Because SOLO patch doesn't mean anything. It's just a name selected for the most right part. In the same way, you are not restricted to play just percussions and drums on the Percussion part.madAhorn wrote:Exactly...why would you choose to play chords with a SOLO patch?
256 voice polyphony?? not as far as i can tell!
Re: 256 voice polyphony?? not as far as i can tell!
-
Bruce Lychee
- Posts: 168
- Joined: 21:56, 24 June 2011
Re: 256 voice polyphony?? not as far as i can tell!
mojkarma wrote:Because SOLO patch doesn't mean anything. It's just a name selected for the most right part. In the same way, you are not restricted to play just percussions and drums on the Percussion part.madAhorn wrote:Exactly...why would you choose to play chords with a SOLO patch?
Then you need to switch off the articulations. Think about the way they work and I think it becomes clear why.
Re: 256 voice polyphony?? not as far as i can tell!
1) because that's the name they gave to the "supernatural" acoustic instruments' emulations.madAhorn wrote:why would you choose to play chords with a SOLO patch?
2) because if I'm soling on a trumpet and suddendly want a second trumpet to counterpoint it - which is not a RARE occurrence in jazz orchestration -, I'd like my synth to "follow my playing style" (as advertised)
Re: 256 voice polyphony?? not as far as i can tell!
I would certainly faceplam if I ever saw a guy on stage trying to counterpoint his own trumpet solo on a synth.
Re: 256 voice polyphony?? not as far as i can tell!
According to the manual each Tone can be assigned to either the solo/percussion or the live set part. The only restriction is that the Organ can't be assigned to percussion/solo because of the effects.
For the rest it means that a piano, e. piano or strings Tone should be assignable to the Solo part. And those are all polyphonic instruments. So, the question about playing chords on a solo part is not really a valid question. Hence even a Marimba, Guitar or Vibraphone is played polyphonically and you should be able to use them on the Solo part.
That's in regard to madAhorn's question.
Let's not forget that we are talking about polyphony issues. I wouldn't have a problem with the fact, that the polyphony is somehow restricted on a per part basis. But it would be fair from Roland to explain it correctly instead of leaving us in the dark. That's unfortunately the way how they behaved before and they obviously don't change it to a better.
For the rest it means that a piano, e. piano or strings Tone should be assignable to the Solo part. And those are all polyphonic instruments. So, the question about playing chords on a solo part is not really a valid question. Hence even a Marimba, Guitar or Vibraphone is played polyphonically and you should be able to use them on the Solo part.
That's in regard to madAhorn's question.
Let's not forget that we are talking about polyphony issues. I wouldn't have a problem with the fact, that the polyphony is somehow restricted on a per part basis. But it would be fair from Roland to explain it correctly instead of leaving us in the dark. That's unfortunately the way how they behaved before and they obviously don't change it to a better.
Re: 256 voice polyphony?? not as far as i can tell!
>I would certainly faceplam if I ever saw a guy on stage trying to counterpoint his own trumpet solo on a synth
Well, it's the 21st century.
I don't want to get into this facepalming stuff, and I'm certainly not an expert on keyboards, but it seems to me some people in this thread may be posting about stuff, even snarking about stuff, without really thinking through the details.
On my arranger keyboard, I can choose if a sound is to be interpreted monophonically or polyphonically. The difference ISN'T just that one can play chords and one can't. A monophonic patch has portamento and glide and other articulation options that don't always make immediate sense in a chord situation.
I sadly don't own a Jupiter-80, but I strongly suspect a so-called "solo" patch means MORE than just the "farthest right" on the keyboard. Mono sounds, typically solo sounds, imply more options going from note-to-note than chord sounds. It doesn't surprise me that a solo patch would create odd sounds played as chords. If people believe their "style" requires playing chords with synth settings designed for monophonic playing, to my eyes such people are not being George Jetson, but more George of the Jungle.
But I wouldn't want to snark because I'm usually pretty out of touch, too.
Well, it's the 21st century.
I don't want to get into this facepalming stuff, and I'm certainly not an expert on keyboards, but it seems to me some people in this thread may be posting about stuff, even snarking about stuff, without really thinking through the details.
