SMSL DO300 DAC Review

[pkane]

Thanks for this litterature. In fact, measuring with REW I do not sée thé HF drop and frequence response AT 48, 96, 192 Filter Off are all the same and do not differ from thé Linear Fast curve

No? Seems like it's 3dB down at 14kHz (@48k sample rate):

 

[tades]

Hi Paul
You're right! I see this attenuation, much more clear when smoothing.
I have no mean to analyse the anti aliasing but distorsion is +3 up to +6dB compared to Linear Fast
This explain the impression I had , that is is more lively but less clean

By the way using REW swap signal (generated at 48kHz), I discoved that the upsampling to 192kHz made by Foobar SoX module is 2dB more distorded than the 48kHz source.

Thank for your measurement

 

[tades]

This mean you are actually filtering the file before reaching the dac? This actually may be the same, but honestly I don't understand why doing so instead of using dac filters

Hi
No DAC is sent to preamp->DCX 2496 -> 3*Amps
Please refer also to my answer to Paul (pkane), Your challenges made me further investigating.
Big Thanks

 

[DeepSpace57]

Hi Paul
You're right! I see this attenuation, much more clear when smoothing.
I have no mean to analyse the anti aliasing but distorsion is +3 up to +6dB compared to Linear Fast
This explain the impression I had , that is is more lively but less clean

By the way using REW swap signal (generated at 48kHz), I discoved that the upsampling to 192kHz made by Foobar SoX module is 2dB more distorded than the 48kHz source.

Thank for your measurement

There is a noticeable boost in sound volume between the filter off and the others. On the other hand, i cannot use +2db and filter off together due to the crackling sound.

 

[tades]

There is a noticeable boost in sound volume between the filter off and the others. On the other hand, i cannot use +2db and filter off together due to the crackling sound.

Hi.
Yes Amirm measured additional 4,3dB
In my ears, it sounds even more 5. or 6 dB
Si notre surprised you cant use +2

 

[Blashyrkh]

Hi
No DAC is sent to preamp->DCX 2496 -> 3*Amps
Please refer also to my answer to Paul (pkane), Your challenges made me further investigating.
Big Thanks

But what is feeding your dac?

That's the key in using no filter, not what's after. You are filtering analog signal that's already been converted wrongly

The use of analog filters may be used to.reach a more gentle approach compared to oversample or digital filters, but still I think the sabre no filter doesn't make much sense if you want to reach a good reconstructed analog signal

 

[Dunring]

I just got a couple of them in today, in case you wanted a laugh. I've never seen one like that either until today.
Edit: Update, I got the firmware from Aoshida-US and flashed it and it's working perfect now. They said it was a firmware problem, but the flash worked and easily fixed.

 

[Smoey22]

I just got a couple of them in today, in case you wanted a laugh. I've never seen one like that either until today.
Edit: Update, I got the firmware from Aoshida-US and flashed it and it's working perfect now. They said it was a firmware problem, but the flash worked and easily fixed.


View attachment 370693

What is the firmware version - I'm guessing v1.015? Could you share it? Thanks.

For the record:
v1.013 (IAP_DO300_1v0_b13) was released June 2023, link, (removed from smsl-audio.com)
v1.014 (IAP_DO300_1v0_b14) was released February 2024, link (not on smsl-audio.com)

The display did a 180 I see

 

[widemediaphotography]

As specified in the ESS manual, "Filter Off" is only for audio signals sampled at least at 44.1x8khz!

 

[tades]

But what is feeding your dac?

That's the key in using no filter, not what's after. You are filtering analog signal that's already been converted wrongly

The use of analog filters may be used to.reach a more gentle approach compared to oversample or digital filters, but still I think the sabre no filter doesn't make much sense if you want to reach a good reconstructed analog signal

Upstream the dac I'm using Foobar, resampling at 192kHz (limit of Wattson Emerson Streamer) with a sharp cut off (96dB/octave) (convolution module generated from Rephase) at 19,5Hz to limit miroring
Meaning I should have no pollution up to at least 96-19.5kHz 76kHz
Thanks for your support in trying to understand what happens

 

[Dunring]

This is the file zipped I got from SMSL (they distribute in RAR format, why I don't know). It's the factory stock used in case yours gets corrupted.

 

[Wild 7]

This is the file zipped I got from SMSL (they distribute in RAR format, why I don't know). It's the factory stock used in case yours gets corrupted.

Any instructions for updating the firmware ?

 

[tades]

As specified in the ESS manual, "Filter Off" is only for audio signals sampled at least at 44.1x8khz!

The best result I got with Filter Off 88,2kHz or 96kHz oversampling (176 or above being slightly polluted by electronic noise Cayin miniCD mkII I2S or foobar).

Slight drop down in trebles approx 1dB at 14kHz, (helping to reach Harmann curve in my room) ... but so lively

Nota Trebles cut off Butterworth 48dB/oct at 20kHz by DCX2496

 

[tades]

This is the file zipped I got from SMSL (they distribute in RAR format, why I don't know). It's the factory stock used in case yours gets corrupted.

Hi, Did you percieve any improvement vs 1.03
Coudl you describe and ponderate? Imperceptible ? Slight ? Nice to have? Significant? Dont miss it,

 

[tades]

This is the file zipped I got from SMSL (they distribute in RAR format, why I don't know). It's the factory stock used in case yours gets corrupted.

Hi. Did you test it? Any Sonic improvement ?

 

[tades]

Hi. Did you test it? Any Sonic improvement ?

Anyone else

 

[Wild 7]

Is the I2S input corresponding to PS Audio pin setting ?

And is it any good ?

 

[Dawsoo]

The DO300 is a nice machine, I used it to measure my "THD measured record" with the my modif Cosmos ADC.

 

[johny_2000]

Here's Multitone Analyzer noise computation for APU+6dB notch into Cosmos ADC 192k (SMSL DO300ES 96k). Without bin subtraction, but with APU compensation curve applied.

A little off topic: What is the purpose of using APU+6dB notch compared to APU+0dB notch?
I found this: *Cosmos APU LNA 34db for the DR test, Cosmos APU notch +6db for the THD+N tests, and THD test used APU notch 0db*, but can't figure out the logic behind each of these usage modes.

 

[pkane]

A little off topic: What is the purpose of using APU+6dB notch compared to APU+0dB notch?
I found this: *Cosmos APU LNA 34db for the DR test, Cosmos APU notch +6db for the THD+N tests, and THD test used APU notch 0db*, but can't figure out the logic behind each of these usage modes.

The notch helps reduce the effect of ADC harmonic distortion, while the +6dB lowers the effect of ADC noise on the DUT measurement.