Asynchronous I2S FIFO project, an ultimate weapon to fight the jitter

[hochopeper]

Originally Posted by iancanada

Thank you so much Chris. I appreciate your efforts to the community!

Now I'm focusing on the Si570 project, I know a lot of friends are waiting for this small board (just got V2.0 PCB ). I'll go through the wiki and add more things.

Any help to the wiki is more than welcome .

Ian

If you want to give it a day or so Ian I'll add a bunch of info to each heading and at least that may give a starting point/framework for you and anyone else wanting to contribute.

I think the big advantage of the wiki is that it can be managed by the community and be one less thing that you need to worry about. We'll manage wiki/index document and the intelligent work is still all done by you


Cheers,
Chris

[Adrculda]

Originally Posted by iancanada

Thank you so much Chris. I appreciate your efforts to the community!

Now I'm focusing on the Si570 project, I know a lot of friends are waiting for this small board (just got V2.0 PCB ). I'll go through the wiki and add more things.

Any help to the wiki is more than welcome .

Ian

Good...
You do that :P

[iancanada]

I did jitter measurement on MCLK of both WM8805 output and FIFO output yesterday. I'll public those screen shots. I agree testing report is more important than telling a story. But I'm afraid I don't have time posting the result today. Will do it as soon as possible.

Another thing I have to mention is, this project is not an open source project, though I have been sharing information with community as much as I can. If there are something I didn't public, that means I reserved them for myself, at least for now. I thank you for your kind understanding.

Ian

[iancanada]

Originally Posted by Adrculda

Good...
You do that :P

Thanks Adrculda for your great suggestion.

Ian

[iancanada]

Originally Posted by hochopeper

If you want to give it a day or so Ian I'll add a bunch of info to each heading and at least that may give a starting point/framework for you and anyone else wanting to contribute.

I think the big advantage of the wiki is that it can be managed by the community and be one less thing that you need to worry about. We'll manage wiki/index document and the intelligent work is still all done by you


Cheers,
Chris

Thanks hochopeper, I will.

But I have to say the new V2.0 Si570 board sounds really great, even better than V1.0 for sure. I'm afraid I have to spend a little bit time on...enjoy my music.

Ian

[qusp]

I wil be more interested in the test setup than the actual jitter measurement, you can confirm its as expected? a simple yes/no will suffice

[TNT]

Originally Posted by iancanada

Thanks hochopeper, I will.

But I have to say the new V2.0 Si570 board sounds really great, even better than V1.0 for sure. I'm afraid I have to spend a little bit time on...enjoy my music.

Ian

Like!

[regal]

Originally Posted by Fridrik

Update: just checked, if I plug my source thru my HDMI switch I'm limited to 44.1khz and 48Khz... If a plug my received with no switch my NVIDIA High definition audio card detect 96khz availability

An even if you send Dolby True HD on the HDMI channel the SPDIF out will probably only send 2 channel 48khz...

One great HDMI compagnon for the FIFO would be the HDfury HDMI to component adapter.

whats the deal with that? I have an old receiver that doesn't decode Dolby True HD or DTS HD, but my computer decodes it fine to LCPM hirez 24/192 using the LAV decoder, we can use our computers to send hirez thru HDMI since the proprietary acttion (decoding) is done prior to the output of the computer in HTPC geek terms not bit streaming.

I can also send 24/96 flac files thru HDMI with most software music players. It is so convenient but has terrible jitter, the fifo would be a nice cure. Really would be awesome to see a multi-channel fifo but can't see it being much use without an HDMI input of some-sort, its just what the industry has gone to.

USB 2.0 Hi-speed audio spec is dead, Intel and M$ are dropping it completely since the USB 3.0 spec includes asynchronous making UAC2 obsolete(mac will have too also with Broadwell). It will die faster than fire-wire driver support will be nill in a few years, but hdmi is here to stay.

[qusp]

sorry I think thats pretty funny, whether or not other interfaces start to be used, USB2 support will continue for many years to come, there is simply far too much hobbiest and professional audio gear that uses it and USB3 connectors can host it afaik (if anything its usb3 that will dodo, as it doesnt compete with truly high speed devices)

vinyl was supposed to be vanquished decades ago too

HDMI is not a replacement due to license and buy in cash and restricted environment and thunderbolt will support all of them at native speeds and protocols via adapters.

enough of the scare-mongering hey? just because windows still hasnt managed to support it with a proper driver after years of other companies/platforms doing so. hell even android supports it better than windows

[hochopeper]

I know very little about USB3 but when it comes to UAC2 over USB3 .... XMOS don't support it.

https://www.xmos.com/download/public...e%281.0%29.pdf

3.5 USB 3.0
Use of the UAC2 Reference Designs with USB 3.0 ports is not supported.
When a high-speed USB 2.0 device is plugged into a USB 3.0 port (blue socket) it is
connected to a USB 3.0 xHCI controller. At this time Microsoft has not yet released
an xHCI driver stack and insteadaA third party bus driver such as the NEC/Renesas
driver will be installed on the host PC. However, there are quite a number of third
party bus drivers and at this time XMOS does not test against any of them.

If future motherboards don't actually have USB2 controllers then ...... who knows what the future may bring. I agree that HDMI is shifting far higher volumes of gear that the rest of us may eventually end up on it or something similar, not tomorrow though and surely not willingly!!

Maybe that is just related to xmos windows driver?? I think more reading is needed.