Introducing Olympus & Olympus I/O - A new perspective on modern music playback
- Thread starter Taiko Audio
- Start date
Christiaan Punter
Well-Known Member
I am referring to an update to the BMS/BPS logic that was implemented in the early days. All Olympus products are up to date.
Christiaan Punter
Well-Known Member
Here is a diagram that illustrates the performance level of local vs NAS vs streaming with and without the Extreme DC Power Distributor and Extreme Router.
Last edited:
Thanks for the helpful chart. I have an Olympus, DCD, router and switch. To clarify, do the "Superlative" benefits require removing the local storage physically from the Olympus, or are the benefits also available by streaming files on NAS / the cloud even if the physical local storage remains in the Olympus? I recall some discussion on the benefits of removing the physical music storage drive from the Olympus altogether but that's not my preference from a practical point of view in my case.
Christiaan Punter
Well-Known Member
The chart remains valid regardless of whether you have internal storage or not. That said, adding an internal storage drive to the Olympus very slightly degrades overall SQ compared to using no storage and relying solely on a NAS. This is a minute difference, far less than the effect of interlinks or the use or absence of the Switch, or even the DCD alone. But you can hear it on a highly resolving system. The difference is so slight that I opted not to include it in the chart, as it would skew the other increments. Emile rates this at a 1-2% impact, and adds that it is not enough to warrant concern.
I'd also affirm what Christiaan has written. I haven't heard any degradation of sound quality with XDMI battery state of charge. I kind of watched/listened for this early on but couldn't detect a difference. So I don't even think about it now.
Steve Z
What is the next step up from superlative? My experience (with others) has been that local WAV playback is *much* better than cloud streaming (which of course is flac based). We need another descriptor.
Indeed! The FLAC vs WAV (or more accurately compressed vs uncompressed) is clearly audible and in some cases very significant in SQ. Some people call it a mystery, but it should not be a mystery. While I have no tech knowledge, intuitively de-compression in real time introduces jitters (in a broad sense) and noise as the compressed files cannot be played without being de-compressed - and such de-compression happens in real time. Some compression level can be super aggressive, FLAC being compressed to 1/4 of the uncompressed WAV size. Not having done any experiment yet with Olympus, I continue to doubt that cloud streaming (which seems to be always via compressed files) can be as good as, let alone being better than, local playback with uncompressed files (WAV). Taiko can reverse its prior position (held in last Oct when O and I/O had been finalized), but please do not chalk it up to readers or users as being "confused" (but for a couple of "confused" words in the email I received, I would not have posted again on the subject of the Router). Nobody should be held to be confused when the words the Router having "no effect for local playback" were used - how could such words not be crystal clear and instead cause confusion?
I completely understand the relative and subjective nature of the diagram above — and if today there is superlative, what will there be tomorrow (transcendent?)
For now it shows that the obtained SQ is respectively superlative or fantastic when the Router is combined with DCD. And even without DCD, it remains fantastic and superb?
This seems to suggest that in both scenarios, with and without DCD, cloud streaming SQ is elevated above local file playback — thanks to the Router.
Last edited:
I would not draw that conclusion without further clarifications.
If both local and cloud streaming are in flac format, that may in fact be true. I have not done that comparison in a long time but they were close a few years ago.
When both are WAV files, it may or may not be true but we don't know because all cloud streaming (that I've used) is losslessly flac compressed and that is well below the SQ of local WAV file playback.
So, based on my listening experiences, I interpret Christian's chart as being for flac playback for all the comparisons.
That’s my understanding as well — cloud streaming refers to FLAC files. However, one must also consider whether the Taiko Router (or full Taiko network stack) and DCD are present or not.
Since according to @ray-dude (Oct. 2024) ‘Skip internal storage on the Olympus Server, streaming music or playing music from your NAS or file server sounds better than music on internal storage’. ‘Stick with Redbook or 96k music for the best sound quality on Olympus’
And @Christiaan Punter ’adding an internal storage drive to the Olympus very slightly degrades overall SQ compared to using no storage and relying solely on a NAS’. Did you try? You use XDMI and not USB when local wav sounds best?
Last edited:
My recollection is that it has to do with the power needed for the internal drive. That power affects sound quality. There is no computer power needed for a NAS. It is external to the system.
