Somebody to help with a good explanation on S/P-DIF Sourc an

A place to talk about whatever Scope music/gear related stuff you want.

Moderators: valis, garyb

Post Reply
User avatar
Nestor
Posts: 6688
Joined: Tue Mar 27, 2001 4:00 pm
Location: Fourth Dimension Paradise, Cloud Nine!

Post by Nestor »

I am doing a clear tutorial about Routing matters...

http://www.planetz.com/forums/viewtopic ... forum=21&9

...and need some help explaning this modules. I have not used them myself, so I have not a deep practical knowlege. I am thinking of beginners, as well as intermediate users of the SFP, so I beg a VERY clear explanation of those modules.

Please, anybody with a good understanding, give me a hand... Post it "here" first, then I'll post it with the proper pictures to match, and the name of the author of the words.

Any help is much appreciated. Thanks :smile:
User avatar
Nestor
Posts: 6688
Joined: Tue Mar 27, 2001 4:00 pm
Location: Fourth Dimension Paradise, Cloud Nine!

Post by Nestor »

And please, cover as much points as you can afford, the conversion, the sample rate matter, the sync, etc., Thank you.
*MUSIC* The most Powerful Language in the world! *INDEED*
jabney
Posts: 98
Joined: Sun Feb 10, 2002 4:00 pm
Contact:

Post by jabney »

S/P-DIF: You can do a lot with S/P-DIF, but you've got to watch the clock.


The first thing one must learn about S/P-DIF is how to pronounce it. I say it "ess pee diff (as in Jiff)." You may imagine other ways of saying it. Perhaps you really do want to go around saying "spee-diff." Sort of like people who think it's cool to say, "boxen" :smile: when speaking of generic PCs.

S/P-DIF comes in two flavors, copper and plastic. These flavors may be used alone, or in combination. Typically:

S/P-DIF copper. RCA connector. e.g. Pulsar Classic I/O whip - yellow Dig In & Dig Out.

S/P-DIF's cousin AES/EBU - XLR connector. e.g. Scope Classic Plus whip - yellow Dig In & Dig Out.

Plastic or "optical" S/P-DIF uses the same connector as ADAT. The Samplerate menu (under Set) lets you choose whether a Pulsar II or newer card will treat one of its ADAT ports as 2 channel S/P-DIF or as 8 channel light pipe.

I will refrain from any participation in the plastic vs. copper debate except to make you aware (in case you were not) that there is a debate. Your search techniques should avail you should you feel this matter of plastic vs. copper is worth investigating.

In theory, S/P-DIF and AES/EBU aren't supposed to communicate, but... If you are curious, try a search using both terms plugged-in to your friendly local search engine (yes, I use Google too, but what about giving the other guys a chance now and then - that'll leave more bandwidth for me back at google :cool:).

There are ways of getting S/P-DIF and AES/EBU to pass information to each other. The simplest method I've heard uses a Drawmer M-Clock to re-clock a CD-Player's S/P-DIF output to the AES/EBU input on a Scope Plus. The M-Clock also controls the timing of the CreamWare environment. The ability to reclock allows more than one S/P-DIF input to be active so the streams can be cleanly mixed. No nasty crackles or pops.

But even absent something like a master clock that can re-clock digital inputs while working with CreamWare's clock board, the SynCPlate; digital output from an S/P-DIF device into the CreamWare environment should still be crackle and pop free. As long as you've chosen S/P-DIF as clock source in the Samplerate submenu under Set.

If you've never changed the clock in sfp before, note that you have a major selection of two buttons at the top and minor selections below. Both major and minor buttons have to be pushed - and even then it can take awhile. Just wait, the clock rate should eventually show in the samplerate window. Until it does, don't waste your time trying to do anything else in sfp.

Oh, and mute anything sensitive downstream from your sfp environment before you change the clock. Give your Creamware device some privacy at this intimate time. Think of it as if your Pulsar or Scope were molting :smile:.

Be aware that the quality of the S/P-DIF device's clock is entirely up to the S/P-DIF device's manufacturer. Myself, I really appreciate having a CreamWare environment with a SyncPlate - and VDAT. Before I got an external clock, I tended not to use the S/P-DIF as clock very often. But you can.

Even if one somehow manages to rig an AES/EBU to S/P-DIF adaptor or even an S/P-DIF to AES/EBU adaptor using readily available parts, there's still that old master-slave thing to deal with.

If there's a choice, typically, the most expensive device supplies the clock. You could try listening, but it does make sense that if there's a big price difference, the more expensive should have better parts - at least you hope so. Of course that is only valid for items made about the same time.

