Sabaj D5 vs. SMSL M500 vs. Topping DX7 Pro: measurement comparison

[RickSanchez]

1. you can use just rca cables right. If you use the dx7 pro only (without seperate amp) you don't have balanced connection aswell
2. Can't you use in this case just the volume knob of the DAC and set the amp to full volume?

1. Yes. But depending on your setup / listening area you may have ground loop issues using RCA interconnects.
2. Yes.

 

[VeerK]

Been doing some of my own reading and I’m wondering if anyone else feels like how I do regarding the M500 vs DX7PRO. I’m thinking that if you want a pure DAC, go with the M500, and if you want a combo unit go with the DX7PRO.

Also, can anyone share the technical definitions of “fake balanced” in the M500. I’m at work so I can’t really look it up right now, apologies if it’s been covered in depth. Will look into the discussions about it more at home.

All things being equal, the lower cost of the M500 makes sense for people like me who already have the THX789 and want a transparent balanced DAC with equally “endgame” performance.

 

[SIY]

Also, can anyone share the technical definitions of “fake balanced” in the M500. I’m at work so I can’t really look it up right now, apologies if it’s been covered in depth. Will look into the discussions about it more at home.

"Fake balanced" is not a technical term, it's a (dishonest) marketing term. Something is either balanced (equal impedances on each leg) or not, there's no "fake" exceptions.

 

[VeerK]

"Fake balanced" is not a technical term, it's a (dishonest) marketing term. Something is either balanced (equal impedances on each leg) or not, there's no "fake" exceptions.

That’s why I’m confused why it’s being associated with the M500. Amir’s measurements wrt to the XLR outputs didn’t indicate anything of concern to me, so I don’t know why M500 has posts accusing it. I was all set to purchase the M500 on 11.11, partially to support a product that handled the ESS hump so beautifully.

 

[SIY]

I don’t know why M500 has posts accusing it.

Some mix of dishonesty and ignorance.

 

[VeerK]

Some mix of dishonesty and ignorance.

Thank you for clearing that up

 

[RickSanchez]

All things being equal, the lower cost of the M500 makes sense for people like me who already have the THX789 and want a transparent balanced DAC with equally “endgame” performance.

If you have a TOTL headphone amp like the THX 789 then buying any one of these three DACs would make for a incredibly transparent stack. Their DAC measurements indicate that all are well-designed and high performing.

After that you look at price, features, build quality, aesthetics, etc. of the DAC. If -- in your case -- price is the key factor then the M500 definitely makes sense.

 

[VeerK]

If you have a TOTL headphone amp like the THX 789 then buying any one of these three DACs would make for a incredibly transparent stack. Their DAC measurements indicate that all are well-designed and high performing.

After that you look at price, features, build quality, aesthetics, etc. of the DAC. If -- in your case -- price is the key factor then the M500 definitely makes sense.

Agreed, it’s basically nitpicking at this point. I don’t love the idea of variable performance based on warmup time on the Sabaj, so it’s always been the M500 vs DX7PRO for me. And with such amazing performance between the two, I was just trying to figure out what reason someone like me with the 789, would buy the DX7PRO. There’s a sale on Drop right now for $520 on the DX7PRO so I am very tempted to grab it, but if there aren’t any purely DAC based advantages, it makes more sense to get the M500 for $350 and just pocket the extra cash for a nice night out with my wife.

 

[Shoaibexpert]

In regards to my note about the filters/FFT measurements, member @Bjoran has provided some new info. The original manual for the DX7 Pro -- which is what Amir used -- listed the filters in the wrong order. This has now been corrected by Topping and there is a new manual.

I don't think Amir has the time to fix the measurement screenshot so please review the original FFT measurements carefully for the DX7 Pro and make the corrections on your own using the new manual ... if that's important to you.

Which filter sounds the best and analogue/musical to people here?

 

[Shoaibexpert]

just pocket the extra cash for a nice night out with my wife.

