Hello:

The parameters in voice edit, from what I’ve seen so far, are midi units.
While these display units are useful in certain situations, some are better suited for me as:
m.sec/sec.
dB
Hz

Has this been brought up before as important to many and has Yamaha commented?

Some things that would be useful:
Envelope times… does anyone know what the Min/Max times are?
Hertz values for filters
dB for levels?

Especially interested in Env max time per segment as I do not have Motif, these are pre-sales questions

Regarding Hz for filters, I don’t think it’s advisable because the filter cutoff settings may just work for a specific note or, at the best, a restricted number of notes. This is because every note has a different fundamental frequency (1st harmonic) and if you set the cutoff frequency to a specific value in Hz, it’s then chained with a specific note. This is an important reason of using MIDI (relative) values and you also get key tracking to alter the way the cutoff frequency will change as the note changes.

For the use of absolute seconds, there are also benefits of using relative values related with the tempo (for instance notes lengths) because you can get better effects that can follow your tempo, no matter which one is.

I think the same applies with dB. MIDI can be a better choice because it enable you to get “relative” values. The downside is that you don’t get “mathematical exact” information… Maybe because musicians are more concerned with the feeling that with the technical details…

BTW I’m engineer. Maybe the best approach would be to keep using the relative values and give details on the formula used to convert the physical units to the relative values…

As a synth programmer of nearly 35 years,
as far as I’m concerned, values in Hertz
and db are meaningless...because I dont
“listen with my eyes”. Move a filter knob, listen
to how it sounds. Change an amp EG attack time,
listen to what happens. Change an LFO rate
and depth, check out how it sounds.

Even if a filter was marked in Hz, you couldnt
determine what would happen by moving
it to a specific value. You still have to
change the cutoff and listen to how it
sounds , you cant predict the result ahead of
time. Threrefore, it makes no difference
whether it is marked in Hertz or just a
numerical value like “64”.

The original Minimoog and Prophet 5 synths
didnt have frequency and db markings on
the knobs. You just moved the knobs to
see what they would do.

You’re not running an acoustical physics
experiment...you’re making sounds to
work in a musical context.  Programming
isnt an engineering exercise...that’s
overthinking it.

Where appropriate parameter values are listed in actual scientific values, as in Hertz, (particularly when adjusting the actual Equalizer, or Filter) you see “parameter units” as you called them when you are editing an “OFFSET” parameter.

For example, the VOICE architecture of the XF can use up to 8 multi-samples (called Elements) simultaneously. Each of these Elements has its own Filter (this filter is set by precise frequency in Hz, bandwidth in Q, and gain in dB) but when you are adjusting the overall CUTOFF and RESONANCE for all 8 Filters, you don’t want to have to adjust 24 parameters - you simply want to make the sound brighter or darker - this is when you view a parameter like Filter Cutoff = +6 or Filter Resonance = -12 (what you call “parameter units” are to indicate adding to or subtracting from the current value of each of the 8 Filters. The values you see are OFFSETS to the actual filter settings. Each Element could have an entirely different filter type. So some of the parameter units are simply OFFSETTING, adding/subtracting units from the actual setting (you can always tell these offsets because they are +/- a whole number, and that number is not followed by “Hz” for Cycles per second, or a Q value like 0.7, or a Gain ratio measured in “dB” (deciBels).

So it depends on where in the XF architecture you are looking. And as Dave says, musicians rarely care about time in seconds or milliseconds, sure you could figure something out with a piece of paper or a calculator, but mostly your ear is king when it comes to times within envelopes. I have no idea how much time it takes for a guitar string’s initial attack to diminish, but I bet I can hear when its right. I usually make the point about music and clock time this way: We have a sense of musical time - such that if I told you the band went to the bridge at somewhere around 1 minute 45 seconds into the song, you might quit this band. But if I told you the bridge occurs at measure 49, you have a better sense of where that is… musically… particularly after I count of the tempo and tell it is in 4/4. Time to musicians is felt, not divided with a clock into seconds and milliseconds - that is non-intuitive (scientific, yes, musical? not here, it is not).