On my arranger keyboard, I can choose if a sound is to be interpreted monophonically or polyphonically. The difference ISN'T just that one can play chords and one can't. A monophonic patch has portamento and glide and other articulation options that don't always make immediate sense in a chord situation.
I sadly don't own a Jupiter-80, but I strongly suspect a so-called "solo" patch means MORE than just the "farthest right" on the keyboard. Mono sounds, typically solo sounds, imply more options going from note-to-note than chord sounds. It doesn't surprise me that a solo patch would create odd sounds played as chords. If people believe their "style" requires playing chords with synth settings designed for monophonic playing, to my eyes such people are not being George Jetson, but more George of the Jungle.
But I wouldn't want to snark because I'm usually pretty out of touch, too.
Re: 256 voice polyphony?? not as far as i can tell!
Isn't the real point here about the design of Roland supernatural scripting? (using that term loosely because we don't know exactly how supernatural has been implemented or programmed....but it "seems" very similar to scripting methods used by LASS and Hollywood Strings/Brass, etc).
We all get that you can take a Kazoo patch, and play 10 finger chords with it if that is what your creativity, inspiration, or necessity calls you to do. And if the JP80 had a boiler plate Kazoo patch, then you would be free to do it with no issues. But that is not what Roland "scripted" the SUPERNATURAL Kazoo patch to do. Rather...they scripted it to replicate the nuance of a Kazoo (which is TRADITIONALLY a solo instrument played in a legato style) to authentically express the traditional articulations and nuance of a Kazoo. In fact, the scripting has some level of 'awareness' of your performance gestures on the keyboard (how hard you press, how rapidly you press, how many notes you press, how you roll from one note to another in legato vs. jump from one note to another...etc...) so that it can intelligently interpret those gestures with the scripted articulations and expressions. Now.....you go and start playing multi-finger chords in real time....and what do you think the scripting is going to do with that!? Its going to try to interpret that as any other performance gestures, based on the underlying expectation that the point of the performance is to authentically express a Kazoo performance....which is a legato solo instrument! And its likely to get confused, over-loaded, or just generally sound like crap!
I don't discount anyones right to use their keyboard anyway they feel or are inspired to do. But perhaps Supernatural is not designed for such abstract performances. Perhaps the expectation is that if you are going to track your own harmonies of the same instrument...that you will create a registration which multiple layers or splits of discrete patches to accomplish this, rather than playing two lines in real time with one patch which is designed to interpret ONE line in real time!
We may be assuming this is Note Stealing....when in fact its nothing more than scripting misinterpretation. And when you think about it.....this makes a ton of sense, and is frankly much more likely than a polyphony issue on a synth with such high polyphony. This supernatural is new technology, so we have not much cross reference to compare to. What we may have discovered here is an aspect of Supernatural tech which is new to us and we just don't fully understand. Lets not jump to the wrong conclusion...IMHO.
We all get that you can take a Kazoo patch, and play 10 finger chords with it if that is what your creativity, inspiration, or necessity calls you to do. And if the JP80 had a boiler plate Kazoo patch, then you would be free to do it with no issues. But that is not what Roland "scripted" the SUPERNATURAL Kazoo patch to do. Rather...they scripted it to replicate the nuance of a Kazoo (which is TRADITIONALLY a solo instrument played in a legato style) to authentically express the traditional articulations and nuance of a Kazoo. In fact, the scripting has some level of 'awareness' of your performance gestures on the keyboard (how hard you press, how rapidly you press, how many notes you press, how you roll from one note to another in legato vs. jump from one note to another...etc...) so that it can intelligently interpret those gestures with the scripted articulations and expressions. Now.....you go and start playing multi-finger chords in real time....and what do you think the scripting is going to do with that!? Its going to try to interpret that as any other performance gestures, based on the underlying expectation that the point of the performance is to authentically express a Kazoo performance....which is a legato solo instrument! And its likely to get confused, over-loaded, or just generally sound like crap!
I don't discount anyones right to use their keyboard anyway they feel or are inspired to do. But perhaps Supernatural is not designed for such abstract performances. Perhaps the expectation is that if you are going to track your own harmonies of the same instrument...that you will create a registration which multiple layers or splits of discrete patches to accomplish this, rather than playing two lines in real time with one patch which is designed to interpret ONE line in real time!
