Log inRegister an accountBrowse CSDbHelp & documentationFacts & StatisticsThe forumsAvailable RSS-feeds on CSDbSupport CSDb Commodore 64 Scene Database
 Welcome to our latest new user Rodrigo Yeowtch ! (Registered 2024-11-24) You are not logged in - nap
CSDb User Forums


Forums > C64 Composing > Sound Design hints & tips
2013-04-10 12:04
Hermit

Registered: May 2008
Posts: 208
Sound Design hints & tips

I looked the 'Composer' category at the Forum and couldn't find a topic specific to sound-design in general.

I think it would be fine to discuss our experiences of designing good sounding instruments for the SID...

This topic would not be about trackers and formats but what you experienced to sound good in waveform-arpeggio table or filter/pulse programs, etc...

topics like:
SID's hardware capabilities and software-based effects (slide/vibrato/pulse-sweep/filter-sweep/keyboard-tracking,etc..)
Spectrum analysis, awareness and know-how of playing with harmonics...
ADSR settings that are suitable for this & that (maybe HR, 1st frame waveform)
Characteristics of waveforms and mixed waveforms (on new/old SIDs)
Ring modulation and channel syncing...

How to make good drums/percussion..
How to make good and fat basses..
How to make fine light sounding arpeggios..
How to make good solo/lead sounds..
How to cleverly combine sounds (e.g. kick+hihat+bass) to save SID channels..

How to mimic real instruments..
(maybe speech with SID, like in 'Pimp My Commodore' or Agemixer's 'Freestyler' cover, whatever )...

Utilization of multi-framespeed....
etc. etc.

I have no time right now to explain these, but maybe some of you have some topic from above to tell stories & share experiences about, that we could discuss here...

Of course one can learn a lot from checking example tunes with SIDdump/SID-Hacker, etc., but it's also good to know the background knowledge & reasoning for the tricks...

 
... 15 posts hidden. Click here to view all posts....
 
2013-04-12 03:37
Soren

Registered: Dec 2001
Posts: 547
I recommend to simply experiment.... Try things out. Worked nicely for me. I never was very technical or theoretic about things. Nowadays I mostly have an idea of how to get a certain sound - the rest is simply to tweak what I have done.
Also remember that the way you use a sound makes a big difference. You can do lead sounds that alone don't sound of much, all the details done when using the sound are important.

SIDwave: Yes you do whatever is needed to get the right sound, even if it means phoning me late at night while I am asleep to ask be about how to do a certain sound... ;-D
2013-04-12 03:40
chatGPZ

Registered: Dec 2001
Posts: 11352
so post your phone# already! =)
2013-04-12 08:59
wacek

Registered: Nov 2007
Posts: 513
I am also in favour of experimentation and something which in the real world is called knob twiddling, and I guess in SID domain is playing with hex numbers ;) I have witnessed Praiser spew out some of his coolest basslines by randomly tapping on the keyboard, so, well... there must be a method to that madness ;)

Too much theory and scientific approach can give the result opposite from the desired one when things sound forced and you can hear it, this is not coding (although to the inexperienced beholder's eye it may look like it due to explicit hex content ;). Let's not forget making music is still art and should be approached as such. Nothing worse than a precalculated, crafted "oldschool" tune with just the perfectly right sounds/harmonies and stuff, just like another perfectly produced album from a famous rock band doing the same shit for the last 30 years.

And yes, we have a lot of extremely technical guys doing SIDs nowadays (I do sincerely respect the effort in achieving that), and they do make insanely crafted sounds, but that's just FORM, and if your CONTENT does not have soul, then you end up listening to the track and thinking to yourself "Yeah, this sounds amazing, but why am I tired in the middle of it and 20 minutes after it's finished can't remember a goddamn thing?". For me MUSIC should never be used as a PROOF OF CONCEPT, I hate that.
2013-04-12 12:36
Hermit

Registered: May 2008
Posts: 208
The title of the topic is 'hints and tips' not 'how to do it exactly'... I also like to experiment with sounds, most of the time I make all my sounds from scratch because of that, though I know my limits. But sometimes it might be better (especially for newbies) to know some common pitfalls before releasing a tune, then afterwards in discouraging comments...but just like in music composing or in life, rules are always there to break to make surprises...

As Flavioweb seems to like the idea I'll continue when I have some free time and post further tips how the leads & basses can be tweaked...the target is people who want to read it, the rest of scene can ignore the topic entirely...