Sounds like a plan...spend your money where it may make a difference... coz I don't think both DACs are audibly different (based on experts opinions on ASR - me not being one of them) Plus, u said you had the 789...that has a SNR of 116...the M500 has an SNR of 117...which means the additional 2db SNR of the DX7 Pro (119db) wouldn't be audibly possible with the 789 at it crosses the ceiling of the amp SNR - at least that's how I think audio chain works. I've got the same amp and an Arya...went with the M500 via Canare balanced cables...upgraded from SU8 (which I sold recovering 65% of what I paid for it...so the SMSL keeps it's value)...should have a blast as soon as it arrives!

For most people here the key to getting the best value is to negotiate the price with reliable vendors like Aoshida HiFi or Shenzhen Audio outside AliExpress or eBay (say on emails) and pay them via PayPal or in Yuan...you will get a price lesser than the $350...don't remember to ask them to adjust the customs value to avoid tarrif and other costs...I hope this doesn't break any rules on ASR

 

[NielsMayer]

No, @NielsMayer is incorrect in his understanding of what "balanced" means. If source impedances are the same for + and -, an output is balanced, regardless of voltage measurements to some other arbitrary point. This is a very common error.

No @SIY , I'm perfectly correct in my understanding of what "balanced" means, and i provided extensive literature explaining what balanced really means in previous posts. There is a difference between "impedance balanced" and "balanced" which is why "fake balanced" can act like balanced in special cases. Issues of symmetry and mismatch between outputs affect many aspects other than noise, such as crosstalk or headroom. "Impedance balanced" just affects noise....

https://www.jhbrandt.net/wp-content...igh-Performance_Balanced_Audio_Interfaces.pdf

