New samplers needed !!!

[dawman]

I would love to see wav file Waldorf style Oscillators.

That form of synthesis is so under estimated, and over looked.

Bowen's RDII Modules has a Waldorf Pair II Oscillator that I am using in Prowave, and more recently SpaceF's Multi-Synth. It is amazing sounding in it's 8bit form, but I can only drool over what a wav file based Oscillator would do.

Whoever builds it should match up the waveforms better though IMO.

For example, start w/ a basic Sine Wave and have it recorded several different times w/ slightly different timbres, then record Triangle, Square, Saw, and then Pulse. This way when using a slow LFO to sweep the wav's the transitions are smoother. I have GB's of unique new waveforms created in Symbolic Sounds Kyma Capybara, and insert them into the Sample Oscillators of Bowen synths, and they sound incredible, but some are too large and can't be used effectively. But I could achieve some incredible sounds if I could morph or sweep through these wav's somehow.

Lagoausenti,...My Brotha',

I do all of my editing of the Giga libraries in GS3, such as layer deletion.

If I were to use the STS series I would then have the GigaEditor convert to 24bit wav., prior to the STS conversion. Translator Pro, or ESC should be able to transfer all of the related layers in their proper place and retain their dynamics and filters. I suppose if they don't transfer totally, the sounds would be there, and a few small tweaks in the STS's could fix all non sounding performance issues there.

I was nervous about HDD's streaming the samples since they are mechanical, and less stable than a RAM based sampler. But for several years, even before I joined our beloved platfrom, HDD's proved to be a better choice for using large libraries.

If you need a great 16bit, 16MB sampled piano. Try to find the Jerry Coakley piano that Alesis, and Ensoniq used in their pianos. It is available as a SF2 I believe, and from what I heard the 48MB / 24bit example of it is quite good.

Truthfully, I wanted to use STS, but Gigastudio / GVI has proven to become a very stable live performer. w/ high quality libraries.

But if I had to have big drum sounds live, I would definately listen to the Akai MPC's next to a DAW, you might agree that the sound is much bigger, but then again to record, it seems that the in the box method is where it's at.

Thankfully we have many choices.

[hifiboom]

[I would love to see wav file Waldorf style Oscillators.

me too! the waldorf oscillator is great for what it is, although the steps running through the patterns are quite hard/digital sounding.(lesser resolution, i guess its to be like that)

Think about what would happen if the sc guys would implement a wav oscillator that does heavy interpolation when you move through the wavetable(s). and much more steps for finer and smoother roll through the different waveforms.
and also more interpolation when you play at different pitch than the sampled waveform. there are some artifacts with the basic wav osc when you play the higher notes.(aliasing)

The waldorf oscillator can even emulate sync oscillator sounds via that technique but simply in an inferior quality.

even waldorf have improved their own osc models a bit. I had a quick view into that blowfeld manual and they now have a brilliance setting on the oscillators for more sound variation.

SC just needs to bring that stuff a bit further. Alll what we have is great, having more of those little gimmicks is just awesome.

A full feature high quality WAV oscillator/wavetable osc would be the ultra hot thing. For modular and even for a complete synth design built around it.
Could be also useful for the solaris synth.

[mr. prawn]

soft samplers do tend to sound thin & the filters are always a letdown. i have an s-760 sitting under my bed which i dont use because its too disruptive to my workflow - even w/ the mouse/monitor set up - dealing w/ scsi & the archaic interface is a bit much (also i have a small attention span). but the sound! its wonderful - i could use all kinds of superlatives but its just nice - the filters will self-oscillate (unlike akais) & your samples will always come out nicer at the other end. anyway im moving houses soon & may attempt to set it up properly.
oh and akai (numark now) have dropped the ball w/ their converters/filters with the newer machines. i use a mpc2500 & the converters are cheap & kind of harsh sounding, the filters are almost unusable. the interface is great though, its a joy sequence on & sample-editing is quick & easy but i wish it sounded nicer (8 analogue outputs + digital so i can run everything through scope to improve things)...
anyway since we're on samplers - how about a harware dsp box with high quality convertes, filters & so-on but editing & sample-management is on the computer. i mean, that way you get the quality of a hardware sampler without the load on cpu (which of course is huge especially w/ clunky apps like kontakt)...