Solo/lead sounds:
-----------------

Expressions: In the past I always wondered how solo lead sounds sounded so good in JT's, and other people's tunes. From the C64 Manual we know the plain waveforms but in BASIC there were no 'vibrato' and 'slide' 'dynamics' and 'sweep' mentioned IIRC. But IMO these are the things that make it sound good, 'pulse-sweep' in 1st place.
These are the things Jeff is talking about, to use a sound stylistically, using expressions. This is true for VST instruments too. A 'violin' is really bad if it's just put into the tune by notes from staff without slides/vibratos/legato/dynamics where needed...
These are the fields where analog synthesis and C64 is strong compared to MIDI-controlled VSTs on modern PCs. MIDI is able to make slides/legato, but in my opinion in a more restricted way than on the C64, I could simulate violin expressions more easily (e.g. Rakoczi Indulo) than on PC....
(It depends on the VST too of course, how it handles CC.)
Often in VST based C64 remixes lack the flexible/variable execution of the leads which we got used to on SID years ago.

solo-ADSRs: Most probably the instruments have short attack and no 'decay' by default. (But the 'decay' can be useful for some kind of solos.) I like to give solo instruments a long 'release' sometimes, so they has a 'reverb'-like feel to them. But that's not good for fast, staccato-oriented solos... Later in patterns the ADSR (especially 'attack' and 'release' can be modified to give variety)
The 'sustain' (i.e. volume of the notes) shouldn't be too harsh, it should be set to be in balance with the rest of the channels...that could make the tune sound more professional if some gives attention to volume ratios.
When using created sounds, you should play much attention to 'sustain' to enhance dynamics of the tune and un-machinarize it a bit, to sound more lively. (A good musician usually feels like by instinct, where to put stronger/weaker/ghost notes.) In music theory there's a rule, often in a hierarchical mode the first sounds of a beat are strong, between them in the middle the sounds are medium-strong, and if there are notes inbetween, they're usually weak. This can change of course (for example with syncopation, as in 'Conga Beat' or 'Garden Party' where bass is a bit before the downbeat..)...
Another useful field of the 'sustain' is to create 'delay echo'-like effect for the lead sounds on ONE channel. What you do there is you don't stop a note with the gate-off ('---') signal in the pattern, but you leave the note running and you decrease the sustain value much lower with pattern-FX... (Then later you can put gate-off also...)..
Backwards it doesn't work on the SID by default, 'sustain' can only be decreased without note-retriggering, cannot be simply increased (will kill the sound)...
To simulate even longer 'delay echo' impressions, you can make the notes change together with decreasing the 'sustain' value, usually to notes 1-2 rows before, but the base note of the musical key can even work in that case...
With delay-tricks you can create phase difference between 2 SID channels, that's how you can create real echoes (Like Robocop3 title beginning solo)... It's up to you and the music-editor how you solve the delaying of the 'echo' channel, where the solo content is mainly the same as on the 'dry' original channel... And of course the 'echo' channel should be more silent, that can be achieved by using another instrument or placing 'sustain' value pattern-effect for the notes. A good 'economical' solution can be seen in Drax's 'Winterbird' tune, where on the 'echo' channel he sets a sound with long 'attack' and the upcoming notes are all played legato....
The delay and volume difference can be really small between the two channels, and that gives a strong 'room' reverb feel, I like that very much...

solo-waveforms:
All kinds of waveforms (except $81 in average players) can be used as lead, it depends on the taste for the tune, or maybe an instrument that should be simulated.

$11 (triangle) - Triangle has half the amplitude, so it can be used as a light flute-like sound. Not too flexible, but some vibratos/slides/legato can make interesting feelings.
A good trick I see in many tunes that the first frame has a stronger waveform, maybe the 2nd and 3rd too, so the sound can be heard, and then it goes light into the $11 waveform.

$21 (sawtooth) - This seems to be the best to simulate violin/trumpet-like sounds. But it stands well for solos in many other cases (e.g. Toggle's SW tunes).. A plain waveform so it needs some vibrato/slide/ADSR-manipulation/detuning not to sound too plain. In contrast, trance-like tunes like the plain vibrato-less waveform but they usually use a lot of other effects in place like detuning...

