What does "snappy" Really. Mean?

Anything modular synth related that is not format specific.

Moderators: Kent, luketeaford, Joe.

bryan
Common Wiggler
Posts: 105
Joined: Sat Sep 05, 2009 10:50 pm

Re: What does "snappy" Really. Mean?

Post by bryan » Tue Sep 29, 2020 11:51 pm

Ok gang, I did some research in the lab. Because I'm a scientist (this is true, actually, but not for sound, trust me nevertheless ;) :party: )

First, I apologize to Grandmother, she does indeed snap. Just less so than the MS-1. However, in the Grandmother sequencer you can't adjust gate time so to get the full Grandmother snap you have to play it on the keys yourself.

I compared the Behringer MS-1 (SH-101 clone) and Moog Grandmother recorded into Wavelab. Here's what I found...

MS-1
attack: 1 ms
decay: 1.5 ms

Grandmother
attack: 1.5 ms
decay: 2 ms

Both have only one simple EG so this makes it easy. Also, I kept the filter open wide to just use the VCA. For the times above, I set attack and decay to 0 and used a high frequency wave from the filter in self oscillation.

I checked different settings for attack and decay but usable range on the low end of the envelope faders doesn't allow for fine scale exploration of this stuff. I found that attack and decay below 6 ms were capable of snap.

My interpretation of snap is pretty simple: a sharp transient. So it is the attack + decay that set the snap, and the sustain level and release time that come after can either highlight the snap or work against it. Thus a nice setting for snappiness is attack and decay at 0, sustain at half, and release at around a quarter second or so. At least, that's what worked for me and I'm interested in what snaps for you all. Here's what those settings look like...

MS-1 and GM.jpg
MS-1 first, Grandmother second. A 0, D 0, S about half, R around 1/4 sec. An interesting thing is that the sustain lasts ~20 ms for the short gate time for both, but the Grandmother's sustain portion of the envelope is around 90 ms when sequenced using the same envelope settings.


Sound examples...

https://soundcloud.com/kabbak/ms-1-4x-g ... xzkR6RzTf5

MS-1 4x, Grandmother 4x, then 4x alternating MS-1/Grandmother, using the settings noted (and screencapped) above.

https://soundcloud.com/kabbak/ms-1-env- ... 20Q6pxj7F4

MS-1 taking things literally. Attack at 0, then lowering the decay.

EDIT: soundcloud files now on private, so I replaced the embeds with links
Last edited by bryan on Tue Oct 13, 2020 1:09 pm, edited 1 time in total.

bryan
Common Wiggler
Posts: 105
Joined: Sat Sep 05, 2009 10:50 pm

Re: What does "snappy" Really. Mean?

Post by bryan » Wed Sep 30, 2020 11:48 am

PS I used triangle waves for the comparison recordings above. For the MS-1, its filter opens all the way beyond 20 kHz. For the Grandmother, its filter does not open all the way so I patched the oscillator directly into the VCA. The point is to have the triangle waves being affected by the VCA only, and not the VCF.

PPS For the MS-1 "snap" recording, I used a triangle wave with some filtering and resonance up high to cut the lower frequencies.

User avatar
sir stony
Ultra Wiggler
Posts: 753
Joined: Mon Feb 25, 2019 6:09 pm
Location: Bavaria

Re: What does "snappy" Really. Mean?

Post by sir stony » Wed Sep 30, 2020 1:51 pm

Great posting, bryan!
Basically the scientific backing for what I said... 8-)

User avatar
SkyWriter
Super Deluxe Wiggler
Posts: 1719
Joined: Fri May 16, 2014 6:18 pm
Location: Absolute Elsewhere
Contact:

Re: What does "snappy" Really. Mean?

Post by SkyWriter » Wed Sep 30, 2020 4:11 pm

bryan wrote:
Tue Sep 29, 2020 11:51 pm
Ok gang, I did some research in the lab. Because I'm a scientist (this is true, actually, but not for sound, trust me nevertheless ;)
Why I invited you. Lol!
MS-1 taking things literally. Attack at 0, then lowering the decay.
Exactly, I got down to 10ms Attack, 10ms Decay, anechoic evironment (headphones), no filter, sine wave (didn't want those pesky triangle high freqs macking anything).

So:
Snappy attack perceptions are in the in the 1-10ms* range, and low pitches the envelope is warping the first derivative of the onset of the first pressure wave. Higher pitches, the attack encompasses more and more full cycles, yet still evidences a similar perception up and down the scale.

Therefore, it's not a single event, nor is it merely an aggregation of events. It isn't isolated to specific pitches, although at very low frequencies it becomes difficult to perceieve exactly wtf is going on as it all sounds less and less like sound and more like somethings broke. :-)

Nor, is it strictly environmental, although environmental effects can eloborate the basic perceptions, they are not necessary, therefore something more fundamental is at play, and evidences a similar sensation.


*-really the 'click-perception' is the lower bound here of the 'snappiness-perception'. And the upper bound is the 'fwoop-perception'. Those occur in the neighborhood of 1ms and 10ms. The exact number are what I hope to get out of this, as well as more rigorous labels befitting a white paper. Lol!
I'm color blind.

bryan
Common Wiggler
Posts: 105
Joined: Sat Sep 05, 2009 10:50 pm

Re: What does "snappy" Really. Mean?

