Syncing feeling ?

MegaHurtz · 22 Jul 2010

The midi mono is pretty tight, I dont know if the poly suffers any problems. My tip would be to use signals as much as you can on the pitch and volume. -A lot of green stuff slows everything down. Plus.. green data becomes part of the way windows would handle data, wich is not in a straight line. a lot of things getting stalled for other processes. And so the timing of green data is not something that can be relied on. Bleu or white data on the other hand is yust getting sent from your processor to your soundcard. This is becouse the "digital signal processor" of the soundcard has to be spot on in giving the correct frequency, and thus it won`t be stalled.(easily). Maybe it`s about time we started using the video card on the side.. ? Yust for graphics.. But untill then use as much blue and white as possible 😉

MichaelBenjamin · 22 Jul 2010

the midi2poly timing is accurate, linear envelopes often sound bland on triggering. and in general its easy to render a synth in sm/se or in c++ if you can code well, but the other half still consist of doing something nice with the numbers, so its the creative dsp that may be lacking. also with audio in general there is a huge amount of vaguely descriptive and mainly indirect emotional usage of words that often reaches into fantasyland in a way that seems to borderline mental retardation while it seems common sense that the quality of the plugin is reflected by its graphical appearance, so that only some experience can give you an authentic impression of the actual dsp quality.

Nu Audio Science · 22 Jul 2010

MichaelBenjamin wrote: the midi2poly timing is accurate, linear envelopes often sound bland on triggering. and in general its easy to render a synth in sm/se or in c++ if you can code well, but the other half still consist of doing something nice with the numbers, so its the creative dsp that may be lacking. also with audio in general there is a huge amount of vaguely descriptive and mainly indirect emotional usage of words that often reaches into fantasyland in a way that seems to borderline mental retardation while it seems common sense that the quality of the plugin is reflected by its graphical appearance, so that only some experience can give you an authentic impression of the actual dsp quality.

Couldnt have said it better myself 😀

NAS

2nvu · 22 Jul 2010

MichaelBenjamin wrote: the midi2poly timing is accurate, linear envelopes often sound bland on triggering. and in general its easy to render a synth in sm/se or in c++ if you can code well, but the other half still consist of doing something nice with the numbers, so its the creative dsp that may be lacking. also with audio in general there is a huge amount of vaguely descriptive and mainly indirect emotional usage of words that often reaches into fantasyland in a way that seems to borderline mental retardation while it seems common sense that the quality of the plugin is reflected by its graphical appearance, so that only some experience can give you an authentic impression of the actual dsp quality.

Hahaha' GeeezzzZZZ "Michael" , back to Uni for me , just to get a degree in Digital pyhcology , so i can understand your quote 😉 lol , hey i am pleased tho, that at least the midi2poly is tight, which concludes my thoery that its all down to getting the exact triggers working & corectly in SYNC: Psttt i do hope this quote was,nt directed @me > borderline mental retardation ? lol 😉

2nvu · 22 Jul 2010

MegaHurtz wrote: The midi mono is pretty tight, I dont know if the poly suffers any problems. My tip would be to use signals as much as you can on the pitch and volume. -A lot of green stuff slows everything down. Plus.. green data becomes part of the way windows would handle data, wich is not in a straight line. a lot of things getting stalled for other processes. And so the timing of green data is not something that can be relied on. Bleu or white data on the other hand is yust getting sent from your processor to your soundcard. This is becouse the "digital signal processor" of the soundcard has to be spot on in giving the correct frequency, and thus it won`t be stalled.(easily). Maybe it`s about time we started using the video card on the side.. ? Yust for graphics.. But untill then use as much blue and white as possible 😉

Cheer's MegaHurtZ . . . aaaahhh Blue & White huhhh ......So i must be nearly there with Resimulator, just add a bit of White & thats the Triggering system sorted? lol u guys realy do make me laugh somtimes, iam talking about triggering & syncing, & you guys are talking about pyhcology & coulours....please keep it up 'it gets funnier by the minuet 😉

Nu Audio Science · 22 Jul 2010

2nvu wrote: iam talking about triggering & syncing, & you guys are talking about pyhcology & coulours....please keep it up 'it gets funnier by the minuet 😉

I think you will find that what Micheal was saying is that they are both related IE if it looks nice it tends to sound nice (To a lot of people) IE maybe there is nothing wrong with the envelope triggering

NAS

MegaHurtz · 23 Jul 2010

lol, I wouldnt even dare to begin ripping on that one. but you get my point. Its as far from true as good looking cows make better milk. LISTEN!!

