The end of Gigastudio!

[Cheez]

Notice: TEAC America, Inc. will cease further development of GIGASTUDIO and GIGASTUDIO related products as of July 21, 2008. Product sales and technical support will continue through the end of the year.

I cannot believe it! All my Giga libraries are going to die!

Tascam certainly killed Giga. Gigasampler when it first came out from a small company (Nemesys) rocked the sampling world. Then Tascam took over. Some bad decisions were made re: copy-protecting developer's samples made it lose the first big crucial battle to Native Instruments (which first introduced Kontakt as the first competitor to Giga). Then it's all downhill from there .

I can still get NI's Kontakt and import my Giga libraries over, but I have to rethink very carefully now. The sampling industry has changed drastically in just the past few years alone. And with the rate it's changing, the future is terribly uncertain. Big time developers are now coming up with their own proprietary engines to save cost, which means they no longer depend on third-party software. The smaller developers would probably still use Kontakt, but how it goes in the future depends on the competition. ARIA and PLAY engines will certainly try to sell their samplers as lower cost 3rd party samplers, which will be way cheaper than Kontakt.

This is a turning point in the sampling industry. The question now is - can we depend on NI? The way to go, I believe, is to go for large sample developers like East West and VSL instead of software developers. It's safer. Software with no samples = zero. The software depends on the developers to produce the libraries - if there are no libraries, then the software is as good as dead. So with sample developers becoming software developers, it's obvious to stay with them!

 

[widdly]

Can't agree with you there cheez. A software sampler without libraries is fantastic. Then you can get be a bit creative and do some actual sampling. Plenty of great musicians wouldn't dream of touching a sample library. Whole genres of music are based on real sampling. Check out Amon Tobim for a good start.

I'm of the opinion that synthesizers shouldn't come with presets either. Blank canvas and let the user create the sounds and learn how to use the thing. Did Wendy Carlos, Suzanne Ciani, Mort Subotnic, Kraftwerk or Delia Derbyshire ever use a preset or a sample library?

For giga libraries I think you will find rompler tools like BFD, sample tank, Atmosphere etc will become the norm. Most of the goodies in kontakt are wasted if all you are doing is loading up someone else's instrument. It's like GM only bigger.

 

[Cheez]

I'm not sure if we are on the same page here. When mentioning about sample libraries, I'm talking about multisamples, not loop sampling. In multisamples, it's like a rompler except that each note is sampled multiple times at different velocities, with different articulations, and chromatically. So, creativity still lies on the composer who has to sequence each note. This is entirely different from loop based sampling in which a person press one button and a continuous loop keeps playing. The beauty of sampling is realism which you can never find in any hardware. By the way BFD is utilising sampling technology. It's not a rompler.

 

[dirtchamber700]

i don't think GS will die. it's name is still there, and i'm sure one of the bigger companies would prob buy it over and it's technologies as well. it's too big a name to let it go to waste. history has shown us that it's quite possible.

on a side note. i don't like putting faith in NI. it's like they are fat and happy and sitting on their ass releasing updates and coming out with nothing new. i do like EWQL tho. heh

 

[Cheez]

Actually, GS is going to die according to Tascam. Tascam bought it and it belongs to them. And they apparently decided to shelf it. The technology is no longer new. So there's no point in people buying it over. The trend now is people developing their own sampling engines. When Tascam first acquired Nemesys, it was different. Gigasampler was groundbreaking technology. It was the talk of the town. Now the situation is we are seeing a dying format.

And it's actually anticipated. There are hardly any new sample libraries out for GS. More and more developers are either developing their own engines or migrated to Kontakt. Their last supporting giant developer was Sonivox. And recently, Sonivox decided to stop dveloping for GS as well. That probably triggered Tascam's decision. What's the point of having an engine with no sample libraries?

Bad marketing, bad management...that's the cause of GS's death.

 

[widdly]

Rompler as in read-only. You don't create or change the sample files.

Plenty of hardware romplers used multisamples.

So since you cant change or create the sample file....

sample libraries = rompler
bfd = rompler

Sampling is about actually sampling a sound ie. recording a new sample.

Some good reading...

Creative Sampling..

http://www.soundonsound.com/sos/dec98/articles/creatsamp.318.htm

The Lost Art of Sampling Series....

http://www.soundonsound.com/sos/aug05/articles/lostscience.htm

Amon Tobim...

http://www.amontobin.com/field/

 

[Cheez]

OK. I see. Semantics. ROM usually refers to media that is very closely tied to specific hardware (chips) and hence the inability to modify the data stored. Hence ROMplers traditionally refers to the hardware keyboards/modules. But yes, you can theoretically classify software as "ROMplers" if the data is locked, although it's not inherently linked to hardware chips and not entirely unlockable.

