Thank you for bringing this to my attention.
I’ve been streaming DSD256 for about an hour and I see actually in practice that the bandwidth hasn’t exceeds 8 MB/s, i.e. 64 Mb/s.
I don’t have a 512 and I probably won’t have any.
At least until The Next Great Thing comes out, eh? 
Also realize if you at some point wanted to use a Mac as the source, you’d need double that bandwidth for DoP. (Unless you got the 8MB/s with DoP.) You could likely still do it because the Lindemann very probably has flow control on the Ethernet input, but not absolutely ideal.
One last thought (and none of this is to discourage the purchase of the Lindemann if it simply sounds great): Starting at higher speeds (10 Gbps), Ethernet specifications provide not only for a maximum limit on how much jitter an Ethernet connection can have, but also that if the source has higher jitter, the Ethernet connection must reduce it to the specified level or below. So more modern higher speed components limit jitter in a way the older slower speed components are not required to.
1 Like
What objective evidence do you have to corroborate the presumption that the output signal waveform of a Taiko box is cleaner than that of my M2 Max Mac Studio running Audirvāna Studio in my playback system amalgamation, in the contextual juxtaposition of the bit-signal lifted from the storage medium?
The problem we always have in these sorts of technical comparisons, is that there is no reasonable objective information to augment the subjective observations… I understand the nature of accumulated jitter and the influence on the final signal audition, however, there is an extreme amount of cognitive bias involved in the subjective assessment at this level… Especially, when the reference software is restricted to Roon and to the synergistic differentials of host platforms running Roon.
What we audition is the result of accumulated jitter in the digital-audio signal as a product of the amalgamated digital signal-path to the D/A circuitry, through to the transducers used in the audition.
AFAIK, Lindemann Audio use for their devices the streaming boards of
Stream Unlimited:
Looking at these specs they seem not to use the latest streaming board (Stream1955) for their devices but the older one (Stream1832) which has the lower ethernet specs.
1 Like
??? … 100mbps = 10MB/s
DSD256 Stereo = 82.1 MB/min / 60 seconds = 1.37 MB/s
Had you noticed this earlier in the thread?
As for maximum theoretical throughput on a 100Mbps link, … you will never see 100mbps. Bear in mind that this is the “signaling” rate of the line. Aside from your data there are packet headers including data such as a preamble, source and destination address, etc., delays between the time data is sent and a response, and other factors, most importantly the efficiency of the firmware that affect the total real world throughput.
When researching WIFI 6E/7 routers I came across several network specific tech sites that do real world tests on throughput of the wired backhaul between routers for comparative purposes (as well as obviously wireless speeds). Despite all my candidates having 10 Gbps ethernet wired backhaul ports, real world testing of throughput ranged between 6.5 and 9.5 Gbps for the various systems. Just as in wireless router speed the stated speed of wired ethernet is a theoretical maximum, the quality of components, quality of system integration design and quality of firmware results in real world speed which is always less…
Though on the other hand with flow control I did in the past happily run DSD512 over a 100mbps connection.
Yes that is correct I believe, based on this question being asked commonly around streaming forums DSD 512 with headers etc should take up +/- 50 Mbps in transmission speed so it should work as I stated previously over 100 mbps link, however real world throughput rates can introduce issues. The more overhead you have to account for realworld throughput the better. Hence the transition in newer streaming devices to installing 1 Gbps wired ports.
1 Like
@Djm1960 @Jud
Stereo DSD512 transfer rate is 164.2 MB/min… approximately 2.7MB/s …
Does it really matter that 100mbps is the bandwidth of the Ethernet interface of the receiving component? The signal is not dynamic…
When I check the bitrate of my DSD files DSD128 is just over 11 thousand Kbps (i.e 11 mega bits per second) DSD256 files are reported as just over 22000 Kbps. I assume a DSD512 file would have a reported bitrate of +/- 44500 Kbps. This seems to tally with the information on forums of +/- 50 mega bits per second for a DSD512 file including headers etc…, 23 Mbps for a DSD256 file and 11 Mbps for a DSD128 file.
I am no expert in this field however… Just tend to believe what I read in white papers and technical sites…
164.7 MB/min = 164.7/60 for MegaBytes per second then convert to Mega bits per second = 2.7 *8 = 22000 Kbps or 22 Megabits per second x 2 for stereo (seems correct to me according to the stated bit rate). Remember ethernet and wireless speeds are reported in bits per second, not Bytes per second…
Got it … 100mbps = 12.5 MB/s transfer rate… (that’s MegaBytes per Second)
1 Like
1 Like
Eyesight not what it was!
I just got new glasses… 
1 Like
Most of the published internet data seems to level out around the following for a stereo DSD file.
“As for the DSD side, stereo DSD64 is 5.6Mbps, DSD128 goes up to 11.2Mbps, DSD256 is 22.4Mbps, and DSD512 goes to a whopping 44.8Mbps”
Contained in the following link.
These figures seem to match the reported bit rates for the various stereo DSD files I have by Audirvana however there is small variance in the various articles on the internet on this topic.
Which is optimal for audio use, since 100mbps ports have significant lower noise levels.
1 Like
1 Hz = 1 PDM (DSD) bit-sample….
- 2.8MHz = DSD64 (64 x 44.14kHz)
- 5.6MHz = DSD128 (128 x 44.14kHz)
- 11.2MHz = DSD256 (256 x 44.14kHz)
- 22.4MHz = DSD512 (512 x 44.14kHz)
Extrapolate into 2 channels…
My calculation
DSD512 Sample Rate: DSD512 uses a sample rate of 22.5792 MHz
One-Bit Audio: DSD is a one-bit audio format, meaning each sample is represented by a single bit of data.
Stereo Channels: Stereo audio requires two channels
Calculation: 22.5792 MHz * 2 (channels) * 1 bit/sample = 45.1584 Mbps.
Adding overhead for data and communication brings this up to around 50 Mbps.
If my logic is incorrect, apologies. Not an expert………
3 Likes
I would be interested in seeing data from currently available equipment if you have it.
Edit: Particularly from 10G and above equipment, including optical (electrical noise of optical-electronic converter).
1 Like