SMSL DL100 DAC and Headphone Amp Review

[Tell]

I don't know if validity was the thing in question, it would be more so the dishonesty involved when the question about the dynamic range was in question. No one is actually posing the question of what the dynamic range of a camera is when multiple exposures are combined into one. They want to know the dynamic range of a single exposure in order to gauge the capability of the sensor in such condition.

The Sony A9III can take a 120 photos in a second, even with slight motion something of this speed can be used to fudge dynamic range results while claiming the camera has a ridiculous amount of dynamic range. (And since dynamic range tests are conducted with still test charts, this will always be something this camera can always Ace). If camera burst rates keep improving, at some point you can take the worst sensors and claim amazing dynamic range (the only side effect being disgusting storage requirements, and battery use required to in-camera combine all the exposures).

The big selling point with the A9iii is the global shutter which among other things give you insane FPS. Not sure how they do it from a technically, but one of the drawbacks is that dynamic range do suffer a bit. Don't think they are "cheating" like in the discussion of this DAC though, instead look at how smartphones do it and you'll see a LOT more "cheating" with their computational photography.

 

[Tks]

The big selling point with the A9iii is the global shutter which among other things give you insane FPS. Not sure how they do it from a technically, but one of the drawbacks is that dynamic range do suffer a bit. Don't think they are "cheating" like in the discussion of this DAC though, instead look at how smartphones do it and you'll see a LOT more "cheating" with their computational photography.

They can cheat all they want, but when a dynamic range figure is posted, the long standing expectation and typical testing methodology requires it to be of a single exposure of the uncompressed RAW file from the camera. "Cheating" isn't the issue, the issue is when someone asks what the dynamic range figure is from a spec sheet perspective - and you reply with a bracketed image, that is simply being ignorant or fallacious. The cheating is great when it comes time to use the camera (like dithering and other such things). What you don't want is the cheating to be the figure when you're constructing the device or evaluating it's technical merits.

This is all a semantic ordeal. Where the question being asked, is being provided an answer that's not answering the actual question itself.

 

[JIW]

Thank you for this review, really interesting unit.

Don't even remember the result? As a prevalent iem user it's one of the measure I'm looking for the most.

Anyway they still sticks to the damn sample rate display instead of volume, what a shame.

You can get an idea from the power and voltage at impedance figures since at 50 mV THD+N is strongly dominated by noise. From the latter it seems to be around 81 dB.

Since 50 mV across a 300 Ohm load gives a power of 0.05^2/300 = 8.33*10^(-6) W = 8.33 uW, the 300 Ohm power figure gives a slightly lower value closer to 80 dB.

Since 50 mV across a 32 Ohm load gives a power of 0.05^2/32 = 7.81*10^(-5) W = 78.1 uW, the 32 Ohm power figure gives a slightly higher value closer to 81 dB.

Thus, it is probably around 81 dB. According to the specifications, the SNR of the headphone output is 126 dB while the output is 7.5 V. Thus, the SNR at 50 mV should be around 126 - 20*log10(7.5/0.05) = 82.48 dB.

 

[GXAlan]

Why wasn't this tested, when 99% eyeing at this DAC is for this input so they can hopefully use volume control from their TV?

The $0 I paid to @amirm for this article does not cover the cost of HDMI ARC and eARC testing.

It’s a good point — but we have to rely on other reviewers to test the ergonomics of a product since the priority is the measurements here, and the APx555 doesn’t do ARC/eARC.

All we can do is to vote with our wallet and tell @SMSL-Mandy that we really want HDMI CEC support for volume control…

 

[Tell]

They can cheat all they want, but when a dynamic range figure is posted, the long standing expectation and typical testing methodology requires it to be of a single exposure of the uncompressed RAW file from the camera. "Cheating" isn't the issue, the issue is when someone asks what the dynamic range figure is from a spec sheet perspective - and you reply with a bracketed image, that is simply being ignorant or fallacious. The cheating is great when it comes time to use the camera (like dithering and other such things). What you don't want is the cheating to be the figure when you're constructing the device or evaluating it's technical merits.

This is all a semantic ordeal. Where the question being asked, is being provided an answer that's not answering the actual question itself.

Tbh if they do some kind of multiple exposure trick to get higher dynamic range but manage to do it without any artefacts I don't have any problem with it being used in spec sheets or measurements. I remember a few years ago RED camera doing something like this when filming, didn't really look it up that much or if it gave any noticeable artefacts, but the few clips I saw back then did look really good.
So yeah, if this DAC is doing tricks to up the dynamic range and it won't give any audible errors I don't have a problem with it. Though tbh I don't see why they should even bother seeing that the performance of their DACs have been far above audible thresholds for a long time. Buut I guess it's all about marketing, they need people to buy their new stuff even though they don't actually need it :\

 

[Bleib]

Possibly the new price performance king

To Topping, SMSL etc, start adding more inputs.

 

[Fleuch]

There is no HDMI/ARC in RAW-MDA1.

Is that a loss, or does the performance of the RAW-MDA 1 benefit from not having the HDMI/ARC interface circuitry? The requirement for quad CS43131 chips compared to dual ESS9039Q2M chips is another detractor.

Amir does not have the means to test the HDMI/ARC output.

Personally, I consider HDMI/ARC to be completely unnecessary in an audio system. Perhaps there is kit on the market that has HDMI/ARC outputs but, at the present time, perhaps the only application is to improve the rather atrocious sound quality found in many modern televisions - which often have an optical output anyway. Some of the implementations on DACs are not full HDMI/ARC, but a form of I²S (ASR : Study: Is I²S interface better for DACs than S/PDIF or USB?)