$31 - I don't use it many times but has interesting sound. The problem is on 6581 old SID, where it's very silent IIRC.
$41 (pulse) - The most useful and most flexible waveform is pulse. That's the secret for good lead sounds, because it can have spectrum from sine-like to sawtooth-like depending on the pulse-width, and is controllable on fine grade (12 bit, from $000 to $fff)..
The pulse is symmetrical with $800 value, that's the fattest sound (maybe a bit more), and as we go towards $000 or $fff it gets sharper, richer, with more harmonics, but looses fatness gradually. I don't hear big difference between very low (e.g. $100) or very high (e.g. $F00) pulse-width values, but their polarity is the opposite. So selecting them might depend on the rest of the channels, not to kill some other waveforms by going into the opposite direction with pulse...
The pulse waveform can also be too plain and machine-like without vibrato/detune/etc. The most common way, and the real strength of C64 solo voices is the pulsewidth-sweep effect, which is handled by the players in software. This modifies (increases/decreases) the pulsewidth in slow/rapid pace, and gives a 'moving' or 'lively' feel to even simple long notes. In my opinion music is about changes, as our brain interprets differences instead of absolute values, that's coming from the neurons' workings. (E.g. you feel smell for a while but you get more used to it after a while. That's coming from nature and the result of a million years' evolution...on music content side, maybe using dissonances, and then resolving them lies around the same principle.)
So the pulse-sweep is a solution to give 'difference'&'movement' to a sound. The starting value of the pulsewidth and the direction/speed should be selected according what you want to hear (needs some experimenting and practice)...
If you want a thin solo, you can start with $100..$400 values, but if you want a guitar-like 'distorted sine' waveform, you can start with $500..$b00 values. The $800 is the fattest as a start, but I usually don't like to start a solo with this bumping 50% duty cycle, instead I start around $700 or $900 which has more harmonics and is less aggressive.
The speed of the sweep depends, I'd classify two kinds:
-Slow sweep ($08..$20 in SW), it's a good technique for 'beautiful' solo sounds.... the majority of C64 tunes uses this technique for solo
-Fast sweep ($20..$70 in SW) - it's advantage is that the spectral distribution of the sound varies rapidly, and a lot of harmonics appear in a short timeframe, often with the feel of detuning. This generates a 'choir'-like effect on ONE channel, and the solo has a space (this is true for bass sounds as well, like Golden Axe's starting). I use this kind of solo mainly in techno/trance like tunes, where that's closer to the style...
The keyboard-tracking (supported with a 4bit resolution in SID-Wizard) can be also a good thing to enhance the variable feel to a solo. If values are selected right (after experimenting) you can reach that for example, what a solo-guitarist does by hand: thinner pulsewidth for deeper notes and fatter pulsewidth (around $800, 50%) for the high-pitched notes...

$51 (pulse+triangle), $61 (pulse+saw), $71 (pulse+triangle+saw)- can produce interesting sounds with mixed waveform. Probably on 6581 old SID not all of them are working well like on 8580. I use $51 waveform with $400..$700 pulsewidth setting for hammond-like sounds (idea taken from Fun Factory solo of Shogoon, I use it in tunes like Arok 2013 invitro)...
In Lenore I also imitate the 'vocal' at the outro of the tune with mixed waveform. These waveforms are to be experimented, they might cause some surprises, haven't been fully utilized yet...
(These waveforms can be used for 'slap bass' imitation too, being similar sometimes...)
Pulse-sweep is not very hearable with the mixed waveforms, but a little bit hearable. After a certain pulsewidth the sound gets silent...
I had some luck with mixed waveforms to produce brass-like sounds (heard in Garden Party cover). Especially thudding them with some low-pass or mixed filter... check out Shogoon's 'Sling' for good brassy sounds.

$81-normally not used as solo (anyway, pitch setting makes sense), but Soundemon's technique with a special player could generate new waveforms from noise. I haven't seen a lot of examples of that yet...

