Is he swapping the clock in the DAC when taking these measurements? Also is he using a Keysight 5052B when capturing these measurements? Like I said, if you’re designing a DAC just to fool the layman, it’s much better to go for higher numbers. Because it’s cheaper and looks much better on paper to the untrained eye. Besides they will never know any different because they will never have an example of what the same DAC could have been with engineers designing it that weren’t targeting laymen who didn’t know any better.
Once again, it doesn’t take extraordinary skills for an engineer to choose a higher jitter clock and support higher frequency data,. It only takes an marketing department telling the engineers that audiophiles are idiots for it to make sense.
Fiber is a great solution to stop the charlatans. They can’t improve on things upstream of glass and light. Unless it’s with DSP manipulation of the signal. Jitter isn’t an issue with Ethernet. Because the data isn’t clocked until the endpoint. Unlike USB. And it’s impossible for noise to travel through the light and glass. All that makes it through is the data.
if jitter mattered with Ethernet, just imagine how bad Tidal streaming would sound passing though 10000 switches and fiber to copper converters between the main servers in Oslo Norway and your home. The folks in Norway would have way better sound than the folks in Australia!
Yes I know they are. But you are not grasping that they are all taken with a low grade scope, and the exact same DAC, with the exact same clock. Yes with that DAC and that clock it may be better. But what you aren’t seeing is the jitter comparisons if that exact same DAC used a much lower jitter clock. The end consumer will never have the privilege of making this comparison. This is why I suggested doing it with the DIYINHK USB interface. Because the DSD 256 version and the DSD 1024 version are identical besides the clock frequency, and the color of the solder mask.
Understood. But I thought you meant earlier that for a given DAC, DSD256 typically out performs DSD512 unless a really good clock is used to help with DSD512’s jitter.
From those measurements DSD512 performs better for that particular DAC no?
So do those measurements suggest iFi have done a good job taming the jitter of DSD512 for that DAC?
Again I’m only asking. I may be looking at things too simply.
I agree, that if you own that IFI DAC, that upsampling to DSD 512 can get you better performance than upsamping to 256 with the same DAC. I’m not questioning that. But DAC building is all about balancing the pros/cons. Once you get to DSD 256, you can push the noise floor down below 120db beyond 100Khz, as well as use filter cutoffs that have phase shifts well above the audible range anyways. So leaning towards the lower jitter of the lower frequency clocks yields a better end result, than pushing the noise further out of the audible range.
Sure if the laws of physics didn’t exist, there would be no compromises going with DSD 512 over 256. But unfortunately reality rears it’s ugly head.
Understood. As you say we need to pay attention with claims of DSD512 support and now DSD1024 support however in this case it appears iFi have achieved measured performance improvements with DSD512 over DSD256 (and my ears certainly agree too).
Once again did you compare the performance with the same DAC equipped with lower jitter clocks? It seems you aren’t understanding what I’m telling you. Surely others reading our conversation have absorbed the clear logic I’ve articulated. But I guess I shouldn’t be so sure.
Apologies, I may not understand. If you mean that using better clocks will result in even lower jitter, I thought that’s intuitive anyway? For almost any product?
You have no way of knowing as a layman end user if the engineering team who designed your DAC could have obtained better performance had they chosen lower jitter clock in the design phase. All you can base things off is the end results of what the engineers did choose. Which were decisions based more on marketing appeal than sound quality. And as you are clearly demonstrating. The marketeers did their homework very well
Anyways I’ve beat this dead horse hard enough. Either folks grasp what I’m saying by now, or sleep easy knowing ignorance is bliss.
I understand all that but you said "It’s not just a simple matter of higher is better. It doesn’t take more expense or knowledge to support DSD 1024. DAC manufacturers just don’t do it because they know the compromises involved. So if you see something with that high of support, keep in mind they only did it for marketing purposes, not for better sound. "
But in this case with my iDSD, DSD512 appears to out perform DSD256.
If DSD256 out performed DSD512 then yes I understand your point about the marketing department perhaps trying to fool people about support for DSD512. But the measurements show it’s not just marketing right? For the iDSD.
[moderated]
I’ll just wait and see what this, indeed interesting new OEM product, will bring to the table and ‘change strategy’ depending on the SQ results out of real world products.
Jitter isn’t an issue? I think so too but telll that to these manufacturers. A lot more of these devices will apear in the near future. Every USB decrapifier device that is out there at the moment is going to get it’s ethernet counterpart.
Wait till you hear a DSD 256 5.1 channel recording through a fully active 3 way 5.1 channel system (with no downsampling to PCM). All you’ll need is a NUC, switch, 6 fiber Ethernet cables, and the 6 speaker boxes (7 with 2 subs), and you’ll be in action. The Zman is the only board on the market that allows a system like this to be possible.
