SMSL DL200 the best ever value DAC/AMP?

[MbphotoX]

I said as much about sound quality. I said it COULD give additional power. My question is what's the cost?

The $200 Fiio K7 desktop dac/amp has true balanced headphone out, for example.

the K7 is awesome, but it doesn't have balanced outputs for studio monitors (where balanced outputs would really make sense, you know?!)

While it does deliver 2W for balanced headphones, it only has 1.22W on the normal output. (DL200 has 1.5W there)

FiiO is a great brand with very reliable products (much better than Topping and SMSL, from what I can tell), but the K7 is a very different device aimed at a different customer.
If you don't need balanced TRS outputs on the back, the DL200 is kinda overkill. There are cheaper options that deliver the same for headphones, afaik.

Also, but that's minor, the K7 cannot decode MQA.. who knows what's in the future for this codec. They messed up big time with their lies about being lossless.. but the codec itself is actually quite amazing (for a lossy compression, that is)

 

[donniekerabatsos]

First time poster.

I've been thinking about a new dac or dac /amp combo or just an amp recently only because I'm having some issues with my Magni. I thought I was done getting mind effed by information overload when I purchased the Modi Magni combo 7 years ago but I peeked into the rabbit hole and it sucked me in. So I've been overanalyzing what I should do. I like dac /amp combos. I've had a couple audio gd products that I really liked. The first one died and I sold the second because it didnt work with windows 10. This SMSL dl200 reminds me of those devices to an extent because it has the inputs and outputs I want. USB for my PC, optical for my TV, a 6.35 out for my headphones and RCA outs for my Kinter driving a pair of Radio Shack Minimus 7's.

I was considering the Topping dx3 but the 3.5mm out for the headphone is a turn off. This SMSL d200 looks interesting. On paper its an upgrade to my current dac but I'm not a lossless or FLAC guy. I'm a Spotify You Tube guy. The other thing that I'm not sure about is the digital menu interface. I dont really want to be clicking through a bunch of things to get from headphone out to RCA out. I want to flip a switch like I could on the Audio GD. I guess what I'm wondering is would I be gaining anything with the upgraded dac given my use. I've got the old push button Modi so it switches between inputs real easy and pair it with the Topping L30 ii. It appears like it's at least as powerful as the Magni if not more powerful and it's got a 3 gain switch and an output switch on the front. Why doesn't Schiit do this?

Anyway tired of fiddling with my headphone jack to get both channels to work on the RCA outs of the Magni.

 

[1mongus]

Anyone got any ideas what the best filter is for soundscape/imaging? I'm using FL3 at the moment.

 

[MbphotoX]

Anyone got any ideas what the best filter is for soundscape/imaging? I'm using FL3 at the moment.

How to pick the best filter setting for your DAC

Does your DAC let you choose the output filter? What do they do. How do the different kinds differ from each one? Which is the best? Which is the worst? We'll help you sort it out.

addictedtoaudio.com.au

fantastic read

sounds like any of FL2, FL3, and FL4 can be chosen.
If your ears are goldensound-level sensitive up to 22 kHz, you probably want FL4 (bc it extends beyond 22khz before dropping off sharply)

 

[Grael]

This graph from smsl is wrong...
The filters are the same as smsl d-6s ones

SMSL DL200 the best ever value DAC/AMP?

Switching between USB 1.1 and 2.0 ALSO worked. What mode is your DL300 set to? Both DACs use the XMOS XU-316 so I would have expected the same behaviour.

www.audiosciencereview.com

 

[ext23]

I got my DL200 recently, loving it! Here's a cheeky pic.

Got a few questions, however.

First, is there a plain English explanation of how the various filters sound?

Next, I have Windows 11 set to 32 bit/44.1 khz. I haven't installed any drivers yet but I want the amp to automatically switch to 48 khz (for example) when gaming, etc. Do I need drivers? The amp itself says USB 2.0.

You can do automatic sample rate switching (=bit perfect playback) without installing SMSL's driver, using the Wasapi Exclusive API.
It's compatible with Microsoft's generic UAC driver.

Lastly, about the RCA output. I have it running into my Klipsch powered speakers as you can see in the pic. But I noticed that the volume is still affected by the knob on the SMSL. This is confirmed by other posts, such as this one:

These outputs are not affected by gain switch but may be adjusted by volume control.

Surely this is wrong? It should be a flat line out, no? If not, why not? Should I set the SMSL to 100% volume and use the volume knob on the Klipsch? Honestly I'm baffled by this. Is there a setting that I'm missing to make it a flat line out?