Most amateur plugins are crap becouse they feel like crap. But they do really sound like crap aswell..(nevermind them runnung like crap) Its yust a game for a few lucky labour of love kinda people. Try to recognise the philosophy, wich I assume 2nvu is quite good at.

Nu Audio Science · 23 Jul 2010

I have absolutely no idea what you mean Megahurtz

NAS

trogluddite · 23 Jul 2010

As MichaelBenjamin said, it's not always down to MIDI timing - it's always worth looking at the effect of the envelopes - there are even real analogue synths that players don't like because the envelopes are too slow.

Also, it's worth knowing that the length of the attack segment affects how your brain perceives the timing - if you use, say, a 50ms attack time, it can be worth moving the MIDI notes earlier by 20-30ms to make the timing sound better, because the PEAK of the attack will always be 50ms after the start of the note. This is one reason I don't like step sequencers very much - very few have the ability to shift the relative timing of the tracks to compensate for sounds with a slower attack. A good envelope generator with exponential slopes can also help to make the attack seem more 'snappy' because the initial build up is faster, then evens out as it nears the peak - so the average level of the attack phase is greater.

BTW, players of real instruments do this all the time, just ask any cello player - they have to start bowing the note 50-100ms early because it takes time for the vibrations of the strings to build up enough to form a clear note.

trogluddite · 23 Jul 2010

Here's some of the science behind my last post... http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.60.9590&rep=rep1&type=pdf ...sorry to be including more psychology - but all music has to be processed by the listeners brain; 'sample accurate timing' is meaningless unless you take into account how a human brain reads the 'groove'. A little understanding of the psychology of perception is just as important as all the techie stuff if you really want to nail a mix. As Joe Meek said; 'if it sounds right, it IS right' - regardless of what your sequencer is showing you on screen.

2nvu · 23 Jul 2010

trogluddite wrote: As MichaelBenjamin said, it's not always down to MIDI timing - it's always worth looking at the effect of the envelopes - there are even real analogue synths that players don't like because the envelopes are too slow.

Also, it's worth knowing that the length of the attack segment affects how your brain perceives the timing - if you use, say, a 50ms attack time, it can be worth moving the MIDI notes earlier by 20-30ms to make the timing sound better, because the PEAK of the attack will always be 50ms after the start of the note. This is one reason I don't like step sequencers very much - very few have the ability to shift the relative timing of the tracks to compensate for sounds with a slower attack. A good envelope generator with exponential slopes can also help to make the attack seem more 'snappy' because the initial build up is faster, then evens out as it nears the peak - so the average level of the attack phase is greater.

BTW, players of real instruments do this all the time, just ask any cello player - they have to start bowing the note 50-100ms early because it takes time for the vibrations of the strings to build up enough to form a clear note.

Hmmm Good point Dude , but iam talking "digi" not acustic/human eror inclusive, in reguards to the timing of course,& since Megahurtz droped a HINT 2 Me? about Triggering from Midi2Poly too / ASDR / Filter /Seq/ + + , i have noticed how the Synth's respond , with far more figure & feeling 😀 but again,it comes back to the "all Important Routing system " if you get that bit wrong 😕 it dose not matter how good lookin your Sythn is, it will sound like Sh*t,,, even if it as real good presets to boot,,,,So someone please point me to a tut? that will show a Newbie like me how to get my House in Order? i say this with Tounge in cheeck, 1 of you Geeks' should put out the easy guide to a Fool_Proof_Midi_Setup ? 😀 Warning inclusive? if you dont follow this guide, forget about it , or go back to "Second life" ? lol 😉 ps i love u all really 😇 hehehe

2nvu · 23 Jul 2010

trogluddite wrote: Here's some of the science behind my last post... http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.60.9590&rep=rep1&type=pdf ...sorry to be including more psychology - but all music has to be processed by the listeners brain; 'sample accurate timing' is meaningless unless you take into account how a human brain reads the 'groove'. A little understanding of the psychology of perception is just as important as all the techie stuff if you really want to nail a mix. As Joe Meek said; 'if it sounds right, it IS right' - regardless of what your sequencer is showing you on screen.

