison in the same machine. The Renderer was always better in this case. That is all I wanted to say and kept away from your regular bashing as I did not want to get into an unnecessary argument, but you just keep at it.
I have nothing really more to add and wish you all the best in your journey. Good luck on your DAC V upgrade and I am sure you will enjoy your music whatever input you finally choose. And let us close this and get back to music and help each other with what we hear. Nothing more on this from me.

 

[audioblazer]

AB,

Is MSB going to upgrade original SELECT purchases (of over a year ago) to the new SELECT (minus new FEMTO33 and Renderer) for "free" ?

I bought MSB Diamond IV USD42k+ retail in late 2013. Less that 2/3 months MSB came out with Select for USD 70k & within a short time to USD75k. 15 months forward , MSB released MSB Diamond V for about USD42-45k which they claimed is much better than MSB Select 2014 selling for USD75k !!!. When the Select program came out in 2014 we were told free upgrade for any top upgrade . Now they came out with a MSB Diamond V which they claimed is better than Select 2014 & it's only retailing for abt USD42-45k !!! The best part is that the free upgrade is not for the best but we have to pay USD9995 for a better clock Femto 33 & USd20k for better power supply !!!. Beside that if they were to upgrade in 2016 to Select 2016 for USD120k , Select 2014 purchasers will have to pay USD (120-75k) = USD45k !!!. Now is that a free upgrade or a loaner for 1 year + that cost USD30k+ ? Bear in mind Diamond V is better than Select 2014 & it's much cheaper at USD40+k. What a suck program . Hence if anybody think it's a great upgrade , u can have mine for 1/2 , brand new , 1 month from now when MSB is ready to process my upgrade . Have been waiting for 3months +

 

[edorr]

Most manufacters of ultra luxurygoods are more subtle about taking their customers for suckers. At least MSB is not pretending there is any relationship between pricing of their product, cost to build it and sound quality, and their pricing is entirely based on their assumptions of how much money can be extracted from their customer base without getting sued.

 

[CKKeung]

I attended the MSB Diamond Select demo at Pacific Audio, the official MSB Hong Kong dealer tonight.

Larry Gullman & his son Daniel were in Hong Kong for 2 days. They brought with them a 2-ch Diamond Select & a multi-ch Diamond Select.

The demo lasted for 1.5 hours. Both 2-ch CD and multi-ch BR discs & SACD were played.

Below are some of the photos taken :

Multi-ch Diamond Select on the rack top.


2-ch Diamond Select. The display screen is active only when the remote is in use.


Series V UMT with two Powerbases.


I was deeply impressed by the performance of the Selects.
They are in another league above the IV & V. (A set of Series V was playing outside the main demo room)
The sound was so natural yet retaining the MSB house sound of great resolution, 3D soundstage and authority.

My Extreme/Pinnacle DAC List now has two members : Trinity DAC and MSB Diamond Diamond Select.
Hope that I can find a chance to compare them directly in future.

 

[Priaptor]

I attended the MSB Diamond Select demo at Pacific Audio, the official MSB Hong Kong dealer tonight.

Larry & Gullman were in Hong Kong for 2 days. They brought with them a 2-ch Diamond Select & a multi-ch Diamond Select.

The demo lasted for 1.5 hours. Both 2-ch CD and multi-ch BR discs & SACD were played.

Below are some of the photos taken :

Multi-ch Diamond Select on the rack top.
View attachment 20145

2-ch Diamond Select. The display screen is active only when the remote is in use.
View attachment 20147

Series V UMT with two Powerbases.
View attachment 20146

I was deeply impressed by the performance of the Selects.
They are in another league above the IV & V. (A set of Series V was playing outside the main demo room)
The sound was so natural yet retaining the MSB house sound of great resolution, 3D soundstage and authority.

My list of extreme/pinnacle DAC List now has two members : Trinity DAC and MSB Diamond Diamond Select.
Hope that I can find a chance to compare them directly in future.

Thanks for the info. Were those production units? Also what sources and were they same on both the SELECT and V

Thanks

 

[CKKeung]

Thanks for the info. Were those production units? Also what sources and were they same on both the SELECT and V
Thanks

They are the finalized engineering units. The formal launch will be at Munich Show on 12 May.

