SPDIF has inherent jitter issues. This was known from the start, it was chosen because of cheap cabling/low cost.
Let's just say that both USB and SPDIF can perform very well with recent techniques. It is not good versus bad.
Last edited:
To me.... I never heard any problems with SPDIF/coax - even optical/toslink.
I believe that USB just moved forward because of the use of computers/streaming and the hype of extreme bit/sampling rate exploration.
Jitter is roughly just a high frequency version of the low frequency "wobble"from tape and LP's. which means - it's simply a minute tiny bit of noise/hiss along with the signal - therfore mostly a wash.
If you just build SPDIF like Allo - then I cant see the problem:
Raspberry Pi audio DACs, Digi and amplifiers
And if in doubt how digital signals work - see this great nerdy video from Christopher Montgomer:
A Digital Media Primer For Geeks by Christopher "Monty" Montgomery and Xiph.org - YouTube
D/A and A/D | Digital Show and Tell (Monty Montgomery @ xiph.org) - YouTube
I for one, would like to see more toslink/optical, now that diodes have become better and switch-mode power supplys are introducing noise in the name of power, space and weight saving.
I believe that USB just moved forward because of the use of computers/streaming and the hype of extreme bit/sampling rate exploration.
Jitter is roughly just a high frequency version of the low frequency "wobble"from tape and LP's. which means - it's simply a minute tiny bit of noise/hiss along with the signal - therfore mostly a wash.
If you just build SPDIF like Allo - then I cant see the problem:
Raspberry Pi audio DACs, Digi and amplifiers
And if in doubt how digital signals work - see this great nerdy video from Christopher Montgomer:
A Digital Media Primer For Geeks by Christopher "Monty" Montgomery and Xiph.org - YouTube
D/A and A/D | Digital Show and Tell (Monty Montgomery @ xiph.org) - YouTube
I for one, would like to see more toslink/optical, now that diodes have become better and switch-mode power supplys are introducing noise in the name of power, space and weight saving.
Last edited:
That you did not hear them does not take away that SPDIF was a troubled interface from the start as the clock signal is multiplexed with the data and needs to be demultiplexed at the receiver side. Jitter and PLL quality more or less explain the differences experienced/measured between SPDIF receivers that came and went. The YM3623 for instance has typical jitter of 5 ns/5000 ps and it was way worse with regards to jitter as the internal DAC in CD players 
Toslink started with deforming square waves and a clearly less good result, then it was SPDIF over coax that was nearly always better. Most DIYers that started early with digital audio can tell details on improvements in the SPDIF interface like the transformers, the signal wave form etc etc. Some devices had absurdly bad wave forms like the CEC expensive transports. Improving this had severe impact on audio quality.
Your vision on jitter is plain wrong, sorry. It is NOT "simply a minute tiny bit of noise/hiss along with the signal and therefor mostly a wash".
Toslink started with deforming square waves and a clearly less good result, then it was SPDIF over coax that was nearly always better. Most DIYers that started early with digital audio can tell details on improvements in the SPDIF interface like the transformers, the signal wave form etc etc. Some devices had absurdly bad wave forms like the CEC expensive transports. Improving this had severe impact on audio quality.Your vision on jitter is plain wrong, sorry. It is NOT "simply a minute tiny bit of noise/hiss along with the signal and therefor mostly a wash".
Last edited:
From what I recall (this is going back to sometime around 1990) if you used it literally according to the DS its jitter was truly horrendous (maybe even into double digits nS). But I found a way to tame it, I think to below 1nS. Still by no means great of course but a huge improvement from 'stock'.
Fair enough, but I simply never heard of the problem, besides decades ago, when the problem was of bigger scale.
I heard and read endless discussions about it - subjective observations. But never heard what it could truly do in practical terms.
So please, what - in my vision - is wrong?
Your vision is wrong as already written. It is inherent with the SPDIF standard and maxing it out makes it OK but not excellent. Anyone having tried I2S knows what a difference that makes.
You could change the attitude of ignoring a well known technical drawback of a certain interface and study it just like the competitors. Then also compare a known jittery SPDIF receiver with a known low jitter SPDIF receiver to the same DAC playing the same material and experience what the issues are both in measurements and in sonical results. Then it will be clear what is meant.
Having said that, SPDIF may be far from ideal, it is still being used extensively also by the people that hate it If one wants to ditch it a source with built in filter/DAC wins hands down compared to having to use USB/SPDIF or whatever interface. So when possible: keep it to I2S over short distances in a 1 box solution with good clocks. One then suddenly realises that less is more.
You could change the attitude of ignoring a well known technical drawback of a certain interface and study it just like the competitors. Then also compare a known jittery SPDIF receiver with a known low jitter SPDIF receiver to the same DAC playing the same material and experience what the issues are both in measurements and in sonical results. Then it will be clear what is meant.
Having said that, SPDIF may be far from ideal, it is still being used extensively also by the people that hate it
If one wants to ditch it a source with built in filter/DAC wins hands down compared to having to use USB/SPDIF or whatever interface. So when possible: keep it to I2S over short distances in a 1 box solution with good clocks. One then suddenly realises that less is more.
Last edited:
That is only because you are discussing tech from 20 years ago. Pass usb and i2s through a fifo with a local oscillator and they become very comparable with some noise advantage on the side of spdif.
I don't personally have much use for spdif but facts are facts. If one does not need resolution above 192/24 spdif is a viable choice. And optical is even better.
Ain't that the truth! Never hated an interface more. Nor could i get why perfectly functional cdplayers needed to be separated into two boxes with this misunderstanding in between.The transport-dac invention never really worked for me, neither with high end commercial boxes, nor with my diy attempts.
It does not harm to be able to play DSD either. Many thought the same about 24/96.... Progressive is progress regardless what one thinks of it. When I see audio guys I know playing high res files on their expensive high res streamers to red book DACs I always smile Often the high res streamer has analog outputs....that have never even been tried out.
Again: a lot more things are possible when things are integrated and USB/SPDIF can be omitted. Way less jitter and better signal quality compared to ingenuous solutions to fight jitter. Forget the separate source/separate DAC model.
Often the high res streamer has analog outputs....that have never even been tried out.Again: a lot more things are possible when things are integrated and USB/SPDIF can be omitted. Way less jitter and better signal quality compared to ingenuous solutions to fight jitter. Forget the separate source/separate DAC model.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.