Very interesting Trog, indeed 😉

trogluddite · 23 Jul 2010

2nvu wrote: but iam talking "digi" not acustic/human eror inclusive

Me too - the point is that often people are looking for an error that isn't really there. The 'digi' timing can be 'perfect', but note on messages that are quantised to the digital grid will not always sound as if they are perfectly in time - the 'human error' is in the listener not the musician. In terms of SMs MIDI routing, if the MIDI input is connected directly to the MIIDI to Poly of the synth, then you can't really optimise any further - once generated, the poly signals will be sample accurate. If you manipulate the MIDI before it connects to the MIDI to Poly (e.g. Channel filtering etc.) then there may be potential for some slight timing errors. Any other 'digi' errors either come from the sequencer/VST-host, or are due to the DSP itself (especially the envelopes) and won't be reduced by changing the plug-ins internal MIDI structure.

2nvu · 23 Jul 2010

trogluddite wrote: 2nvu wrote: but iam talking "digi" not acustic/human eror inclusive

Me too - the point is that often people are looking for an error that isn't really there. The 'digi' timing can be 'perfect', but note on messages that are quantised to the digital grid will not always sound as if they are perfectly in time - the 'human error' is in the listener not the musician. In terms of SMs MIDI routing, if the MIDI input is connected directly to the MIIDI to Poly of the synth, then you can't really optimise any further - once generated, the poly signals will be sample accurate. If you manipulate the MIDI before it connects to the MIDI to Poly (e.g. Channel filtering etc.) then there may be potential for some slight timing errors. Any other 'digi' errors either come from the sequencer/VST-host, or are due to the DSP itself (especially the envelopes) and won't be reduced by changing the plug-ins internal MIDI structure.

Hmmm Trog, i am a tad lost here? i am the Musician/not the listener, & if my tools aint workin properly from the onset, it really does,nt matter how good of a player i am, & i do believe that midi data can be optimised futher? here's where i am at a loss? from Midi in - to Midi to Poly - & now i must trigger the Filters/ASDR/AmpEnvolopes + + +this list could go on? that for me is where the problem is,,,, if not resolved, will lead to a poorly runing Synth 😉 thats all iam asking, Try the latest upload of PimPOscRiffWurX,,,, it now feels like its really trying to be a Synth & not a Fony? lol Peace out Bruv.. 😀 O & thank you for your input Trog,,,, am tellin thi 😉

Acrobat · 24 Jul 2010

MichaelBenjamin wrote: so that only some experience can give you an authentic impression of the actual dsp quality.

I hate to say, but I think you are too much sarcastic with noobs at times, and sometimes being not so pro as you think you are, as "DSP quality" does NOT exist, actually, it is kind of an oxymoron to my ears. Quality is always referred to products, not simulations or computations. 😉

MichaelBenjamin · 24 Jul 2010

maybe you are right. best i can say is that it was not meant in a sarcastic way (i looked it up again what it means), or to judge and demoralize anyone. if that is your impression, maybe i can think about some better words to come up with. on a second thought i agree that the term dsp quality can be regarded as nonsense at first sight. maybe a better term to explain would be the dsp visionarity, where the numbers express rules generated by humans resulting in gazillions of anchorpoints for that inspiration thing that humanity needs to survive. or in another way. the next minimoog clone wont do that. thats stagnation. in this sense - why should quality, as a term to describe the utility factor for humanity only be attached to 4d matter? and tell me where i am "not so pro as i think", you know this is a stab from behind i can not tolerate without some hard evidence. i am curious. (on the other side, being not so pro as one thinks is probably natural, so i also could agree with this, but just for the fun - give me a link)

btw: just because i dont use the smilies does not mean i am raging arrogant or an emotionally devoid cynical while typing this (atleast i hope so). while i agree that it can be hard at the internet to always assume a friendly undertone, i decided some years ago not to use any emotional expressions with additional signs in text that i write. maybe i should put a signature "dont take the above serious at all" but i think it also works this way.

aliasant · 24 Jul 2010

MB.

Start using smileys ffs! 🙂

I used to run a gamers community and most common problem we, as admins, had to deal with was people not clearly understanding others caused by jokes or similar that didnt have smileys. Some became enemies for life but after a lot of admin hours of mediation from admins became best friend for life.... 😉 until the next poorly written joke slipped into a thread.

Its easy in the real world to see if another is joking or not. You look at the face, gestures etc.

We have to help eachother out here. Past weeks has been way to many silly arguments between good, intelligent members here at SM and many of them could have been avoided if we only undewrstood eachother correctly. In the northern hemisphere heat might be blamed as well as to little vacation time.

