CHORD M-Scaler Review (Upsampler)

[amirm]

OK, I spent a bit of time and figured out the incompatibility problem with Topping D70s. Here is the eye pattern of the Audio Precision APx555 S/PDIF output (configured for "consumer" levels):

At the bottom the scope has computed the peak to peak voltage of about 0.55 volts. S/PDIF spec calls for 0.5 to 0.6 volt so we are compliant.

Now see what happens when I measure the output of M-Scaler:

It is three times higher at 1.84 volts! Yes, this is terminated into 75 ohm on the scope side.

The much higher output voltage is overdriving the Topping D70s S/PDIF input causing problems. Some DACs will not care (I tested Gustard and it was fine). But others will not.

This is a clear compatibility problem. Here are the allowable levels for professional and consumer serial digital audio: https://en.wikipedia.org/wiki/S/PDIF

1.84 volts exceeds that of AES3 unbalanced. For balanced, it is actually too low. So it doesn't comply with anything.

I went to look to see if Chord documents output levels and this is all I find under specifications:

Can you believe it? The type of FPGA is more important than mentioning output levels for the digital output! Marketing fluff over needed specs.

Conclusion
The Coax output levels of M-Scaler are not in compliant with any standard and will surely overdrive some number of consumer DACs as I have shown with Topping D70s. Company needs to provide a voltage divider for better compatibility with third-party DACs. Until then, it needs to warn users both in website and manual about this issue.

EDIT: I simulated the same high output voltage with my analyzer and it doesn't impact the D70s. So level is not the problem.

 

[voodooless]

Looks like it’s taken straight from the 1.8V logic…

 

[Blumlein 88]

OK, I spent a bit of time and figured out the incompatibility problem with Topping D70s. Here is the eye pattern of the Audio Precision APx555 S/PDIF output (configured for "consumer" levels):

View attachment 217230

At the bottom the scope has computed the peak to peak voltage of about 0.55 volts. S/PDIF spec calls for 0.5 to 0.6 volt so we are compliant.

Now see what happens when I measure the output of M-Scaler:
View attachment 217231

It is three times higher at 1.84 volts! Yes, this is terminated into 75 ohm on the scope side.

The much higher output voltage is overdriving the Topping D70s S/PDIF input causing problems. Some DACs will not care (I tested Gustard and it was fine). But others will not.

This is a clear compatibility problem. Here are the allowable levels for professional and consumer serial digital audio: https://en.wikipedia.org/wiki/S/PDIF

View attachment 217232

1.84 volts exceeds that of AES3 unbalanced. For balanced, it is actually too low. So it doesn't comply with anything.

I went to look to see if Chord documents output levels and this is all I find under specifications:
View attachment 217233

Can you believe it? The type of FPGA is more important than mentioning output levels for the digital output! Marketing fluff over needed specs.

Conclusion
The Coax output levels of M-Scaler are not in compliant with any standard and will surely overdrive some number of consumer DACs as I have shown with Topping D70s. Company needs to provide a voltage divider for better compatibility with third-party DACs. Until then, it needs to warn users both in website and manual about this issue.

Maybe this sounds better than standards compliance at -300 db.

 

[Jimbob54]

OK, I spent a bit of time and figured out the incompatibility problem with Topping D70s. Here is the eye pattern of the Audio Precision APx555 S/PDIF output (configured for "consumer" levels):

View attachment 217230

At the bottom the scope has computed the peak to peak voltage of about 0.55 volts. S/PDIF spec calls for 0.5 to 0.6 volt so we are compliant.

Now see what happens when I measure the output of M-Scaler:
View attachment 217231

It is three times higher at 1.84 volts! Yes, this is terminated into 75 ohm on the scope side.

The much higher output voltage is overdriving the Topping D70s S/PDIF input causing problems. Some DACs will not care (I tested Gustard and it was fine). But others will not.

This is a clear compatibility problem. Here are the allowable levels for professional and consumer serial digital audio: https://en.wikipedia.org/wiki/S/PDIF

View attachment 217232

1.84 volts exceeds that of AES3 unbalanced. For balanced, it is actually too low. So it doesn't comply with anything.

I went to look to see if Chord documents output levels and this is all I find under specifications:
View attachment 217233

Can you believe it? The type of FPGA is more important than mentioning output levels for the digital output! Marketing fluff over needed specs.

Conclusion
The Coax output levels of M-Scaler are not in compliant with any standard and will surely overdrive some number of consumer DACs as I have shown with Topping D70s. Company needs to provide a voltage divider for better compatibility with third-party DACs. Until then, it needs to warn users both in website and manual about this issue.

Good work! So its a dud another way.

 

[Jimbob54]

Well... it was clear where the discussion was going to go and for what reason that person was here.