that said - i dont think SC should pursue a sampler but some more advanced wavetable stuff would be brilliant (i like everything hifi is saying). its kind of the missing link in modular (some better drawable, multi-seg envs would be nice too) - a proper wavetable osc w/ user wavs & wave scanning/morphing, with flexors shapers/filters/granular stuff should produce some pretty unique & 'modern'-sounding results.

[mr. prawn]

[I would love to see wav file Waldorf style Oscillators.

me too! the waldorf oscillator is great for what it is, although the steps running through the patterns are quite hard/digital sounding.(lesser resolution, i guess its to be like that)

Think about what would happen if the sc guys would implement a wav oscillator that does heavy interpolation when you move through the wavetable(s). and much more steps for finer and smoother roll through the different waveforms.
and also more interpolation when you play at different pitch than the sampled waveform. there are some artifacts with the basic wav osc when you play the higher notes.(aliasing)

if someone developed this, it would be nice to be able to switch off the interpolation too so you can get grungy ppg style sounds. would it be difficult to make such a device? im sure adern could put something together - the wave shaper module is almost there except you cant sequence its waveforms (you kind of can but its rough & steppy).
you can get pseudo wavetable synthesis using the mod mixers to morph b/w different wavetable oscillators or wave shapers but it would be nice to have some dedicated modules.
i quite like the way wave sequencing is handled in the 'oki computer' patch from reaktor:



each of those 16 boxes underneath the waveform has a different wave & you drag the mouse through them at whatever speed (you see the waveform evolving) & then the sequence you made is played back when you hit a note. or you can use the little table sequencer. in scope i guess you could use the 'draw' lfo or one of flexors sequencers to achieve something similar, if only we had this magical wavetable osc w/ a few mod inputs...yesh

[dawman]

I actually have a pretty FAT sound I learned from using the QWave from Bowen.

But that beast is so DSP hungry I decided to use the RDII Waldorf Pair II Oscillators is SpaceF's WS4-MKII.

I was using them in ProWave, but in WS4-MKII I can use 6 pairs !!

You only need 2 of SpaceF's Slots for the RDII's and the other 4 or 6 slots can use more Adern, Bowen or SpaceF Oscillators.

I especially like the S-MOS Oscillators as they are complete synths.

With WS4-MKII the modulations, adders, and routings were made for this.

I could post a small snippet here as a demo if someone would explain how to do it for me.

I'd rather not put a snippet in the Music thread. That's for short performances.

Oh the hell with it, here's my style of wavesweeping.

http://www.planetz.com/forums/viewtopic ... highlight=

[lagoausente]

I do all of my editing of the Giga libraries in GS3, such as layer deletion.

If I were to use the STS series I would then have the GigaEditor convert to 24bit wav., prior to the STS conversion. Translator Pro, or ESC should be able to transfer all of the related layers in their proper place and retain their dynamics and filters. I suppose if they don't transfer totally, the sounds would be there, and a few small tweaks in the STS's could fix all non sounding performance issues there.

If you need a great 16bit, 16MB sampled piano. Try to find the Jerry Coakley piano that Alesis, and Ensoniq used in their pianos. It is available as a SF2 I believe, and from what I heard the 48MB / 24bit example of it is quite good.