Use smileys. 😠 Help us understand 🙂

trogluddite · 24 Jul 2010

2nvu wrote: Midi to Poly - & now i must trigger the Filters/ASDR/AmpEnvolopes + +

Do you have some small example .osm of how you are doing this that you can post? or maybe audio example of what you hear as bad and good? - sounds like the kind of problem where words don't describe it too well. For most synths, connecting the MIDI to Poly 'Gate' output to the gate of the envelopes is the only kind of 'triggering' needed; Osci's, Filters etc. just start automatically when the MIDI to Poly creates a new voice. Because the MIDI to Poly outputs and all of the envelope/osc/filter are all streams (white), all of this happens on exactly the same sample - even if the schematic is very complex or untidy. So, I'm not sure there really is such a thing as a 'bad' way to arrange the 'triggering' of a new note - it will always happen on the same sample when the MIDI to Poly receives the Note On message. OTOH, there are many poor synths (digital and analogue) with 'sluggish' envelopes or filters which 'smear' the sound - they can make for the poor playing experience you describe, even though there is nothing technically wrong with the note triggering itself. Think you must be lucky to have the 'golden ears' that can hear these things - I am an old punk rocker, so bad timing and tuning sounds quite normal to me! 😉

2nvu · 24 Jul 2010

trogluddite wrote: 2nvu wrote: Midi to Poly - & now i must trigger the Filters/ASDR/AmpEnvolopes + +

Do you have some small example .osm of how you are doing this that you can post? or maybe audio example of what you hear as bad and good? - sounds like the kind of problem where words don't describe it too well. For most synths, connecting the MIDI to Poly 'Gate' output to the gate of the envelopes is the only kind of 'triggering' needed; Osci's, Filters etc. just start automatically when the MIDI to Poly creates a new voice. Because the MIDI to Poly outputs and all of the envelope/osc/filter are all streams (white), all of this happens on exactly the same sample - even if the schematic is very complex or untidy. So, I'm not sure there really is such a thing as a 'bad' way to arrange the 'triggering' of a new note - it will always happen on the same sample when the MIDI to Poly receives the Note On message. OTOH, there are many poor synths (digital and analogue) with 'sluggish' envelopes or filters which 'smear' the sound - they can make for the poor playing experience you describe, even though there is nothing technically wrong with the note triggering itself. Think you must be lucky to have the 'golden ears' that can hear these things - I am an old punk rocker, so bad timing and tuning sounds quite normal to me! 😉

Hahah thanks , i have posted in examples an Osm , PimpOscRiffwurX. triggering , hmm i thought that u could route out from midi2ploy say? velocity to Ampenv/ same Velo 2 filter, so as to refresh , sorry bud, this whole triggerin thing piss's me offffffffffffffff 😠 😠 😠 😠

trogluddite · 24 Jul 2010

Ah now I see, think we are each using the word 'triggering' to mean slightly different things. You are right, the velocity, frequency etc. outputs from the MIDI to Poly can be used for many things - I would have called that 'modulation' rather than 'triggering' - but even in SM itself, the word 'triggering' doesn't always seem to mean the same thing all the time. 😕 Found your .osm, so I'll see what I can find in there and get back to you soon...

2nvu · 24 Jul 2010

trogluddite wrote: Ah now I see, think we are each using the word 'triggering' to mean slightly different things. You are right, the velocity, frequency etc. outputs from the MIDI to Poly can be used for many things - I would have called that 'modulation' rather than 'triggering' - but even in SM itself, the word 'triggering' doesn't always seem to mean the same thing all the time. 😕 Found your .osm, so I'll see what I can find in there and get back to you soon...

Hey thank you Trog,,,, by the way thats the only word i can think of , Triggering,,,, Only cos its on the Mod Midi2Poly,,,but the more i experiment with the out_puts of the velo from M2P >>>> /filters/ADSR/ & other Mods, the better the repsponce, 🙂 & i think that will all change when building a different synth? yeh 😉 thanks for your time of lookin into the Mod, its nice to know people are willing to help.........Jamie............

trogluddite · 27 Jul 2010

Had a chance to look through your synth at last; not had time to delve in much detail, but here's my first impressions.

Sorry to sound like your Parent/Spouse/Policeman etc., but it could really do with a bit of a tidy up - all the links that criss-cross and go over/under modules make it tricky to see what's going on. You'll most likely see a lot of the problems yourself once the routing is neater. It also makes it easier to get help on the forum coz people can find their way around easier.

Most of your routiing problems seem to be similar... All the white poly connections in SM look the same, and SM treats them all the same way - but for a synth designer there are really two kinds of poly signal and it's the designer's job to know which is which. Type 1) The synth's audio - any poly link carrying an actual sound. On most modules the top input/outputs (usually just called 'in' and out') carry the sound itself. Type 2) Control Signals - these aren't sounds, but are the signals telling the synth what to do to the sound. For example, the output of an Knob, LFO or Envelope, or the signals from the MIDI to Poly. An input on a module that you might normally connect a knob to is expecting a control signal.