THE BALANCED INTERFACE The purpose of a balanced audio interface is to efficiently transfer signal voltage from driver to receiver while rejecting ground noise. Used with suitable cables, the interface can also reject interference caused by external electric and magnetic fields acting on the cable. The true nature of balanced interfaces is widely misunderstood. For example “Each conductor is always equal in voltage but opposite in polarity to the other. The circuit that receives this signal in the mixer is called a differential amplifier and this opposing polarity of the conductors is essential for its operation.” [3] This, like many explanations in print (some in otherwise respectable books), describes signal symmetry – “equal in voltage but opposite in polarity” – but fails to even mention the single most important feature of a balanced interface. SIGNAL SYMMETRY HAS ABSOLUTELY NOTHING TO DO WITH NOISE REJECTION — IMPEDANCE IS WHAT MATTERS! A good, accurate definition is “A balanced circuit is a two-conductor circuit in which both conductors and all circuits connected to them have the same impedance with respect to ground and to all other conductors. The purpose of balancing is to make the noise pickup equal in both conductors, in which case it will be a common-mode signal which can be made to cancel out in the load.” [4] The impedances, with respect to ground, of the two lines is what defines an interface as balanced or unbalanced. In an unbalanced interface, one line is grounded, making its impedance zero. In a balanced interface, the two lines have equal impedance. It’s also important to understand that line impedances are affected by everything connected to them. This includes the line driver, the line or cable itself, and the line receiver. The line receiver uses a differential amplifier to reject common-mode voltages. The IEEE Dictionary defines a differential amplifier as "an amplifier that produces an output only in response to a potential difference between its input terminals (differential-mode signal) and in which output due to common-mode interference voltages on both its input terminals is suppressed." [5] Since transformers have intrinsic differential response, any amplifier preceded by an appropriate transformer becomes a differential amplifier. The basic theory of the balanced interface is straightforward. (For purposes of this discussion, assume that the ground reference of Device A has a noise voltage, which we will call "ground noise," with respect to the Device B ground reference.) If we look at the HI and LO inputs of Device B with respect to its ground reference, we see audio signals (if present) plus the ground noise. If the voltage dividers consisting of Zo/2 and Zcm on each of the lines have identical ratios, we’ll see identical noise voltages at the two inputs of Device B. 3 Since there is no difference in the two noise voltages, the differential amplifier has no output and the noise is said to be rejected. Since the audio signal from Device A generates a voltage difference between the Device B inputs, it appears at the output of the differential amplifier. Therefore, we can completely rejects the ground noise if the voltage divider ratios are perfectly matched. In the real world, we can’t perfectly match the voltage dividers to get infinite rejection. But if we want 120 dB of rejection, for example, we must match them to within 0.0001% or 1 part per million! The ground noise received from Device A, since it exists on or is common to both wires, is called the common-mode voltage and the differential amplifier provides common-mode rejection. The ratio of differential or normal-mode (signal) gain to the common-mode (ground noise) gain of the interface is called the common-mode rejection ratio or CMRR (called "longitudinal balance" by telephone engineers) and is usually expressed in dB. There is an excellent treatment of this subject in Morrison's book. [6] If we re-draw the interface as shown here, it takes the familiar form of a Wheatstone bridge. The ground noise is “excitation” for the bridge and represented as Vcm (common-mode voltage). The common-mode impedances of the line driver and receiver are represented by Rcm+ and Rcm!. W hen the + and ! arms have identical ratios, the bridge is “nulled” and zero voltage difference exists between the lines — infinite common-mode rejection. If the impedance ratios of the two arms are imperfectly matched, mode conversion occurs. Some of the ground noise now appears across the line as noise. The bridge is most sensitive to small fractional impedance changes in one of its arms when all arms have the same impedance. [7] It is least sensitive when upper and lower arms have widely differing impedances. For example, if the lower arms have infinite impedance, no voltage difference can be developed across the line, regardless of the mis-match severity in upper arm impedances. A similar scenario occurs if the upper arms have zero impedance. Therefore, we can minimize CMRR degradation due to normal component tolerances by making common-mode impedances very low at one end of the line and very high at the other. [8] The output impedances of virtually all real line drivers are determined by series resistors (and often coupling capacitors) that typically have ±5% tolerances. Therefore, typical line drivers can have output impedance imbalances in the vicinity of 10 Ù. The common-mode input impedances of conventional line receivers is in the 10 kÙ to 50 kÙ range, making their CMRR exquisitely sensitive to normal component tolerances in line drivers. For example, the CMRR of the widely used SSM-2141 will degrade some 25 dB with only a 1 Ù imbalance in the line driver. Line receivers using input transformers (or the InGenius IC discussed later) are essentially unaffected by imbalances ® as high as several hundred ohms because their common-mode input impedances are around 50 MÙ — over 1000 times that of conventional “active” receivers. Note that this discussion has barely mentioned the audio signal itself. The mechanism that allows noise to enter the signal path works whether an audio signal is present or not. Only balanced impedances of the lines stop it – signal symmetry is irrelevant. W hen subtracted (in the differential amplifier), asymmetrical signals: +1 minus 0 or 0 minus !1 produce exactly the same output as symmetrical signals: +0.5 minus !0.5. This issue was neatly summarized in the following excerpt from the informative annex of IEC Standard 60268-3: “Therefore, only the common-mode impedance balance of the driver, line, and receiver play a role in noise or interference rejection. This noise or interference rejection property is independent of the presence of a desired differential signal. Therefore, it can make no difference whether the desired signal exists entirely on one line, as a greater voltage on one line than the other, or as equal voltages on both of them. Symmetry of the desired signal has advantages, but they concern headroom and crosstalk, not noise or interference rejection.

However, there may be confusion remaining from the historical legacy of 600ohm balanced telephony transmission lines. Audio equipment no longer has a 600 ohm input impedance and works in "voltage mode" as opposed to "power mode"....