So where and when the scientific values can help you, musically, you will find them in the Motif XF, and where they are not really helpful, you will not. Also when working with the envelopes “velocity” can influence how quickly a stage happens so it is not a fixed set of time for an envelope. It is influenced by the note-on velocity. In the real world as an instrument is played with more vigor, not only does the loudness reationship of the harmonics change but things like a quick percussive attack cause a difference in the speed of the overall rest of the sound. So may things are in play in addition to just a fixed time for each segment. The faster a key is pressed could cause the attack to be more aburpt, the slower a key is press could cause the attack to slower. So time is only one variable in an envelope. Again your ears will be a better judge than the values in milliseconds.

Many of LFO’s that can be set to specific values, and can be synchronized to tempo by musical sub-division. Many of the Effect processor parameters are quite accurate in terms of time (delays, etc), but when they are TEMPO driven, you will see them in musical values: quarter note, dotted eighth note, 8th note triplet, etc., etc. Again when you want the actual time you can find an effect that let’s you set them with actual time values, when you want to sync it to tempo, the XF list these in musical sub-divisions… because it’s the right thing to do and is completely musical and logical. It depends on where you are looking. If you are asking did Yamaha think about this - I think it is fair to say, yes, a lot of thought went into it.

Hope that helps.

MarPabl - 25 November 2011 01:59 AM

Regarding Hz for filters, I don’t think it’s advisable because the filter cutoff settings may just work for a specific note or, at the best, a restricted number of notes. This is because every note has a different fundamental frequency (1st harmonic) and if you set the cutoff frequency to a specific value in Hz, it’s then chained with a specific note. This is an important reason of using MIDI (relative) values and you also get key tracking to alter the way the cutoff frequency will change as the note changes.

For the use of absolute seconds, there are also benefits of using relative values related with the tempo (for instance notes lengths) because you can get better effects that can follow your tempo, no matter which one is.

I think the same applies with dB. MIDI can be a better choice because it enable you to get “relative” values. The downside is that you don’t get “mathematical exact” information… Maybe because musicians are more concerned with the feeling that with the technical details…

BTW I’m engineer. Maybe the best approach would be to keep using the relative values and give details on the formula used to convert the physical units to the relative values…

Thanks MarPabl:
I see uses for three parameter types:
The one you recommend, “midi” is a valid one. I’ve used it.
Kurzweil, where applicable, uses primarily “actual” units: Hz, sec/m.sec, dB and sometimes both. I’ve used this also.
Of course most synths offer in addition, where applicable, “tempo” units, etc.
Another type, “Note”, where applicable, is interesting because its the normal language (parameter unit) for conductors, instrumentalists, etc.
Midi and note are already related.
I can work with any “units” as provided.
Motif is not an issue regarding programming units.

I like using “actual” types a lot.
Anyway, just curious what motif users felt and whether it was ever discussed at length here.
Would be interesting to see various types available in one system. I’ve used a few of those types.

DavePolich - 25 November 2011 02:51 PM

As a synth programmer of nearly 35 years,
as far as I’m concerned, values in Hertz
and db are meaningless...because I dont
“listen with my eyes”. Move a filter knob, listen
to how it sounds. Change an amp EG attack time,
listen to what happens. Change an LFO rate
and depth, check out how it sounds.

Even if a filter was marked in Hz, you couldnt
determine what would happen by moving
it to a specific value. You still have to
change the cutoff and listen to how it
sounds , you cant predict the result ahead of
time. Threrefore, it makes no difference
whether it is marked in Hertz or just a
numerical value like “64”.

The original Minimoog and Prophet 5 synths
didnt have frequency and db markings on
the knobs. You just moved the knobs to
see what they would do.

You’re not running an acoustical physics
experiment...you’re making sounds to
work in a musical context.  Programming
isnt an engineering exercise...that’s
overthinking it.

I’d agree with most of what you say, additionally no matter what the markings/display, etc., might say, the mind determines the value.
However, I see use for for variation in metering and adjustments.
...

You’re not running an acoustical physics
experiment...you’re making sounds to
work in a musical context.  Programming
isnt an engineering exercise...that’s
overthinking it.