This is rather similar in many ways to the ongoing debate about whether MQA is a positive addition. Better to use the available budget to improve component quality, as found in the RAW-MD 1 or, as suggested, provide a 12V trigger.

 

[Robbo99999]

Isn't this the best value / best measuring balanced DAC / amp combo unit that we've ever measured on here?!

 

[mc.god]

It does show you the volume when you adjust the level it will momentarily show you the volume gain level and then it goes back to the bitrate data.

Yes, i took it as obvious, it would have been totally nonsense if volume didn't show up at all But seriously I don't understand why SMSL still stick to this default sample rate view when having volume on check at a glimpse would be much more useful and, I add, safer for the ears.

 

[AdamG]

Yes, i took it as obvious, it would have been totally nonsense if volume didn't show up at all But seriously I don't understand why SMSL still stick to this default sample rate view when having volume on check at a glimpse would be much more useful and, I add, safer for the ears.

Go to their support web site and put in a product/firmware improvement suggestion request. It probably can be done fairly easily with a firmware update. It has enough real estate room on the screen to display both.

 

[mc.god]

You can get an idea from the power and voltage at impedance figures since at 50 mV THD+N is strongly dominated by noise. From the latter it seems to be around 81 dB.
View attachment 388307

Since 50 mV across a 300 Ohm load gives a power of 0.05^2/300 = 8.33*10^(-6) W = 8.33 uW, the 300 Ohm power figure gives a slightly lower value closer to 80 dB. View attachment 388308

Since 50 mV across a 32 Ohm load gives a power of 0.05^2/32 = 7.81*10^(-5) W = 78.1 uW, the 32 Ohm power figure gives a slightly higher value closer to 81 dB.
View attachment 388309

Thus, it is probably around 81 dB. According to the specifications, the SNR of the headphone output is 126 dB while the output is 7.5 V. Thus, the SNR at 50 mV should be around 126 - 20*log10(7.5/0.05) = 82.48 dB.

I was thinking about it, but all the graphs here are for high gain mode only, while low gain would be the right mode for 50 mv performance with efficient iems and should give much better than 81 db SNR. If this unit doesn't have a low gain mode then it's useless for iem users.

 

[mc.god]

Go to their support web site and put in a product/firmware improvement suggestion request. It probably can be done fairly easily with a firmware update. It has enough real estate room on the screen to display both.

I'll think about it even though it was just a little runt since I'm not on the market for this unit at the moment, I'm already plenty ok with what i own.

 

[JIW]

I was thinking about it, but all the graphs here are for high gain mode only, while low gain would be the right mode for 50 mv performance with efficient iems and should give much better than 81 db SNR. If this unit doesn't have a low gain mode then it's useless for iem users.

It doesn't have different gains.

 

[JIW]

Go to their support web site and put in a product/firmware improvement suggestion request. It probably can be done fairly easily with a firmware update. It has enough real estate room on the screen to display both.

It can only display three digits so it's either or. The right half of the display indicates DSD, MQA and stand-by.

 

[mc.god]

It doesn't have different gains.

Thank you, it's too bad then, I'm going to lower my vote to fine.

 

[Mr Swing King]

Thank you, it's too bad then, I'm going to lower my vote to fine.

I voted for fine as well. With 1.4 watts in 32 ohm this is not only going to be difficult finding iems you can plug in, but there is also a bunch of modern headphones with the same problem.
Aune AR5000, the newer Beyerdynamics, Austrian Audio cans, Meze, Focals and so forth.
I predict poor volume control in a lot of cases.

This is actually something that has puzzled me for a while; headphone amplifiers are getting beefier by the minute whereas headphones are getting much easier to drive. I get that a lot of people equalize and need a bit of extra headroom, but this is still a weird trend imho.

 

[BR52]

This is an ugly duck (DAC) both inside and outside.
You know that something must be wrong when they start doubling the number of DAC chips.
This DL100 has 4×CS43131 chips, the DO100 Pro had 2xES9039Q2M chips, and the D-6S just 1xES9039Q2M chip.

And by the way, what is the "job to be done" of an HDMI ARC circuit or a Bluetooth circuit in a DAC? Is it precision, or is it convenience?

In electronics, when you add circuitry for convenience, you add more components, and more components add more noise, and more noise reduces precision.

It's common engineering practice to reduce noise by paralleling devices (OpAmp's, DAC's etc.).
With n devices, you will get root n less noise. For more info: https://www.analog.com/en/resources...ers-improves-signal-to-noise-performance.html. For example.

 

[TonyJZX]

It can only display three digits so it's either or. The right half of the display indicates DSD, MQA and stand-by.

the screen reflects the price

the SMSL RAW-MDA isnt too much more

I probably have a dissenting view in that I dont look at the DAC display too much and so the 44.1 or 96 or 384khz figure is fine by me and i like outputing max voltage to my preamp

BUT I guess some people love to use this as a digital preamp?

SMSL should take the hint and make an option in the menu to display what it is supposed to be the default???

 

[JIW]

the screen reflects the price

the SMSL RAW-MDA isnt too much more

I probably have a dissenting view in that I dont look at the DAC display too much and so the 44.1 or 96 or 384khz figure is fine by me and i like outputing max voltage to my preamp

BUT I guess some people love to use this as a digital preamp?

SMSL should take the hint and make an option in the menu to display what it is supposed to be the default???

I have the display on my DL200 turned off. It turns on when I change the volume, go into the menu or change the sample rate and then turns off shortly after.

 

[Esprit]

Same thing here with D-6s.