Am I missing something here? What softsynths don't respond to program change MIDI messages? Seems to me that I do it all the time using my K2661 as a controller and spmetimes I even have to remember to set the Kurz not to send program changes externally when I rotate the alpha wheel in a program select mode.

Hi GC,

I threw 12Ghz out as a number. It doesn't mean anything in real world terms. I just picked a number and ran with it. I was referring to more processing power and, as you mentioned, dual-core systems address part of the concerns. But even with higher processing speeds we'd still be facing the small memory footprint that's available with today's PCs. Streaming from disk is inefficient and cannot solve the aforementioned program change issues(more on that later).

Receptor is a fine, but it's no more useful than another computer. Basically, I have the equivalent of Receptor now. I run one PC, 3GHZ P4, 2 Gbytes RAM for my Synthogy Ivory and the other soft synths. Actually, come to think of it, I have more RAM now than what a maxed out Receptor allows(from what I've read it only goes up to 1GB). So Receptor isn't the answer for me. Heck, if I am going to pay $1600 for Receptor I might as well buy three P4s on Ebay and stack them up and network them via ethernet or MIDI. Then I would be able to run TONS of plug-in synths! No brainer there.

Please note that I am not saying that solutions cannot be found for the shortcoming of SW synths. If money is not an issue, then throwing it at a problem will find you a solution. You can build a cabinet and put half dozen 3GHZ PCs, maxed out on RAM, each one capable of running ten or twelve instances of SW synths and you'll be set to go! However, I don't have the kind of money for such a massive project. Doing something like that would negate one of the premises of SW which is to eliminate or reduce the need for more HW. Having to buy more PCs or constantly having to upgrade them in order to accomodate new SW synths kinda makes the whole excercise a moot point. Instead of stacking up HW synth racks, I'm stacking PCs! LOL!

The final solution that will put HW synths on EBay, at least for me, will be when RAM is so cheap that we can have dozens of Gigs that will allow instances of multiple SW synths to sit in memory, whereby processors can swap out samples at higher speeds. Of course, the proper OS would have to be designed/revamped to allow for such incredibly high memory addressing and management. But that isn't going to happen any time soon.

As for Korg to start dropping hardware in favor of SW synths is questionable. The company released the OASYS, and frankly, it's THE MOST powerful synthesizer to date. Korg is banking that this platform will be a step toward a modular system that will trickle down to more affordable units(maybe an OASYS rack module?) This monster runs on Linux and that in itself is a HUGE move in the right direction. I'm guessing that within the next two or three years we'll begin to see HW synths with several Gigs of memory(the OASYS has as a standard 500+MB of sampling RAM and more for the program area). Basically, Korg is addressing some of my concerns. It's just not affordable yet.


Hi John,

When I mention program/bank changes I am speaking specifically about making these changes to the soft synths WHILE the sequencer is running. In other words, let's say that I have this song/sequence running in Sonar and I need to change the program that is running on MIDI channel 5, which is assigned to my first instance of Dimension Pro. I want to change from a splash cymbal to a crash. How would I do that? Currently, there is no way to automate that through Sonar. I cannot send a MIDI program-bank change to DPro to tell it to change programs WHILE Sonar is running.

In contrast, if I had MIDI channel 6 assigned to my Korg Triton and it was running a synth patch and in the middle of the song/sequence I wanted to change it to an organ patch, I'd simply insert a program change via Sonar on that MIDI channel. I'd assign the bank and patch in Sonar that corresponds to my Triton patch and whala! The program would change mid song. I can do this toggling of patches an infinite amount of time and each time there would be NO glitch on Sonar and none on the Triton. Why? Because Sonar is simply sending a MIDI control message which takes next to no CPU cycles to process and the Triton, having ALL patches easily accessible via ROM/RAM, can change the patch in a blink of an eye. The Triton has no hard drive to contend with.

This basic, and may I say old, MIDI principle cannot be sustained with Dimension Pro and other SW synths. Supposedly some can do it by having the user predetermine the patches that will be needed during the project and putting these on a queue. Once running, the sequencer can swap these patches in and out of memory. But, again, we are talking about MORE MEMORY to load MORE PATCHES.

Why can't DPro make program changes on the fly? Because DPro uses huge samples that reside on disk. The ones that are already loaded and used in Sonar are in RAM, but all others are on disk. Hence, any program changes would force a load of the new samples from disk to memory. This, sadly, is one strong short coming of SW synths. It negates the premise of SW doing away with HW. It tells me that I need a LOT of RAM in order to exploit the power of my SW synths. So, am I really doing away with HW?

Of course, you could ask me: "Why would you want to change patches midsong?!" Well, if I recorded directly to audio each time I played a patch, then I wouldn't need to make program changes at all. In essence I'd be working in the audio domain an not with MIDI. But I don't work that way. I ONLY commit to audio when I am fully satisfied with my composition. I may try a dozen bass patches and a dozen synth patches etc for each track before I commit any to audio. Once in audio, I cannot change the timbre. But in the MIDI domain I can take my time and change things MANY times.

This is a good article on DPro that mentions the MIDI program/bank change issue:
http://remixmag.com/synthesizers_and_samplers/remix_cakewalk_dimension_pro/

"...
(Note that because Dimension Pro's programs depend on large samples stored on the hard disk that you designate upon installation, Dimension will not respond to MIDI Program/Bank change commands.)
..."