We may be assuming this is Note Stealing....when in fact its nothing more than scripting misinterpretation. And when you think about it.....this makes a ton of sense, and is frankly much more likely than a polyphony issue on a synth with such high polyphony. This supernatural is new technology, so we have not much cross reference to compare to. What we may have discovered here is an aspect of Supernatural tech which is new to us and we just don't fully understand. Lets not jump to the wrong conclusion...IMHO.
-
doctorrock
- Posts: 22
- Joined: 13:29, 17 September 2011
Re: 256 voice polyphony?? not as far as i can tell!
ron,
what you say in your last post about scripting makes perfect sense and undoubtedly what people are sometimes interpreting as note stealing is in actual fact roland's scripting algorithms just not being able to keep up with what they would be programmed to think of as "inappropriate" playing, e.g 10 note chords on a kazoo!
i am not about to debate the merits of roland putting these, what some people on here are referring to as "solo," sounds in the middle of factory patches (registrations and live sets) without any warning in the manual about the resultant mess when playing chords, fast passages etc. but, maybe that wasn't the wisest move on their part....
however, there is undoubtedly a polyphony issue too, which you keep denying. i'm 99.999% sure that i've proved that there is a 64 voice per layer limit, having run my (thorough!) test on 3 different machines and come up with the same results every time. if anyone doubts this just go back to the beginning of this thread, read what i did and try it for yourselves.
what you say in your last post about scripting makes perfect sense and undoubtedly what people are sometimes interpreting as note stealing is in actual fact roland's scripting algorithms just not being able to keep up with what they would be programmed to think of as "inappropriate" playing, e.g 10 note chords on a kazoo!
i am not about to debate the merits of roland putting these, what some people on here are referring to as "solo," sounds in the middle of factory patches (registrations and live sets) without any warning in the manual about the resultant mess when playing chords, fast passages etc. but, maybe that wasn't the wisest move on their part....
however, there is undoubtedly a polyphony issue too, which you keep denying. i'm 99.999% sure that i've proved that there is a 64 voice per layer limit, having run my (thorough!) test on 3 different machines and come up with the same results every time. if anyone doubts this just go back to the beginning of this thread, read what i did and try it for yourselves.
Re: 256 voice polyphony?? not as far as i can tell!
doctorrok,
I appreciate your post. And I appreciate your position on this matter. I am not here to disagree with you or make you wrong or challenge your findings....I am just sincerely struggling to verify what you are saying on this topic.
Out of respect for your most recent post, I (reluctantly, because its 5 pages of sometimes extraneous material) re-read (more like scanned, so please don't jump on me if I missed a particular detail) the first few pages of this thread. What I gleaned is this....
1. I theorize that the "lag" you sometimes experienced may have more to do with scripting than with anything else. But who knows. Lag is lag....could be anything from your individual machine, to a partially corrupt OS, to a power issue at your location, to heat, to processor overload, or the Roland Jupiter 80 being inherently slow. But, I have not once in my experiences with the JP80 experienced lag in any form.
2. I am truly struggling to completely follow some of your posts which explain your tests and findings. I mean no disrespect, but some of the language and detail gets confusing, and I am not able to 100% follow your reasoning or results. I know you were tired when writing a few of those posts (cause you said so)...but if its a scientific determination that we can all agree upon that we seek....I am just struggling to *completely* grasp the testing and the findings.
3. I find it particularly hard to comprehend or reproduce a test which plays over 64 audible voices, and listens for 1 voice being stolen. Audibly, its impossible for me to make a qualified determination if I hear a voice being stolen or under what conditions it happened.
4. If I were to try to verify your results, I would do a few things differently....firstly, I might use different PCMs, rather than the same PCM, for the voices, for the first notes played, so they stand out if one gets stolen. But more aptly, I would use a midi sequencer to trigger the notes, one at a time, all from a single tone with a single partial on an init synth patch. I would make a sequence that consecutively plays >64 notes, each one consecutively beginning on a quarter note bar marker, adding new notes one at a time, holding each note across the timeline, and with a velocity of 100 on the first note played, and with a velocity of 1 on each subsequent note. As the sequence adds notes across the timeline, I would listen for the first (and only audible) note to drop. Not knowing much about Roland voice allocation or other aspects of the workings inside the JP80, its a more scientific test than loading up 12 voice notes and manually playing 10 finger chords. I don't have time to do this today...but maybe you do? It surely would be interesting.