Ringmod/sync effects can come handy in solo-sounds occasionally to give the sound some 'roaring' feel, Jeroen Tel uses this in the 'Eliminator' tune I guess but didn't check so far if this is what exactly happens there...
There are some examples where ringmod is used extensively throughout the whole tune, Necropolo is a master of it. His 'engine' sound in 'Cadmium' is unforgettable...
But as I've said before I'm honestly not really aware yet which channels do what with ringmod/sync, so I cannot give advices. I'm experimenting most of the time with ringmods, with less luck...some helps from others would come well, coz I couldn't really understand ringmod from example tunes yet.

solo-pitches:
An automatic delayed vibrato can be good, vibratos can be good to 'un-machinarize' the plain simple pitches, but they shouldn't be used everywhere. That's why a delay of 8..20 frames can be handy, so the fast note-changes won't suffer from intonation issues due to vibrato... Usually the vibrato sounds good for long notes, but sometimes I can use them to cause some extra dynamics (!) as well, especially when 'sustain' cannot be bigger.. The vibratos can bend the pitch into both directions on analog synths and SID, but in SID-Wizard they can be set to bend only upwards (like on guitar without tremolo-arm) or downwards....Increasing the vibrato-amplitude for a long note over time even gives a special feel (like in old SID tunes, or on real violin)...
Other commonly used trick for solo sounds is to give an octave up/down shift for the 1st frame of the sound, which cheats the ears to hear a doubled sound. Non-octave intervals in 1st frame are also good (like a third/fourth) and can give a totally modified mood to a solo voice (e.g. beginning solo of Pimp My Commodore tune)...
A $81 noise waveform with a certain pitch is also applicable for 1st frames of solo-sounds, for example pressing a key on the real Hammond organs has this little white-noise for some milliseconds before the real musical note starts (I use this in Arok 2013 invitation tune).
The detuning can come handy to make 'choir'-like effect on solo sounds, it needs 2 channels, one in normal pitch, the other with the same note but pitched up/down with some cents. Most current trackers support this function. A $21 with good detuning sounds like a real string-ensemble in many SID tunes.

Finally, if you have already mastered the above mentioned techniques to tweak a good sound you want to hear (or you create a good one by accident), you can make your solo sound even more interesting by placing a filter/filtersweep on it... As SID has only one filter to share between the 3 channels you have several options:
-Make only the solo channel filtered.
-Use the solo channel as primary filter-controller, and the others will follow it.
-Or an other (e.g. bass) channel controls the filter and the solo channel uses the same filter...
Filter sweep is a bit similar to the pulsewidth-sweep, can be controlled in 11bit resolution. On 6581 the fast filter-changes (cutoff/type/resonance) are heard as popping sound depending on the degree of change, on 8580 it's much less audible. So on 6581 the wild filter-programs should be avoided...These clicking sounds can be heard in my 'Deep Though' tune's intro for example. Increasing frame-speed can make the transitions smoother and these clicks less audible...
Anyway, when I decide to use filter for solos I usually use gentle, smooth filters... Jeroen Tel's solos using filter-sweep (and echo) are good examples, like Myth...
Generally filters (pulse/filter) should be two-directional to avoid overflow when reaching $00 or $ff values, but sometimes it's used as advantage... Sometimes the sweeps are into only one direction but the notes are short enough not to wait for this range-overriding...

In more restricted trackers musicians who care about 'expressions' by timbre, used separate instruments in parts of the solos. For example, the same sound, but with less 'release' in parts of the solo, where 'staccato' sounding was desired...For an example to this solution: I used several instruments of different, timbre (pulsewidth) in 'The Loner' cover, mainly to simulate the solo-guitar sound (described above with 'keyboard-tracking)...

Many trackers (like DMC,SW for example) support a setting to disable the 'sweep-reset' on sound start, so a new note won't start the sweep from the starting point but let it continue. I use this too in Pimp My Commodore solo, this gives even more variety to a solo...
This is also usable to simulate filter-cutoff automation in techno-like tunes, if there are no pattern-effects for that...

That's for now, if there's need for it I can tell how I try to make the bass sounds to sound fat and arpeggios to sound smooth...
The article about this topic (after valuable discussions here) is a good idea for codebase64 ;)
I believe we can learn a lot from each other and open the doors for the newbies too into SID composing...

2013-04-12 13:49
SIDWAVE
Account closed

Registered: Apr 2002
Posts: 2238
Dont forget this good guide to the SID
SID Guide

There you have all possibilities, so its just to try them out!
2013-04-12 15:10
Hate Bush

Registered: Jul 2002
Posts: 461
masturbation \o/
i'll only speak of a 'saving channels' matter, as i don't feel competent enough lately ;)

1) mixing kick with bass
you don't really need more than a half of a tick row (3 ticks when tempo is 06 and so on) for a kickdrum, because the 'tail' of it is often unhearable in the mix anyway. place the usual 81-41-41 - or 81-41-40 - in the wavetable first and follow it by your bass sound of choice.
tip: this sounds best when filtered, with different filtertypes for kick and bass.
hint: this is when decay comes in handy; such combinations scream for DS like 6A or such. why? because the bass would probably be too loud with max volume (sustain=A then), whilst drums need it badly (decay=6).

