I need help - Better DAC or NEW CD player?

I am in the middle of purchasing some components for my new system, unfortunately, reading posts on a rainy night on Audigon only makes my "upgradeatitis" syndrom more acute.

The problem is that some time ago, I bought a squeezebox Duet, to use as transport for lossless files. The idea obviously was to get a good DAC to go along with it.

The one I was almost set on was the Dacmagic, which gets very good reviews here and almost everywhere you look. But, on the other hand, I am willing to spend a little more......Now, the thing is my current CD player is a Rotel RCD 1072, and was wondering whether this Dacmagic would give me any improvements if I use the rotel as a transport for it.

Or, I could get for example a Wyred DAC or a Benchmark DAC1 and use the Rotel for transport, but I dont know the limits of the Rotel as transport alone, I dont want to spend 600 more bucks on a better DAC to be fed with 2 "not so good" transports (the Duet and the Rotel).

So I am kinda stuck!! On the other hand I am contemplating getting the Dacmagic to use only for the Duet, and getting a new better CD player altogether, which puts me into another doubt, I dont know if the little money I can for the rotel + lets say 1000 more will really give me any improvements in the system if I purchase a whole new CD player since the Rotel is said to be very capable.

FWIW my amp will be a W4S STI500 and speakers are B&W 803D.

Let me know what you think about this and your suggestions on what you would do.... thanks in advance for all your help!
Demianm - Asynchronous upsampling DACs like Benchmark DAC1 have very strong jitter rejection and transport should not make any difference as long as it is "bit transparent" (no DSP, no volume control etc.). With Benchmark you can use even cheap DVD player (great traction) and cheap coax or Toslink. Rejection of jitter (that is already at about -60dB) is at frequencies of interest (kHz) over 100dB - impossible to detect. In fact Benchmark is clean/transparent to point of being too clean (sterile). I like it but many people prefer more "lush" presentation. Listen first before you buy (Benchmark had free 30 day evaluation program). If you decide Benchmark - get new one to avoid problems of early revisions. Step above Benchmark (and 2.5x price) is Bel Canto DAC3 (with remote control).
I second the Benchmark, but I strongly disagree with:

"Asynchronous upsampling DACs like Benchmark DAC1 have very strong jitter rejection and transport should not make any difference as long as it is "bit transparent" (no DSP, no volume control etc.). With Benchmark you can use even cheap DVD player (great traction) and cheap coax or Toslink. Rejection of jitter (that is already at about -60dB) is at frequencies of interest (kHz) over 100dB - impossible to detect. In fact Benchmark is clean/transparent to point of being too clean (sterile)."

Bits are not just bits. The more involved I become with digital, the more I discover it is like analog. Even the net effect of "getting it right/better" has similar traits to having a turntable/arm/cart set up properly.

The Benchmark usb and up versions have not change to my knowledge. Furthermore, the Benchmark is sensitive to cords, power and fuses...

I think the thing is that no two DACs sound the same, much like phono setups, but that is where the similarites end.

I think Kijanki is correct in asserting that with the BEnchmark, its sound will not vary much based on transport within certain wide and easily met (with more modern transports) limits, due to the BEnchmarks specific design which essentially eliminates external dependencies that can affect the sound, mainly the clock signal needed to convert the bits to analog accurately.

How one takes to the sound of the BEnchmark or any other DAC is still largely a matter of preference, but at least with a DAC like the Benchmark, as Kijanki indicates, you should get pretty much the same sound with most any transport up front. That may not be so much the case with other DAC designs that rely on an external clock signal provided by the source in order to convert the bits to an analog soundwave.
The rest of your system is quite good. I think that you want to get as good of a source as you can and as good as the DacMagic is - and it will do a better job than the DAC in the RC-1072 - with the budget you are suggesting you can do better. There are at least two ways I think you can go:

1) a CD player that can function effectively as a DAC, like the Cambridge Audio Azur 840C - very good CD player and very good upsampling DAC with a nice array of digital inputs

2) stand alone DAC like the Benchmark you mention, and using your existing gear as transports. There are other more expensive DACs I would also consider, even if it means waiting a bit, like

-Bryston BDA-1
-Bel Canto DAC3vb
There is one thing I dont like about the Benchmark - There is only one coaxial input.

