-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. 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 filters mislabeled on ES9039 DACs
- Thread starter JIW
- Start date
AnalogSteph
Major Contributor
In that case, either one of the (actual)
* Linear Phase Fast Roll-Off
* Minimum Phase
or
* Minimum Phase Fast Roll-Off
filters make for a good compromise.
Slow Roll-Off filters are only good for 88.2 kHz and up.
I would not recommend the Apodizing filter at 44.1 kHz either, this one is for 48 kHz minimum.
- Thread Starter
- #23
For the most accurate reproduction, use a fast linear phase filter. On the DL200, FL3, FL4 or FL5. FL4 is probably the best compromise.
- Thread Starter
- #25
Minimum phase has different delay for different frequencies. Linear phase has the same delay for all frequencies.
AnalogSteph
Major Contributor
It all depends on your priorities. A minimum phase filter will shave off about 0.69 ms of latency at 44.1 kHz or 0.635 ms at 48 kHz in this case, making it ideal for live monitoring. The uneven group delay in the treble will make it non-ideal for measurement applications though, even if you're not going to hear this. In all other parameters, the standard minimum phase filter is a very good performer.
I was arguing against the apodizing filter (FL3) at 44.1 kHz mainly because its bandwidth is sub-20 kHz then (-3 dB at 0.44 fs = 19.4 kHz). That's the price to pay for a filter that reaches stopband at exactly fs/2.
The "normal" fast rolloff filters extending to 22+ kHz still allow for some alias components to get through between 0.5 fs and 0.55 fs (~22.05 to 24.2 kHz), but since these are mirroring the 0.45 fs to 0.5 fs range (~19.8 to 22.05 kHz) and there generally isn't much going on up there to begin with, this generally is quite an acceptable tradeoff for audio.
FL5 should be the Linear phase fast rolloff low ripple filter, I think? I would actually prefer that to the regular fast rolloff filter (FL4) - it's got lower passband ripple and associated pre-echo, and a stopband rejection of 90+ dB still is plenty for a DAC. Historically, it wasn't uncommon to see values in the 40s in run-of-the-mill CD player DACs, or the 60s in fancier ones. You just need to keep ultrasonic levels low enough so that following stages do not give any trouble.
I was arguing against the apodizing filter (FL3) at 44.1 kHz mainly because its bandwidth is sub-20 kHz then (-3 dB at 0.44 fs = 19.4 kHz). That's the price to pay for a filter that reaches stopband at exactly fs/2.
The "normal" fast rolloff filters extending to 22+ kHz still allow for some alias components to get through between 0.5 fs and 0.55 fs (~22.05 to 24.2 kHz), but since these are mirroring the 0.45 fs to 0.5 fs range (~19.8 to 22.05 kHz) and there generally isn't much going on up there to begin with, this generally is quite an acceptable tradeoff for audio.
FL5 should be the Linear phase fast rolloff low ripple filter, I think? I would actually prefer that to the regular fast rolloff filter (FL4) - it's got lower passband ripple and associated pre-echo, and a stopband rejection of 90+ dB still is plenty for a DAC. Historically, it wasn't uncommon to see values in the 40s in run-of-the-mill CD player DACs, or the 60s in fancier ones. You just need to keep ultrasonic levels low enough so that following stages do not give any trouble.
nanook
Active Member
No, not here in ASR nor via PM.
Very disappointing......
Roland68
Major Contributor
But unfortunately neither unusual nor unexpected.
Bleib
Major Contributor
- Joined
- May 13, 2021
- Messages
- 1,885
- Likes
- 3,360
Some new firmwares are out, are the filters now named correctly?
SMSL Audio Official -Professional audio product manufacturer
Foshan ShuangMuSanLin electronics Co., LTD. founded in 2009 which located in Shenzhen, China. And our branch company VMV Technology(HK) Company Limited registered in HK in 2013. Our company is specialized in the audio DAC, stereo headphone amplifier, power amplifier, and with own R&D...
www.smsl-audio.com
binaryblaze
Member
- Joined
- Jun 29, 2024
- Messages
- 15
- Likes
- 12
looks like it on the DO400:
Attachments
- Thread Starter
- #32
As mentioned in the DO400 thread, filter 4 is ESS #4 instead of ESS #3 it is labeled as.
- Thread Starter
- #33
D-6S has no new firmware.
DO100 Pro manual has not been updated.
This is from the updated manual of the DL200. It is correct apart from it being unclear if filter 7 is ESS #7 or ESS #8.
However, this is from the product page. Filter 1 and 6 have the right name but the wrong colour and placement.
Finally, the SU-X. This is from the manual. Filter 4 is mislabeled the same as for the DO400. It is Linear phase fast roll-off low ripple (ESS #4) not Linear phase fast roll-off (ESS #3).
olds1959special
Active Member
I have the DO100 Pro. What's filter setting #1? It sounds the smoothest, at least compared to 4.
AnalogSteph
Major Contributor
Per the above post, that's "Filter off". Since it has some pre-20k rolloff and actually doesn't filter an awful lot, you are expected to use that with software upsampling to a higher sample rate. 88.2/96k minimum, I would probably go to 192, maybe even 384. Leave appropriate extra digital headroom in the player. Are you doing so now? Keep in mind ESS DACs tend to not like overs very much.
- Joined
- Aug 18, 2020
- Messages
- 2,450
- Likes
- 3,802
Absolutely! It will make a complete mess of CD at 44.1kHz as well as content sampled at 48kHz. 88/96 are marginal as well.
- Thread Starter
- #37
The manual has not been updated about the filters. So it should still be ESS#8 - Minimum phase slow roll-off low dispersion.
Similar threads
