Master/Slave using adat and spdif

[Conqueror's Reign]

Hello all back again with more questions, as large as this forum is this has probably been asked, so please forgive the neophyte.
I have 3 PC's that i want to connect digitally.
PC 1(DAW) this has a Scope project card with the Classic I/O option
PC 2 (Gigastudio) this has 2 maudio audiophile 2496's for a total of 4in/4 out analog and 1 Spdif available because the other sdif pair has to be used to sync the two cards to work as one.
PC 3 (Synth, FX Box) this has a Korg Oasys PCI with 2in/2out analog, spdif, 1 adat in/out and word clock I/O
My question is if i use spdif(coaxial) and slave PC1(DAW) to PC2(Gigastudio) for recording would i be able to use the adat out of PC3 (Synth) into the adat in of the Scope card PC and be able to record from eithier or both pc's at the same time?
Can the Scope card be slaved to two different devices without having any conflicts or problems having two different sources?
Last thing, out of curiosity is it possible within SFP to connect a spdif source to a spdif destination directly within SFP and create something like a sdif thru?

Thanks in advance!!!!!!!!!

[garyb]

only one device can be master. all others must be slaves.

[Conqueror's Reign]

Thanks Garyb, thats what i thought but i wasnt clear on adat, if it needed to be master as well when sending audio. Question, so how would for example a studio with multiple adat equipped devices be configured to record on a single adat equiped computer for recording, would you have to record a single device at a time and switch each to master when your ready to record it? Is this where a device like ad/da converter comes into play?

[garyb]

no, most ad/da can be master as well as slave.

typically, i make scope the master and slave everything else to adat unless a device only wants to be master and then i make scope the slave. if two devices only want to be master, then you're scr*wed. you'll have to use one device at a time like you suggested. many devices will use word clock, so that's another option(slave one or more devices to wordclock).

[Conqueror's Reign]

So when using adat the scope can be master even though the other devices are sending audio?
So adat is more flexible and not a one way path like spdif?

I dont have any experience using the adat format only what i have read but it is very appealing to be able to send and receive 8 channels at once.

[garyb]

yes.
actually, sp/dif works that way also. it's just that most devices using sp/dif are consumer playback devices and it's cheaper to make them master only since for them to be slave they'd need inputs and consumers are to self-absorbed to fiddle with master/slave relationships...

[Conqueror's Reign]

Wow now i feel like willie lump lump, so if a device can be set to slave it can still send a audio signal spdif out to a master device? All of this time i thought spdif had to work in the reverse only. Is this achieved with any spdif device that can be set to slave or do they need that aes/ebu option?

[garyb]

nope, if it can be slave, it can sync. samples will arrive in the proper order at the proper rate.

of course to be slave, there must be a clock signal at the devices input. if you only connect the output and set the device to slave, no sync is possible.

that's assuming that the engineers of the device didn't make the operation impossible by making the device unable to see the clock signal when in playback, but that wouldn't be because sp/dif wouldn't allow it. it would be because the designer of the device was not thinking deeply enough.

[Conqueror's Reign]

So both of the spdif connections are used to make this work, spdif out from the master then spdif in to the playback device thats sending the audio you want and from that same playback device the spdif out is sent to the master so it can be recorded, am i correct?

[astroman]

as GaryB wrote: it entirely depends on the implementation - how the engineers designed it.
A Pulsar One (for example) can NOT do s/pdif send/receive simultaneously, a Luna or a Scope card can.

you have to verify that individually for each (outboard) device you'd like to use in that way.
before I got my Pulsar I never heard about that sync stuff, too (and it was a complete mystery). You get used quickly to it, and also to the various 'missing or bad sync' errors in sound. Silence and crackling is easy to detect, bad cables or an instable clock are more difficult.
Just don't think that a digital audio transfer is comparable to a data transfer from a harddisk - it's a completely different approach.

cheers, Tom

[Conqueror's Reign]

More research is in order i see. It does looks like adat would be a better option with the scope card being master.
I guess it is misleading when you think about digital audio you think perfect with absolutely no faults from device to device but common wisdom says better, nothing is perfect. Which leads to another question.

So how much would you lose in the quality of the signal by going a analog route into scope for example using something like the A16? I know everyone one will vary in opinion but in general are their definite advantages using analog over digital or vice versa besides the sync ability?

[garyb]

there's no absolute there either. not all digital transfers are better sounding. most of the time, going in analog is easier and sounds great. it really depends on your needs and purposes. experience will be your best guide.

basically, for a device to be a slave, there must be clock at the input. if a digital device has no input it must be master. if that device has a bad clock, you're probably better off using the analog outs rather than clock the entire system to an inferior clock.

the best clock would be an high-quality external clock generator(word clock) but that requires an expensive clock, a syncplate, proper cables and a bnc input on any other digital device that will use digital connections.

[astroman]

the quality of an anlog path with additional AD/DA conversion is defined by the quality of the analog parts, conversion noise and artifacts and finally by the stability of the clock driving the converter.

on a digital path (A/D) it's the clock, the cable quality and the error correction.

that's why GaryB's statement ... digital is usually better... applies - you have less possibilities to mess things (if all is setup properly).
Nevertheless it just needs a wrong cable WILL to set a digital connection below the quality of even a midclass AD conversion
There is more than one (technical) way to handle the details of a digital transfer, therefore the hint not to mess it with a transfer from hardisk. The latter reads a sector and will re-read it automatically if any bit doesn't match, it's a non-realtime application.

A dgital audio transmissin IS realtime and there's only a limited amount of time for the error correction (which is omnipresent) to do it's job.
If the transfer of a certain data word fails within that timeframe, it has to deal with the incomplete data, and will try it's best (interpolation etc), but of course the datastream is corrupted partially.
The degree of this corruption may vary between definetely not noticable up to a THD+ noise ratio of (say) 25% , like a well driven guitar amp...

my Pulsars are on their own box with no network connection and so I often listen to mp3 through the digital out of the onboard AC97 codec of my devloping machine (Gigabyte P4).
The chain is digital out to a Philips DCC recorder for conversion (a high end device at it's time) to a NAD 1020 preamp driving my monitors.
The SFP machine's is also connected to this preamp (a convenient switchboard and volume controller, maybe a tiny bit too sweet for hard analytics...)

So you wouldn't expect much difference between a 16bit mp3 played back via any of the machines, eventually it's all digital...
WRONG the difference is stunning (or rather shocking) , it's in no way subtle
...I have to transfer via usb stick to the Scope machine for a detailed listening.

cheers, Tom

[Conqueror's Reign]

I want to thank you both again for your knowledge, this has been a great study for me and hopefully it will help someone else just as much. I think what i will do eventually is replace the Maudio cards in the Gigastudio Pc with a scope card.

[garyb]

good choice.