I would like to use it for my Squeezebox, my CD player/transport and Oppo BRP. I know I have the coaxial and toslink available, but most people say coaxial is better over the toslink. I also could use the Oppo for CD but I think it sucks for that use, mainly due to loading times, I´d rather use it only for video but nonetheless somewhere in the future upgrade my CD transport.
Demianm - Toslink creates more jitter than coax but with Benchmark it doesn't matter. Benchmark tested DAC1 using 500' of cat5 cable instead of coax and there was no audible difference. I use, with Benchmark, Toslink from Airport Express and coax from DVD player and cannot tell the difference.

I stand by my original statement. I agree that all DACs sound different. I also assert that the connections make a difference, and that the Benchmark is not as immune to inputs as they claim it to be. The overall signature remains consistent, but the inputs sound different. I hear differences with USB cables, spdif and toslink. Furthermore on different cables of the same type and usb converters such as the hiFace- not to mention transports especially. I am not knocking the units at all- I own two, a USB version and the HDR one. I just feel it does a disservice to others perpetuating the myth (IMO) that bits are just bits. Bits are only just bits if they are there at the right time IME.

Bottom line, as with most things in audio, everything matters. Jitter is real, and you are best off avoiding it in the first place, as opposed to "dealing" with it after the fact.
" I hear differences with USB cables, spdif and toslink. Furthermore on different cables of the same type and usb converters such as the hiFace- not to mention transports especially. "

That is interesting and surprises me a bit. I'm not sure exactly how to account for those kinds of differences with the BEnchmark specifically but I imagine there is an explanation. Perhaps there are bits being dropped or something like that in certain cases. maybe someone else can offer a technical explanation for what you hear.
"bits are just bits. Bits are only just bits if they are there at the right time IME."

That's what we're talking about - jitter (noise in time domain). Jitter creates sidebands at -60 to -80 dB still audible since not harmonically related to root frequencies (noise with complex signal). Benchmark's jitter bandwidth is just a few Hz and at kHz frequencies suppression is over 100dB (overall -160dB).

Tricky part is to use the same transport to compare Toslink to coax. Once you have different transports all bets are off. It is not only that one might be not "bit transparent" but might have weaker laser, dirty lens etc. It is not related to price of CDP (cheap DVD players have great tracking) or jitter but just not reading information properly. Some people use green markers etc but I ripped CDs as data (CDP cannot do it working in real time) using MAX program and have perfect long lasting version on my server.

In addition to above Toslink breaks ground loops and therefore in certain situations overall system might sound better (but Benchmark has nothing to do with it)

Easy way to test is to connect transport that has both coax and Toslink and listen on the headphones switching inputs. I don't hear any difference that way and theory says it shouldn't be any but there is still placebo effect. If one believes that there should be some difference then he will be able to hear below -160dB easily.
I have the bel canto DAC3 and I cannot tell the difference between sources except for the usb and toslink connections, which I don't use anyway. AES/EBU, coax and SPDIF all sound the same to me. I cannot tell the difference using my high end CD player using balanced inputs vs. my squeezebox running flac files using a basic digital coax cable--and that is switching immediately from one source to the other using the DAC3 to switch sources. The DAC3's asynchronous upsampler makes it all sound great, and comes highly recommended. The bonus is you can also run the DAC3 as a preamp using the digital volume control, but I find I like a preamp (I run a Pass XP10) for dynamics and an even better soundstage.
I rarely play CDs directly anymore. It only takes a few minutes to rip a new CD and scan my music library on my Squeezebox server. A DAC with 2 digital inputs, one for your Squeezebox and one for your CD player, would be ideal, but you may find the CD player is just takes up shelf space after a while.

The Benchmark HDR has 3 spidif (coax) inputs with an analog in.


The Bel Canto in not asynchronous USB (I do not know what async upsampling is).
Their site: "The USB input goes through a 3rd de-jitter stage to insure optimum playback from computer based sources."


I am not sure where to start. Jitter- Duh, that is what I was talking about. The Benchmark is not really async, but adaptive if I understand it. Regardless you contradicted yourself. First you said that transports won't make a difference. Now you are saying "Once you have different transports all bets are off." I agree with the latter, and also agree that there are cheap DVD players that make good transports.


