How does Korg sample its sounds for the Kronos

methusala · Post by **methusala** » Wed Jul 22, 2015 11:45 am

Just a bit of background here. Just curious. How does K go about this ?

Where are they recorded---in a studio or sound stage etc.

Are all non-synth instruments recorded live.

Are the pianos, for example, multiple miced and for lid variances.

How many same brass instruments are involved in a recording--one horn or 3/4 etc

Where is the post-recording processing done?

Patrick

LZ · Post by LZ » Wed Jul 22, 2015 2:57 pm

I don't know the answers, but I DO know that with the exception of HD-1 sounds, they are not straight samples. For instance, as I understand, the pianos are a combination of samples and modeling.

EvilDragon · Post by **EvilDragon** » Wed Jul 22, 2015 3:05 pm

SGX-1 is samples only AFAIK, SGX-2 only models sympathetic resonance, but the core sound is still samples.

EP-1 uses samples at its core, but employs some sort of resynthesis to mold the final sound from multiple elements within those samples. Other engines are not sample-based (except in MOD-7 you can use samples as a soundsource instead of a generated waveform, and in STR-1 you can use them as exciters for the model).

michelkeijzers · Post by **michelkeijzers** » Thu Jul 23, 2015 10:35 am

[quote="EvilDragon"]SGX-1 is samples only AFAIK, SGX-2 only models sympathetic resonance, but the core sound is still samples.
quote]

Afaik SGX-1 is a mix of samples and modeling, or at least the selection of what samples to be used.

EvilDragon · Post by **EvilDragon** » Thu Jul 23, 2015 11:04 am

michelkeijzers wrote:or at least the selection of what samples to be used.

That doesn't fall into physical modeling, though. Sympathetic resonance modeling of SGX-2 does (well, more like additive synthesis in practice, I guess).

jimknopf · Post by **jimknopf** » Thu Jul 23, 2015 11:20 am

EP-1 is the Kronos engine which most clearly seems to use samples (or sample slices), while at the same time calculating seamless transition between velocities and the overall dynamic behaviour, based on these samples/sample slices.

It's no pure physcial modelling, but a mixture of sampling and calculating, like the big three use it in various ways.

EvilDragon · Post by **EvilDragon** » Thu Jul 23, 2015 11:30 am

More of a resynthesis method of sorts, it seems to me. So that'd be additive synthesis underlying that tech.

burningbusch · Post by **burningbusch** » Thu Jul 23, 2015 3:22 pm

The SGX-1/2 pianos appear to be mic'ed with two mics (stereo) but there might be more to it than that. It's always possible they mixed more than two mics and brought them down into a stereo mix. The lid variations are EQ/filter, not addition micing. So too with play position which just reverses left/right channels. I make these statements based on my ear, not any additional info.

Most sampled instruments in workstations were recorded in mono (single mic) and with little to no room ambience. I believe anechoic chambers were used. A few instruments such as ac. piano, ensemble strings and ensemble brass were recorded in stereo. This was generally the case with the Kronos until the last year or so.

Korg has been using Kurt Ader's sound libraries for additional sounds like EXs11 Strings and EXs16 Funk and Soul Brass which are available for purchase at the Korg Web Store and a subset is included with the Kronos 2. I believe all of these were recorded in stereo. There are variations of the string ensembles with different sizes. Also the brass ensembles includes a wide variation of combinations:

• Ensembles
• Trumpet x 2
• Trombone x 2
• French horn x 2
• Trumpet + tenor sax
• Trumpet + trombone
• Trombone + bass trombone
• Trumpet x 2 + trombone x 2

Busch.