Actually, GS libraries are the only ones that allow end-users to edit the raw wav form. This is also why GS died, because developers want to protect their materials but GS wants to give the end-user greater creativity and flexibility. So they refused to lock the raw data. Good intention, but developers don't quite like it and started migrating to alternatives, namely Kontakt. And now, it's everybody's game since they are making their own engines.

Sampling from zero is not common with composers. DJs may do it, some may do it for one-off hit samples or creating attain unusual sounds. THAT is still common, although it is dying. But when we talk about multisamples, it's neither realistic nor feasible to sample (for example) an entire 24-strings section - need to pay for the recording time of the string players, not to mention the rental of a concert hall etc. If one has the time and money to multisample an instrument, that person is probably a developer and will eventually sell the product.

But in a way, I agree with you. We are spoilt nowadays. Gone were the days when we had to use Emus and Akais to sample. But those sampling are very different from the ones we get from the large multi-GB libraries today. Which reminds me....the way GS is dying has a tint of how Emu and Akai died...

 

[iansoh]

Not everyone here is into synthesis and creating our own sounds. I see that Widdly comes from a more 'electronic music' perspective, but not everyone here is talking about electronic music. I don't know much about electronic music techniques, but I can identify why Widdly would say that synthesizers should come blank, samplers should not have sample libraries etc

However, some of us just want to play and sound like a real orchestra without having to hire one and sample them. Hence romplers. Whole genres of music are based on that, great musicians use them too.

We should put the discussion into context - IMO the subject isn't the sampler per se, it's what are we gonna do with those sample (or rompler, whatever) libraries. those strings, woodwind, piano etc libraries.

I sense a turning point of a differnet kind - to physical modelling. Not sure if it'll happen across the board for all, but there are already a few good ones - Pianoteq with their piano modelling, and from soundfonts.it their VB3, MrRay73, Miles'tone; Hammond organ, fender rhodes and trumpet synths respectively. I particularly love the products from soundfonts.it. The best part about modelling - small software footprints, low RAM requirement. Pianoteq is just 16MB, VB3 is 2MB and MrRay is 1MB if i remember correctly, or about there lah. cheap too - 89 Euro for hammond + rhodes + something else can't remember.

As I said, can't really see it happening for other sounds, e.g strings, if they can even be modelled to the same quality and standard as Pianoteq and VB3 (i.e. good enough to replace samples).

 

[widdly]

Of course I can see that some people like to use orchestral libraries and are more interested arranging and composing than creating sounds. My point was that samplers are great tools even without huge sample libraries.

Physical modeling is a great idea. Milestone is pretty fantastic. The hammond models are great. Even for non realistic stuff it is pretty cool. The steampipe instrument in Reaktor is pretty awesome. I've written a few physical modelled vsti myself although I wouldn't describe them as realistic. More to play around with the ideas. If you are interested, you can check them out here... http://dmi.smartelectronix.com/

 

[Cheez]

Yes, merging modeling and sampling is the new thing now.

Anyway, back to the main topic - just got Komplete 5 today. Can't believe I'm migrating... Still mad at Tascam.

 

[widdly]

Free, open source giga compatible sampler...

http://www.linuxsampler.org/

..despite the name it runs on windows too. Not really ready for the primetime but worth keeping an eye on.

I'm surprised gutturalpiss isn't pimping it.

How did Kontakt cope with your giga libraries cheez? The import in Kontakt always seemed a bit flakey to me.

 

[Cheez]

Yes, linuxsampler has been around for a little while. But since giga libraries are going to be extinct with no new developers making them, it's best to steer clear of giga.

I've not converted my gig files. So far, reports are that conversion are OK. As long as there are no complex programming and scripting like complex cross-fades. I'm glad I'm using Kirk Hunter. He lets me cross-grade without problems. I'm not a fan of East West and certainly not EWQLSO. With Kirk Kunter layered with Giorgio's Trumpet and Garritan's Strad (hopefully ARIA version coming out soon), I should be able to get pretty good sound that's more unique than the standard sounds you get.

 

[gutturalpiss]

I'm surprised gutturalpiss isn't pimping it.

I don't think it'll be of further help since the discussion is leaning more towards the libraries rather than the software. It's just an alternative glass for the juice. The format will not die, the standard will. However, the format will become deprecated. But yes, that is why it's interesting to see where it'd end up because GIG was and still is the primary focus.

I believe physical modelling is the present and future. Unlike sampling, you don't work with result - you work towards it. Replicate, simulate, practise the actions before the result, and you have a finer result than the person before you who took the original and just reimplemented it in a different way.