Also, from a tech layperson's perspective, I would guess that NAS going through some cable connection or the innards of a router is not a 100% transparent passage and such extra connection or passage would introduce some jitters itself. The only way is to rely on one's own ears (and I may run tests later). My own experience is that WAV sounds noticeably better than FLAC, PCM noticeably better than DSD (a bit dull and bottom heavy), and hi-res files in excess of 24/96 are invariably worse than 24/96 or 24/48 (I listen to lots of large-scale complex orchestral and choral music; 24/192 is never good for me). If a hi-res album which is re-mastered (many classic classical recordings were made in the heydays of late 1950s to 1970s) sounds much better in a recent reissue than the first CD remastering (in the early 1980s) it is, I guess, because of the more careful remastering (via original tapes), not because of hi-res format (which is hyped for marketing reasons).
yes, xdmi and usb both showed the wav/flac delta in SQ, but the delta is not slight, and in both cases the full router/switch/dcd stack were used. the same conclusion was found in extreme and olympus based systems.
One notable SACD/DSD exception. Somehow Esoteric's extremely careful remastering of a classic recording (none better than Klemperer's classic recording of Ein deutches Requiem by Brahms recorded by EMI in 1962 or so) is better than EMI's own remastering at 24/96. Other recent SACDs (ripped to DSD files) are disappointing, such as Beethoven's String Quartets cycle by Tokyo Quartet (its 2nd recording shortly before disbanding) on Harmonia Mundi; I ended up buying the Red Book versions (which sound more sparkling than the much more expensive SACD/DSD versions). Also disappointing is Mark Wigglesworth's cycle of Shostakovich symphonies issued by BIS on SACD for another example.
Attachments
Last edited:
File transfer from local drives is inferior to TCP/IP transfer on the Olympus when combined with the router.
This is due to a protocol limitation—we cannot modify the local drive protocol to a sufficient degree. However, we can manipulate the TCP/IP data stream to a more meaningful extent.
A NAS acts like a local storage to tcp/ip transfer converter.
The DCD is mentioned because it appears to be a critical component for achieving audiophile-level sound quality with the vast majority of “audiophile” power supplies on the market. If your power supply is of high enough quality, the DCD may not offer an upgrade and can be omitted. That said, I have yet to encounter one that meets that threshold. I would have expected several companies to have incorporated this technology into their designs by now. It seems we are among the most copied companies in the space—but oddly, not in this particular area, despite the fact that it’s not rocket science. Clean power is however more or less a requirement with SFP copper DAC (Direct-Attach) links, which is why we keep mentioning it.
The switch is not mentioned, as it’s not a critical component in achieving this goal—it has no intelligence that affects the data stream.
By the way, this mechanism is why XDMS performs worse than Roon on the Olympus + Router/DCD setup. Bypassing this mechanism would require a significant rewrite of one of XDMS’s core systems. As an alternative, we’ll be offering UPnP support soon. It’s actually ready, but we’re holding off on release to create some breathing room in our current workload, which continues to shift. At the moment, I’m spending most of my time sourcing critical components that went out of stock in many regions after the tariffs were announced.
This is due to a protocol limitation—we cannot modify the local drive protocol to a sufficient degree. However, we can manipulate the TCP/IP data stream to a more meaningful extent.
A NAS acts like a local storage to tcp/ip transfer converter.
The DCD is mentioned because it appears to be a critical component for achieving audiophile-level sound quality with the vast majority of “audiophile” power supplies on the market. If your power supply is of high enough quality, the DCD may not offer an upgrade and can be omitted. That said, I have yet to encounter one that meets that threshold. I would have expected several companies to have incorporated this technology into their designs by now. It seems we are among the most copied companies in the space—but oddly, not in this particular area, despite the fact that it’s not rocket science. Clean power is however more or less a requirement with SFP copper DAC (Direct-Attach) links, which is why we keep mentioning it.
The switch is not mentioned, as it’s not a critical component in achieving this goal—it has no intelligence that affects the data stream.
By the way, this mechanism is why XDMS performs worse than Roon on the Olympus + Router/DCD setup. Bypassing this mechanism would require a significant rewrite of one of XDMS’s core systems. As an alternative, we’ll be offering UPnP support soon. It’s actually ready, but we’re holding off on release to create some breathing room in our current workload, which continues to shift. At the moment, I’m spending most of my time sourcing critical components that went out of stock in many regions after the tariffs were announced.
Similar threads
- Sticky
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |