Emu XL-7 & MP-7 User's Group

Hi
> I've recorded some loops with my audio card trought s/pdif at 24 bit
> and all seems to be ok and in sync.
> but my doubt was about the possibility of passing a "real" 24 bit
> signal throught a s/pdif. is it that possible?

Looks like the S/PDIF output is 20 bit, so you recording in at 24 bit 
will not really yeild any real advantage / or disadvantage.
Maybe if your multi-tracking and everything else is 24bit you will 
have to stay in this format.
The extra low order bits will just get padded out.
I had issues in the past when trying to use the S/PDIF, it does seem 
to depend on how the receiveing device handles the S/PDIF signal.
 
> another doubt was about the internal audio resolution of the xl-7: 
>is the xl-7 capable of delivering a 24 bit true signal?

Well looking at the specs in the manual..
S/PDIF specified @ 20 bit(AES is just another format, not a differant 
bit depth, ie not 24bit)
It dosent usually matter which format you use as the differance is 
only copy protection in consumer S/PDIF which the Pro version use's 
these same bits for other purpose's.
Also the AES standard is a balanced connection, which the output on 
the command station isn't.
All sample data is 16 bit, 20bit D/A converter on the main outs and 
18 bit D/A on the sub outs...
The only mentioned 24 bit processing is the FX engine.

> maybe these are stupid question, but I've never used s/pif so please
> be kind : )

No there not stupid questions, you learn these things as you come 
across them. Theres loads of resource on the web on the subject.
Go as deep as you want...

>if you record a kick alone, velocity  at 127, track volume at 127, 
>you get  a signal of about  -10, -12 db, and less for hi hat or other
>sounds. Then I normalize it at - 3 db, but I was wondering if it 
>would be better to record it throught the analogic output that 
>is "hotter" and I easily get good level without normalizing.

-10db or -12db isn't a bad thing. Dont forget your in the digital 
domain so theres no really any noise being introduced
If you were recording a band into your PC you would have to leave 
about 12db (maybe more) of headroom for the peaks (theres no headroom 
above 0dbFS)if you didnt have any compression on the inputs. 
Also as soon as you start doing any processing on your wav's you have 
to leave headroom for that as well.
Its a compromise between good signal to noise ratio and having 
enough 'bits' encoding your audio to not create to many errors.
I always used to bring everything upto peak -6db, do all the eq'ing 
etc and then average normalise it upto 0db for that wall of sound 
punchy dance sound... (psy trance at the time..)
Looking at the analogue world, most true pro gear has a headroom of 
+24db of headroom before it clips.
Much gear has a lot less than this..

>the analogic output that is "hotter" and I easily get good level 
>without normalizing.

Also you might prefer the sound of the analogue outputs?
And yes the less messing about you do with your recorded audio the 
less errors start creeping in, though this will be less when encoding 
at 24bits (its a bigger number representing your audio) 
Cheers
Gonz