jitter free s/pdif for imac

[Nikos]

Hi,

I have a nonos dac with s/pdif input.

I would like to use my imac as a music server and feed the dac with "jitter free" signal via spdif.

My dac is very sensitive to jitter (the imac optical with optical to spdif converter sound is not acceptable...)

I would like to be able to use 24/96 as well.

I am looking for sugestions up to 1000$ 1200$

I have an esoteric sa-60 that I use right now as a cd transport and I hope the computer server will be at least as good.

thanks alot

Nikos

 

 

[Audio_ELF]

There are various interfaces you could consider ...

 

At the low priced end many people are having success with the HiFace USB to SPDIF converter - I think this is around £100.

Higher up people use interfaces such as the TC Konnekt 8 (think this is now discontinued) which cost around £300. Also the SonicWeld interface mentioned recently or (soon to come) Wavelength's Async USB to SPDIF interface.

At the top end you could look at the Weiss INT202 - this is probably pushing the top of your budget.

 

Eloise

 

[Nikos]

A lot of people are using the hiface and other similar products with great success but they have dacs capable of eliminating jitter. I don't. I could use a reclocker but I prefer to use something that is not produce jitter instead of produce and then eliminate it.

when the wavelength asynch usb to spdif will be available?

Any estimation of it's price?

Is the firewire interface and the clocking of int202 at least equal with the asynch usb of wavelngth/ayre?

thank you for your help

 

 

 

 

[Wavelength]

Nikos,

 

There is and never will be a jitter free SPDIF. Just from the way it works there is no way it can be. Not even if the clock is feed back from the dac to the transport.

 

Anytime the master clock has to change or vary because of the interface then that in itself makes jitter. SPDIF requires a variable clock and therefore cannot have zero jitter. Well really nothing can have zero jitter.

 

The toslink jitter varies with several of the macs I have from 1100ps to 1700ps. If you use some external USB or Firewire converters you may get it below 1000ps.

 

I have measured about 80 devices with my Prism and even the best devices cannot get below 300ps of jitter PP on SPDIF.

 

Your results depends on how well your dac get's rid of the jitter.

 

Thanks

Gordon

 

[Nikos]

Mr Rankin,

Thank you so much for your response..

I know that "jitter free" spdif is not possible.

your comment about 300ps is quite interesting.

I am very pleased with my nonos dac and I would like to keep it (at least for now...)

please could you suggest which product could give me a low jitter spdif signal?

thank you again

Nikos

 

 

[Elprior]

Hi Gordon,

 

John Atkinson did measure better values.

On the Nagra Dac for instance (quoting his article) :

"Driving the Nagra DAC with S/PDIF data from a PS Audio Lambda CD transport gave a measured jitter level of just 140 picoseconds peak-peak, which is superbly low."

The BelCanto USB Link he measured at 5850ps, for another example.

 

What's so different in the way the measures are done (including yours) ?

 

Thanks,

Elp

 

[Wavelength]

Elp,

 

But that is the dac response not the SPDIF jitter.

 

Guys... look it maybe a little confusing about jitter and what John has measured now and in the past. When John used the Miller unit it was capable of a numerical amount of jitter by looking at the spuriae (those vertical spikes) on the response to a test tone called the JTEST.

 

The Miller was unable to test SPDIF jitter at all as it was merely a DAC and ADC with some good software on a PC.

 

Now John has at his disposal the Audio Precision 27xx series with digital option which does the same thing as the Miller did but it also allows for testing SPDIF directly. Now the big problem with the 27xx series is the finite amount of jitter or the floor for which it can test is only like 1000-1200ps.

 

The other issue with the 27xx series and really any other test set other than the Miller is that it will not give you a numerical number for jitter by looking at the spuriae. You can calculate this using Juliann Dunn's method's outlined in a tech note available on the Audio Precision site.

 

The unit I use is the Prism dScope III which is capable of less than 300ps of jitter measurement on any of the SPDIF ports. The new Standford SR1 is also capable with the digital option of testing to 500ps of jitter.

 

Thanks

Gordon

 

[Audio_ELF]

I think we can all agree that truely ZERO jitter is not possible (with SPDIF?) and where marketing claim "zero jitter" they mean jitter at a level which will not (to their belief) affect the sound quality at the analogue output.

 

However I think MANY DAC manufacturers have shown it is possible to get good sound using the SPDIF interface (contry to what Gordon and others with a vested interest in other technologies would have you believe) and what we are looking for when we say "jitter free" is something which has a low enough level of jitter at the SPDIF output to enable these DACs to shine.