Post by bryan » Thu Oct 01, 2020 12:15 pm

sir stony wrote:
Wed Sep 30, 2020 1:51 pm
Great posting, bryan!
Basically the scientific backing for what I said... 8-)
Hey thank you sir! I also agree with what you said before about backing down sustain levels to get the "snap" really happening.

bryan
Common Wiggler
Posts: 105
Joined: Sat Sep 05, 2009 10:50 pm

Re: What does "snappy" Really. Mean?

Post by bryan » Thu Oct 01, 2020 12:44 pm

SkyWriter wrote:
Wed Sep 30, 2020 4:11 pm
*-really the 'click-perception' is the lower bound here of the 'snappiness-perception'. And the upper bound is the 'fwoop-perception'. Those occur in the neighborhood of 1ms and 10ms. The exact number are what I hope to get out of this, as well as more rigorous labels befitting a white paper. Lol!
Let's write up that paper! ;)

In my case, I couldn't narrow it down further than ~6ms attack/decay being the threshold for snap because there isn't enough resolution in the 1-6ms range in the synths I used. (but this reminds me of some further Maffez MS-1 mods I was thinking of doing)

It was a fun exercise to check out the limits of the MS-1 and Grandmother, though. Hey, now I know their fastest attack and decay times. Another anecdote... I was looking at the original Roland SH-101 manual and was surprised to see that they listed the envelope times! Considering that the SH-101 is known for snap, these are decent benchmarks for fast envelopes..

Verbatim from the Roland SH-101 manual:
Attack Time knob (1.5ms~4s)
Decay Time knob (2ms~10s)
Sustain Level knob (0~100%)
Release Time knob (2ms~10s)

And, again in comparison, what I found on the Behringer MS-1:
Fastest attack time 1ms
Fastest decay time 1.5ms

So in the end, the Moog Grandmother actually matched the SH-101 fastest envelope speed (see previous post for more on the Grandmother times). Well, at least if we can assume that Roland reported the envelope SH-101 accurately (this is likely, I would think). And the MS-1 is even faster than the OG SH-101... interesting.

BTW, I did the same envelope tests on the MS-1 with its noise source. Interesting that I couldn't perceive the snap so well with the noise. Like, the exact ms range for "snap" was unclear with the noise instead of the triangle wave. Just gonna guess that a transient is best heard when it is on a clean continuous wave (sine, for example) because the break in continuity (aka snap) is clearly perceived.

Ok, now maybe the best way to proceed is with a DAW that can put exacting fades, in the sub-10ms range, and use a digital sine wave. That would be a way to narrow snap perception down to the ms. Will have to check if the DAWs that are on-hand can do such tiny fades.

User avatar
SkyWriter
Super Deluxe Wiggler
Posts: 1719
Joined: Fri May 16, 2014 6:18 pm
Location: Absolute Elsewhere
Contact:

Re: What does "snappy" Really. Mean?

Post by SkyWriter » Thu Oct 01, 2020 1:25 pm

LOL!

Indeed, white noise wasn't as helpful as I thought either. I used a Mutable Instruments Tides as it presents continuous variable attach and decay profiles from a concave exponential to a convex exponential - not great descriptions but you get the idea; sort a left leaning sharks tooth to a right leaning one :-) with ms control. It's spiffy!

PS: Pots are coming today, so hopefully positive results later today for 911 tests. That will make it a better tool for that than it currently is.

@bryan, note the starting phase of the waveform wrt to attack time. If they're not synched, you'll get spurious spikes and the glitchies - those don't count :-) only repeatable events. Ha!
I'm color blind.

User avatar
mome rath
a dubious cube of margarine
Posts: 1777
Joined: Mon Jun 08, 2009 8:23 pm

Re: What does "snappy" Really. Mean?

Post by mome rath » Thu Oct 01, 2020 6:49 pm

bryan wrote:
Thu Oct 01, 2020 12:44 pm
SkyWriter wrote:
Wed Sep 30, 2020 4:11 pm
*-really the 'click-perception' is the lower bound here of the 'snappiness-perception'. And the upper bound is the 'fwoop-perception'. Those occur in the neighborhood of 1ms and 10ms. The exact number are what I hope to get out of this, as well as more rigorous labels befitting a white paper. Lol!
Let's write up that paper! ;)
one of my students picked a paper about skrillex disrupting mosquito behavior for the infectious disease journal club

i spent half the seminar popping off about audio production and recording techniques
:mad:
Image
soundcloud, dear friends

good trader cred
instagraM
guest wrote:
Fri Jan 03, 2020 2:45 pm
basically, as the dude put it "no walter, youre not wrong, youre just an asshole".
zoogoo wrote: your stupid

bryan
Common Wiggler
Posts: 105
Joined: Sat Sep 05, 2009 10:50 pm

Re: What does "snappy" Really. Mean?

Post by bryan » Fri Oct 02, 2020 8:26 pm

Here's a simple test. Reserving my judgment to let you all weigh in...

