ethernet cable audioquest vodka
In general, use Sc/STP or S/FTP in place of STP/FTP which in turn better than UTP.
CAT7 is slightly better than CAT6 which is slightly better than CAT5e
--------
please check the post by klimax users, but I could tell one truth, Linn's digitl player compared with their CD, too many detail lost... CD12 should be their best product.
you might need reference good enough equipments at lease like GENELEC 8050A, best is use 15" woofer speaker hifi system with Bryston AMP(AMP depends,but speaker is the most important thing)...
somebody should say I'm tryingto sell those things above...
As I said, I can't remember what the maximum backoff delay in Ethernet is. I seem to recall that you are allowed up to 16 retries after collisions, with an individual delay of up to 2^10 times the minimum delay. Anyway, let's guess that it could be up to 5 secs. The receiver therefore needs enough storage for 5 secs of audio. Actually, make that 10s just in case we get two problems back-to-back. 44.1kHz x2 (16 bit data) x2 (stereo) x10 is 1.764Mb, so we need 2Mb of buffer. A dedicated LAN with no other devices might get away with much less, but should then make that clear in the instructions.
How much buffer space do real Ethernet-based DACs actually have?
How much buffer space do real Ethernet-based DACs actually have?
Why does marce need a 15" woofer in order to understand Ethernet?powerpan said:
Last edited:
most affection by this weird phenomenon will show up in mid-low frequency range, that's suggestion, not a must.
because no matter how many smaller woofer you use, still could not represent what a bigger woofer could do.
DIY stuff always has no calibration by reference equipments
It is always nice when you lay a trap, put a big notice on it saying "This is a trap!", but the prey still walks straight into it.
I'm not sure I know how to put this more clearly (to someone who might not understand Ethernet). If the Ethernet delivered the packets in the wrong order it still would not affect the audio!! Unless, of course, the device was designed by somebody who doesn't understand Ethernet so made assumptions which are not true; in that case it would seem a little unfair to blame the cable!
I'm not sure I know how to put this more clearly (to someone who might not understand Ethernet). If the Ethernet delivered the packets in the wrong order it still would not affect the audio!! Unless, of course, the device was designed by somebody who doesn't understand Ethernet so made assumptions which are not true; in that case it would seem a little unfair to blame the cable!
I knew a little about TCP/IP potocol, it will send the data untill it's correct or raise an error to OS, while LINN's machine is using UPNP protocol, that's HTTP file server, so, in theory, no error should happen during the transmission. but, it affects the sound.
I'm not here selling something, just wanna talk about this weird thing, should never happen in theory but it did affect the hearing by double blind test.
when I switch the file using a pad in the customer's lab, the engineer knows nothing I'm playing, but he could tell immediatly which is from network, which is from USB stick (2 same song, one is in the USB stick, one in NAS)
before that, I was definitely on your side, dude, because I never believe this kind of saying at all, especially after many stage of FIFO !!!
and, I'm tell everybody here that's attacking my points now, I don't know every detail but I know almost every design detail because our engineer will share all this together and discuss all possibilities to meet the requirement of our customers.
you might be an expert in you major, but we need to know all detail in this kind of products. and hard for software engineer to know hardware, anaalog to know digital, that's all we need to overcome.
and again, please read what I have posted, none of the business of errors during transmission but something else. we have some clue but not sure, need to prove.
you might be an expert in you major, but we need to know all detail in this kind of products. and hard for software engineer to know hardware, anaalog to know digital, that's all we need to overcome.
and again, please read what I have posted, none of the business of errors during transmission but something else. we have some clue but not sure, need to prove.
Ethernet, whatever protocol is used, simply delivers data (and may sometimes fail to do even that). It is up to the receiver to sort the data into the right order, request any missing data, and then use it in whatever way it wants to. If changing an Ethernet cable changes the final result, then there is something wrong with the receiver. It could be grounding, or power supplies, or far too small a buffer but it must be clear that the problem is not caused by the Ethernet cable. All that is doing is exposing a problem elsewhere. Just as cable sensitivity is a sign of poor analogue design, it is also (arguably more so!) a sign of poor digital design or poor software. It saddens me when so many people seem to think the opposite.
Last edited:
Er I use the OB's these days, based on Martin J Kings designs and info, and actually have 4 15" two per side, have a look at the last link, acrive crossover, all the speakers I have built my self over the last several years, before that I had some Kef 104/2 and some tdl MONITOR 1M
http://www.diyaudio.com/forums/full...ge-speaker-photo-gallery-190.html#post2847520
Posts #1894 #1895
Trouble is I enjoy the music, thats why I have a system, I do listen when I change somthing, but can I trust my brain...
http://www.diyaudio.com/forums/full...ge-speaker-photo-gallery-190.html#post2847520
Posts #1894 #1895
Trouble is I enjoy the music, thats why I have a system, I do listen when I change somthing, but can I trust my brain...
I think you should read up "network layers" so you don´t confuse them.
Hint: It´s most likely within the first ten pages of any book about TCP/IP.
you mean physical layer? we know this for sure.
magnetic isolation has no use at all. while you changed the transformer of S/PDIF output transformer, the sound will change.
always twisted wire will transfer data combind with clock(coding), and the receiver will pick the data out of clock, that makes payload low cause header / footer something was added in the package, right?
I was a software engineer and go from top all the way down to IC design and made a VFIR infered transceiver PHY IC...
While S/PDIF is completely unrelated to TCP/IP this claim is nonsense as well.
This really gets annoying.
what I mean is : transformer does affect the data transfer, but to your thinking, TCP/IP will make sure the data is bit perfect, so the sound will not change. at this point, FIBER will transfer the same data as coax does, but sounds different
so, same data will not for sure get the same sound, it depends on "timing"
let me give you another spec of prism sound sound card, it will enhance phase noise by about 60dB (not infinity) from coax input,
analog LDO will have PSRR, will weaken input noise like 60-90db, that means, no matter digital or analog, signal will transfer to the next stage no matter how you "isolate" it, including jitter something, but for sure, affection will weaken but still there.
this problem cause because it includes so many industries, so, that's why hard to setup the theory model.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.