 

To this end, devices such as Lynx AES16, RME 9632 and HDSPe card, FireWire interfaces such as Weiss INT202 and AFI1 and USB interfaces such as the HiFace and SonicWeld can provide a perfectly acceptible alternative to throwqing away your DAC and buying from Metric Halo or WaveLength's one of their (undoubtably good) DACs.

 

Eloise

 

[Elprior]

Right, thanks for the explanations.

I could not believe there was so many disparities within the different measurements.

 

I guess there's no right or wrong here, but rather knowing how to read those numbers (within their environments).

Since I am not able to do so, I'll stick to what I have been doing so far, that is, not reading the numbers (apart from the price tag)

 

Elp

 

[EDIT] Eloise, agreed. The WaveLink does qualify too

[EDIT] Gordon I can't resist it : how does the WaveLink perform with your measurements ?

 

[topher_m]

Hi

 

Original poster asked the same thing but I don't think I saw a response.

 

Can you fill us in on the availability of the new interface?

 

Thanks

 

[bordin]

Hi Nikos,

 

I just had a chance to compare my friend's Hiface (Async usb-2-spdif) and Konnekt 24D (Firewire) with ones I have, i.e. Lynx AES16e PCIe, M-Audio Audiophile USB, and Metric Halo ULN-2 (Firewire Mac only).

 

The S/PDIF output was fed through my Oritek X-1/D cable to my OMZ DAC/HEADAMP + HD800 w/ APS V3.

 

The following list is the rank, from my personal judgement, corresponding to the performance level (transparency, tonal balance, extension, speed, dynamic range & contrast, etc) of computer's S/PDIF audio interface - better to worse.

 

1. Metric Halo ULN-2

2. Lynx AES16e

3. Konnekt 24D

4. Hiface

5. M-Audio Audiophile USB

 

It is very apparent that the AES16e and the Konnekt 24D sound better on Mac OSX.

 

Weiss INT202 and RME Fireface are also among your choices.

 

My 2 cents.

 

[Mr.C]

I wonder if the BNC HiFace would have an advantage...

 

[JR_Audio]

As I have stated also at some other posts here at CA and as Gordon stated also here, that there are some many kinds of jitter numbers, that one should really be more exact, what number of jitter he is referring to.

 

What is right, that the SPDIF (or also AES/EBU) signal can’t be jitter free in theory, so numbers of broadband eye pattern jitter below 300 ps are not easy to get and also not easy to measure. It is right, that the Audio Precision systems are not very good in measuring this jitter number and the prism sound dScope is also not the perfect way to measure this jitter, but does this really matter?

 

This number does not take into account, at what frequency this jitter occurs. Every SPDIF signal is received and read by a PLL and this PLL have a low pass filter to filter out the jitter products above 10 kHz with first or second order low pass filter, but leaves every jitter product below 10 kHz. So there can be 2 units, with both 2000 ps of broad band eye pattern jitter, but they can sound totally different. One unit can have mostly HF jitter products at around 2 – 20 MHz, that ware suppressed by the PLL and the other my have mostly LF jitter products at around 200 Hz – 2 kHz, that were not filtered out with the PLL.

 

So about 15 years ago or so, I was also looking mostly to reduce this jitter number but after some years came to the conclusion, that there is no relation between this simple number and the sound, so in the actual years, I do an FFT analysis of the SPDIF signal and calculate the jitter values within the audio band. So with the Audio Precision, I can calculate the mean value of the input SPDIF signal and take the variations (the jitter) with an FFT. So even with an high measurement limit on the broadband eye pattern jitter of the AP System of about 1000 ps, I am able to measure the FFT values on the SPDIF sideband to an resolution down to 1.5 ps (@ 1kHz) and this gives me a much better assumption of the audio related quality of the SPDIF signal, because the audibility of the jitter varies with the frequency and this is related also to the measurement results after an DAC, where the influence of the jitter on the THD is also frequency and level dependent.

 

To come back to the question in the original post, I can recommend, and I have also used it in the past, the no longer build Audio Alchemy DTI Pro, which has a very good analog 3 stage VCXO PLL with a low pass filter of below 2 Hz, so suppressed the SPDIF sidebands above 2 Hz very good and gives good results at the broad band eye pattern jitter and also (what’s more important) an very good result of the FFT analysis of the jitter signal. It also has an I2S Output, which is far inferior concerning jitter, as the SPDIF signal, but then you need a DAC with an I2S Input. I know this device can only 44.1 and 48 K, but has also an optical TOSLINK input to isolate from the computer.