 

[staticV3]

First, is there a plain English explanation of how the various filters sound?

None based in sound (heh) science.

I wouldn't expect audible differences in the first place and it turns out, many people can't hear any.

Next, I have Windows 11 set to 32 bit/44.1 khz. I haven't installed any drivers yet but I want the amp to automatically switch to 48 khz (for example) when gaming, etc. Do I need drivers? The amp itself says USB 2.0.

If you want automatic sample rate switching, then your application needs to use the Wasapi Exclusive, ASIO, or Kernel Streaming API.

Wasapi and KS work plug-and-play, for ASIO you need to manually install the SMSL driver.

The crux though is that virtually no game uses these APIs for audio, as doing so would for example block you from hearing your friends via Discord etc.

Instead, they use Wasapi Shared or DirectSound, which go through the Windows resampler and mixer.

Surely this is wrong? It should be a flat line out, no? If not, why not? Should I set the SMSL to 100% volume and use the volume knob on the Klipsch? Honestly I'm baffled by this. Is there a setting that I'm missing to make it a flat line out?

What the output should be is subjective.

Many people for example depend on the DL200's ability to control Line out volume as their main volume control.

From what I can tell, the DL200 lacks the ability to set the outputs to fixed volume.

You can set the DL200 to 100% and leave it there, or use it for volume control. Whatever is most convenient to you.

 

[ext23]

If you want automatic sample rate switching, then your application needs to use the Wasapi Exclusive, ASIO, or Kernel Streaming API.

Wasapi and KS work plug-and-play, for ASIO you need to manually install the SMSL driver.

I'm happy with Wasapi and would prefer not to install any extra drivers. What should my Windows audio settings be? I can't see an "auto sample rate switching" setting or anything like that...?

What the output should be is subjective.

Many people for example depend on the DL200's ability to control Line out volume as their main volume control.

This still confuses me. So it's not a true line out? It's a pre-amp with gain control? That is absolutely not how it is advertised.

If these people are using passive speakers then they would need a separate speaker amp too, right?

 

[Qu@rk]

This graph from smsl is wrong...
The filters are the same as smsl d-6s ones

SMSL DL200 the best ever value DAC/AMP?

Switching between USB 1.1 and 2.0 ALSO worked. What mode is your DL300 set to? Both DACs use the XMOS XU-316 so I would have expected the same behaviour.

www.audiosciencereview.com

For me it seems that the filter list for the d-6s is also not correct for the DL200.
It seems the list (not the graph) of the DL200 manual is correct.

FL1 (Minimum phase)
FL2 (Linear phase apodizing fast roll-off)
FL3 (Linear phase fast roll-off)
FL4 (Linear phase fast roll-off low ripple)
FL5 (Linear phase slow roll-off)
FL6 (Minimum phase fast roll-off)
FL7 (Minimum phase slow roll-off)

I tried to compare these to an RME ADI-2/4 PRO and Qudelix-5K.
But I am not 100% certain, measurements would be necessary.

The "Minimum phase slow roll-off low dispertion" (8th filter in ESS datasheet) is missing.

 

[staticV3]

What should my Windows audio settings be? I can't see an "auto sample rate switching" setting or anything like that...?

There is no such setting.

If your Application uses Wasapi Exclusive, ASIO, or Kernel Streaming, then sample rate will switch automatically.

If your Application doesn't, then sample rate won't. That's it.

If these people are using passive speakers then they would need a separate speaker amp too, right?

An actual, fixed volume Line out is just inconvenient nowadays.

Many people have switched to active studio monitors, in which case the DL200's volume knob and remote may be the only physical volume controls in reach.

Even with passive speakers, the DL200's ability to control volume enables customers to use either Int Amps, or Power Amps.

There is no disadvantage really.

 

[JIW]

For me it seems that the filter list for the d-6s is also not correct for the DL200.
It seems the list (not the graph) of the DL200 manual is correct.

FL1 (Minimum phase)
FL2 (Linear phase apodizing fast roll-off)
FL3 (Linear phase fast roll-off)
FL4 (Linear phase fast roll-off low ripple)
FL5 (Linear phase slow roll-off)
FL6 (Minimum phase fast roll-off)
FL7 (Minimum phase slow roll-off)

I tried to compare these to an RME ADI-2/4 PRO and Qudelix-5K.
But I am not 100% certain, measurements would be necessary.