I think that the technical explanation is that my systems are probably more revealing than yours. I mean no offense, but it is pretty bold to think that because you can't hear it, it doesn't exist. You really seem to like your Ohms, and I see you mentioning them frequently in the forums. They have an interesting presentation and a place amongst audiophiles, but I wouldn't say that there are known for being the most revealing speaker out there. My original digital analog "analog" loosely is:

Transport > turntable

DAC > cartridge

Jitter/timing > VTA, azimuth, anti skating etc-Possibly just as elusive, hence the use of atomic clocks for instance.

Sorry for hi jacking this thread...
Dont worry, I am getting a ton of information nonetheless. Just to add a bit more confusion, does anyone have anything to say about the W4S dac-1 which is coming out now? It is said to be better than the stage 4 PS audio DLIII which was indeed very good.....It is also one of my options.

My comments were regarding the Benchmark DAC design and what differences might be heard there and why. My system really is not even relevant to that discussion.

That aside, I think you underestimate the ability of the OHMs to resolve detail for whatever reason. I hear significant differences with the OHMs whenever I change something, including ICs and wires so my experience is not consistent with your assessment.
4 EST, as I said, I don't run the DAC3 at all using usb. For you to nitpick my observations, while simultaneously saying you don't know what asynchronous upsampling means, pretty much makes your comments unhelpful. That said, Bel Canto (and others) make USB asynchronous adapters that convert the USB to SPDIF. There's a review in the latest Stereophile of some new entries on that front. USB is not a great interface due to inherent jitter issues, which has been covered ad nauseum elsewhere. Since the OP asked for advice about DACs and Squeezebox, I offered my firsthand advice on that topic. Go coax or SPDIF if at all possible into your DAC from Squeezebox.
PS, Demianm, I bet the new W4S Dacs are going to be interesting, and a good value with the Sabre 32 bit DAC. I have thought about comparing the DAC2 vs the Bel Canto DAC3...someday!
"The Benchmark is not really async, but adaptive if I understand it"

- You don't. Benchmark resamples data with new asynchronous clock.

"you contradicted yourself. First you said that transports won't make a difference. Now you are saying Once you have different transports all bets are off".

- No I did not. Transports don't make any difference as long as they are "bit transparent". Cheap DVD player might deliver the same performance with Benchmark as expensive transport - not true with other non-jitter rejecting DACs where sound quality is proportional to transport quality (jitter). CDP with dirty or weak laser (or digital volume control or DSP processing etc) might not be "bit transparent". I was merely trying to understand how you can hear -160dB.

I will also restate that spdif cable quality does not make any difference with Benchmark.
Sorry Swanny,

That wasn't meant as a nitpick, but a by the by sort of thing. Async USB is new, and it gets thrown out there a lot. I was just trying to stay accurate with my posts.


It is entirely relevant when you say that you do not understand or can't hear it. Benchmark is a great product, but you do the OP and others a disservice leading them to believe it is perfect or that transports don't matter. They do. A poor transport will cripple any DAC. A poorly set up server can suck too- even if it is bit perfect. Itunes is bit perfect, Pure music sounds better and it is bit perfect too. Presumably Amarra is better yet, but I don't want to spend that much on software. There is complicated math that goes on in the conversion processes that mates with important clocks. It is amazing anything close to music comes out at all. I don't pretend to understand it all, but I have repeatable evidence. Different systems with similar results on various DACs. YMMV

You do no understand async then my friend. For it to be asynchronous the dac would have to control the cpu and it does not. I am not saying it is bad. Async USB is Ayre, dCS, Wavelength and some adapters give or take a few. Firewire is intrinsically async. I don't want to get into a battle over the sounds of cables or whether or not they matter or wheter or not async usb is better than adaptive. It is pointless. If I may use your own words from your virtual system:

"I enjoy the sound a lot, especially Hyperion speakers but my experience in Audio is too limited to make any judgments or claims."

Mine is not. I was around when fancy cables came out. I couldn't hear it then, but I can in today's designs. Quit trying to be an armchair digital engineer by quoting specs. Sonics trump numbers, besides Elias of Benchmark is plenty capable of defending his product (and helpful to boot)!