 

Juergen

 

 

[Elprior]

Thanks for clarifying the situation about Jitter measurement.

 

Josh from Sonicweld is more or less saying the same, that absolute measurement - FWIW - does not tell the whole story, and that a classification is to be made.

 

This is so cool to get numerous comments from experienced people, what an exciting time for computer audiophiles !

 

Elp

 

[Nikos]

thank you all for your contribution to this thread.

I know that the jitter measurements are not so easy and accurate and that there are many type of jitter/jitter measurements.

But my original question remains the same...

which mac to s/pdif interface will be the best for a dac that has no jitter reduction circuits?

Has a firewire to spdif solution equal results to the infamous wavelength asynchronus usb products?

24/96 is mandatory (my esoteric is a good redbook transport).

Nikos

 

 

[Nikos]

bordin thank you very much for your list of devices.

If I had a powermac an internal lynx would have been a great choice.

 

 

[Audio_ELF]

As mentioned above - one of the best devices is the Weiss AFI1 and by extension the INT202 should suit you well.

 

The Wavelength device (for SPDIF) isn't yet available so is a bit irrelivant to speculate on it.

 

Eloise

 

[Elprior]

Well you're speculating on the int202.

 

Elp

 

[Audio_ELF]

ELP ... you're right ... I am speculating on the INT202 ... but people (outside of the manufacturers) have seen and heard the INT202 in use ... the WaveLink is still not even officially announced. And I mentioned the WaveLink in the first place :-p

 

Eloise

 

PS. does your DAC have a Word Clock input or output?

 

[bordin]

Adding a photo. ;-)

 

Click to enlarge.

 

[Elprior]

Yes, you're right

 

Wordclock :

The upsampler has both input and output.

The dac has just an output.

 

Any precise idea in mind ?

 

Elp

 

[Nikos]

elp you need a word clock input on your dac in order to synchronize it.

My dac has no input.

(question for digital experts... is it possible a non os dac based on ad1865n-k to have a word clock input?)

 

I forgot to mention that i have an apogee duet firewire dac and I am not very happy with it. (poor analogue stage).

I can afford a weiss int202 but I would like to know that it's firewire is at least equal to asynchronus usb interface.

thank you all for the interesting discussion

Nikos

 

 

[ar-t]

Somebody posted this:

 

"However I think MANY DAC manufacturers have shown it is possible to get good sound using the SPDIF interface (contry to what Gordon and others with a vested interest in other technologies would have you believe) and what we are looking for when we say "jitter free" is something which has a low enough level of jitter at the SPDIF output to enable these DACs to shine."

 

Where do you people get these ideas from? When has Gordon or anyone else ever said "it is not possible to get good sound using SPDIF"???? It is not impossible, just takes work.

 

No jitter? Impossible.

 

Low jitter? With work, possible. But only at the DAC out.

 

Good sound? Something totally different. Depends on lots of things.

 

No jitter at the SPDIF output? Impossible.

 

Lower jitter than the competitor at the SPDIF output? Possible, but the RX chip in your DAC is going to add a whole bunch of its own. Which may or may not be cleaned up inside.

 

[ar-t]

Another manufacturer posted this:

 

"This number does not take into account, at what frequency this jitter occurs. Every SPDIF signal is received and read by a PLL and this PLL have a low pass filter to filter out the jitter products above 10 kHz with first or second order low pass filter, but leaves every jitter product below 10 kHz. So there can be 2 units, with both 2000 ps of broad band eye pattern jitter, but they can sound totally different. One unit can have mostly HF jitter products at around 2 – 20 MHz, that ware suppressed by the PLL and the other my have mostly LF jitter products at around 200 Hz – 2 kHz, that were not filtered out with the PLL."

 

Everyone here needs to read this a second time. And a third and a fourth, etc.

 

[JR_Audio]

If my explanation is too short, you can find some good measurements and explanations of what I mean on the Wolfson Microelectronic website, where they have done similar thing that I have done. Measured and compared the broad band jitter on the eye pattern of the SPDIF signal (SPDIF out) and compared it to the FFT jitter analysis of the Masterclock after a typical receiver (I2S Out) and also similar to the jitter on the analog side, after D/A conversion (Analog Out).

 

Juergen

 

PS: Concerning the language, please keep in mind, that I am not a native US.