96k - 002

[rick]

has anybody ever sent a 96k file out of a digi 002 digitally ....?
apparently not !

if you have - i would love to know how
actually i would love to know how 5 hours ago

[PeterR]

I dont have one but I know that the 002 only supports 96k on the Spdif out - not the ADAT.

[Howard Jones]

I don't doubt what you say and haven't had a chance to check for myself. However, if Digi have done this, it's pretty poor. SPDIF is a consumer-only format which definitely does not run at 96K. If Digi are sticking 96K out the SPDIF port than it represents a blow to the development of well- understood and implemented standards.
Mind you, it wouldn't be the first time that a manufacturer has done something silly like this.

[Linear]

to an outsider (ie i don't own protools or any protools hardware), Digidesign are starting to resemble microsoft in how they are abusing their market power and forming their own standards. i still can't believe that you cannot use third party hardware with Alsihad, unlike logic or cubase. is that true?



Chris

[Kris]

Don't even get me started. Digidesign/AVID have a long history of anti-trust like behaviour. If you have a technical support contract with Avid Australia, which you pay good money for yearly, it still doesn't mean they will help you. Why? For example, if you're not using AVID badged hard disk drives and you have a system failure of some sort they will not help you, even if you have a maintenance contract, because they're not AVID drives attatched to your sysetm.

They will have to wake up to themselves sooner or later and ammend their corporate policy to refelct the needs and demands of their clients. If they don't they will lose significant market share. Apple are making very large inroads with Final Cut Pro and Logic 7. There are plenty of other alternatives. If they don't change their ways soon it may be too late.

I've always wondered if it's because Microsoft has a 9% stake (or thereabouts) in AVID. Maybe it really is a case of the apple not falling far from the tree?

[PeterR]

I think you may have misunderstood my post. Re-reading it - my wording is ambiguous and can be interpreted 2 ways. What i meant was:

The 002 will do 44.1, 48, 88.2 and 96k on the spdif I/O.

It does not support SMUX channel interleaving on the ADAT I/O - it will only do 44.1 and 48k with ADAT.

If rick was trying to do 96k on the ADAT outs - its not possible. It is only possible on the Spdif out.

[rick]

err 96k is not actually coming out of any digital outs in any format

on the duc in very tiny print in the answer section it says the 002 spdif optical or coaxial doesnt not support 96k !

its a drag that the concensis is /was so sure it plays back 96k digitally
i have spent my ten hours on that ship , then i jumped off.

again i ask has anybody actually done it ...?

[Henry]

The 002 is complete ripoff, I've been saying that for ages. What's the point if its the same as a 001? I hate digidesign! Extra for mp3, digitranslater, no cd write, no latency compensation they just plain suck!

[PeterR]

something is wrong somewhere - I'd give digi a call if you still care? I wouldn't after 10 hours!

from the answerbase

http://answerbase.digidesign.com/detail.cfm?DID=28584

Answer Details

Do the S/PDIF (SPDIF) I/O connectors on the 002 support 96 kHz?

The optical connectors on the 002 support either 8 channels of ADAT I/O (up to 48 kHz) or 2 channels of Optical S/PDIF I/O (up to 96 kHz).

The BNC (RCA) connectors on the 002 support 2 channels of S/PDIF I/O, up to 24-bit, 96 kHz.


Platform: n/a ID: 28584 Created: 06/25/2004


PS. not a fan of digi either - but can't resist that nice simple audio editing

[Howard Jones]

They may be RCA connectors but, if they are running at 96kHz, they're not S/PDIF as we know it.

[chris p]

All right Mr O'Neill - you've got my attention - what exactly were you trying to do with a 96K bitstream in the first place, and what workaround did you come up with when the Digi didn't do it?

Howard, you're right of course in saying that SPDIF doesn't spec 96K, but it wouldn't be the first time that a spec has taken on a life beyond its standardised paper origins, and Digidesign is far from the only culprit out there - many pieces of gear claim SPDIF at 96K, but RO'Ns experience has caused me to ponder the claim.

It again reminds me why 44.1kHz is a nice number, and to focus on word length quality (it may only be 16 bits after dithering, but hey - they're the RIGHT 16 bits!) rather than sampling rate quantity.

[Howard Jones]

It is simply impossible for S/PDIF to be 96kHz - I will dig out more on this at work tomorrow and post it here (96kHz wasn't even conceived of when S/PDIF was implemented*).
If Digidesign or any other manufacturer want to run 96kHz on an RCA connector, I don't necessarily have a problem with this - but such a manufacturer needs to avoid at all costs calling such a connection 'S/PDIF'.
S/PDIF is also limited to 16 bit.

