Page 1 of 1

XITE-1 RAM

Posted: Sun Sep 06, 2009 4:39 am
by dawman
Let's call this the Telethon Project.
4 x AUX, 4 x stereo BUSS for live FOH mixing, and 4 x BUSS live recording.
AUX # 1 is gated and used for Solo'd instruments.
The other 3 x AUX's have different Decay length's and sizes so I can place the instruments according to depth/distance or the percieved seating of an Ensemble.

This is why I am thinking my motherboard RAM is still being accessed.
I have tons of RAM on the XITE-1, and 4 x PT-96k's will get me this dialog.
I am trying to assign all 4 Reverbs to a single DSP chip.
These Verbs are older than the SC/Timeworks versions.
I can get 2 or 3 of the SC Plate-X's on a single chip and I never see this dialog...ever.
Is it because the Paul Tanti Reverbs are much older and cannot access the RAM of the XITE-1?
untitled.JPG

Re: XITE-1 RAM

Posted: Sun Sep 06, 2009 9:24 am
by cortone
Yes.

That type of updating will likely require SDK 5.0. I've never heard a timeline for its release.

Paul Tanti has long since left the Scope world, I doubt that these will ever be updated, unless he is kind enough to release the source code to one of the devs who are active now.

Maybe some one here knows him and can inquire?

Cory

Re: XITE-1 RAM

Posted: Sun Sep 06, 2009 11:33 am
by dawman
I prefer the P100's but the PT-Free-96k has a great sound and small footpint.
The PT-2012 uses the same amount of DSP as the P100 but falls short of the massive editing and precise tail-oring the P has.
But I think these were made way back when, and definately were meant to be used with a few cards.
The A100/P100/Inverse and Chorus Delay haven't been optimized yet, but they work flawlessly, except for the CDelay still with its non-insertable status where presets cannot be used........ :x
In Time.

Re: XITE-1 RAM

Posted: Sun Sep 06, 2009 11:39 am
by netguyjoel
Are there actual RAM slots in the XITE, or did they put the chips right in the board? thanx JV..:wink:

Re: XITE-1 RAM

Posted: Sun Sep 06, 2009 12:14 pm
by Warp69
Nope - I don't think XITE access the RAM on your motherboard.

The PT-Free-96k is a reverb that is optimized for 96Khz samplerate - Are you operating at 96KHz? Remember that we (reverb developers) use twice as much RAM for 96Khz compared to 48Khz.

Im not even sure that XITE is able to transfer data over the PCIe interface like the old PCI boards - is the PCIe board and the cable a real PCIe extension with full bandwidth (2Gbit/s interface)? With a 10m long cable? I would imagine that they have allocated 80 streams full duplex for ASIO, WAV etc.

Re: XITE-1 RAM

Posted: Sun Sep 06, 2009 12:36 pm
by cortone
Paul T. made those many years ago. He was a CW employee in NA, who got squeezed out during the first round of CW's financial troubles. A sad loss for CW and us.

The XITE does access system RAM, at least based on my email conversations wilth Ralf D. Per his explanation, this is why the STS samplers are problematic at this point, they have not been optimized either for the PCIe bus (or the internal RAM?). PCIe is much greater than 2GB/S, so there is bandwidth to lose if they so choose.

It wouldn't make sense to add the cost of RAM sockets, and a human to install the RAM during manufacturing, to a closed box like the XITE, plus it would be hard to find sticks to match the 768M that is allegedly on board. Ralf D. would know...

Signed, JV :wink:

P.S., has anybody heard from the Body of Ralf recently?

Re: XITE-1 RAM

Posted: Sun Sep 06, 2009 12:48 pm
by Warp69
cortone wrote:The XITE does access system RAM, at least based on my email conversations wilth Ralf D. Per his explanation, this is why the STS samplers are problematic at this point, they have not been optimized either for the PCIe bus (or the internal RAM?). PCIe is much greater than 2GB/S, so there is bandwidth to lose if they so choose.
Well - 1 PCIe V1.1 lane is able to transfer 2Gbit/s and I believe that the PCIe board included in the XITE package is a PCIe 1x board.

Can any one confirm that the interface on the XITE can handle 2Gbit/s?