HISTORY AND 600 Ù LINES The first widespread users of balanced circuits were the early telephone companies. Their earliest systems had no 4 amplifiers yet needed to deliver maximum audio power from one telephone to another up to 20 miles away. It’s well known that, with a signal source of a given impedance, maximum power will be delivered to a load with the same, or matched, impedance. It is also well known that "reflections" and "standing waves" will occur in a transmission line unless both ends are terminated in the line’s characteristic impedance. Because signal propagation time through 20 miles of line is a significant fraction of a signal cycle at the highest signal frequency, equipment at each end needed to match the line’s characteristic impedance to avoid frequency response errors due to reflections. Telegraph companies used a vast network of had a huge installed base of open wire pair transmission lines strung along wooden poles. Early telephone companies arranged to use these lines rather than install their own. Typical lines used #6 AW G wire at 12 inch spacing and the characteristic impedance was about 600 Ù, varying by about ±10% for commonly used variations in wire size and spacing. [9] Therefore 600 Ù became the standard impedance for these balanced duplex (bi-directional) wire pairs and subsequently most telephone equipment in general. Not only did these lines need to reject ground voltage differences, but the lines also needed to reject electric and magnetic field interference created by ac power lines, which frequently ran parallel to the phone lines for miles. Balanced and impedance matched transmission lines were clearly necessary for acceptable operation of the early telephone system. Later, to make “long distance” calls possible, it was necessary to separate the duplexed send/receive signals for unidirectional amplification. The passive "telephone hybrid" was used for the purpose and its proper operation depends critically on matched 600 Ù source and load impedances. Telephone equipment and practices eventually found their way into radio broadcasting and, later, into recording and professional audio — hence, the pervasive 600 Ù impedance specification. In professional audio, however, the goal of the signal transmission system is to deliver maximum voltage, not maximum power. To do this, devices need low differential (signal) output impedances and high differential (signal) input impedances. This practice is the subject of a 1978 IEC standard requiring output impedances to be 50 Ù or less and input impedances to be 10 kÙ or more.[10] Sometimes called "voltage matching," it minimizes the effects of cable capacitances and also allows an output to simultaneously drive multiple inputs with minimal level losses. With rare exceptions, such as telephone equipment interfaces, the use of matched 600 Ù sources and loads in modern audio systems is simply unnecessary and compromises performance.

In other words, please don't be confused about "impedance balanced" versus actual balanced operation.

 

[NielsMayer]

"Fake balanced" is not a technical term, it's a (dishonest) marketing term. Something is either balanced (equal impedances on each leg) or not, there's no "fake" exceptions.

Wrong again bob!

"Fake balanced" is an anti-marketing term -- and the truth. The dishonest marketing term is "impedance balanced" which gets rid of the anti-marketing truthiness of "fake" and instead tells half-the story by calling it "impedance balanced".

Again, I provided links in previous posts about this topic and how "impedance balanced" is actually a marketing term used to cover up the use of a resistor going to ground on the (-) pin instead of an active output. It's from Mackie marketing specifically:

https://web.archive.org/web/2002061...ndweb.com/install/sac/n27/rick/balout-1.shtml

The people at Mackie Designs have played with a scheme that they called “Impedance Balanced” for phone jack outputs. At first, I gasped and sputtered. But after thinking about the problem, I’ve come to embrace this as the best solution, especially when you don’t have dual output connectors.

Simply stated, you use a TRS jack for the connector, and you connect the ring contact to ground through the same resistance as used for the build-out resistor on the tip contact. From the line’s standpoint, the circuit is balanced, driven from a grounded center-tap source. Yes, only half of the circuit is driven from an audio stand point. It doesn’t matter what sort of plug the user inserts, either way is optimum for that case. For the unbalanced case, no ground current flows.

From the balanced input’s standpoint, there are equal impedances from both sides of the line to ground. This results in the best common-mode rejection (CMR) performance, especially if close-tolerance resistors are used at the output side. The circuit noise is actually 3 dB less than either of the other two balanced line drivers. There is no level difference if used with a balanced or unbalanced input.

Note again the assertion that "fake balanced" gives 3dB less circuit noise than "real balanced" because it halves the noise and distortion of the output op amps (aka 3dB).

 

[SIY]

If impedances are balanced, it's balanced. If they're not, it's not. Period.

You can find the references to Bill Whitlock and Bruno Putzey's papers on this all over the forum. AFAIK, neither one works for Mackie.

 

[NielsMayer]

Some mix of dishonesty and ignorance.

How do you explain the large 5dB improvement in dynamic range on XLR on the M500

• Dynamic range:
  • XLR 125dB
  • RCA: 120dB
• (can't find XLR vs RCA SNR readings).

On DX7Pro, only 4dB improvement in SNR

RCA SNR 122dB
XLR SNR 126dB

FYI, I now have the DX7Pro and IMHO it sounds better than the M500, though similar (feeding Hypex NC400 -> ELAC Adante AS61)

That "absolute sound" article about midrange or nonlinearity of FR in M500 vs DX7PRO is IMHO a mis-analysis, but based on a clearly perceptible difference. They're both equally flat sounding but there is something more "interesting" sounding about the midrange on the DX7PRO whereas the M500 sound was "not as exciting"... At least for listening to regular 44.1/48 material off Tidal. Maybe it's the lower jitter specs of the DX7Pro, given all other things are mostly equivalent.