As I hoped, Wavelab was fine for precise digital envelopes, with the sinusoid fade in and exponential fade outs appearing to most closely match the envelopes from the synths I tried before.

I used a high frequency sine and at 0s you hear the unfaded sound. Then, starting at 2s and then every 2s after that it's the same tone with equal length fade in and fade out. So, starting at 2s there is a 20 ms tone with 10ms fade in/fade out, then at 4s there is a 18ms tone with 1ms fade in/fade out, and so on, until the last one at the 20s mark, a 2ms tone with 1ms fade in/fade out.

Where do you feel the snap? :foul:

https://soundcloud.com/kabbak/snap-judg ... 3LJmI1pTwJ

time location - duration of fades (tone length is double the fade time)
2s 10ms
4s 9ms
6s 8ms
8s 7ms
10s 6ms
12s 5ms
14s 4ms
16s 3ms
18s 2ms
20s 1ms

EDIT: soundcloud file is now on private, so embedded player now replaced with a link
Last edited by bryan on Tue Oct 13, 2020 1:10 pm, edited 1 time in total.

bryan
Common Wiggler
Posts: 105
Joined: Sat Sep 05, 2009 10:50 pm

Re: What does "snappy" Really. Mean?

Post by bryan » Tue Oct 06, 2020 1:10 pm

Just bringing this up again.. to conclude, for me, at 3ms envelope response (attack/decay) it sounds snappy. 4ms, no.

So it makes sense to me if 2.5 ms is a max threshhold of attack time that has been used in synths.

That said, it's nice to have even more snap than 2.5 ms, especially for making drum sounds or dialing in an 808-style kick with click on the attack.

User avatar
SkyWriter
Super Deluxe Wiggler
Posts: 1719
Joined: Fri May 16, 2014 6:18 pm
Location: Absolute Elsewhere
Contact:

Re: What does "snappy" Really. Mean?

Post by SkyWriter » Tue Oct 06, 2020 3:02 pm

Yeah, it's hard you really have to go over and over it. Once you do, it's not that easy to describe what you're 'hearing'.

There's distinct progression to these 'attack only' reproductions.
"Clicky" < 1.0 ms
"Snappy" < 2.5 ms
"'Some other third thing thing' of noise" < 4.0 ms
"Fwoopy" from there on out :-)

I don't like the clicky sound as part of an envelope. It's too close to 'The Bad' noise to my ear; digital noise.
Snappy is pleasant, as is some other third thing of noise I hesitate to call "Pop". Then there's our old friend "Fwoopy" - He's so cute!

Personally, I had a hard time hearing it on my iPad or the headphones running off it. Probably a shitload of compression somewhere along the line either in the cloud or my iPad. I take it right off the VCA output. Dynamics are much different.

Again, once you practice your ear on a specific tone, it's different animal than selecting a historical example - anyone's history is colored by context - if I may go so far.
I'm color blind.

User avatar
KSS
Super Deluxe Wiggler
Posts: 7211
Joined: Mon Jan 25, 2016 7:28 am

Re: What does "snappy" Really. Mean?

Post by KSS » Tue Oct 06, 2020 5:23 pm

SkyWriter wrote:
Tue Oct 06, 2020 3:02 pm
I don't like the clicky sound as part of an envelope. It's too close to 'The Bad' noise to my ear; digital noise.
Generations of Hammond players might disagree. ;)

Though they might agree that the digital representations *of* their click *is* "The Bad noise"
Again, once you practice your ear on a specific tone, it's different animal than selecting a historical example - anyone's history is colored by context - if I may go so far.
Yes. See just above.

-------------------------

Everybody still seems to be focused only on the attack time? What about *after* that initial peak is achieved?
Is "snappy" really *only* the initial transient?
Last edited by KSS on Tue Oct 06, 2020 5:49 pm, edited 1 time in total.

User avatar
SkyWriter
Super Deluxe Wiggler
Posts: 1719
Joined: Fri May 16, 2014 6:18 pm
Location: Absolute Elsewhere
Contact:

Re: What does "snappy" Really. Mean?

Post by SkyWriter » Tue Oct 06, 2020 5:34 pm

KSS wrote:
Tue Oct 06, 2020 5:23 pm
Generations of Hammond players might disagree. ;)
Oh no, I had a B3 and a leslie, that's a different click. This is an like uncorrectable data symbol in an otherwise perfect recording kind of click.

There's certainly a lot of territory to cover after the beginning. Then that's what beginnings are all about, yeah? :-)
I'm color blind.

DJ Tap Water
Common Wiggler
Posts: 242
Joined: Sun Dec 30, 2018 9:35 am
Location: London
Contact:

Re: What does "snappy" Really. Mean?

Post by DJ Tap Water » Tue May 25, 2021 1:49 am

Snappy is onomatopoeiaic

User avatar
KSS
Super Deluxe Wiggler
Posts: 7211
Joined: Mon Jan 25, 2016 7:28 am

Re: What does "snappy" Really. Mean?

Post by KSS » Tue May 25, 2021 5:27 am

^ Says the one who taps water! ;)

Post Reply

Return to “Modular Synth General Discussion”