The "Minimum phase slow roll-off low dispertion" (8th filter in ESS datasheet) is missing.

The figure says the same as the manual on page 17.

 

[JIW]

The filters of the DL200 from the manual.

For comparison, this is from the D-6S manual.

It is identical. The filters of the DL200 are the same as those of the D-6S.

However, at least some are mislabeled for both. This is evident from the data sheet for the ES9039Q2M chip. Firstly, the names and properties.

From this, it is evident that SMSL #3 is ESS #2 and SMSL #4 is ESS #3. The former due to full rejection at fs/2 and the latter due to the highest stop band attenuation.

For the other ones, since they are distinguished in large part by stop band ripple and level, the frequency response will reveal better which are which. Note that the ESS figures go to 200 kHz whereas the figure from SMSL only goes to slightly above 40 kHz which is the third vertical line after 10 kHz on the ESS figures. Also note that the smaller figures in the figures do not all have the same vertical scale.

ESS #1 is down about 3 dB at 20 kHz and has some ripple in the stop band where the level falls in frequency. However, this is not compatible with the property of -3 dB at 0.55 fs. Thus, the figure for ESS #1 is incorrect. However, it may be correct for ESS #8 since it is also displayed for it but also for ESS #7. Also, the figure for ESS #7 is incorrect since it is clearly showing the same as for ESS #6 where it seems correct.

Another slow filter is ESS #5 which has significantly more ripple in the stop band where the level also falls in frequency. Further, the first dip is closer to 30 kHz than for ESS #8. Thus, ESS #5 is SMSL #6 and ESS #7 or ESS #8 is SMSL #1 due to the slightly darker hue of blue in the SMSL figure. In Amir's figure the colors are clearer. This leaves SMSL #2, #5 and #7 to be identified among ESS #1, #4 and #6.

Due to the high stop band ripple, ESS #4 is probably SMSL #5. So far, SMSL #3 is ESS #2, SMSL #4 is ESS #3, SMSL #5 is ESS #4, SMSL #6 is ESS #5 and SMSL #1 is ESS #7 or #8. Due to the smooth stop band where the level also falls in frequency, ESS #6 is probably SMSL #7. This would leave ESS #1 for SMSL #2. While there is no figure for ESS #1, the slightly higher stop band level than ESS #6 with the same pass band and stop band makes this probable.

In summary,
SMSL FL1 - Minimum phase slow roll-off or Minimum phase slow roll-off low dispersion (ESS #7 or #8)
SMSL FL2 - Minimum phase (ESS #1)
SMSL FL3 - Linear phase apodizing fast roll-off (ESS #2)
SMSL FL4 - Linear phase fast roll-off (ESS #3)
SMSL FL5 - Linear phase fast roll-off low ripple (ESS #4)
SMSL FL6 - Linear phase slow roll-off (ESS #5)
SMSL FL7 - Minimum phase fast roll-off (ESS #6)


SMSL DL200 manual: https://www.smsl-audio.com/themes/p...?file=/upload/portal/download/DL200Manual.pdf
SMSL D-6S manual: https://www.smsl-audio.com/themes/p...ml?file=/upload/portal/download/D6SManual.pdf
ESS ES9039Q2M data sheet: https://www.mouser.de/datasheet/2/1082/ES9039Q2M_Datasheet_v0_1_3-3168946.pdf

 

[MbphotoX]

I'm happy with Wasapi and would prefer not to install any extra drivers. What should my Windows audio settings be? I can't see an "auto sample rate switching" setting or anything like that...?


This still confuses me. So it's not a true line out? It's a pre-amp with gain control? That is absolutely not how it is advertised.

If these people are using passive speakers then they would need a separate speaker amp too, right?

I use Tidal and set it to Exclusive mode. That way, Tidal controls the sampling rate and the DL200 receives whatever sampling rate I'm playing.

upside: you don't get any re-sampling done by Windows.. kinda bothered me to buy an expensive DAC only to have Windows change the signal beforehand...
Downside: switching sampling rate in the DL200 causes a crackling sound... if you let Windows do the resampling, this doesn't happen, afaik.


It's a true lineout, but one that doesn't have a fixed voltage! I'm very glad it's that way, because the balanced TRS outputs are clearly aimed at active studio monitors and the likes, which don't have volume controls.
if you use passive speakers, you can just hook the DL200 up to a power amplifier and use the DL200 as a preamp and adjust the volume via the remote/knob.