During the demo last night, the multi-ch Select paired with a new Series V UMT.
The 2-Ch Select paired with a metal tray Signature Data IV (or V?).

No preamp was used.
The 2-ch Selects directly drove a MSB S201 poweramp for 2-ch demo.
The multi-ch Select drove five Pigreco poweramps.

Speakers were from Rosso Fiorentino : Florentia for 2-ch and five Siena for multi-ch.

 

[Elberoth]

Yesterday, I had a chance to talk to Larry Gullman and his son (who is a digital engineer at MSB) about the new DAC Select. It indeed is a completely new product, which has nothing in common with the old DAC IV Select.

Here are some interesting design details that I have learned about during our conversation.

1. The DAC Select uses 8 DAC modules vs 4 in DAC IV. The modules are balanced and run in parallel, meaning there are 8 DACs per channel (16 DACs total).

2. The DAC modules are run at max speed of 12MHz, meaning they can process Quad DSD natively, without any PCM downconversion. The DAC modules in the DAC IV were run at 1.5MHz, and could be run at 3MHz max (that was barely fast enough for DSD 64x). Mr. Gullman Jr. was not sure at what rate PCM is run through those DACs.

3. The new DACs do not have an opamp based output stage (or any output stage for that matter). 8 of them gives enough current to drive the outputs directly. Output impedance is 75 Ohms from balanced outs. There is no analog filter and there is only one resistor in the signal path to perform passive I/V conversion. So just like the Trinity DAC, this is a sophisticated, yet extremely simple design.

4. No opamps in the DAC modules and no extra voltage regulators means DAC modules do not run as hot as in the DAC IV. They reckon the heat output from DAC modules was reduced by 70%. As a result, the whole DAC Select runs barely warm - which is in a stark contrast to the hot running DAC IV.

5. The part of the DAC they are most proud of is the new passive volume attenuator. It is a constant impedance design (impedance does not change with volume setting) and has no capacitors in the circuit. The passive volume attenuator module is equipped with one balanced input. Signal from this input IS routed via an active buffering stage (a preamp if you will) before going through the passive attenuator.

6. The DAC will be equipped with Renderer module (Ethernet input) like the Analog DAC. The module is not ready yet. Digital input modules that are ready are I2S, various SPDIF options and Quad USB. Renderer module should be ready in 4 months time.

7. The Select DAC comes with the Galaxy Femto Clock (also called the 77 femto clock) as standard. For $10k extra, you can get their latest 33 femto clock. The 33 femto clock can be bought semapately for $20k and works with all DAC IVs and DAC Vs. All femto clocks use two OCXOs (oven controlled) - one for each freq family.

8. A lot of time was spent to isolate the Quad USB input from the computer noise. MSB uses a combination of optoisolators and microtransformers.

9. One of the very innovative features of the new DAC Select, is the clocking system. The DAC Select uses just ONE clock to run all circuits. The DAC IV - like most of the DACs - used multiple local clocks, which generated more noise on the ground plane than a single clock (one clock for audio, one clock for the display and one clock for the USB input). Since the common clock drives both the DAC and the display, they claim there is no sonic penalty for having the display on.

10. There will be next components housed in the same chassis as the DAC Select - before the CES we can expect a matching transport and a poweramp (loosely based on the S201 design).

11. It is my understanding that all DAC IV Select dacs will be exchanged free of charge for the new DAC Select (with the standard 77 Femto clock).

12. DAC IV Select and DAC V Diamond are exactly the same in terms of hardware. The only difference between the two is an upgraded software in the DAC V Diamond.

All in all - it looks like a very interesting design. The price is crazy though.

 

[Elberoth]

Some images from yesterday meet:

PS> I have also learned an interesting family story. 4 out of 11 children of Mr. Gullmann are engineers. Two of them work for MSB. One is a digital engineer and another is a mechanical engineer. It is the last one who designed the new DAC Select exterior.

 

[Priaptor]

Yesterday, I had a chance to talk to Larry Gullman and his son (who is a digital engineer at MSB) about the new DAC Select. It indeed is a completely new product, which has nothing in common with the old DAC IV Select.

Here are some interesting design details that I have learned about during our conversation.

1. The DAC Select uses 8 DAC modules vs 4 in DAC IV. The modules are balanced and run in parallel, meaning there are 8 DACs per channel (16 DACs total).