I already can load any gig piano in STS. Just was asking you for corious. The converters like ESC and CD-extrac seem to work ok, unless with the aditional folders, like release samples, pedal samples, etc.
The best way I found for converting, and editing, is importing the gig into Halion. Halion has a nice drag and drop editor what allows direct selecting a zone of samples with the mouse, even a layer zone (y-axis), or along the key range (x-axis). Just can simply select, copy, paste, and edit all easy.
The other point is in the right column, where the samples are clasified in folders, so is very easy to delete all aditional folders, and save the new halion program with samples to a new folder.
ESC allows converting from Halion to STS.
A big Gig piano library can be loaded this way easily in STS, and the feeling playing there is very fine.
Just a note, the better thing that STS have is not latency, what is still good. A modern computer can give a very low latency with a native sampler, but what fails (have measured it in diferent tests), is the time accurracy.
If you run a metronome to use as trigger of the sampler, and record the sampler audio output, in the sofware based, the clicks never have the exact position between one and the next. There is a jitter, probably due to the queing in the cpu processes.
In the same test with STS, the result is exact, it has no jitter.
The jitter in the Native, is no so evident to the ears, so much people don´t take care, but gives a diferent feeling when live playing.

[dawman]

Great Explanation My Brotha' !

I have a free copy of Halion that came w/ Cubase 4.1. I will definately check out your findings, they are quite unorthodox but obviously very effective.

It seems like you are the go to guy for STS. I might be contacting you on some ideas if it's O.K.

I will be upgrading to 64bit this year, there's no avoiding it. When this happens I cannot see why STS could not take advantage of the increased memory addressing that inherently will occur.

I do hear latency in all soft samplers except for Giga and STS. Gigastudio 4 will still have kernel level MIDI for that same fast response.

When I first opened up Brainspawns Forte, I actually could hear and feel the latency. It was unacceptable. I would love to try a nice piano in STS though. Latency is never noticable on Gigastudio thank God, but I still would love to check out your piano for STS.

Some of my new GVI BETA's use extreme amounts of memory and CPU. I would love to know if I could use 8GB's of RAM and have leftover RAM for an STS instrument or 2.

A very interesting disscussion, thanks for your great tips w/ samplers and latency.

I would be very interested in seeing and playing a Piano library that you have converted to STS.

O,K,.........let me Beg You...................................PLEASE,........Pretty Please ?

[Herr Voigt]

Jimmy, the STS has a 64 voice polyphony, that means a 32voices stereo piano. And the RAM seems to be limited, that's bad for non-looped samples with many velocities. Is that really enough for such an advanced piano player like you?
Giga has up to 160 voices, that should better match to you.
And: Do your STS 3/4/5000 run really stable? If yes - what a lucky man you are.
Your opinion would be very interesting for me.

Greetz, Thomas

[lagoausente]

I would be very interested in seeing and playing a Piano library that you have converted to STS.

O,K,.........let me Beg You...................................PLEASE,........Pretty Please ?

Well, I´m not the type of guy that play a lot with sampler parameters (Envelope, filters etc..) but I will be happy with a small contribution for anyone use more the Scope soft, so feel free to contact, or PM.
My msn, is jose_ferreira_coira@hotmail.com

The way that Halion arrange the diferent layers is very fine to can edit it.
Once you import a Gig or any other library, halion converts the library to it´s own format (Halion program). The nice thing here is that it creates a folder with each layer (release, note on, note off,damper pedal.. etc).
You just go to the column in the right, expand the program folder, select the subfolder we like to save (for example note on), right-click, appears a dialog, go for archive/save folder with samples and voila, you have a new halion program with that layer.
Can to the same for any of the layers. You can convert then separately each layer to a diferent STS program, and so decide then if you want to use all of them or not depending of the size.
You can then edit the velocity layers for each of the previous layers/folders.
Load for example again in Halion the "note on" program. Select the desired samples to delete, in the keyzone. If you want to set a new velocity range, select the samples, hold shift, and move with the mouse
the range up/down.
You can change any of the parameters (filters, envelope, etc), to "all" the samples, or to the "selected" samples. Can select that option in the upper right corner.
I found Halion is very intuitive to this type of editing, and easy. It would be a good thing STS could have.
I think that STS in the right thing to "play", and record, the type of use that a hardware sampler has, but with the fine beneficts of be integrated in the Scope platform, and with a sequencer.
In drums is specially fine. You can send 14 outputs with a STS4000 to the sequencer, and so have multiple audio tracks in your sequencer (kick, snare, hithat, etc etc) jitter free.
If I have time I´ll try to make a little web with some tests and some reviews of the pros of this awesome dsp based platform.