S/P-DIF sources could be a CD player, Mini-Disc player, DAT player (recorders) or another (computer with) Pulsar I or II Classic.

The S/P-DIF source device should be master and CreamWare the slave. (Clock is possible exception.)


S/P-DIF dests could be a DAT recorder or another (computer with) Pulsar I or II Classic. And you can have as many S/P-DIF dests as you have physical CreamWare Digital Out I/O connections.

It's a lot easier going out, by the way. In that case, your clock should come from the Creamware device unless another clocking arrangement makes more sense.

The tricky part in going out to an S/P-DIF device is making sure you've chosen the right clock speed (for example 44.1 for CDs or 48k for DATs). Believe it or not, there's a debate on that, too. As always, good search techniques will get you through - if that's where you want to go :smile:.

john

<font size=-1>[ This Message was edited by: jabney on 2004-02-26 00:50 ]</font>

<font size=-1>[ This Message was edited by: jabney on 2004-02-28 14:05 ]</font>

<font size=-1>[ This Message was edited by: jabney on 2004-02-28 19:01 ]</font>
User avatar
Nestor
Posts: 6688
Joined: Tue Mar 27, 2001 4:00 pm
Location: Fourth Dimension Paradise, Cloud Nine!

Post by Nestor »

Thank you Jabney, for your contribution... That's great and very uncouraging also, that some people start to participate.

Nevertheless, I will ask you a little more effort information this time, please, :smile: and this would be the USE of it, in terms of routing posibilities or uses. Think about you yourself, being using it for the first time, and try to make him, (i.e., you) understand how to connect devices and why. Thank you again...
jabney
Posts: 98
Joined: Sun Feb 10, 2002 4:00 pm
Contact:

Post by jabney »

S/P-DIF Continued:

Once you have all your clocking issues dealt with, then use S/P-DIF in the same way you would Analog In or Analog Out.

The most direct way is to drag the S/P-DIF In and S/P-DIF Out icons from the Hardware I/Os section of the menu onto the surface of the CreamWare sfp RoutingWindow. This provides a two channel (typically stereo) input from, and a two channel output to some digital device located outside the computer box.

An Example:

Let's say you want to listen to the digital output of a Tascam DA 30 Mk II DAT recorder. That model DAT machine has both S/P-DIF and AES/EBU inputs and outputs. If you have both a Scope or Pulsar Classic card and a Scope or Pulsar Plus card in your computer, you could connect using either S/P-DIF or AES/EBU. Which? I'd say use the one for which you have the better cable.

Please be aware that expensive digital cables do exist and not all of them are snake-oil. Still, I'd hesitate before spending thousands (or even hundreds) on a digital cable. At least make sure that you are using a digital - as opposed to audio, mic' or line - cable. (If you want to get technical, the difference between digital and audio cables has to do with impedance, and here it's a sort of 'sideways' measurement. Not at all like a simple resistance measured from one end to the other wherein a 1000 meter cable would have more resistance than a 1 meter cable of the same gauge. A search engine can provide more answers for the curious.)

If you insist, you could listen and try to pick whether AES/EBU or S/P-DIF sounds best. Good luck trying to hear a difference, though. A primary reason for using any digital connector is to assure that the signal is passed along without any alteration. So S/P-DIF, AES/EBU, ADAT lightpipe over a few feet should all sound exactly the same - in theory. For this example we'll stick with S/P-DIF.

info added 2/28/04

I'd like to thank Tom (known on PlanetZ as Astroman) for pointing out a 'fly' in the ointment of digital perfection. The error correction applied in a CD player for example, is not as robust as the error correction applied in a computer to a file. If you need to get as close to the original as possible, then a grabber program (e.g. cdparanoia) to convert the CDDA file to a .wav file is the way to go. Then, you probably wouldn't be using S/P-DIF, instead you'd just use your computer's data CD reader.

One other thing: it is not always a bad thing to go through a digital to analog/analog to digital cycle. In fact the conversion hardware may add or subtract just enough to give you the sound you are looking for - in a mix, for example. Conversion may be (in fact, has to be) less faithful than pure digital, but a little color can be useful now and then. For now, we'll persevere with the S/P-DIF, though.

Once you've made the physical connection between the DAT machine's S/P-DIF output and the CreamWare's Digital In, turn on the DAT machine. Load the DAT tape you want to hear and hit Play on the DAT machine. When you hit Play, an already recorded DAT tape will transmit its clock rate to the DAT machine. You should be able to tell the clock rate on the DAT by seeing either 44.1 or 48k on the DAT machine's display. Once the DAT's clock is displayed, you should be able to rewind the tape and hit Pause. The clock info should still be there. (A CD player gets its clock information from a valid CD and except for the parts about loading the DAT tape _ use a CD instead _ and 48k _CDs only do 44.1, the above information is also applicable.)