Again, if I missed some points in the content of this 5 page thread, please forgive me....and enlighten me (or just point me to it). RIght now...*I* am tired after a long Monday at work. :-)
thanks for the discussion....
Ron
I appreciate your post. And I appreciate your position on this matter. I am not here to disagree with you or make you wrong or challenge your findings....I am just sincerely struggling to verify what you are saying on this topic.
Out of respect for your most recent post, I (reluctantly, because its 5 pages of sometimes extraneous material) re-read (more like scanned, so please don't jump on me if I missed a particular detail) the first few pages of this thread. What I gleaned is this....
1. I theorize that the "lag" you sometimes experienced may have more to do with scripting than with anything else. But who knows. Lag is lag....could be anything from your individual machine, to a partially corrupt OS, to a power issue at your location, to heat, to processor overload, or the Roland Jupiter 80 being inherently slow. But, I have not once in my experiences with the JP80 experienced lag in any form.
2. I am truly struggling to completely follow some of your posts which explain your tests and findings. I mean no disrespect, but some of the language and detail gets confusing, and I am not able to 100% follow your reasoning or results. I know you were tired when writing a few of those posts (cause you said so)...but if its a scientific determination that we can all agree upon that we seek....I am just struggling to *completely* grasp the testing and the findings.
3. I find it particularly hard to comprehend or reproduce a test which plays over 64 audible voices, and listens for 1 voice being stolen. Audibly, its impossible for me to make a qualified determination if I hear a voice being stolen or under what conditions it happened.
4. If I were to try to verify your results, I would do a few things differently....firstly, I might use different PCMs, rather than the same PCM, for the voices, for the first notes played, so they stand out if one gets stolen. But more aptly, I would use a midi sequencer to trigger the notes, one at a time, all from a single tone with a single partial on an init synth patch. I would make a sequence that consecutively plays >64 notes, each one consecutively beginning on a quarter note bar marker, adding new notes one at a time, holding each note across the timeline, and with a velocity of 100 on the first note played, and with a velocity of 1 on each subsequent note. As the sequence adds notes across the timeline, I would listen for the first (and only audible) note to drop. Not knowing much about Roland voice allocation or other aspects of the workings inside the JP80, its a more scientific test than loading up 12 voice notes and manually playing 10 finger chords. I don't have time to do this today...but maybe you do? It surely would be interesting.
Again, if I missed some points in the content of this 5 page thread, please forgive me....and enlighten me (or just point me to it). RIght now...*I* am tired after a long Monday at work. :-)
thanks for the discussion....
Ron
Re: 256 voice polyphony?? not as far as i can tell!
In my opinion doctorrock's test is an absolutely valid approach of testing polyphony/voicestealing.
I'd be surprised if this is a faulty unit/OS issue - would be very helpful if someone who's not experiencing this issue did the same tests and posted an audio file of the result.
-Freddie
I'd be surprised if this is a faulty unit/OS issue - would be very helpful if someone who's not experiencing this issue did the same tests and posted an audio file of the result.
-Freddie
Re: 256 voice polyphony?? not as far as i can tell!
A couple of thoughts regarding the lag. Maybe it appears on certain kind of tones or live sets, mainly the brass patches. This is what also happens on the arx brass card.
Since the processor has to determine which keys you play and how to distribute the different solo instruments to different parts, there is a lag/latency between pressing the keys and the sound appearing. The jp80 certainly uses a faster processor but under high load it could be that those lags become more apparent.
Since the processor has to determine which keys you play and how to distribute the different solo instruments to different parts, there is a lag/latency between pressing the keys and the sound appearing. The jp80 certainly uses a faster processor but under high load it could be that those lags become more apparent.
-
Bruce Lychee
- Posts: 168
- Joined: 21:56, 24 June 2011
Re: 256 voice polyphony?? not as far as i can tell!
I was able to replicate the note stealing test on A7, but not Marc Strings.mojkarma wrote:A couple of thoughts regarding the lag. Maybe it appears on certain kind of tones or live sets, mainly the brass patches. This is what also happens on the arx brass card.