[lagoausente]

Jimmy, the STS has a 64 voice polyphony, that means a 32voices stereo piano. And the RAM seems to be limited, that's bad for non-looped samples with many velocities. Is that really enough for such an advanced piano player like you?
Giga has up to 160 voices, that should better match to you.
And: Do your STS 3/4/5000 run really stable? If yes - what a lucky man you are.
Your opinion would be very interesting for me.

Greetz, Thomas

Well, I´m not sure the opinion of Jimmy.
I think that 160 voices are simply unnecesary. You have 10 fingers, that means 10 voices at time. When sound the release samples, the note on are not sounding, and viceversa. And probably is better just note on jitter free that a 2gb library with a drunkend midi track. I just use the note on.
Unless you are using 16 pianos at time..

[dawman]

Yes Herr Voigt,

You are correct about the needed polyphony. It comes in handy w/ sostenuto, and sustain pedal work definately. But an Allman Brothers / Lynard Skynard rockin' piano will use very little sustain, and that's what I am interested in.

I already have a rockin' Upright for that type of work, but want to delete several of the velocity layers and have 3 for unsustained parts, and one for the pedal down that would have embedded harmonics pre sampled.

I will see what happens and report back about it. But this won't even be looked into until SFP 5.0 becomes available, and I can dedicate extra RAM for a 64bit Gigastudio 4 set up using XP64.

SonicCore has to be the crux of my DAW, and everything had better fit into that. Once you use Scope mixers / effects live, any other solution would not suffice. If Gigastudio 4 needs a separate DAW, I will even be happier, as STS will be used within the Scope enviroment w/o any interferrance in respect to RAM addressing.

I am prepared either way.

Thanks Jose, I appreciate your help understanding STS better.

[lagoausente]

Just a note. Have in account that 95% of midi controllers doesn´t have
"note off velocity", and so doesn´t take care about how quick do you release the key.
In some piano library maybe find velocily layers for release samples. If your keyboad doesn´t have "note off velocity" , will be totally unusefull.

In the other side, I´m not sure if understand you about poliphony Scopelive.
When you hold down the pedal, Gigastudio will use the "note on pedal on" samples. And if you release the pedal, will play the "note one". Not all them at time.
The same for release, the release will sound when you release the key, and so "note on" and "release" samples will not sound at time either.
And once you release the pedal, all "note on pedal on" stop sounding, so I think you´ll never need such poliphony.

[dawman]

The Polyphony when immatating a real piano comes frrom having to use the sustain pedal too much just to try and immatate the way an acoustic piano would carry over harmonic content by " Half Pedalling " and pressing the pedal at the beginning and ending of musical phrases.

I end up using twice the amount of pedaling to get a closer emulation of the acoustic instrument. An inherent flaw with digital recreations of the real thing.

This means that each note must not be re triggered from the lack of polyphony, like envelopes of a synthesizer re trigger when the polyphony runs out. The problem w/ DAW based samplers is the audio cuts out. A hardware sampler would simply steal a note from the beginning of the phrase, which I would prefer actually. But each note on a piano sample would be 2 multiplied by the number of velocity layers. 3 velocity layers is the minimum, and a simple octave run on an acoustic would require pedal down one time for the octave, On a sampler that means one pedal down for two octaves worth of notes, the math then becomes a little more understandable, as one octave sustained would be approximately 7 notes. times 2 or 4, as you can see the polyphony becomes multiplied very fast. That would cause an audio dropout.

I hate it when the audio just goes blank. Very embarrassing. But as I said 32 stereo voices is accepatable for jamming styles like Orleans, Blues, Southern Rock, Country & Western ( Gag ).

But for pretty music like classical and the kind of piano styles one would use to convince others that he's well trained, high polyphony counts are crucial.

