I got my TEMPEST again some days ago, and as some of you who have looked in the TEMPEST forum already know, I managed to brick it pretty quickly, trying to update it to the latest OS... the TEMPEST obviously do not have a bootloader function, so right now I'm waiting for a replacement board to arrive from Sequential, so that i can get it up and running again... the time off, has made me think a bit about how I want to go about designing sounds for the TEMPEST, and how i want to be using it... how to organise the BEATS in it...
I know very well from my earlier dealings with the TEMPEST, that you have to really work around some "things", especially if you want consistency in the transients, so I want to be planning ahead with this in mind, and also a few "problems" regarding individual processing of sounds...
To be able to achieve consistency, each drumsound need to be assigned to a single voice... this improves it quite a lot, but you still have some inconsistency when more than one sound use the same voice... i bet this is because the TEMPEST has to reset every sound parameter when two voices that use the same voice, has to be switching between the two sounds... that is; if you hit the same pad all the time, you have no inconsistency, but if you then hit another pad, it will on the first hit sound a little bit different to any following hits of the same sound... simply because CVs has to settle to new values etc...
The only way to remedy this is if yuo asure that one sound alone is allocated to a particular voice... in other words, you limit yourself to 6 drumsounds in a beat, all of them using their own voice... in that case you cannot get more consistent really... this sounds like a total waste of up to 32 sounds... but I intend on expanding the number of sounds to 12... double up so to say... each drumsound will have an "accent" sound in addition... in that case, those two use the same voice, and since they are usually rather close in nature, the inconsistency would be a lot less... say: Kick and Kick with long decay... long/short snare... Open/Closed Hat... High/Low tom etc. etc...
I intend on placing these six sounds on the lower left row of pads, and their accents exactly above which seems most logical... the two pads left over at the lower and upper row at the right side I do not know what to do with yet, but I'll figure that out.
Another reason to do it this way is to overcome another problem... you do not have 32 individual outputs... when you start to assign more than one sound to the same output, the harder it gets to individually FX any drumsounds. With my plan, having only 6 sounds each with an accent, any of the sounds will get it's own separate output, not sharing it with any other sound... in essense, I'm creating a kind of "hardwired" drummachine... almost like a 6 track/sound drummachine with individual outs for every sound... on top of this I get the best consistency I can get, and really... you do not need polyphony in a drummachine anyway.
This means that if you want to use the built in distortion and compressor, you simply do not plug any cable into the individual outs, and when you need external processing, you simply plug in a cable for whatever of the 6 drumstrack sounds you want.
I know it's a waste of the 32 sounds, but I intend on using sound bank B for a variation kit on all my BEATS... that leave only 8 wasted sounds, but I can live with that.
In my point of view, the TEMPEST should have been made a lot more simple... i actually think it would have been better if it was made with this "one voice per sound, on it's own output"... it would have made it much more logical in it's layout... the synth engine should have been left out completely... that task is for real synthesizers, and it only takes up codespace, and complicates the code even further... but hey... TEMPEST is what it is, and it's a cool sounding drummachine, especially when you know how to work around the quirks and remaining bugs.
Back to my plans... earlier I wanted to just make sounds... these days i can see the usefulness of dropping the individual sound approach and instead focus on crating BEATS instead (or rather KITS, which is the right term for this). I want to be able to have kits where all sounds fit together at the get go, where levels and BEAT FX has already been set up... sure you may want to change something, but then it's easier to just tweak the BEAT sounds, or load individual sounds if you want something else, instead of having to start from an initialized BEAT every time, and load every single sound from scratch... SO I'm really appreciative of the way Dave and Roger has organised the memory of their structures... that a single BEAT has all of the 32 sounds "built in", and not just pointing to sounds... that's a HUGE plus with the TEMPEST!
I also want the KITS (I'll call them kits from now on because i really find the name "BEATS" confusing...) to be as modulateable in reatime as possible, so I want to make full use of the beatwide control the TEMPEST offer... also there will be NO sequences programmed because everyone would want to make their own rhythms anyway, and having to constantly erase the sequence tracks is just gong to be a pain in the neck... therefore; EMPTY! ... you be the creative part with the sequencing, I'll do most of the sound design
I'm not sure that my kits will be minded solely on Ambient because there really is not much rhythm in ambient music anyway... there would quickly be a lot of kits with thumping sounds anyway, so I'll probably focus these kits more at the experimental area...but certainly I'll focus on the weird and the wonderful more than on the boring cliches (think 808, 909 etc...) ... it's time to broaden peoples sense for percussion other than just the usual "four on the floor" type of approach... there is already plenty of options in that department out there... I'm certainly going to "rape" that AudioMod and AudioFeedback parameter for sure
that was a bit about my current plans for the TEMPEST.