Re: XITE-1 RAM

Posted: Sun Sep 06, 2009 1:16 pm
by cortone
Warp,

It is an x1 card. Do you know if it is only Gen 1.1? Gen 2.0 has been out since 2007...it doubles the 1.1 throughput.

It would be great if someone could test the actual bandwidth. Some serious equipment involved there! There are apps to check video card bandwidth, maybe one of those could work? Beyond my skills...

Cory

Re: XITE-1 RAM

Posted: Sun Sep 06, 2009 10:50 pm
by siriusbliss
netguyjoel wrote:Are there actual RAM slots in the XITE, or did they put the chips right in the board? thanx JV..:wink:
I can't open the chassis to see how they did RAM, but my presumption is that the RAM are distinct chips on the main board.

We'll have to track down Jim's photo from NAMM and see...

Greg

Re: XITE-1 RAM

Posted: Mon Sep 07, 2009 4:01 am
by dawman
Here Ya' Go...

It's only a pic of the DSP chips showing.
But the bottom 12 chips are the new ones, and on the back of the PCB for each new DSP chip there is a 64MB RAM chip attached.
Pretty serious shit.
That's why someone who is knowledgable w/ DSP programming of Reverbs and time based effects should be frothing at a chance to turn this box into a full fledged 5.1 / 6 x algo'd reverb unit, while still having mixing capabiliuties.
I can imagine this becoming a serious mixing & mastering solution for 5.1/96k.
If a really well designed 5.1 reverb could be used this box would rival or surpass the current PCM96........ :o

I don't know about you guys, but I am as excited about our little 1U as much as the day I saw it at NAMM.
Every gig or rehersal and practice session is just a joy to do.
To me, when something inspires me to practice, it is worth it's weight in Gold.
I never seem bored with music beacuse I am fortunate enough to be in Vegas where we have built in crowds even on a weeknight. XITE-1 is as good as it gets......I wake up everyday and run to my racks before I even turn on the Cappacino machine...!!
My recent gig playing by the misters on the sidewalk was at first a bull shit job as I had to compete with the dancing showgirls who entice everyone, and as soon as they would split and run back inside everyone would leave, and I had to play the stupid Piano parts I was hired to do.........fuck that shit. I started playing Hendrix and heavy Guitar & Bass grooves and now the dancers stick around and even the people stay and dance and buy drinks.
If I had VDAT for the backing drum tracks and other key instruments it would be even better.
Today I add 3 more Van Halen tunes ( pre Sammy ) so I am always happy to go to work.
If I only had a Motif or a receptor full of weak VSTi's, I would be miserable and ashamed to play. But I have a job to do...to make people stop and buy drinks after the half naked ladies leave......without XITE-1, I would be another chump on a fake Piano. But now I compete w/ a famous DJ who performs on the sidewalk, and he even comes down to hang out on breaks.
So I took a shitty losers job and turned it into a festive gathering with free drinks and top shelf trim because of the XITE-1....Ankyuvarymush.
IMG_1688a.JPG
OJ is still excited too. I send him a few mp3's and pics of the gigs as he was so impressed at the show, he begged me to keep him informed on it's progress.


OJ_XITE-1.JPG

Re: XITE-1 RAM

Posted: Mon Sep 07, 2009 9:05 am
by netguyjoel
Makes PERFECT sense! Especially with single chip assignment capabilities! :)

Re: XITE-1 RAM

Posted: Mon Sep 07, 2009 9:50 am
by siriusbliss
Having done PCB layout and chip design for 28 years (on the side while I'm busy NOT being a rock-star), I am still simply flabbergasted that they were able to get that much DSP routed on one PCB.

I see those famous A/D's and even some big juicy PFETs there, so we can see that even the power supply was well designed.

I still haven't had time to really check out those mic pres yet :roll:

Greg

Re: XITE-1 RAM

Posted: Mon Sep 07, 2009 10:17 am
by cortone
It's been rumored that they are using a 24 layer board (here on the Z, never heard that from SC). That would be luxurious for the layout efforts! If true, it's an interesting trade-off between materials/manufacturing and layout/validation costs, and a smart one. Put the effort into the layout and sink the cost into the mat/man, and drop the validation costs to the minimum (because they designed effectively enough to remove the signal integrity issues). Engineering brilliance...all for us!

Cheers,
Cory