* S/PDIF is an acronym for Sony/Philips Digital InterFace. The Sony & Philips companies developed workable consumer CD technology and the associated CD Redbook standard between them. They realised early on that consumers might want to make a digital connection between CD players and other equipment i.e. amplifiers, so they came up with S/PDIF which specified an unbalanced signal running on short-length cables to enable comsumers to keep connections in the digital domain.

So, the historical background to S/PDIF is as a connection to allow 16 bit, 44.1kHz transfers over short distances on coax cable. Anything else is not S/PDIF - let's call it Digi-DIF or Akai-DIF or whatever. Calling it S/PDIF causes confusion and is simply incorrect.

These aren't the only guilty parties - even Sony themselves have been guilty of transgressions. On the PCM-2700, Sony implemented their AES digital output by way of an unbalanced RCA connector - completely stupid (a pathetic measure which must have saved them all of $2). Naturally, this confused the hell out of anyone who just took a look at the back panel, saw an RCA connector and assumed an S/PDIF connection. This is a prime example of why we have standards and why manufacturers should not ignore them.

[rick]

i had a session done on the 002 that the client wanted left as streams
i figured i would just pipe it into any number of converters in my room that handle 96 k instead of using the digi onboard converters
but with a digital interface on the back of the 002 that will not , send out 96k things got a bit perplexing.
i jerry rigged every version of a spdif to aes cable, moved every version of adat light pipe and 2chn optical implementation around and generally just swore a lot

i can easily ( or slowly at least) dump any 96k data file into my sadie and send it around the room at those sample rates
i can go down the hall grab a hd rig , and get on with it but thats not the point really

the 002 would not output a 96k digital stream .
it records 96k through its converters , makes , plays and labels 24 bit 96 k files and i would just like to find out if it plays 96k files out its digital outs if so how so ...?
so after much confusion i have asked the question , here there and everywhere
HAS ANYONE ACTUALLY DONE THIS...?

i know it looks like it should be possible in some reviews , i know it is impossible via the actual specs as howard has confirmed
i know digi should be able to do it out the adat lightpipe cable but they dont..
i know they shouldnt be able to do it from the spdif but it is claimed they do

the record i was doing did not have anything like the recording quality that would justify this waste of time , but people love those 96 k numbers dont they.

our preferred sample rate here at turtlerock is 24bit - 44.1k
for a host of reasons the least being how hard it is to move 96 k around
so again who has done this ...
how..?

[Howard Jones]

Okey-dokey, Rick,

The following info is copied from a Digidesign PDF publication entitled 'Getting Started with Digi 002 & Digi 002 Rack' pp54-5:

"The Sony Phillips Digital Interface Format (S/PDIF) is used in many professional and consumer CD players and DAT recorders. These
S/PDIF input and output jacks are unbalanced 2-conductor phono (RCA) jacks that utilize a full 24-bit, 2-channel digital data stream. To avoid
RF interference, use 75-ohm coaxial cable for S/PDIF transfers and keep the cable length to a maximum of 10 meters.

"Digi 002 and Digi 002 Rack support consumer mode S/PDIF output format (IEC-958 Type 2) at sample rates of 44.1 kHz and 48 kHz, and professional mode S/PDIF output format (IEC-958 Type 1) at sample rates of 88.2 kHz and 96 kHz. For information on which S/PDIF formats are supported by your S/PDIF-compatible gear, refer to the manufacturer

[Adam Dempsey]

Hi, yes to my knowledge Type II meets AES and SPDIF supports 24bits (though perhaps only IEC958 typeII).
Not having ever used a 002 though, no (balanced) AES/EBU outs? It's not an 88/96k dual wire issue? Am I full of Friday crazy-talk?

[Howard Jones]

Hi Adam,

I wondered about the dual-wire issue also but the 002 doesn't appear to handle 96k in this way. The trouble from Rick's POV is that none us knows how a 002 handles 96k. The Digi manuals aren't very helpful on this.

[mfdu]

rick

i'm just wondering a little thing.

why didn't the client consolidate and bounce out all files at 96k, providing you with (ideally) one multitrack full-FX copy on optical media (cd or dvd), and one copy on magnetic (hdd)?

the client knows the project and spends the time to consolidate so you can do what you do best.

not meaning to sound like i'm teaching you to suck eggs. i just would'a thought having a clients 002 brought in to the studio would give you prosumer-rash.

that said, i am afflicted with 002, and this is the only way i would ever take projects in for mastering . . .

chris.