Actually Dave, sometimes I am- even with a workstation!
Sometimes I like to make acoustical physics
experiments, in a musical context. Sometimes I enjoy programming
from an engineering perspective.
I like applying a lot of thought to both the music and engineering, but I do agree that no matter what method or mindset, listening intently is primary.

Bad_Mister - 25 November 2011 03:28 PM

Where appropriate parameter values are listed in actual scientific values, as in Hertz, (particularly when adjusting the actual Equalizer, or Filter) you see “parameter units” as you called them when you are editing an “OFFSET” parameter.

For example, the VOICE architecture of the XF can use up to 8 multi-samples (called Elements) simultaneously. Each of these Elements has its own Filter (this filter is set by precise frequency in Hz, bandwidth in Q, and gain in dB) but when you are adjusting the overall CUTOFF and RESONANCE for all 8 Filters, you don’t want to have to adjust 24 parameters - you simply want to make the sound brighter or darker - this is when you view a parameter like Filter Cutoff = +6 or Filter Resonance = -12 (what you call “parameter units” are to indicate adding to or subtracting from the current value of each of the 8 Filters. The values you see are OFFSETS to the actual filter settings. Each Element could have an entirely different filter type. So some of the parameter units are simply OFFSETTING, adding/subtracting units from the actual setting (you can always tell these offsets because they are +/- a whole number, and that number is not followed by “Hz” for Cycles per second, or a Q value like 0.7, or a Gain ratio measured in “dB” (deciBels).

So it depends on where in the XF architecture you are looking. And as Dave says, musicians rarely care about time in seconds or milliseconds, sure you could figure something out with a piece of paper or a calculator, but mostly your ear is king when it comes to times within envelopes. I have no idea how much time it takes for a guitar string’s initial attack to diminish, but I bet I can hear when its right. I usually make the point about music and clock time this way: We have a sense of musical time - such that if I told you the band went to the bridge at somewhere around 1 minute 45 seconds into the song, you might quit this band. But if I told you the bridge occurs at measure 49, you have a better sense of where that is… musically… particularly after I count of the tempo and tell it is in 4/4. Time to musicians is felt, not divided with a clock into seconds and milliseconds - that is non-intuitive (scientific, yes, musical? not here, it is not).

So where and when the scientific values can help you, musically, you will find them in the Motif XF, and where they are not really helpful, you will not. Also when working with the envelopes “velocity” can influence how quickly a stage happens so it is not a fixed set of time for an envelope. It is influenced by the note-on velocity. In the real world as an instrument is played with more vigor, not only does the loudness reationship of the harmonics change but things like a quick percussive attack cause a difference in the speed of the overall rest of the sound. So may things are in play in addition to just a fixed time for each segment. The faster a key is pressed could cause the attack to be more aburpt, the slower a key is press could cause the attack to slower. So time is only one variable in an envelope. Again your ears will be a better judge than the values in milliseconds.

Many of LFO’s that can be set to specific values, and can be synchronized to tempo by musical sub-division. Many of the Effect processor parameters are quite accurate in terms of time (delays, etc), but when they are TEMPO driven, you will see them in musical values: quarter note, dotted eighth note, 8th note triplet, etc., etc. Again when you want the actual time you can find an effect that let’s you set them with actual time values, when you want to sync it to tempo, the XF list these in musical sub-divisions… because it’s the right thing to do and is completely musical and logical. It depends on where you are looking. If you are asking did Yamaha think about this - I think it is fair to say, yes, a lot of thought went into it.

Hope that helps.

Hi Bad_Mister:
Thank you for the excellent details.
Yes I can see, thanks to your time, where Yamaha chose to switch accordingly as per their arrangement, its certainly a workable system and would no complaints with synth engine or effects units.
As I’ve said in my earlier responses, I actually enjoy working with Kurzweil’s approach as well. Kurz and Yam both provide no hindrances to meaningful uses of parameter units.

Your reply was very informative regarding a quick and clear guide to how these things are handled. A nice and concise view.
An entire manual written by you would be excellent.