Start the CreamWare sfp environment, and adjust the SampleRate (under Set) so that CreamWare is a slave to the appropriate S/P-DIF input stream. Then just connect the output pads of the S/P-DIF In icon to a couple of channels on your VDAT or STM-2448 or what-have-you.

An Alias:

Scope and Pulsar (II) Plus users with AES/EBU digital connectors will still use the S/P-DIF icons (at least as of sfp 3.1c) since there is not a dedicated AES/EBU input or output icon. There is a workaround if you think this could cause a problem, say you have a clueless, yet stubborn, techno-snob client who might think the AES/EBU signal path was somehow polluted by having even the label "S/P-DIF" in the path. You can create your own AES/EBU-labeled icon. (This has more practical applications, too. So read on.)

Perhaps you say, "I have so many things connected to my CreamWare environment, the back of my computer looks like a porcupine at the beauty parlor. How can I possibly remember what goes where?"

Not to worry. CreamWare has the Tools to solve your problem. In fact, the Tools menu is where the "Source S" icon is found. For our DAT example, we will drag the Source S icon onto the CreamWare RoutingSurface and double click to open the icon's surface.

The first thing to do is to change the icon's name from "Source S" (which stands for Source Stereo) to something more meaningful, perhaps "DAT Play." It's as easy as changing any other text field in CreamWareLand, but this one actually displays as the name on the icon.

Below the name you will see two fields to the right of some potentially confusing labeling: Output then Left and Right. "But aren't we working with inputs?" you ask. Yes, but only into the DAT Play (formerly known as Source S) device. The output of the DAT Play icon is the stream it gets from the hardware (S/P-DIF) Input which in turn becomes ... well, the good thing is, you can't mess it up. When you click in the Left or Right field with the Right Mouse Button, a list of Sources is displayed. Choose an unused S/P-DIF Source then choose L Out. Repeat for the Right channel. Then you are almost done.

I recommend setting-up a separate directory for the icons you create using Source M, Source S, Destination M and Destination S tools (as well as the similar but more complex External Effects of various flavors under Effects/Stereo (or Mono)/Others). Then, you can save your new icon (using the save sub-icon on the device's upper-right-corner) so you don't have to do it all over again the next time.

There is one "gotcha" lurking behind these alias icons: the icon for the hardware I/O you want to connect must be physically present on the CreamWare RoutingSurface. I like to hide the physical icon below the bottom of the screen, because it does not have to be connected to anything. The connection is made with the alias icon.

Oh, and sometimes the alias icon dialogue may show that a physical connection is "occupied." The best way to deal with that is to experiment. A hint, before you experiment with stealing connections for alias icons, make sure that you can see the connections that are "occupied" then watch what happens :smile:

What We've Done:

Let's review. You decided that you want to use the S/P-DIF Out of a DAT player. You made a physical connection from the DAT player to your CreamWare card's Digital In and you made sure to use digital cable. You started playing the DAT tape so that it would furnish the clock.

On the CreamWare RoutingSurface (or using your favorite method) you connected the right-facing pads of the S/P-DIF source icon to two left-facing audio pads of some other icon. You made sure that CreamWare was slave to the appropriate S/P-DIF stream using the SampleRate dialogue. You set the levels in software, knowing that the DAT player's output levels have no control over the S/P-DIF stream. If you were comfortable with the idea of S/P-DIF, perhaps you set-up an alias icon. Now what are you waiting for? You're in CreamWareLand: make some music!

A Messy Post Script:

At some point, you may want to use an external device with S/P-DIF In and S/P-DIF Out. "Nifty," you think, "Now that device can stay entirely in the digital realm." Perhaps. The reason it gets messy is that sometimes it works, sometimes it doesn't. In general, later generation devices (CreamWare and external) tend to be more cooperative. And strict compliance to S/P-DIF standards may make a highly compliant device appear to be less capable than a more loosely-spec'ed device. Experimentation done at your leisure is the answer. A live show (or with an anxious client breathing down your neck) is not where you want to experiment with double (or even single) sided S/P-DIF.