[rick]

some people want to leave their mixes as "almost finished"
and make changes in mastering.

this is not the way i want to do things, and not the way its generally done
but the guys across town encourage it , so from time to time we do what we are asked.

to be clear Turtlerocks preferred format is 1/2 inch
then 24 bit 44.1 k data
then 16 bit CD audio
then whatever comes along. (48k,96k , cassette ...?)

untill somebody makes a format where 96k is the standard file type
is always going to be a hassle.

the 001/002 is capable of remarkable things i am just not sure that 96k digitally is amonst them

[mfdu]

thanks for that rick.

i am very interested to hear what you say about 96k (in general and re. the 002)

in my *studio* i swear blind i can hear the openess and honesty of 96 over 44.1 - that's purely in relation to tracking the best that i can, of course.

no dispute whatsoever about the benefits of the 24 bit wordlength, but when it comes to samplerates, am i just fooling myself???

so - is it because you can't hear the difference b/w 44.1 and 96, or is it because the final stage through to cd is made simpler and more controllable if maintaining a 44.1 stream?

not dissin ya bro - i really am interested!!!

:)
chris.

[Jason Dirckze]

There are benefits of recording at higher sampling rates, but mainly these are to do with moving the quantisation errors into higher frequency range, well above our range of hearing. When recording at 44.1k, the quantisation errors can reside in the 20Hz-20kHz range, sampling at higher frequencies effectively moves the quantisation errors out there above 20kHz.

Have I just said the same thing twice?

I cant really go into more detail, cos I cant remember alot of the theory, perhaps someone here can give a better answer, but IMO there is definately SOME benefit of recording at higher sampling rates, the question is whether or not it is WORTH it.

[Howard Jones]

Quantisation errors are a function of sample depth although, at a higher sampling rate, each error occupies a smaller amount of time.

[Adam Dempsey]

Main benefits are largely a combination of:
the anti-aliasing A-D filters working >20K,
some dsp operations running better at higher s rates,
plus any noise shaped dithering shifted >20k.

While on the topic, all things being equal & in the absence of super-duper sr convertors, keep it at 24/44.1 for CD... or mix to tape!
(our preferred:
15 ips with Telcom NR or
30ips without, at OVU=520nW/m)

[wez]

FWIW, i've recently done some very unscientific tests, comparing 44.1 to 88.2, by switching the sample rate on my ADC, and listening post-convertor in sequoia. i flick the switch a few times (not looking at it) until i don't know what it's set to, then i go back and forth, listening for differences. not a good AB test methodology, because i am expecting to hear differences. and i'm expecting to hear extended top end more than anything, but other than that i've tried to keep an open mind and not make too many assumptions.

it's an ongoing test, and i can't make any conclusions yet... other than to say i am picking the higher sample rate about 6 or 7 times out of 10. that's not a lot better than random, but it's an interesting exercise in critical listening. when i'm right, and i was sure i was right, it all seems, umm, right. but then you get it wrong and it's back to square one.

anyway, it's not something i'm going to obsess about just now. when you consider that it's hard enough to hear the difference pre-SRC, then the difference once it's on CD... anyway, accumulated during tracking, it might be more noticeable. but really, i hate to agree with everyone else, but.... there are so many other things to get right first. and if you get them right you won't care what sample rate you used. and neither will i, nor the a&r guy, or the punter, or the radio programmer.....

[Jason Dirckze]

It all comes back to whether or not the difference in sound quality is worth it in the end. From what I know, there is a difference, but not one that is good enough for me to record anything above 44.1k, especially in my humble project studio, where I have more to worry about (like the neighbours dog, birds, my damn G5 making a whole lotta noise....)

[Howard Jones]

Agreed. The other aspect yet to be mentioned is the increase required in storage capacity & handling. Obviously, going to 96k will double the storage requirements and going from 16 bit to 24 bit will increase them by 50%. This is on top of all the other very good reasons already stated for wondering whether you will be able to hear the difference...

[mfdu]

and then of course there's the question of

"have i found the right piece of air?" . . .

since my last comments, i realised i forgot to say to everypne that i love recording, and i'd do it to tape, 16/44.1, 24/96, minidisc or mobile phone speaker.

i'd record on the sides of thick wax candles if that was all we had.

$0.02

chris.

[Damien]

i just found this on the tools web site. not sure if this would have helped.

"You can't have the optical port and RCA port both trying to be used for S/PDIF.. you need to make sure the optical = ADAT and RCA = S/PDIF.

Rail

[JulienG]

"You can't have the optical port and RCA port both trying to be used for S/PDIF.. you need to make sure the optical = ADAT and RCA = S/PDIF.

Eh, that's the same as the digi001, not surprising. This whole thing is a case of marketing adding another item to the feature box without actually ensuring that the hardware is fully capable of it.

[wez]

so where are we at, vis

[mfdu]

how many tracks are you wanting to get out at 24/96? stereo or more?

i'd be inerested in hearing what the tech lads have to say tho, being saving my pennies for a decent extra two channels of ada through s/pdif (and a better clock into the bargin please - i hear it makes a difference . . .)

chris.