24 bit source material
Quote from mikronom on 15 June 2025, 8:05 PMHello Pedja,
I saw Mathew's review of the S5b, and I am intrigued. I like the simplicity and that it's single ended by design. It would fit quite nicely into my chain. The one thing I am concerned about is the question, how future-proof a 16 bit dac is in 2025. Mathew mentions the need to dither any 24 bit source material to 16 bit before sending it to the dac to avoid truncation. If I understand correctly, is neither my own streamer (Musical Fidelity MX-Stream) nor Mathews Holo Audio red capable of doing that. Did I miss something? How does the S5b handle 24 bit source material?
I am using Tidal and local files on a NAS through Roon and I see more and more 24 bit material, that is why I am wondering. In case I missed the blog that would answer my question, just point me towards it 🙂
Thanks and best regards
Miro
Hello Pedja,
I saw Mathew's review of the S5b, and I am intrigued. I like the simplicity and that it's single ended by design. It would fit quite nicely into my chain. The one thing I am concerned about is the question, how future-proof a 16 bit dac is in 2025. Mathew mentions the need to dither any 24 bit source material to 16 bit before sending it to the dac to avoid truncation. If I understand correctly, is neither my own streamer (Musical Fidelity MX-Stream) nor Mathews Holo Audio red capable of doing that. Did I miss something? How does the S5b handle 24 bit source material?
I am using Tidal and local files on a NAS through Roon and I see more and more 24 bit material, that is why I am wondering. In case I missed the blog that would answer my question, just point me towards it 🙂
Thanks and best regards
Miro

Quote from Pedja on 16 June 2025, 6:08 PMHello Miro,
The TDA1541A chip is, of course, a 16-bit converter, and the S5b DAC is made to convert the first 16 bits and truncate longer binary words. So, there is no dither inside the S5b.
The S5b USB input is a bit of a "special case", since the USB interface is bidirectional, and the source receives back the information about the DAC. To avoid forcing the source to dither down the longer binary words, i.e. to avoid the dither applied unintentionally, the S5b USB stage enumerates itself as a 24-bit unit, and thus the 24-bit source streams also go intact right to the DAC.
So, to have the bottom 8 bits of 24-bit sources included in the D/A conversion by this 16-bit DAC, you do need to perform the dither on the source side.
I am unfamiliar with both Musical Fidelity MX-Stream and Holo Red, however since the dithering comes down to the software, and should not be that complicated, I would expect the streamer manufacturers to provide some dithering options, or to allow some workarounds. But, without dithering, and with 24-bit music sources, how does 16-bit DAC fare?
First, is dithering itself beneficial or not, I am not sure if there is a consensus. Also, there are different dithering techniques, and they do sound different. The dither was all the rage when it was introduced, but later people got more cautious about it, so its use has its history. Ultimately, it is on you to make your choice.
I use entirely PC and NAS-based solutions myself, and dithering might be easier with such hardware than with dedicated streamers, yet I usually do not use dither with S5b to play 24-bit sources, generally preferring bit-perfect (bit-transparent) approach, in all areas. So, I truncate them instead.
But, since I am now regularly receiving questions of this kind and this topic is recently apparently becoming more intriguing, it might be the right time to reconsider this approach here, and do more dedicated listening, compare different dithering types in a controlled way, and characterize them also against non-dithered playback.
I will try to do this in the next few weeks, and get back with the results sometime in July.
Regards
Hello Miro,
The TDA1541A chip is, of course, a 16-bit converter, and the S5b DAC is made to convert the first 16 bits and truncate longer binary words. So, there is no dither inside the S5b.
The S5b USB input is a bit of a "special case", since the USB interface is bidirectional, and the source receives back the information about the DAC. To avoid forcing the source to dither down the longer binary words, i.e. to avoid the dither applied unintentionally, the S5b USB stage enumerates itself as a 24-bit unit, and thus the 24-bit source streams also go intact right to the DAC.
So, to have the bottom 8 bits of 24-bit sources included in the D/A conversion by this 16-bit DAC, you do need to perform the dither on the source side.
I am unfamiliar with both Musical Fidelity MX-Stream and Holo Red, however since the dithering comes down to the software, and should not be that complicated, I would expect the streamer manufacturers to provide some dithering options, or to allow some workarounds. But, without dithering, and with 24-bit music sources, how does 16-bit DAC fare?
First, is dithering itself beneficial or not, I am not sure if there is a consensus. Also, there are different dithering techniques, and they do sound different. The dither was all the rage when it was introduced, but later people got more cautious about it, so its use has its history. Ultimately, it is on you to make your choice.
I use entirely PC and NAS-based solutions myself, and dithering might be easier with such hardware than with dedicated streamers, yet I usually do not use dither with S5b to play 24-bit sources, generally preferring bit-perfect (bit-transparent) approach, in all areas. So, I truncate them instead.
But, since I am now regularly receiving questions of this kind and this topic is recently apparently becoming more intriguing, it might be the right time to reconsider this approach here, and do more dedicated listening, compare different dithering types in a controlled way, and characterize them also against non-dithered playback.
I will try to do this in the next few weeks, and get back with the results sometime in July.
Regards
Quote from mikronom on 17 June 2025, 10:14 AMThank you, Pedja, for the clarification!
In that case, truncation seems like a viable option that I am not too worried about. A comparison of different dithering methods with non-dithering would be interesting, of course.
I will need some time to consider my options. I'll get back to you after the summer.
Best regards
Thank you, Pedja, for the clarification!
In that case, truncation seems like a viable option that I am not too worried about. A comparison of different dithering methods with non-dithering would be interesting, of course.
I will need some time to consider my options. I'll get back to you after the summer.
Best regards

Quote from Pedja on 30 July 2025, 6:56 PMHello Miro,
I opened a separate topic on dither, and I will add more posts to it in the next few days. Hope it will be useful.
https://www.audialonline.com/community/topic/about-dither-on-a-practical-level/
Regards
Hello Miro,
I opened a separate topic on dither, and I will add more posts to it in the next few days. Hope it will be useful.
https://www.audialonline.com/community/topic/about-dither-on-a-practical-level/
Regards