MQA on the M500 sounded really nice but i got the impression that the sound was being hyper-enhanced above and beyond any bitrate or sample depth extensions -- as if MQA internally used that cheap effect of "audio separation expansion" like you see on boomboxes .... But for the majority of audio material (whch is not MQA) regular sound files sounded better and more engaging and more interesting through the DX7Pro than the M500. YMMV.

My main complaint about the DX7Pro is that the display is very dim, compared to the DX7S it is replacing (which now becomes my bedroom DAC).

Anybody else notice the display is dim on the DX7Pro, or is my unit defective??

 

[NielsMayer]

If impedances are balanced, it's balanced. If they're not, it's not. Period.

You can find the references to Bill Whitlock and Bruno Putzey's papers on this all over the forum. AFAIK, neither one works for Mackie.

WRONG AGAIN BOB!

https://www.rane.com/note165.html

 

[Shoaibexpert]

No @SIY , I'm perfectly correct in my understanding of what "balanced" means, and i provided extensive literature explaining what balanced really means in previous posts. There is a difference between "impedance balanced" and "balanced" which is why "fake balanced" can act like balanced in special cases. Issues of symmetry and mismatch between outputs affect many aspects other than noise, such as crosstalk or headroom. "Impedance balanced" just affects noise....

https://www.jhbrandt.net/wp-content...igh-Performance_Balanced_Audio_Interfaces.pdf



However, there may be confusion remaining from the historical legacy of 600ohm balanced telephony transmission lines. Audio equipment no longer has a 600 ohm input impedance and works in "voltage mode" as opposed to "power mode"....



In other words, please don't be confused about "impedance balanced" versus actual balanced operation.

@SMSL-Mandy and @SMSL_Liu any technical help would be appreciated on your M500 around this topic. In essence, most people would want to know if M500 is fake balanced, would it theoretically impact the performance adversely...(noting that none of this would be audible for most)

 

[SIY]

The improvement is irrelevant to the question of balanced versus unbalanced and everything to do with implementation.

 

[SIY]

WRONG AGAIN BOB!

https://www.rane.com/note165.html

The Pin 1 problem is also irrelevant to what balanced means. It is an implementation fault.

 

[Shoaibexpert]

I now have the DX7Pro and IMHO it sounds better than the M500

Oh! The objectivity patrol on ASR should take notice of this...throw in those 'All DACs sound the same' comments...

 

[Shoaibexpert]

How do you explain the large 5dB improvement in dynamic range on XLR on the M500

• Dynamic range:
  • XLR 125dB
  • RCA: 120dB
• (can't find XLR vs RCA SNR readings).

On DX7Pro, only 4dB improvement in SNR

RCA SNR 122dB
XLR SNR 126dB

FYI, I now have the DX7Pro and IMHO it sounds better than the M500, though similar (feeding Hypex NC400 -> ELAC Adante AS61)

That "absolute sound" article about midrange or nonlinearity of FR in M500 vs DX7PRO is IMHO a mis-analysis, but based on a clearly perceptible difference. They're both equally flat sounding but there is something more "interesting" sounding about the midrange on the DX7PRO whereas the M500 sound was "not as exciting"... At least for listening to regular 44.1/48 material off Tidal. Maybe it's the lower jitter specs of the DX7Pro, given all other things are mostly equivalent.

MQA on the M500 sounded really nice but i got the impression that the sound was being hyper-enhanced above and beyond any bitrate or sample depth extensions -- as if MQA internally used that cheap effect of "audio separation expansion" like you see on boomboxes .... But for the majority of audio material (whch is not MQA) regular sound files sounded better and more engaging and more interesting through the DX7Pro than the M500. YMMV.

My main complaint about the DX7Pro is that the display is very dim, compared to the DX7S it is replacing (which now becomes my bedroom DAC).

Anybody else notice the display is dim on the DX7Pro, or is my unit defective??

I wanna ask if you have any linearity (vs Frequency) response curve to prove that M500 mid range is recessed...