JSmith

Urgh- that video is horrible. Yes- it can upscale streaming (true) SO it improves sound (eh?)

 

[JSmith]

Urgh- that video is horrible.

Yep... and apparently FLAC is not good enough either.


JSmith

 

[sq225917]

So the spdif output is out of spec, which is a shame as the eye pattern is blindingly good otherwise. The M2Tech Hiface similarly had a high spdif output years ago and caused issues with some dacs, remedied with a couple of resistors.

Props for going the extra mile and digging deeper Amir, otherwise some might have laid the blade at Toppings door.

 

[Jomungur]

My post wasn't really directed at you, more so to few individuals who are likely trolling looking at their post history. I'm glad you went through the trouble of sending stuff to Amir, the more gear he can test the better for us all.

Oh and I'm not really against people using their money however they please, I won't call anyone an idiot because he bought some fancy cables for example. I'm more against BS claims in spite of evidence, they should be rightfully dismantled.

No worries, thanks for clarifying. And agreed on the BS part.

 

[mccririck]

Every digital file we listen to has a mix of resolutions. The reason is that the sample rate - whether 44.1kHz, 96kHz, or any other sample rate - is constant, while the frequency of sounds in a musical recording varies.

So for example, a kick drum at around 50Hz is always going to get sampled 10 times as much as a guitar strum or vocal harmonic at 500Hz, and 100 times as much as a high hat at 5kHz.

Digital sampling theory tells us that a sound has to be sampled only twice in order to be properly reconstructed - this is not a perceptual truth or a "most of the time" truth but rather a mathematical truth that makes digital sound possible at all (in cell phone and many other applications, not just hi-fi audio). That's what the Nyquist frequency is: the highest frequency that can be accurately encoded by a given sample rate, and that's why Nyquist is 1/2 the sample rate - you need to sample every frequency twice.

So if sampling a frequency more than twice improved fidelity, then we'd all have easily heard it 1000s of times in every digital recording, because the lower frequencies would be "higher res" and "more refined" than the higher frequencies. That's not what any of us experiences, though, and we don't experience that because digital sampling theory is correct.

There's a lot wrong with your post there. Human ears are less sensitive to lower frequencies. Amir has talked about this in one of his videos.

When you say "a sound has to be sampled only twice in order to be properly reconstructed" I don't think you understand what "properly reconstructed" means. It does not mean it's going to be perfect.

I'm not a sound engineer or electrical engineer and I know those two things. Don't be ignorant.

 

[Purité Audio]

Thanks Amir, ASR is a veritable beacon.
Keith

 

[Jomungur]

I really don't get arguments about money. Businesses will charge what they can or want, and I presume most people here are over 18 and are grown up enough to make sensible decisions based on their financial resources. It's good that there are now so many sites like ASR and all the headphone sites that check that the budget stuff is, or is not, ft for purpose. The M-Scaler is not intended for people on a tight budget, and clearly it was an inconsequential cost (3-5% of his budget) to the person who loaned it to ASR. I wear very cheap, old clothes and very expensive shoes. Don't know why, but makes me happy. If I were to spend $12,000 on a cable, I would only have myself to blame. Rest assured, I haven't, and won't.

Strangely, a couple of weeks ago I was at a wedding siting next to a homeopath. She had worked as a fully qualified medical doctor for 10 years and, dealing with her own health, homeopathy helped her where conventional medicine didn't. She also practised existential psychology and a few other things. I must admit she was as fit as a flea and looked 10+ years younger than she probably was. I listened to her with interest, without being judgemental.

I think the money factor is important. An industry where you can make large mark-ups on products with little accountability attracts frauds, hucksters and even just well meaning inventors/designers with bad ideas.

It's an open secret that almost no one pays list price with these high end products. So it's hard to know what the mark-ups are, but that just adds to the overall murkiness of the industry. One of the dealers I've gotten with has no engineering background at all, and says some strange things (music sounds better at night because of the lower usage of the power grid), but he is very well connected and seems to get better discounts than others on the same products.

But yes, agreed, about being non-judgmental until you actually experience or test something. I'm not immediately skeptical of homeopathy, incidentally. It's used quite a lot in certain European countries. The problem with assessing health and well-being, though, is that there so many factors it's really difficult to ascertain what's causing what. Too many possibilities for confounding and collider bias. So with the person you met it could be the homeopathy, or it could be her genes, or a combination, etc. You can only sort this out with widescale studies.

Audio has a similar issue. If the best we can do when assessing products is X "sounds better" than Y, then it's hopeless. Might as well argue whether vanilla ice cream is better than chocolate. Objective measurements don't capture everything, but if you can to a point where you can subjectively distinguish how products sound based on their objective measurements, then that's a big step. When I look at the products Amir has reviewed, I don't put much value in the listening tests, although they can be interesting. What I do is scan the measurements of the products I know I like (like many Hifiman headphones) and see what they have in common. That helps educate me on the measurements I might pay attention to for my next purchase.