Of course, if you don't care about MIDI program change then all of this talk is superfluous!

I think, as some posters have mentioned, that both have their merits. That said, I'd always own at least one 'real' keyboard i.e hardware synth, so that when music is the first priority, mucking around with computers doesn't have to destroy creative impulses.

As I see it, there are several directions that each of us can go in at any different point in time.

1. Working on musical ideas. Use a hardware synth and something immediate like a minidisc recorder to capture ideas as a sketchbook. Because when you're in a musical plane, you're really using a different part of your brain and you - well, certainly I - don't want to faff around loading stuff or mucking around in any way.

2. Gigging. Have one real keyboard and any amount of backup computerised stuff. If the computer stuff crashes you can at least busk on the keyboard - that, in my opinion, is vital - and something that will restart almost instantly after a power glitch is much preferred.

After that most of the stuff we are doing will accommodate computers and some time spent thinking - so soft synths are fine, but I would never relinquish all hardware. I love having knobs and lights in front of me - it makes me feel I'm interacting in a tangible way with the technology.

I think analog is 20-40% better than digital.

ORIGINAL: jinga8
Look at ROMplers like Dimension Pro, the IK Multimedia thingy, and that Colossus thing. For synth emulations, check out Arturia, (Moog, Minimoog, ARP2600, CS-80) or Native Intstruments (FM7, P-53, etc) or for some innovative stuff checkout Virsyn, Cakewalk's Rapture and z3ta+, etc.

Dimension Pro is not a ROMpler, it's a sampling synthesizer. It can be used as a straightforward sampler, or you can create your own patches in .sfz code with a simple text editor. Likewise you can just drag and drop single-shot .wav file samples and it will map them automatically up and down the keyboard, without any aliasing.

Softsynths have a home in our studio, but we just bought a Motif ES8 too!

It's hard to beat a great sounding hardware synth, analog or otherwise, running through a great analog preamp.

stratton, you raise another valid point, and one that I didn't want to bring up(for the sake of not wanting to open a can of worms!), namely, the actual SOUND of hardware synths vs software.

I can only speak from my personal experience, but I have not been able to match the DEPTH and WIDTH of my Triton and Trinity patches with Dimension Pro or the CS-80V. Specially the Korg organs, strings and synths. Those patches are simply amazing. The depth and thickness of the B3 organs don't get much better than on the Korgs. In contrast, the B3s on DPro's expansion pack 1 are pretty weak. They are thin and lifeless and adding chorus and reverb to them only makes them sound worse.

I don't know if it's the patches themselves or the D/A converters on the Korgs(I use M-Audio audiophiles 24/96 on my PCs, which are not shabby sound cards), but the sounds seem, to my ears at least, richer and more complex on the Korgs. Of course, it's all subjective. But I've tried long and hard to fiddle around with my soft synths, tryng all kinds of combinations and effects stacking and I cannot match the depth of the Korgs.

Something else that adds to the depth and richness of my hardware synth sounds are the TC Electronic M300s I have running in series with each synth. These units allow me to set compression AND reverb to the signal path and having this combination far exceeds anything I am able to generate via software. I have yet to match the reverb quality of the entry level M300s with anything I have in software; and I have Soundforge reverbs as well as the Lexicon plug-in that comes with Sonar, but neither one of these can enhance the signal as well as the M300s. The differences are dramatic! I am not talking about taste here. It's a discernable difference that most anyone will pick up on. Say what you want, but I am not convinced that software reverbs can stand up to their HW counterparts. Even entry level ones like the M300s.

I have the Orchestral collection expansion card in my Triton rack. The combis offered by that card are FAR superior to any of the Garritan strings and orchestral stuff that comes with DPro(and the samples are smaller on the Korg!). There is no comparison. For classical purists the combis on the Korg expansion card will not work. For example, the patches on the stacked strings combis all play in unison. This is a no no, since each patch should be playing a distinct note of a given chord. So there is no way to fool a purist when playing the combis. However, if one were to take the individual patches and compose using those then realistic approximations can be achieved. I wouldn't attempt to use the strings on DPro on any piece that featured them prominantly. They are simply flat and lifeless.

I just cannot find an area where DPro or the CS-80 can match the richness of the Korgs. True, the CS-80 has a unique PERSONALITY that the Korgs can't match, but I often find myself stacking Korg patches on top of DPro and CS-80 patches to add some depth to them. I hardly, if ever, do the opposite.

Yep, I agree jrfernan.

I track our Korg KARMA and the Motif with no effects through a Neve Portico pre, then s/pdif aux out from SONAR to an Eventide Eclipse and record effects on their on tracks. In terms of effects sound quality, nothing in software or the synths compares to the Eclipse.

I usually record audio and midi at the same time so I can stack a softsynth or two if I want, or bring in another hardware synth.

I had some old roland analogue synths and some string ensembles in my studio the other day. And I am actually pretty happy that i dont have to deal with them every day... tuning, strange behaviour... turn up a knob and then down again, and the sound isnt the same as it was last time... which is probably part of the charm, but kinda annoying and time wasting for me at least. and they are noisy, and have to be well kept or the knobs are noisy when you turn them etc.
I'm perfectly happy with my (almost) exclusively soft studio :)

I asked the very same question over on the KVR forum a few years ago:

"Should it be Hard or Soft?"

Well, I was advised to go Soft and I have never looked back...

SilkTone