To get back to your original question, your Rotel is a pretty decent CD player. I haven't heard the DAC Magic but suspect that it's likely a lateral move from the Rotel. Maybe someone who has heard both can comment much more intelligently.

If I were you, I would get a DAC as opposed to a new CD player because it gives you the opportunity to both play CD's (through your Rotel CD player via a digital link to your DAC) and to player computer-based audio. Thus, it's much more versatile, unless you get one of those relatively rare CD players that does have digital inputs.

Now, the question of which DAC to get is much more nebulous. You can search the archives here and you will be given a number of different choices, and that includes the Benchmark and Bel Canto mentioned above, and numerous others.

If you're happy with your W4S amp, then you might consider trying the W4S DAC. They've built a reputation as a high value manufacturer. And they do offer a 30 day money-back guarantee, but with a 15% charge according to their website. So you could at least have the peace of mind that it is returnable (and likely easily sellable on Audiogon) for not too much less than the cost of the DAC. I say go for it.

"For it to be asynchronous the dac would have to control the cpu and it does not."

I'm not sure where you're getting this stuff from. Asynchronous sampling just means that two clocks are not related. I modestly mentioned limited experience in audio but I know electronics.

Read this datasheet: http://www.analog.com/static/imported-files/data_sheets/AD1896.pdf - (upsampler used in Benchmark DAC1) and look at the title. It says "Asynchronous sample rate converter".

"Quit trying to be an armchair digital engineer by quoting specs." - you really puzzled me with this one. Now I don't know what to admire more - your ignorance or your rudeness.
Where am I getting this stuff? I am not pretending to know all of the answers. I even claimed that I didn't know what async upsampling was. You are the one pretending that it is not possible to hear the jitter in the Benchmark or differences in digital cables on various sources as long as it is bit perfect. What is a bit perfect transport? Do you have a spidf bit perfect meter I could borrow or show me the bit perfect specification on a data sheet on a transport. Your claim that a bit perfect player shouldn't matter is useless info because they are all basically bit perfect or a program wouldn't load or DVD wouldn't play. Sure , you can alter the bits with DSP, but that is a different issue. My assertion (and may others mind you) is that digital transmission of music more delicate than you are presenting it as. As for the rudeness and ignorance, you should admire both. I held my tongue considering...
back to the OPs question, I would go for the DAC in his case rather than a new player.
Forget my comments about rudeness or ignorance - I didn't mean to offend you. As for bit perfect or not - it is "bit transparent" and no, I don't have meter but reviewers often verify it. CDP will play scratched CDs because it has very loose checksum checking and error correction and but I was talking about CDP processing alone. Some of them have digital volume control - that would make CDP not bit perfect. As long as output of two different CDPs is the same bit for bit the only difference is jitter. Benchmark supresses jitter and allows to use cheaper CDP and digital cables.
Sufentanil - That was exactly my reasoning. I have now DVD player and Airport Express (computer across the room) connected to Benchmark DAC1. According to Stereophile review Bel Canto DAC3 is better (and has remote) than Benchmark DAC1 but costs 2.5x more.
Kijanki and 4est,

The Benchmark DAC and the Wavelength/Ayre DACs are both asynchronous, and neither is "adaptive." Meaning that in both cases the d/a conversion process is accomplished with a clock that is independent of the computer's clock and that is not periodically "adapted" or re-synchronized to the computer's clock.

However the techniques used in the Asynchronous Sample Rate Converter (Benchmark) approach involve some very small degree of interpolation as part of the jitter elimination process, as mentioned on the first page of the datasheet Kijanki linked to, while the Asynchronous USB (Ayre/Wavelength) approach does not.

Charles Hansen's well-written white paper on the Ayre DAC addresses this further and provides some good additional background.

Under what circumstances and to what degree that interpolation may be audibly significant, I have no idea.