Since the processor has to determine which keys you play and how to distribute the different solo instruments to different parts, there is a lag/latency between pressing the keys and the sound appearing. The jp80 certainly uses a faster processor but under high load it could be that those lags become more apparent.
I definitely can't reproduce the lag.
-
doctorrock
- Posts: 22
- Joined: 13:29, 17 September 2011
Re: 256 voice polyphony?? not as far as i can tell!
of course ron! nobody - not even mozart! - would be able to hear one voice stolen from 64 if they were all audible! what you have to do - and i did explain this earlier - is turn the volume down (but leave switched on!) of all parts (or layers if you're testing a registration) except for one, consequently giving you a whole lot less notes to listen to and therefore making the voice stealing MUCH easier to spot!RonF wrote:3. I find it particularly hard to comprehend or reproduce a test which plays over 64 audible voices, and listens for 1 voice being stolen. Audibly, its impossible for me to make a qualified determination if I hear a voice being stolen or under what conditions it happened.
talking of mozart, i was playing through a piano concerto of his this morning using the INIT registration that roland has put into all the free registration slots consisting of 1514 grand piano and 1805 stringssect1 and the strings were cutting out all over the place! this is a good example of what i feel is the clumsy, thoughtless programming by roland which i sometimes come across that i hinted at in my last post. to me, it makes sense that any string patch that is layered as an accompaniment to a piano (especially one in the INIT registration) should be able to sustain for a reasonable amount of time - especially bearing in mind the frequency which the sustain pedal is used when playing piano parts - yet due to the polyphony limit, 1805 stringsect1 is hampered in it's ability to do this, it being a live set with a full quotient of 4 active layers. not to mention that 2 of the layers are solo violin and solo cello with the attendant scripting that may well get thrown into confusion when these "solo" instruments are being played like a piano. to my mind, 1805 stringsect1 (and therefore the INIT registration) would be better served by these 2 voices being switched off by default - like the solo part in the INIT registration, 0042 violin, which, when switched on works astoundingly well in being able to somehow pick out a melody from a complicated piano part being played, especially when stringsect1 is switched off leaving solo violin and piano. give it a try!
Re: 256 voice polyphony?? not as far as i can tell!
Ok...I will give MY test suggestion a try as soon as I can get the time to set it up to see if your contentions bear out.doctorrock wrote:
of course ron! nobody - not even mozart! - would be able to hear one voice stolen from 64 if they were all audible! what you have to do - and i did explain this earlier - is turn the volume down (but leave switched on!) of all parts (or layers if you're testing a registration) except for one, consequently giving you a whole lot less notes to listen to and therefore making the voice stealing MUCH easier to spot!
An assumption, spoken like its fact.doctorrock wrote:yet due to the polyphony limit,
That was hard to follow. But I "think" I got ya!doctorrock wrote:not to mention that 2 of the layers are solo violin and solo cello with the attendant scripting that may well get thrown into confusion when these "solo" instruments are being played like a piano. to my mind, 1805 stringsect1 (and therefore the INIT registration) would be better served by these 2 voices being switched off by default - like the solo part in the INIT registration, 0042 violin, which, when switched on works astoundingly well in being able to somehow pick out a melody from a complicated piano part being played, especially when stringsect1 is switched off leaving solo violin and piano. give it a try!
Re: 256 voice polyphony?? not as far as i can tell!
Voice reserve parameter is telling me there are 64 voices available to my live set or solo/perc layer. It doesn't take much to completely tax a 4 layer live set using Jupiter as an external instrument in Logic.
Half glass full opinion incoming: Maybe its not necessary to program a 4 layer string patch when I need to play 8 note chords. Most of that phatness gets lost in the mix with other instruments. Get creative use 2 layers of the live set and one of the solo/perc layers to extend poly. Maybe put the reserve parameter to use so there are more voices available to your primary tone.
I would liked to see poly dynamically allocated but that just isn't the case with Jupiter 80.
Half glass full opinion incoming: Maybe its not necessary to program a 4 layer string patch when I need to play 8 note chords. Most of that phatness gets lost in the mix with other instruments. Get creative use 2 layers of the live set and one of the solo/perc layers to extend poly. Maybe put the reserve parameter to use so there are more voices available to your primary tone.
I would liked to see poly dynamically allocated but that just isn't the case with Jupiter 80.