a 2nd PC as a sampler question

jilhead · Post by **jilhead** » Thu May 10, 2001 10:18 am

Has anyone a solution for the following...

I'd like to set up a second PC as a sampler (the first PC being the sequencer). I was thinking of using the STS4000. I believe that the STS4000 can handle 64 voices? How can you send enough midi information to access all of these voices? The obvious limitation is that the Pulsar has but one midi in port with only 16 channels.

Perhaps there is a midi card available that will provide midi information to the Pulsar via a 'Sequencer Midi Source', without having Logic Audio running.

Gregory · Post by **Gregory** » Thu May 10, 2001 3:00 pm

I don't understand the question.

The STS4000 can only handle the standard 16 channels/instruments of MIDI and among those 16 instruments it can play up to 64 total notes at once (i.e. 64 notes of polyphony). So if you want to run the sequencer on one PC and Pulsar on another you should be fine.

But why have a Sequencer and Pulsar on separate PC's? A standard 1 Ghz+ PC should handle both well. You probably need to describe your studio setup and how you work in a little more detail. If you were running Gigasampler then you would REALLY need a separate PC.

I'm not quite understanding your questions. Like I said, the readers here would probably appreciate a little more information on how you work in order to help out.

algorhythm · Post by **algorhythm** » Thu May 10, 2001 3:56 pm

I think that 64 voices means that the sampler is 64 voice polyphonic, and that it is multitimbral in that it can play 16 different timbres at a time in a program.

A single MIDI cable is adequate for this, so I do not see the problem of a multi PC setup.

However, If you are using the STS4000, I do not see the point of a multi PC setup either, as the dsps are doing the work.

Buy more RAM

MeloManiac · Post by **MeloManiac** » Thu May 10, 2001 5:36 pm

I am considering buying the STS3000... How much DSP does these samplers use? I am hoping for a very low DSP usage....

Nestor · Post by **Nestor** » Fri May 11, 2001 1:33 am

Using the STS4000 and 8 Sampler Players and placing the instruments in a way that they can all be played individually in my controller keyboard, I can handle about 30 instruments at once! Yes… And I have just the Pulsar I, with 4 sharks. Of course I've got quite a lot of ram nevertheless: 512 MBs PC 100.

Everything works fine, the only problem I sometimes have is the opening of the project. Because loading so many samples at the same time gets the system stacked and keeps forever in "Loading samples". But all I have to do is to close Pulsar and restarting it once or maybe twice, which is not so terrible when you think you can get so many instruments, and it works fine with an amazing polyphony number. I would use between 5 to 12 notes polyphony in each of the Sample Players and about 50 or 60 in the STS!

BUT!!! And please, remember this! Make back-ups of all your projects! For instance, if your song is called let's say: "Active Motions", save a second one "Active Motions, back-up" so it appears besides your project in the list of projects. As you can understand, working and fine-tuning about 30 instruments can take hours so it is sometimes tragic to lose all those precious parameters. With a back-up you can save the situation.

When there is many samples loading if your sequencer crashes, your project may well become useless.

The STS is a fabulous instrument.

MeloManiac · Post by **MeloManiac** » Fri May 11, 2001 2:32 am

Scuse me, but all those problems you talk about doesn't sound 'amazing' to me. And that you can play 30 instruments with ONLY 4 DSP's.... doesn't impress me either as I can play around 64 instruments and 160 voices poly taking up up to, let's say 20 GB of space, with Gigastudio without any DSP and just a 1Ghz P3 or TBird.... The two reasons I have to move my drum programming from Gigastudio to STS are these: 1) I will get all digital - no recording of the drums like I have now from one computer to the Pulsar machine. 2) Better timing. When I stack bassdrums on top of each other it shows that either the timing of Gigastudio or the MIDI timing is not too good - I get small phasing problems and each 'hit' of samples don't sound the same. I was hoping that I could just put my drums in the STS-3000 and get 1-2ms latency and very tight timing. And that without using much DSP. I can't see how sample playing can take up a lot of DSP - I don't want to use any filters or anything.

Zer · Post by **Zer** » Fri May 11, 2001 2:44 am

Sorry but the timing lacks in the creamware sts-4000 ... I haven´t been able to use this device because the timing is so bad...any sampleplayer or the EXS-24 does a better job for me.

[ This Message was edited by: Zer on 2001-05-15 10:31 ]

Mo · Post by Mo » Fri May 11, 2001 3:59 am

On 2001-05-11 03:32, MeloManiac wrote:
Scuse me, but all those problems you talk about doesn't sound 'amazing' to me. And that you can play 30 instruments with ONLY 4 DSP's.... doesn't impress me either as I can play around 64 instruments and 160 voices poly taking up up to, let's say 20 GB of space, with Gigastudio without any DSP and just a 1Ghz P3 or TBird....

äähm... you should not forget, that the sharc´s on the board just run at 60mhz. and on p1 it´s just four of these... so... what´s up with that 1000mHz-cpu...?
i would say the sharc´s are some more performant.

MeloManiac · Post by **MeloManiac** » Fri May 11, 2001 8:19 am

The difference is that DSP's were made for sound purposes only, so you can't just say 4*60mhz = 240 against 1ghz CPU. I would expect sample playing to take up close to NO DSP time, I'd even prefer it to use CPU and not DSP for whatever calculations there are. Using DSP for sample playing seems like a waste to me...

subhuman · Post by **subhuman** » Fri May 11, 2001 10:19 am

STSx000 timing is rock solid. If you're having problems then perhaps you've installed something like Hubi's Loopback device or some other virtual midi driver that is messing it up.