 

[pma]

More in deep

https://www.st.com/resource/en/application_note/an5073-receiving-spdif-audio-stream-with-the-stm32f4f7h7-series-stmicroelectronics.pdf

 

[A Surfer]

Well... it was clear where the discussion was going to go and for what reason that person was here.


JSmith

I get that, and clearly it isn't my place to decide what should be tolerated here, that is up to the host and staff, but I still think as long as possible, people should talk. They say that you don't make peace with your friends, rather your enemies, and for me there is a similarity here. The enemy isn't the person who is staunchly defending the value of subjective appreciation of audio gear as a valid means of testing, the enemy is the emotional state and psychological mechanisms that prop up such entrenched views/ideology, IMO. Transitively by staying engaged in conversation, open public conversation those entrenched views are exposed to scrutiny.

I get that many who hold such views will never come off of them or even meaningfully engage in questioning; however, some do. That alone for me makes keeping conversations going worthwhile. So if I look at it using that lens, a ban after two posts feels like a step in the wrong direction. Now perhaps there is more that I do not know and this person has had multiple accounts and been trolling aggressively,

Another consideration is that when the inflexible subjectivist is kept in conversation other people who may be early into their exploration of the hobby read the back and forth and that gives communities such as this the opportunity to plant the seed that scientific testing and objective listening tests are the best way forward for anybody who cares about the hobby. We may be able to force the odd mind more open, and dare I say sometimes remind ourselves that we might sometimes need to be more understanding. Now I am not saying that with a product such as this where even if it worked as advertised the claimed benefits are not very significant at all, making it a rather dubious creation to be sure.

Anyway, moderating a community is tough work, I did it for a search engine optimizing community back in the day so I appreciate that not every decision is cut and dry and perhaps a ban is the correct response. Not being in the situation I really can't judge, but I do hope that here at ASR we don't come off as petty and closed-minded either.

 

[Jimbob54]

I get that, and clearly it isn't my place to decide what should be tolerated here, that is up to the host and staff, but I still think as long as possible, people should talk. They say that you don't make peace with your friends, rather your enemies, and for me there is a similarity here. The enemy isn't the person who is staunchly defending the value of subjective appreciation of audio gear as a valid means of testing, the enemy is the emotional state and psychological mechanisms that prop up such entrenched views/ideology, IMO. Transitively by staying engaged in conversation, open public conversation those entrenched views are exposed to scrutiny.

I get that many who hold such views will never come off of them or even meaningfully engage in questioning; however, some do. That alone for me makes keeping conversations going worthwhile. So if I look at it using that lens, a ban after two posts feels like a step in the wrong direction. Now perhaps there is more that I do not know and this person has had multiple accounts and been trolling aggressively,

Another consideration is that when the inflexible subjectivist is kept in conversation other people who may be early into their exploration of the hobby read the back and forth and that gives communities such as this the opportunity to plant the seed that scientific testing and objective listening tests are the best way forward for anybody who cares about the hobby. We may be able to force the odd mind more open, and dare I say sometimes remind ourselves that we might sometimes need to be more understanding. Now I am not saying that with a product such as this where even if it worked as advertised the claimed benefits are not very significant at all, making it a rather dubious creation to be sure.

Anyway, moderating a community is tough work, I did it for a search engine optimizing community back in the day so I appreciate that not every decision is cut and dry and perhaps a ban is the correct response. Not being in the situation I really can't judge, but I do hope that here at ASR we don't come off as petty and closed-minded either.

I really doubt the user was banned for the content of public posts. I would speculate the mods found something odd about the nature /origin of the poster.

 

[DSJR]

It's an open secret that almost no one pays list price with these high end products. So it's hard to know what the mark-ups are, but that just adds to the overall murkiness of the industry. One of the dealers I've gotten with has no engineering background at all, and says some strange things (music sounds better at night because of the lower usage of the power grid), but he is very well connected and seems to get better discounts than others on the same products.

Mark-ups are better than they used to be 'back in my day' but if they approach 100% on the electronics (except Linn if you do as you're told in the contract), I'd be surprised.

Back in my naive late 70's, I became fond of the first issue Naim amps (bolt up and different internal layouts with back plate output devices to the post 1980 models). DC on the mains caused transformer snarling in most samples and I have to say after ten thirty to eleven at night, after the News finished, on my pal's active Isobarik system (three NAP250's), you could hear the soundstage opening up quite palpably. I admit that there may have been other factors involved (I have never smoked anything and even if alcohol was involved, it would have been consumed far earlier in the evening..), but what was once regarded as sign of a better amp because it 'showed up' bad mains, I now regard it as a weakness in the amp's design, which was cobbled together by a non EE engineer.