I am notorious for killing my hard drives w/ acoustic piano solo work. I tend to get lazy and use too much sustain pedal, and even my mighty Raptors can be brought to their knees when I do Chopins Fantasie Impromptu. That's w/ 384 voices of Polyphony.

I am waitng for Garritan's Steinway which is suppose to have implemented the una corde sample like a real sostenuto pedal can create. In this way a chord can be sustained w/ the sos pedal and all other notes can be pedalled, and not pedalled separately.Just like they do on a real piano.

Anyway I grow weary of all of this scientific talk about imperfect ways of immatating an instrument, and barely getting close to the real feel of it. Such a waste of time IMHO.

I am just an old hardware cowboy trying to stay in his saddle in this world of virtual everything.

My honest opinions are invalid, as many I try to explain things to, have no real instument to reference themselves to, so it is usually all moot at the end of the day.

I just try to play the instruments in a sampler that sound bareable to me, and pretend that I am really enjoying myself, knowing that in the end, I have played all of the real ones before, so I just shut up and play. Thankfully I am a real Ham when I play so it helps.

No matter how many scientific explanations of virtual instruments we discuss, there are certain instruments that just cannot be copied effectively. Thank fully pianos are somewhat acceptable.

But slap bass, saxophone, REAL strings, and many others require physical nuances that the virtual world will never be able to capture. I am grateful I can play my keyboards and immatate many instruments with a modicum of realism, but many more I wouldn't even attempt to try, as it is unbareably weak, and causes too much laughter while performing.

But I still would like a great little jammin' piano sample library for the STS, as that would become my jammin' piano and they are so much fun to play and require little effort w/ those styles.

[lagoausente]

where come 384 voices? I didn´t understand the math.

[garyb]

384 voices is all about the strange meter in gigastudio....

[dawman]

I started at 400 and worked my way back to 384...........the ppq of all of my hardware sequencers.

I am a little supersticious, afterall, when a hardware sequencer works for 24 years w/o a single crash on or off stage, you tend to get a little tingly allover, and try to preserve it's memory somewhere in other endeavors.

Just think of the MIDI Limitation # and multiply it by 3.

[YiannisK]

Scope4live,

You probably are familiar with this but I'll mention it just incase you are not.
A good piano for the STS is the "Holy Grail" by Q up Arts.

Nice and light .

[dawman]

I was familiar, but forgot that little beauty.

I quickly became overcome by Gigasampler back in '99, and soon forgot about the 128MB Piano. Thanks for the reminder.

I am looking right now at deleting several layers of a great sounding upright from Vinataudio called the Clinton. I have a Steinway Chickering Upright also, but the Clinton is so natural sounding. It's perfect for what I want, and it also has the embedded pedal down harmonics that really makes any Acoustic Piano quite realistic w/o using reverb to bathe it in. But the pedal up ( unused ) needs a splash of the 91 on it, or even the nice sounding Plate from hifiboom. It's quality is nice, especially when sparsely used. He must have created it with percussion samples as it has a very controllable high end.

I am so lame on computers but from creating samples for years, trading, and editing, I am right at home w/ this stuff. I doubt I could even see a small LCD anymore.

Thanks 4 The Tip. I remember that that piano was the talk of the town for hardware samplers, especially my beloved old Emu's, R.I.P.

[lagoausente]

You can load 730 mb in STs, so from 128mb to 730 you have a good margin to create a more than decent jamming piano.
If you want to use the release samples I think would need to convert the note off messages to note on, and send to STS in a diferent channel then the other samples.
I´m not sure if exist other way to do that.
You can experiment with more o less velocity layers and extra pedal/release layers, seems a good exercise to find out how much of that big size libraries you really need.

[dawman]

Brotha' Man Laguoesente,

Gigastudio 4 will arrive before months end.

Once that's arrived and eLicensed, I shall contact you on these matters.

I am sure we can do business together.

I think you might like the Upright sample sound. No phony chorus muddying up thr sound, just a front and center natural Honky Tonk.