You can't really compare the STSx000 with EXS, since it's 2 entirely different ways of working. STSx000 is like an instrument that you play in real time with no latency (essential for drums or anything with sharp attacks), and EXS is not, there will be 2-20x as much latency (only 13ms with older CW products, "as low as 2ms" with new cards, ULLI, killer processor and 96khz). If you don't play in realtime, then you may not notice this distinct, but very important difference.

MeloManiac · Post by **MeloManiac** » Fri May 11, 2001 10:28 am

subhuman, how is the DSP usage on STS? Does it depend on how many samples you load/play or does it just eat up one chunk no matter what? I want to know if I should hit that PURCHASE button on CW's site or not!!!:)
I want LOW DSP usage.... Lets say.. playing 32 voice poly should take up no more than half a SHARC or so...

subhuman · Post by **subhuman** » Fri May 11, 2001 10:37 am

It's a Sampler, not a Sample Player

So, for each voice, you get LFOs, envelopes, and a filter. You can do a little bit of modulations here, assign things to the envelopes (like velocity), and control one single parameter externally (like filter cutoff).

On 4 DSPs (single PulsarI), you will just be able to fit 64 voices, and maybe a mixer and a few (small) effects. Remember that you get a bunch of extras per voice and that is where a lot of the DSP is eaten. But, I think it is very DSP efficient. In fact, I wish it was more fully featured on modulation routings, (and took more DSP), then I wouldn't need a hardware sampler in addition to do this. I guess I could have just bought a SCOPE board and plugged a modular patch behind each output (but, still I wouldn't have the modulations per voice but at least per timbre).

MeloManiac · Post by **MeloManiac** » Fri May 11, 2001 11:54 am

So, you can't turn off the DSP intensive functions.... sigh

Nestor · Post by **Nestor** » Fri May 11, 2001 3:42 pm

You are right MiloManiac, but for me 30 instruments at once it's amazing! My Pentium it's just a 550 III. I'm really impressed that you can play so many instruments with Gigastudio with such polyphony. But even so, I'm still satisfied with the STS4000. It takes some DSP no doubt about it, but hey, it is still very efficient.

Anyway, a question… Where do you place so many instruments, does Gigastudio has a special implementation for midi rutting or something like that? I've never played nor tried Gigastudio, so I don't know.

MeloManiac · Post by **MeloManiac** » Fri May 11, 2001 4:04 pm

GigaStudio has 4 MIDI ports = 4x16 instruments at once.

jilhead · Post by **jilhead** » Fri May 11, 2001 5:29 pm

Well thankyou for the answers... it seems that I don't need a second PC, but I'm a little confused. Is there a verdict on why one should use the STS4000 v's, the EXS24, or gigasampler.

The main issues seem to be:
- timing isn't tight on any of the samplers (STS4000, EXS24, or gigasampler)
- there is a difference between 'sampler' and 'sample player' (effects, filtering, modulation, etc)
- STS4000 can be played as an instrument because of low latency of the pulsar.
- gigasampler can play from a larger selection of samples

Is this a fair representation?

Thanks again for the great information,
jilhead

MeloManiac · Post by **MeloManiac** » Sat May 12, 2001 4:47 am

Probably a good presentation yes. GigaStudio has the advantage of very advanced programming capabilities with keyswitching changing patches etc., and of course that samples can be up to 4gb. But I noticed that when I 'fire' 3 bassdrums at the same time they don't hit at the same time in Gigastudio. Maybe it is a problem of MIDI communication from my 1st machine to the Gigastudio machine, maybe MIDI just can't bee tighter? I don't know.. I was just HOPING for a way to be able to play totally tight drums for dance-music and such. I thought STS was the answer, but I am not sure.

Nestor · Post by **Nestor** » Sun May 13, 2001 3:20 am

Thanks for your answer MeloManiac. Now I fully understand! 16 MIDI X 4? Wow!!! This is just too much. I have to try that. The problem is that GigaStudio is quite expensive, isn't it?

Anyway, how does it work with the rest of the system, is it stable and works fine?

MeloManiac · Post by **MeloManiac** » Sun May 13, 2001 3:32 am

It is quite expensive, but when I got it there was a cheap upgrade path from Gigasampler to Studio. And it works fine. I had some trouble with weird hanging MIDI notes and such when I used it on my P3 system, but on my Athlon it has been rock solid. There can be clicks and pops in the sound sometimes if you stretch your harddrives performance and go up to very high polyphony, but I can live with that. It has some incredible sound libraries, and as I do mostly orchestral stuff, it is really the best solution available with a new incredible string library coming out soon, and I know people are working on a brass-library too... Also some guys called DS Soundware are doing fantastic stuff for it, like Ultimate Orchestral Percussion, 4 CD's with timpani, bassdrum, glockenspiel etc. - the tubular bells patch alone takes up 300mb or so. And they just released 6 CD's with timpani only....

jilhead · Post by **jilhead** » Sun May 13, 2001 2:16 pm

Does gigasampler play akai samples too?

I like the idea of having lots of instruments 'on tap'.

Regarding sound libraries. Has anyone heard of Miroslav (I'm not sure how it's spelled). I'm not sure where to look with regard to sample libraries. I haven't found anyone who stocks any kind of range in Ireland yet.

I have been looking on and off and on for a while now. The answers I have received so far are:
- we don't do those
- we've got xyz... (low budget scratchy stuff), and I don't know what I'm selling, so I can't answer your questions

How much are these gigasampler libraries going for? I take it there's no brass libraries yet then?

Thanks,
jilhead