2. The DAC modules are run at max speed of 12MHz, meaning they can process Quad DSD natively, without any PCM downconversion. The DAC modules in the DAC IV were run at 1.5MHz, and could be run at 3MHz max (that was barely fast enough for DSD 64x). Mr. Gullman Jr. was not sure at what rate PCM is run through those DACs.

3. The new DACs do not have an opamp based output stage (or any output stage for that matter). 8 of them gives enough current to drive the outputs directly. Output impedance is 75 Ohms from balanced outs. There is no analog filter and there is only one resistor in the signal path to perform passive I/V conversion. So just like the Trinity DAC, this is a sophisticated, yet extremely simple design.

4. No opamps in the DAC modules and no extra voltage regulators means DAC modules do not run as hot as in the DAC IV. They reckon the heat output from DAC modules was reduced by 70%. As a result, the whole DAC Select runs barely warm - which is in a stark contrast to the hot running DAC IV.

5. The part of the DAC they are most proud of is the new passive volume attenuator. It is a constant impedance design (impedance does not change with volume setting) and has no capacitors in the circuit. The passive volume attenuator module is equipped with one balanced input. Signal from this input IS routed via an active buffering stage (a preamp if you will) before going through the passive attenuator.

6. The DAC will be equipped with Renderer module (Ethernet input) like the Analog DAC. The module is not ready yet. Digital input modules that are ready are I2S, various SPDIF options and Quad USB. Renderer module should be ready in 4 months time.

7. The Select DAC comes with the Galaxy Femto Clock (also called the 77 femto clock) as standard. For $10k extra, you can get their latest 33 femto clock. The 33 femto clock can be bought semapately for $20k and works with all DAC IVs and DAC Vs. All femto clocks use two OCXOs (oven controlled) - one for each freq family.

8. A lot of time was spent to isolate the Quad USB input from the computer noise. MSB uses a combination of optoisolators and microtransformers.

9. One of the very innovative features of the new DAC Select, is the clocking system. The DAC Select uses just ONE clock to run all circuits. The DAC IV - like most of the DACs - used multiple local clocks, which generated more noise on the ground plane than a single clock (one clock for audio, one clock for the display and one clock for the USB input). Since the common clock drives both the DAC and the display, they claim there is no sonic penalty for having the display on.

10. There will be next components housed in the same chassis as the DAC Select - before the CES we can expect a matching transport and a poweramp (loosely based on the S201 design).

11. It is my understanding that all DAC IV Select dacs will be exchanged free of charge for the new DAC Select (with the standard 77 Femto clock).

12. DAC IV Select and DAC V Diamond are exactly the same in terms of hardware. The only difference between the two is an upgraded software in the DAC V Diamond.

All in all - it looks like a very interesting design. The price is crazy though.

MORE IMPORTANTLY how did it sound. I understand it is hard to compare given the lack of control of all variables but for the system how did it sound?

Yes the price is crazy. As a high end buyer I could more justify the price if the 10 year upgrade option was free for 5 years and a delta between cost of new and old for the next 5 years rather than just one year for the former

 

[Elberoth]

No idea. That was distributor's system I was not familiar with. So I didn't bother. If I have a chance to borrow one (right now thaere are only 5 Selects, all beeing preproduction samples, so noone has the real product), I will.

 

[asiufy]

We're going to have one of the those units next week, in one of our 3 rooms at the Newport Beach show

Needless to say, I can't wait !!!


alexandre

 

[Priaptor]

We're going to have one of the those units next week, in one of our 3 rooms at the Newport Beach show

Needless to say, I can't wait !!!


alexandre

Well let us know what you think. I am going to hold off on my IV --->V upgrade until the dust settles; although I did do the Quad USB upgrade which I think made a big difference.

 

[Brucemck2]

Well let us know what you think. I am going to hold off on my IV --->V upgrade until the dust settles; although I did do the Quad USB upgrade which I think made a big difference.

The Signature V with the Quad USB is very, very nice

 

[Priaptor]

The Signature V with the Quad USB is very, very nice

And that is PROBABLY the direction I will take UNLESS I do something utterly idiotic which I am prone to do and get blown away by the Select and decide to hold off on retirement or make some other completely irrational justification for going into hock to have one.