Phase Issues with STS

scary808 · Post by **scary808** » Tue Feb 10, 2004 4:46 pm

I've been working on a project that includes 1 STS4000 & 1 STS5000. As the track has grown, I sampled some of the drums, hits, etc...

I'm nearly finished with the track getting ready to mix & to my horror, 3 of the mentioned samples are phasing. Here's the weird thing; I removed all effects, I made sure all of the Auxs were turned off & it's still there. The samples collectively are 5MB & I have 512MB of RAM. Another thing, one of the phasing samples is stereo & the other two are mono.

Anyone else run into this? Anyone fix it? It would be a shame if not, I've actually grown to love these samplers.

marcuspocus · Post by **marcuspocus** » Tue Feb 10, 2004 7:26 pm

Are you mixing in STM1632? There is a phase compasation button, maybe that's it?

scary808 · Post by **scary808** » Tue Feb 10, 2004 7:52 pm

I'm using the stm 2448. The phase comp button to my knowledge is there to automatically adjust timing differences due to effects. Even though SFP is super speedy, even EQs cause a slight delay if inserted on a channel. Phase comp just delays all channels to match up to the channel that is delayed most from signal processing.

Anyway, I think I've determined that this problem is a Logic/RAM problem. If I play the sample from the keyboard it doesn't phase, If Logic plays it back via MIDI it does. It also phases when I click on the key in the matrix edit window. I've already made sure there aren't doubled notes so that isn't it. Hmmmmm...

[ This Message was edited by: scary808 on 2004-02-10 19:52 ]

scary808 · Post by **scary808** » Tue Feb 10, 2004 8:19 pm

I found it! It was a MIDI feedback loop. I had the MIDI I/Os connected incorrectly. The STS' are sounding beatifully again!

deejaysly · Post by **deejaysly** » Mon Apr 05, 2004 9:30 am

Hi, just picking up on this one again...

I noticed a phasing problem when using any of the STS samplers and stereo samples, in my case drum samples. I got a phasing effect but it also seemed to vary. When I had one track on it's own it seemed ok. Oh yeah, this was noticed when I switched the mixer output to mono (STM4896) and phase compensation was no help. Tried orbitone's(?) phase fix module and it seems i have no phase issues through the mixer. It more seems like that the STS sampler is not playing each channel back at the same time consistently????? I got rid of the Emulated MIDI drivers in Cubase SX aswell thinking it might be this but no joy either!

I'm not sure whether I am doing something wrong as I am using Cubase SX which I am new to (used Cubase 5 VST32 before this).

Scary, can you tell me where your MIDI feedback loop was? in your SFP project? or was it something in the sequencer?

I think MIDI thru is enabled on Cubase SX but I am not feeding anything into the Sequencer Dest module so I can't see how a loop is being created????

I've had to do with using mono samples for my drums which seems to be ok. I could use Audio tracks but I try to contain everything within the SFP project.

If scary or anybody knows I apologise in advance if it turns out that I am doing something completely stooopid!!!

cream · Post by **cream** » Tue Apr 06, 2004 6:09 am

I recently started a thread about this subject(stero image->problems) but i didn't find the solution either. i converted my drums into mono, at that time it disappeared ofcourse..i'm also questioning myself if cubase is teasing me but i've never had this before..Not even with the bigmixer...Probably there's something i did wrong (nobody has this problem on planetz)..i'm working on it..

[ This Message was edited by: cream on 2004-04-06 07:12 ]

scary808 · Post by **scary808** » Fri Apr 09, 2004 12:24 pm

So I still have occasional pahse issues. Here is my theory:

My PC is an older off-the-shelf unit. It came with 128MB of SDRAM standard. I added 384MB of SDRAM. I believe I was able to install faster RAM sticks than those which were supplied with the PC. If the samples are stored indiscriminantly within all available RAM, some audio information could possibly be lagging in turn causing phasing.

I plan to get a new machine soon & give my old one a gangland beating. When I do, I'll update this thread with my findings.

Anyone else with odd RAM combos?

scary808 · Post by **scary808** » Fri Apr 09, 2004 12:57 pm

Oh yeah. The feedback loop was from a redundant connection I made. It partially solved the problem(as you can see from my previous post). The seq source was connected to the pulsar midi dest. I was using external midi gear in a previous incarnation of the project.

Basically what was happening was some midi was being triggered twice beacause it would travel from the (deep breath) seq source to pulsar dest. Then through my JP8000 to pulsar midi source to seq dest. The midi info would be doulbed on any midi track that was armed for record(as though I were playing along with the sequence a millisecond or two behind). This is a bit difficult to word but I hope this helps. I think the issue I discuss in my previous post is the culprit for me though.

valis · Post by **valis** » Fri Apr 09, 2004 2:46 pm

Scary, are you using any of the sampler filters?

scary808 · Post by **scary808** » Fri Apr 09, 2004 4:02 pm

Not in this particular instance.

valis · Post by **valis** » Sat Apr 10, 2004 2:21 am

I would actually wonder if some of the voices are winding up on another dsp chip. You might try to optimize your usage of voices by using only what you need.

deejaysly · Post by **deejaysly** » Sat Apr 10, 2004 7:15 am

Thanks for the info Scary. I understand what you are saying. Unfortunately, I haven't got a connection like that in my project. All I got is seq midi source into the STS and I get this phasing problem.

cream · Post by **cream** » Mon May 17, 2004 10:59 am

have the solution!

