I received the SU-2 today, measurements and review soon! 
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
SMSL SU-2: a dumb, wasted opportunity
- Thread starter Human Bass
- Start date
311626030434
Roland68
Master Contributor
You definitely have to test both versions; perhaps the version with the flatter button sounds flatter, too.
I also received my SMSL SU-2 today. I'm using it with the AES out of my Lynx E44 sound card, balanced out into a Topping L50. One thing I noticed right off the bat is a slight fade in whenever audio is played from digital silence, even if digital silence is part of the audio file. Here are measurements of the RCA out at the start of a 1 kHz tone:
48 kHz
96 kHz
192 kHz
The fade in is ~20/10/5 milliseconds at 48/96/192 kHz, so the fade is probably a fixed number of samples. For casual listening this is probably unnoticeable most of the time, but I feel since SMSL is targeting this DAC for production work this is absolutely not ideal. The digital audio frames are constantly streaming so why does it need to fade in from digital zero?
Another thing I don't like is that the "LOCKED" LED doesn't actually represent whether a valid digital signal has been locked onto. If digital silence is playing, the LOCKED LED constantly blinks even though it's receiving valid frames. It acts kind of like the L and R LEDs but it appears that its threshold for constant illumination is lower.
The SU-2 does sound good, however. No complaints there. A quick RMAA test showed nothing egregious but I am looking forward to NTTY's review for the deep dive.
48 kHz
96 kHz
192 kHz
The fade in is ~20/10/5 milliseconds at 48/96/192 kHz, so the fade is probably a fixed number of samples. For casual listening this is probably unnoticeable most of the time, but I feel since SMSL is targeting this DAC for production work this is absolutely not ideal. The digital audio frames are constantly streaming so why does it need to fade in from digital zero?
Another thing I don't like is that the "LOCKED" LED doesn't actually represent whether a valid digital signal has been locked onto. If digital silence is playing, the LOCKED LED constantly blinks even though it's receiving valid frames. It acts kind of like the L and R LEDs but it appears that its threshold for constant illumination is lower.
The SU-2 does sound good, however. No complaints there. A quick RMAA test showed nothing egregious but I am looking forward to NTTY's review for the deep dive.
This fade in is a feature on ESS DACs, as I understand, that has varied implementations controlled by the firmware. But I don’t believe you would hear it as a fade in as it’s way too fast. Instead you may hear a softening of attack on a transient. But since it’s happening only at the very instant of the beginning of music, I can’t imagine when this would factor as important unless a song starts with a sharp transient, I understand this is done to reduce pops when DACs lock into sample rates. I also wondered about the functionality of the locked LED and if you look back a few posts, this was explained by another poster. The pop I hear with the unit comes on from standby seems to me to be unrelated to this ramp. That said, I also hear a fade in some times but it’s much slower than what was measured. It’s very bothersome, at least to me, but I haven’t figured out if this is caused by me WiiM Pro Plus streamer or the DAC.
Yes, I believe I’ve turned that off on the WiiM, but it feels like the WiiM still fades in sometimes with Roon. The approx 1/60th second fade in measured above is way to fast to sound like a fade in. I bought a couple of these to drive surround and height channels from a Trinnov Altitude CI. I don’t think there would be much content in those channels starting from a digital zero. I don’t know what they were thinking when they came up with the locked and L/R leds, maybe a poor translation from mandarin when the labeled one as locked? These really are more like signal strength indicators to alert you something is playing. Locked blinks at >-30db and L/R at >-15db.
I believe essentially every modern DAC chip has this feature, and although it’s configurable in firmware, I would think many manufacturers will enable it to avoid clicks and pops on sample rate changes and when tracks start from digital zero. Perhaps someone can chime in on how common it is for manufacturers to enable this fade in.
Last edited:
Thanks for the feedback. From my initial testing, I confirm all of what you said. Since the fade in is not part of my standard testing, I will reference your post in the review.
By the way, since I’m there, here’s a sneak peek of the review. This is the SINAD measurement exactly as @amirm runs it (4Vrms, ie -1.5dBFS, span of 10Hz-22.4kHz):
This is a SINAD of 122.8dB (left channel, if I remember). Distortion is stupidly low
Full review next Friday!
Any thoughts on whether this fade in is typical of most DACs? I can’t imagine a 1/60 second fade in would sound like a fade inThanks for the feedback. From my initial testing, I confirm all of what you said. Since the fade in is not part of my standard testing, I will reference your post in the review.
By the way, since I’m there, here’s a sneak peek of the review. This is the SINAD measurement exactly as @amirm runs it (4Vrms, ie -1.5dBFS, span of 10Hz-22.4kHz):
View attachment 551912
This is a SINAD of 122.8dB (left channel, if I remember). Distortion is stupidly low
Full review next Friday!
Roland68
Master Contributor
I think @Ruediger was showing his non-flat button to go in your directionThat was an ASR inside joke
We definitely need our own sarcasm emoji.
Look's clean enough i would sayThanks for the feedback. From my initial testing, I confirm all of what you said. Since the fade in is not part of my standard testing, I will reference your post in the review.
By the way, since I’m there, here’s a sneak peek of the review. This is the SINAD measurement exactly as @amirm runs it (4Vrms, ie -1.5dBFS, span of 10Hz-22.4kHz):
View attachment 551912
This is a SINAD of 122.8dB (left channel, if I remember). Distortion is stupidly low
Full review next Friday!
- Joined
- Nov 27, 2020
- Messages
- 883
- Likes
- 764
Which begs the question: are there any songs with an sharp opening transient where this fade-in is detectable to a human listener?
Similar threads