2) mixing snare with bass
this is very similar and all the above applies, however the snare isn't as straightforward as the bassdrum, for it contains full range of frequencies. normally, we would follow the THUMP of 81-41-41 with 80 or something. this time, i'd recommend complimenting the THUMP with some 81s (51-81-81 for example) on any other channel available. they will serve as a high frequency layer of sorts. can yield great results or sound weak and shit, it all depends on the context and details.
tip: nothing stops you from creating some other sound (arpeggio for example) after the noise layer. this way you can program full drums, bass and arps lines on just two channels.
hint: experiment with filter settings AND attack. remember that you can filter two or three channels simultaneously.

3) mixing (closed) hihat with bass
if you got a hold of 1) and 2), this is too simple to be explained ;)
tip: you can of course mix drums and hihats with arps or even melody.
hint: could be tricky in singlespeed. in general, the above technique benefits from speed multiplication, as a resolution of a sound is increased, and so is a number of possibilities.

4) mixing bleeps and blops with arps and melody
Wacek's trick, can't remember using it myself. in short - take the instrument definition, copy it a few times and make it preceded by some other sound, but at different intervals. this way you can play with two melodies on a single channel, and much more.
tip: the intervals shouldn't be less than an octave (both ways)?

5) rapid legatos
with faster tempos, it is possible to create a distinct, very short sound by a direct legato to a higher pitched note and back.
like this:
c3
.
.
.
g5 300
c3 300
tip: works best with low pass filter.

that is all i can remember for now. happy SOUND DESIGNING :D
2013-04-12 16:23
NecroPolo

Registered: Jun 2009
Posts: 231
- ringmod on everything, sync on everything and watch the SID spraying blood

- being aware of what other folks do and avoid most of that

- if sound kills smaller animals than 2cm in a 10m area and gets on the nerves of girlfriends, you're on the right direction

- blown out speakers

- if you can do anyone's sound that's good but if anyone could do your sound, you're wrong

- from that point, all the other details are about tweaking numbers I guess :)
2013-04-12 16:34
Stainless Steel

Registered: Mar 2003
Posts: 966
Awesome thread guys. If i had any clue about the matter i'd post something, but alas i wont :-D

2013-04-12 17:09
Soren

Registered: Dec 2001
Posts: 547
Groepaz: I better not. :)
2013-04-12 18:37
Stinsen

Registered: Feb 2012
Posts: 76
What Stainless said, really interesting thread.
Previous - 1 | 2 | 3 - Next
RefreshSubscribe to this thread:

You need to be logged in to post in the forum.

Search the forum:
Search   for   in  
All times are CET.
Search CSDb
Advanced
Users Online
Matt
Twoflower/ΤRIΛD
serato/Finnish Gold
jmagic
Sokrates
sLASH
v3nt0r/ibex-crew
Guests online: 79
Top Demos
1 Next Level  (9.7)
2 13:37  (9.7)
3 Coma Light 13  (9.7)
4 Edge of Disgrace  (9.6)
5 Mojo  (9.6)
6 Uncensored  (9.6)
7 Wonderland XIV  (9.6)
8 Comaland 100%  (9.6)
9 What Is The Matrix 2  (9.6)
10 No Bounds  (9.6)
Top onefile Demos
1 Layers  (9.6)
2 Party Elk 2  (9.6)
3 Cubic Dream  (9.6)
4 Copper Booze  (9.6)
5 Libertongo  (9.5)
6 Rainbow Connection  (9.5)
7 Onscreen 5k  (9.5)
8 Morph  (9.5)
9 Dawnfall V1.1  (9.5)
10 It's More Fun to Com..  (9.5)
Top Groups
1 Performers  (9.3)
2 Booze Design  (9.3)
3 Oxyron  (9.3)
4 Nostalgia  (9.3)
5 Censor Design  (9.3)
Top Coders
1 Axis  (9.8)
2 Graham  (9.8)
3 Crossbow  (9.8)
4 Lft  (9.8)
5 HCL  (9.8)

Home - Disclaimer
Copyright © No Name 2001-2024
Page generated in: 0.081 sec.