 

[Qu@rk]

@JIW
Thank you very much for your incredible effort.

My listening experience and comparisons with other ESS devices is different.
But that's not a very scientific approach.

Please correct me if I am wrong but your assumption is, the graph in the DL200 manual is correct and not just a copy from one manual to another?

Unfortunately I have no Audio Precision Analyzer or professional software but I could use my RME ADI-2 PRO + REW to get some filter measurements from the DL200.

Is the precision/resolution of this setup sufficient to obtain evaluable results?

But still, I would have to find the time.

 

[staticV3]

Unfortunately I have no Audio Precision Analyzer or professional software but I could use my RME ADI-2 PRO + REW to get some filter measurements from the DL200.

Is the precision/resolution of this setup sufficient to obtain evaluable results?

Definitely.

 

[icy20]

Hey guys new here. I have a question. Did anyone try to see if the DL200 can play the same source from pc to the headphone jack and the line out simultaneously? I have a goxlr and i want the sound to go there as well except from playing at my sennheisers? If dl200 doesnt do that does anyone know of any other similar dac/amp that does that? A friend said fiio k7 does but it doesnt have a trs output. Thanks in advance

 

[Qu@rk]

@staticV3

Thank you, that's nice.

As said I will try to find the time for research and how to do this correctly.
I have some basic knowledge but I could need a little bit of advice.

Is there a website or a forum thread that I could study, maybe with a similar setup with REW?

My plan would be to feed the DL200 over USB a 48 Khz signal from REW and record the response with the RME ADI-2 PRO at 192 kHz over USB. This should eliminate the A/D filter effects of the RME ADI-2 PRO and provide enough bandwidth.

But I have to learn how to do these measurements with REW. But that shouldn't be a big problem, I guess or better I hope (hihi). The problem here is my limited time, but I am also very curious to see how this would work out.

Windows 11:
Should I use the ASIO drivers for the DL200 from SMSL or from the XMOS website?

I would prefer to do this on Linux, no driver mess. But I am not sure how well REW would work on Linux. I used it on Linux several years ago.

BTW, crappy Windows 11:
Yesterday on a test system I performed a firmware (BIOS) update.
Suddenly Windows 11 Home encrypted all my NTFS partitions on this system (BitLocker) without any notification whatsoever. AND therefore no BitLocker Recovery Key backup!

I only noticed it because I am using dual boot with Linux on this test machine and Linux wouldn't boot (mounted an NTFS partition at boot in fstab).

It was a test system but still!

Unbelievable! If something would have happened to my drives I would have never gained access to my data again without a backup of the BitLocker Recovery Key.

Please inform everybody, friends, colleagues etc. that they have to check if their partitions are encrypted (BitLocker). If this is the case, immediately create a backup of your BitLocker Recovery Key (privacy & security settings) or disable the encryption if you don't need it (mobile devices should always be encrypted).

Without the Recovery Key you could lose all your data if something bad happens.

HINT: Do not store the Recovery Key on an encrypted drive.
Print it out on paper and put it in a safe place.

I usually use Windows 10/11 Pro. I don't know if the Pro version is also affected by this dangerous nonsense.

 

[staticV3]

As said I will try to find the time for research and how to do this correctly.
I have some basic knowledge but I could need a little bit of advice.

Use REW's Generator->Save to file to create a Random white noise test tone at 44.1kHz. I'd do Mono, 24bit, 60+ seconds.

Use foobar via Wasapi Exclusive or ASIO to play this test tone on your DL200 on loop. If you have another bit-perfect audio player that you prefer, use that.

In REW, select your ADI-2 as I/O (use Java EXCL!) and set the sample rate to 96 or 192kHz, then open the RTA window.

RTA settings should be small FFT length (≤32k), Flat top window, Forever averaging, and 90% Max overlap.

With the DL200 playing white noise into the ADI-2's Line in, start the RTA capture and relax.

For clean results, REW will need to capture a lot of averages (1k+).
I once did like 13k AVGes for a super clean end result.

If you're happy, click "Save current", rename the measurement with the DL200's oversampling filter used, then change the filter and repeat.

In the end, you should have a graph like this:

 

[Qu@rk]

Thank you very much! What a great guide! I like the "... and relax." (hihi)
I will try that as soon as possible. I am very curious.

 

[Qu@rk]

With the explained setup I measure the frequency response.

Would it be complicated to measure also the impulse (and/or step response)?

An additional impulse response would make it easier or more clear to determine the filter type.