So, don't be too hard on your dealer pal. I'm suggesting the gear he sells is suspect in certain ways, as the better designed amps do seem pretty much immune to the mains these days, unless you and they live in a really suspect mains area with massive voltage swings (I think the EU directives have minimised that over these parts, but can't be sure) and even current Naims don't appear to be quite as silly-sensitive as they were back then.

I had the company presented dem of the M-Scaler and definitely 'heard a difference for the better' with it doing its thing (Dave dac but can't remember the source other than the iPod used for music choice - Tidal?). Apparently the plastic mains block in between the racks makes an improvement too... Dreadful phone photo but I'm too fumbly to take good pics these days. Speakers were Dynaudio Confidence 60's which sounded a lot 'calmer' than they did via a Naim Statement amp (Naim digital source) in the same room a few weeks earlier. Bottom left a two box Melco source/library?

 

[Tks]

It is plug and play. Turn it on and in one or two seconds it is ready to use. With software you have to have a computer, maintain it, deal with crashes, viruses, etc.

Now if you mean as part of the pipeline of what you already use such as my Roon player, then it works but sucks up fair amount of CPU.

For the cost of this, the sort of computer you can get, won't be sucking any appreciable amount of CPU ever, and you'll basically have a flagship prosumer desktop computer.

Doesn't make sense at all to me. Where the convenience of plug and play for such a niche purpose somehow warrants this cost. This isn't something simple like a mono-toggle coming with your device (where even a layman would benefit). Anyone messing around with filters is already in the weeds deep enough, so I can't see them trying to pay out the nose to rectify this issue. If this is how much they pay for filters, an amp would cost more than a mansion by my order of operations when thinking about it analogously.

 

[Tks]

Not everyone uses a computer as a source, and not all of those who do feel both competent and inclined to deal with installing/configuring HQPlayer or the like.

I get this general stance. But I can't get $5,000 worth of "competency" or "inclination" can't be mustered up for something such as filter performance..

Like, I appreciate what you and others who say this actually mean (I use an RME DAC simply because of the quick toggles I can fire off, when in reality I don't need more than a Topping DX3 Pro). I just can't understand how something this seemingly insignificant would lead someone to shell out this sorta money for the sake of some convenience. It's like paying someone to deliver your food in double the speed, but then paying 50X more than the cost of the food itself. Oh and also the food place is literally right across the street from your home.

It just strikes me as odd due to price. This isn't a situation of knocking on high cost things (a La Ferrari may be a ridiculous purchase for someone who just wants to drive it to work, but if you want a car that looks like that, you have no choice). But this thing? It's like paying for a La Ferrari that ONLY goes to work, yet you paid like 50X the actual car's cost.

 

[raif71]

I get this general stance. But I can't get $5,000 worth of "competency" or "inclination" can't be mustered up for something such as filter performance..

Like, I appreciate what you and others who say this actually mean (I use an RME DAC simply because of the quick toggles I can fire off, when in reality I don't need more than a Topping DX3 Pro). I just can't understand how something this seemingly insignificant would lead someone to shell out this sorta money for the sake of some convenience. It's like paying someone to deliver your food in double the speed, but then paying 50X more than the cost of the food itself. Oh and also the food place is literally right across the street from your home.

It just strikes me as odd due to price. This isn't a situation of knocking on high cost things (a La Ferrari may be a ridiculous purchase for someone who just wants to drive it to work, but if you want a car that looks like that, you have no choice). But this thing? It's like paying for a La Ferrari that ONLY goes to work, yet you paid like 50X the actual car's cost.

I suppose people will eventually learn from their mistakes. It's a process one makes and go through the trials and tribulation. Look at how "captain jack sparrow" went through the last few years..

 

[Geert]

There's a lot wrong with your post there.

No there's not.

When you say "a sound has to be sampled only twice in order to be properly reconstructed" I don't think you understand what "properly reconstructed" means. It does not mean it's going to be perfect.

Yes it is.

Don't be ignorant

Indeed. So take a look at the first 8 minutes of this video:

 

[Blumlein 88]

There's a lot wrong with your post there. Human ears are less sensitive to lower frequencies. Amir has talked about this in one of his videos.

When you say "a sound has to be sampled only twice in order to be properly reconstructed" I don't think you understand what "properly reconstructed" means. It does not mean it's going to be perfect.

I'm not a sound engineer or electrical engineer and I know those two things. Don't be ignorant.

I don't see anything wrong. Yes a bit over 2 samples and a waveform is properly reconstructed as perfectly as other factors allow. As perfectly reconstructed as with 10 samples.

Watch Monty Montgomery's Digital Show and Tell video. GEERT beat me to it.