Best regards,
-- Al
Al - pretty good explanation of Benchmarks operation can be found here: http://www.soundonsound.com/sos/jul05/articles/benchmark.htm in chapter "Technology". Of course 50GHz frequency is not real and involves mathematical manipulations.
Thank you Al for the clarification. The Benchmark is a great DAC and can be very revealing- worth the trouble. After experimenting with one for the better part of three years, I keep finding ways to get more out of it, hence my initial post about transports mattering. I purchased a hiFace usb>spdif for hi rez (async type like ayre requiring drivers) computer use and noticed it took the usb up a notch. This lead me to experimenting with different connections on it including using my Weiss DAC as a firewire to spdif converter. The only reason I posted in the first place was that if I hadn't tried different transports (and then servers), I would have sent my first Benchmark back. The first transport I tried was an Oppo DVD player that I was looking to replace the DAC section w/. Well, that Oppo (the original), and my current Oppo (983) both suck as tranports, and yet are very servicable with video and as cd players. The same thing with an higher end 50# copper clad SACD player. Of the five transports I tried, 3 were "dull", and the best was a $40 cheapie DVD from Best Buy with an awful picture that lasted 6 months before it quit. I have a hard time imagining that only one was "bit transparent". I also know that it isn't just galvanic isolation as I have tried optical USB and it sounds different than toslink from the same server. In all of these I never had any clicks, spits or sputters (until I tried Pure Music- another story).

The Benchmark was a ground breaking product in my mind, but is not infallible. It can be sensitive to power and its input- both of which Benchmark claims isn't. Ultimately what does one expect for $1000 in this hobby anyway?
Regarding Sufentinal's comment:

"...unless you get one of those relatively rare CD players that does have digital inputs."

This is why I recommended the Cambridge 840. It has an excellent transport, power supply and chassis, better DAC's and electronics than the DacMagic, and the ability to play disks or serve as an upsampling DAC for external transports. In it's price range, it is a "relatively rare CD player" in terms of sound quality, versatility and value.

The company has been on a roll lately, and my guess is that they will be coming out with the successor to the 840 soon, probably called the "850". So the 840's may be available new at discounted prices soon, and a new and likely more capable player will also available to purchase at a price point near the previous model.

To provide one person's opinion on the DacMagic versus the DAC performance in the Rotel 1072 - there is no comparison. The DacMagic is much more musical, with better PRAT and overall resolution. Not a horizontal moove in my opinion.

That said, the Cambridge 840 is more natural and "analog" sounding than either.
4est is correct; the Benchmark DAC1 I reviewed was sensitive to cables and transports. I have never encountered a component which is immune to cable influences.
The cambridge 840C is also one of the options I had considered, I have both a better CD player than the Rotel and a "seemingly very good" DAC looking at the specs, the chipset used and the fact that it also upsamples the signal. Considering the selling price of my almost new Rotel, it would be a very little amount of money to get an used 840C, maybe 500 bucks. To get a good DAC I am looking at 1000 bucks and still will have a "questionable" CD transport, so I am looking at 1000 bucks for the DAC and and possible CD transport change. Now it´s where the rule of diminishing returns hits me hard in the head and I question how much of a difference could I notice between the 840C and a transport with a DAC such as the benchmark or the wyred4sound one.....

Some people have pointed out that the 840C is a little thin sounding, but I also plan to experiment with a tube buffer, last week I ordered the Grant B283 MKII buffer just for kicks (to use it between the CD player or the DAC, whateven I should choose)

On the other hand, I am thinking that for only 500-600 bucks I could upgrade my Rotel in the meantime, use the 840C as standalone player and later get a better DAC and use the 840C 24bit upsampled output to the new DAC.....does this sound reasonable?
The Cambridge did not sound thin to me when I heard it. That might be partially due to the quality overall of the system I heard it on. It impressed me as much as any player I have heard.

It fruitless to assert that all source devices are created equal in any case. That's not what I'm saying.

Timing can be an issue with transports or optical drives reading optical media in real time fast enough to apply error correction and keep the bits flowing. Music servers that avoid real time reading of optical media to provide the audio datastream and devices like network music players that cache data locally before sending to the DAC help assure that all bits available from the source media get captured and sent in time.

In the case of the BEnchmark, I suspect that there are many source devices at all price points that sound essentially the same, more so than otherwise. Which ones specifically do and which ones do not? Try specific units and find out I suppose.

But I think I would apply the strategy that I would not assume I need to spend a lot or even more in order to find a suitable source with the Benchmark specifically as perhaps might be more the case with some other DACs.
Are you in a position to go listen to the 840 and compare it to other CDPs or DACs? It didn't sound "thin" to me.