Watch the Clock (or, Messy Post Script Part II):

A single active S/P-DIF stream appears to provide a stable clock to the CreamWare environment. Why not let CreamWare default to be slave to a DAT or CD player? Here's why not (in case you missed it above): the S/P-DIF clock that you might think of as coming from the DAT player or CD player is actually coming from the DAT tape or CD disc. If you don't have a DAT tape or a CD loaded, then you don't have clock.

john

<font size=-1>[ This Message was edited by: jabney on 2004-02-28 14:00 ]</font>

<font size=-1>[ This Message was edited by: jabney on 2004-02-28 18:57 ]</font>

<font size=-1>[ This Message was edited by: jabney on 2004-03-01 19:28 ]</font>
User avatar
astroman
Posts: 8455
Joined: Fri Feb 08, 2002 4:00 pm
Location: Germany

Post by astroman »

tnx John for that huge effort, but unfortunately it almost ends when it starts to become interesting :wink:
Imho the pdf doc is pretty good at explaining the basics of digital routing as long as it concerns a single device and only in one direction.

But I can't even remember the case reading from and writing to a device on the same port is covered in the manual.

It's stated though that digital fx 'roundtrips' are impossible on generation one cards (4 DSP Pulsar), but do work in all 2nd generation boards (from Luna on) by seeing the SPDIF specs a bit non-specific (as you mention).

So what finally happens is specific to the manufacturers implementation and there seem to be many different viewpoints - you cover this topic by the suggestion to try it out in non-pressure situations, which seems the only way to go.

It's important for the beginner to know there are specs, but they are of little use when it comes to practical applications - and it's not always one's own stupidity when a setup doesn't operate.

Maybe I didn't search extensively enough, but I do miss all technical details (in the original documentation) about Pulsar's clock specs, how this and that is related, what has priority and the pure signal specs.

Then come the more complicated setups with at least 2 external digital units, like an ADAT IO-box and an fx unit.
A digital mixer, ZLink and fx unit to make it even worse.
What gear works good together and what is to avoid ?
When does an external clock source provide a noticable improvement and what units have a good reputation (that stuff is expensive).

Imho we need to collect some hands-on facts from existing setups, preferably from the more experienced users - who have probably the least time :wink:

One import fact I'd like to contribute is error correction, which isn't mentioned at all yet.
Everyone tends to interpret the digital audio streams the way data is read from a harddisk, but it's not like that - it's more like modem data sent down the phone line.

It's important to realize that an identical copy of such a stream does not mean the data was read identically (and reliably) from disk or tape - it just means that the error correction has a consistent way to deal with it.

cheers, Tom

ps: I found the last sentence :cool: somewhere on this site http://www.tnt-audio.com/int.html
some understandable (!) facts about digital signals, clocks, jitter and aliasing
http://www.tnt-audio.com/clinica/convertus1_e.html

<font size=-1>[ This Message was edited by: astroman on 2004-02-28 18:43 ]</font>
User avatar
dehuszar
Posts: 619
Joined: Wed Mar 27, 2002 4:00 pm
Location: Chicago, IL United States of Amnesia

Post by dehuszar »

While we're on the topic, has anyone Wordclocked their A16 Ultra to the sync plate? Is there any benefit or does Z-Link do it all?

I'm on the verge of picking up some wordclock cables, but I may have to mod my Magma chassis to get it all going.

I'd appreciate any thoughts.

Sam
jabney
Posts: 98
Joined: Sun Feb 10, 2002 4:00 pm
Contact:

Post by jabney »

Sam,

I'm a big advocate of having a good wordclock. When you think about it, other than the analog circuitry feeding the AKM A/D converters that practically everybody and their uncle uses ('cause they sound great and have reasonable pricing), the big difference-maker in the sound of a digital system is the clock.

I went with the Drawmer M-Clock because it provided different output clock rates for up to 8 different devices. I was comfortable ordering without hearing because I had never heard a Drawmer device I didn't like. Drawmer did not let me down with the M-Clock. The conversion of up to four different digital streams all clocked and ready to go click-free into the CreamWare environment was an unexpected bonus.

There are other clocks with varying degrees of utility. The new Apogee looks interesting (having only 6 BNCs took it off my list), the Lucent seems affordable (remember, clock and clock distribution are not always one and the same) and new universal Audio and Mytek clocks join the venerable AardSync in units worthy of consideration.

I've tried ways around the clock. I tried using an RME ADI-8 Pro as master clock. I tried pumping that clock through a video amplifier which had one 75 ohm input and a dozen 75 ohm outputs. Always with the CreamWare SyncPlate connected. But the addition of a dedicated external clock made the whole CreamWare system feel more stable.

Do folks think we need a tutorial on clocking?

john
Post Reply