I see a few places in your synth where the two are being mixed in odd ways... 'To Amp Env' on the sequencer is plugged into the 'In' of the ADSR - the 'In' on the ADSR should take only the audio that you want to do an envelope on - 'To Amp Env' is a control signal, most likely for the envelope amount (multiply it with one of the VCA Levels inside the ADSR module). The Sequencer module takes an input from the filter - but the filter output is an audio signal, so not really something that you can directly control sequencer parameters with.

The way I would tackle it would be to arrange all of the modules/links that carry the audio together in a line, from Osc>>Filter>Amp Env>>Combiner>>Delay All the remaining modules then hang off the side of the line of audio modules, feeding them with control signals. Here's a link to the best Synth resource I've ever found - there's no programming tips or SM stuff, it just really explains what everything inside a synth is there for (Part 3 is a good explanation of control signals and modulation) http://www.soundonsound.com/sos/allsynthsecrets.htm

Of course, like all rules, there are plenty of grey areas and exceptions - so don't stop the experimenting, some very weird and wonderful things have been discovered that way too...

2nvu · 27 Jul 2010

trogluddite wrote: Had a chance to look through your synth at last; not had time to delve in much detail, but here's my first impressions.

Sorry to sound like your Parent/Spouse/Policeman etc., but it could really do with a bit of a tidy up - all the links that criss-cross and go over/under modules make it tricky to see what's going on. You'll most likely see a lot of the problems yourself once the routing is neater. It also makes it easier to get help on the forum coz people can find their way around easier.

Most of your routiing problems seem to be similar... All the white poly connections in SM look the same, and SM treats them all the same way - but for a synth designer there are really two kinds of poly signal and it's the designer's job to know which is which. Type 1) The synth's audio - any poly link carrying an actual sound. On most modules the top input/outputs (usually just called 'in' and out') carry the sound itself. Type 2) Control Signals - these aren't sounds, but are the signals telling the synth what to do to the sound. For example, the output of an Knob, LFO or Envelope, or the signals from the MIDI to Poly. An input on a module that you might normally connect a knob to is expecting a control signal.

I see a few places in your synth where the two are being mixed in odd ways... 'To Amp Env' on the sequencer is plugged into the 'In' of the ADSR - the 'In' on the ADSR should take only the audio that you want to do an envelope on - 'To Amp Env' is a control signal, most likely for the envelope amount (multiply it with one of the VCA Levels inside the ADSR module). The Sequencer module takes an input from the filter - but the filter output is an audio signal, so not really something that you can directly control sequencer parameters with.

The way I would tackle it would be to arrange all of the modules/links that carry the audio together in a line, from Osc>>Filter>Amp Env>>Combiner>>Delay All the remaining modules then hang off the side of the line of audio modules, feeding them with control signals. Here's a link to the best Synth resource I've ever found - there's no programming tips or SM stuff, it just really explains what everything inside a synth is there for (Part 3 is a good explanation of control signals and modulation) http://www.soundonsound.com/sos/allsynthsecrets.htm

Of course, like all rules, there are plenty of grey areas and exceptions - so don't stop the experimenting, some very weird and wonderful things have been discovered that way too...

Hahahha wooohhh , i have a lot to learn, but i must say thankyou very much for pointing out some very interesting fundimentals, & as it happens i have pulled out the sends from M2P/Velo/freq , that i had runing to ADSR & Filter, it did not make sence ,since they are already being trigger'd 😀 but your conformation is good news for my experimenting 😉 & big thanks for this article, looking forward to a good read 😀 Cheer's Jamie 😀 not bad for an first impression 😉

trogluddite · 27 Jul 2010

No Problem - it's definitely headed in an interesting direction; I've always had a soft spot for those kind of polyphonic phrase-arpeggiator type thingies.

2nvu · 27 Jul 2010

trogluddite wrote: No Problem - it's definitely headed in an interesting direction; I've always had a soft spot for those kind of polyphonic phrase-arpeggiator type thingies.

Hahah Same here Dude,,, Cool how they just do all the Work ? lol 😉 Psst ' Hope you have give my "Resimulator" a look? & a whirl , she's running like a dream,,,,but of course still needs some additional Tweaks, under the hood.... 😀 its only 2&half Megs desktop/vsti. . . plus there's a Bit of a tune Demo'ing the Synth 🙂 no worries if your busy,tho,,, J:)