Did you hear the 840C? I dont know if it is really thin, I just read that from some other review but also doubt it sounds "thin" considering the great reviews for it.

I could listen to it, but the dealers where I´ve seen it do not have a good listening room to really be able to notice any difference, it´s like listening to it at bestbuy, no good at all. I´d rather jump into the pool and listen to it at home.....
Yes, I have heard it. It is very good at its price point. For comparison, most cheaper Cambridge players have a forward sonic balance with fairly flat frequency curve, meaning that bass and treble are not attenuated in favor of mid range tones, as some other manufacturers of budget players prefer to do (NAD comes to mind here). The other Cambridges tend to put their empahsis on resolution and timing over soundstage width and depth - so you get musical details and tunefulness, but not necessarily the best "hiFi" effects. The DacMagic is at the very musical end of these characters, having essentially the DACs and upsampling electronics from the CA 740C, which is one model and several hundred bucks down from the 840C.

For comparison with the player you are familiar with, I find the Rotel sound to be fairly accurate and flat in frequency resoponse, but somewhat sterile or dry (lifeless is probably too strong a term) overall in comparison to say, the Cambridge 640C, the Dac Magic or 740C.

The 840C on the other hand to me sounds warmer and more liquid through the mid range than its littler siblings. It keeps the tunefullness and resolution, but wraps it in a tropical breeze, and does a better job of building a believable soundstage. Soundstage size is still not top o0f class for the Cambridge compared to other players in its price range, but the other things it does well more than compensate in my mind.

The reason for the leap in capability for the 840 over its sibilings probably results from the fact that it uses a completely different set of DACs and I believe different algorithms for upsampling among other upgrades. In any case, your amp has balanced inputs, and if you fit the 840C with good balnaced ICs and a capable and compatible power cord like the Analysis Plus Power Oval 2 or Shunyata Diamondback Platinum or better, you can maximize the positive traits (including expanded soundstage) for both CD playback and use of the 840 as a stand alone DAC serving other sources. I think you would enjoy it.
I heard the 840c on ARC pre-amp and Rogue poweramp running PSB Synchrony speakers. The results were top notch, one of the best I have heard.

I've heard DCS and other high end units on other systems. The sound in each case was different but top notch still. The system the player reides in, particularly amp, speakers and room obviously are major factors in determining the resulting sound. My conclusion was I could live easily with teh Cambridge in the right system for me with no reservations if needed. I would not look at the way more expensive units without some kind of direct a/b comparison possible which I have not heard.

Which other CD player except the Raysonic would you recommend at this price point ($1000)?

Does anyone happen to know if it is possible to use the 840C with both outputs (unbalanced and balanced) wired to the amp? I would like to be able to stuff my tube buffer in between the amp and CD/DAC to be able to improve the sound of MP3 playback, but use the balanced output for regular CD playing.
That has the capability to act as an external DAC for other sources if necessary, or just a good $1000 +/- CD player?
For just as good, I´d rather get the 840C. I mean a really better one!
??? A really better one? The 840C is better. Nothing for $1000 is better, or likely just as good.

The DACs I directed you to earlier are "really better" than the Rotel, but they cost $2K to $2.5K.

I can think of other CD players for about $1500 that are better than the Rotel - like the Naim CD5i or the Moon CD1, but in about the same league or perhaps not as sophisticated as the 840C.

"Really better" for $1000 is "really" tough.

Good luck
You might get a different and perhaps "better" sound with different sounding DACs at that price but nothing clearly better in a player IMHO.

I have been very satisfied with both mhdt DACs I run ( SS constantine and tubed Paradisea). These can be had used for much less than a new 840c. They both work extremely well with the all the modest sources I drive them with ($400 Marantz DVD, $600 Denon recorder/player, Roku Soundbridge network players).

The differences in sound between SS and tube mhdt DACs are quite significant, as are the tube Paradisea with different tubes. Which one sounds best would largely be determined by personal taste, the rest of the system and to some extent possibly kinds of music listened to.

Just so there is no confusion, I think the Cambridge 840 is one of the best sounding and most flexible CD players available for the money - any criticisms leveled are in the absolute sense comparing it to reference players costing many dollars more.