Quite simple actually..
If you have a program filled with stereo samples you should check the polyfony. Every stereo sample = 2 voices. When you set the polyfony to 3 or 5 or any other unequal number this phasing problem occurs.
so if you layer 3 (drum)sounds on top of eachother, make sure the polyfony is set to an equal number.
This may be the first lesson in sampling but i've never thought of it before. Hope this is the answer to your problem..

[ This Message was edited by: cream on 2004-05-17 11:59 ]

valis · Post by **valis** » Mon May 17, 2004 12:35 pm

Nice to see you resolved this

scary808 · Post by **scary808** » Tue May 18, 2004 12:01 pm

On 2004-05-17 11:59, cream wrote:
have the solution! Quite simple actually..
If you have a program filled with stereo samples you should check the polyfony. Every stereo sample = 2 voices. When you set the polyfony to 3 or 5 or any other unequal number this phasing problem occurs.
so if you layer 3 (drum)sounds on top of eachother, make sure the polyfony is set to an equal number.
This may be the first lesson in sampling but i've never thought of it before. Hope this is the answer to your problem..

[ This Message was edited by: cream on 2004-05-17 11:59 ]

I tried this one a while back. It still didn't correct my problem. When my poly is set too low, samples merely are cut short. I can have my poly boosted all the way & I still get phasing.

alfonso · Post by **alfonso** » Wed May 19, 2004 4:26 pm

The problem is the slow MIDI protocol. If you send 3 "contemporary" midi note on messages, they are not contemporary at all, they are sent in sequence, and very very far from the audio sync signals, even their timing is not predictable, so not fixable with offsets. So if you hope to get, f.e., a nice kick playing 2 or 3 sounds together via midi, forget it, this can't be done this way.

cream · Post by **cream** » Thu May 20, 2004 3:32 am

The problem is the slow MIDI protocol. If you send 3 "contemporary" midi note on messages, they are not contemporary at all, they are sent in sequence, and very very far from the audio sync signals, even their timing is not predictable, so not fixable with offsets. So if you hope to get, f.e., a nice kick playing 2 or 3 sounds together via midi, forget it, this can't be done this way.

Well, this wasn't my problem..I know layering 3 kicks on top of eachother cannot work.(unless you sample it).The problem was that layering different samples gives phasing. For example:kick,snare,hihat.

alfonso · Post by **alfonso** » Thu May 20, 2004 10:20 am

On 2004-05-20 04:32, cream wrote:
The problem is the slow MIDI protocol. If you send 3 "contemporary" midi note on messages, they are not contemporary at all, they are sent in sequence, and very very far from the audio sync signals, even their timing is not predictable, so not fixable with offsets. So if you hope to get, f.e., a nice kick playing 2 or 3 sounds together via midi, forget it, this can't be done this way.

Well, this wasn't my problem..I know layering 3 kicks on top of eachother cannot work.(unless you sample it).The problem was that layering different samples gives phasing. For example:kick,snare,hihat.

Sorry but what do you mean for "phasing"?
If you mean that every time the combination sounds different, that's exactly the same problem...to properly speak about phasing you have to compare two identical waveforms and see how one is shifted from the other, but obviously percussive sounds have some common freqs., and the fact that the trigger is not simultaneous and not time fixed gives you that effect.

Another situation could be that you hear the "phasing effect" on a single sound, and that can happen on sounds that receive a double midi note on message, when for example they get both the message from the keyboard and from the sequencer midi through...

STS's work perfect here under this aspect, it's just MIDI that's not perfect.

deejaysly · Post by **deejaysly** » Fri May 21, 2004 2:51 am

When I was experiencing a "phasing" problem, I had stereo samples for kick, snare, open and closed hat in STS (either one, even tried the old sample players). I would play the 4 MIDI track drum sequence in cubase. I could hear something funny going on so I put the mixer into mono mode. I then experienced this "phasing" effect on the snare every other hit and I could also hear the hi hats phasing almost like every other aswell. If I muted the other MIDI tracks and just left the snare going the phasing would not happen. If I remember righly it appeared relate to the amount of MIDI activity so it is possible that alfonso is on the right track with our problem. Also, I have looked for the obvious doubled up notes in my sequencing and didn't find any but I don't get this problem with mono samples so it can't be that.

Also the polyphony thing, the above problem was set to an EVEN number of voices as cream suggested.

I understand alfonso's explanation but is the MIDI timing really that BAD!!!? The problem I experience is like I have a stereo sample and the MIDI timing is different for each channel of the sample. I thought seeing as one MIDI note is triggering the sample, it's like STS is not playing back both channels of the samples at the same time??

[ This Message was edited by: deejaysly on 2004-05-21 03:51 ]

cream · Post by **cream** » Fri May 21, 2004 10:03 am

@ Alfonso: with phasing i meant ''out of phase''...

[ This Message was edited by: cream on 2004-05-21 11:03 ]