Hardware or Software sampling?

[ckett]

Hello,

I am looking into establishing a sampling system for my music composition. Debating whether to go with an Akai S6000 or stick with a software sampler.

I will be using a sampler for triggering all sorts of sounds for an ambient world music style. I don't need hard disk streaming for using super large orchestral libraries. Many of the samples will be self created and manipulated in a variety of creative ways.

So really the question is this: Has software sampling been reliable enough to use without too many hang-ups or should I look into an Akai S6000 that has a proven reliable OS?

I just see so many issues with every upgrade in the software world that it appears that tools such as Kontakt and Gigastudio are more of a headache than a reliable tool.
Upgrade your sequencer, then your sampler version X has problems. Upgrade sampler x and then it doesn't run on your current sequencer version x.


Here is my DAW computer system:
Cubase SX 3
Asus A8V Deluxe with AMD X2 dual core chip
2GB Corsair RAM
RME Digiface Hammerfall sound card

Thanks and look forward to your interesting experiences!

[keyman_sam]

If you're not doing stuff live, i'd say go with Kontakt.

[M'Snah]

For the price of a hardware sampler you can probably buy all the major current software samplers . I'm very happy with Kontakt 2, and haven't run into crashes or bugs yet.

[kuniklo]

For the price of a hardware sampler you can probably buy all the major current software samplers . I'm very happy with Kontakt 2, and haven't run into crashes or bugs yet.

You can get a very good hardware sampler for ~$300 or so. High-end E-Mu or Akai. Roughly the cost of Kontakt 2.

There are tradeoffs either way. Software is more convenient and powerful, but IMO software developers have somewhat lost sight of the art of sampling and resampling as a sound design tool.

I haven't had any showstopper issues with Kontakt 2 but I still prefer my MPC for a lot of routine sampling functions and I'm still tempted to pick up an E-Mu E6400 or Akai S6000 for a different way of working.

[ckett]

Are most of you using Kontakt as a VSTi within your sequencer or are you running it on a dedicated machine?

It seems that with a dedicated machine you would avoid a lot of the upgrade issues with matching the best working version of Kontakt with the best working version of your sequencer.

[jupiter8]

Have you ever worked with a HW sampler ?
If not, it might not be as straightforward as you think.
There are lots of benefits to software you might not appreciate. Timing and total recall to name a few.

It could just be you're just exchanging one set of problems with another.

[kuniklo]

Have you ever worked with a HW sampler ?
If not, it might not be as straightforward as you think.
There are lots of benefits to software you might not appreciate. Timing and total recall to name a few.

That's what I meant when I said software was more convenient.

I've worked with an Akai S5000 and extensively with the MPC. I much prefer the MPC to any software sampler for percussion, but only for the interface.

The thing almost all software samplers miss is actual *sampling*. You can do all kinds of crazy things to a sample once you've loaded it, but the kind of sampling and resampling you can do on old hardware units is way more work.

[jupiter8]

Have you ever worked with a HW sampler ?
If not, it might not be as straightforward as you think.
There are lots of benefits to software you might not appreciate. Timing and total recall to name a few.

That's what I meant when I said software was more convenient.

I've worked with an Akai S5000 and extensively with the MPC. I much prefer the MPC to any software sampler for percussion, but only for the interface.

The thing almost all software samplers miss is actual *sampling*. You can do all kinds of crazy things to a sample once you've loaded it, but the kind of sampling and resampling you can do on old hardware units is way more work.

I only skimmed the thread and see now you more or less said what i said.A simple +1 would have sufficed.

[Distorted_Mastermind]

They both have bonuses for me.
Hardware is cool because it's fast and intuitive up to the point of tracking. Doing most things with a sample once it's been sampled on hardware is much easier imo and is getting the sample there.
Software has bonuses like nothing needs tracked out...it's all already there. The editing abilities of software are more advanced, and the sound quality is going to be a bit better since it isn't being converted as many times.
Of the two I don't really have a preference overall. I seem to go through phases where I prefer one to the other, but couldn't pick one.

[hollowsun]

The editing abilities of software are more advanced, and the sound quality is going to be a bit better since it isn't being converted as many times.

Eh??

A hardware sampler has A-D convertors to get the sound in and D-A convertors to get the sound out. What other 'conversions' are being used?

If anything, a hardware sampler is likely to have better sound quality because it will have dedicated ASICs and hardware (coded at the 'machine level') to provide higher quality interpolation for transposed samples and it will have hardware dedicated to accurate loop playback, etc.. This is particularly true of the S5/6000 which has excellent interpolation, etc..

In fact, a hardware has dedicated bits of electronics throughout that are optimised for the specific task they have to perform.

Software samplers, on the other hand, either have to 'fudge' these in software. Or they ignore them completely (a transposed sample without interpolation - yeeeeugh!!!)!

I have seen comparative tests of aliasing, etc., on the net and h/w samplers inevitably win out.

That said, I agree that there is a place for both h/w and s/w samplers and both have their pros and cons. I know which I prefer but it would be foolish to deny the advantages that s/w samplers bring to the table.

To the original poster, if you can afford it, get an S6000 anyway and use both. The fact that the S6000 uses standard WAVs means that the samples are interchangeable with a s/w sampler (and there are convertors that make a pretty good job of converting in either direction).

Steve

[Sascha Franck]

Hm, I don't know. Steve has a valid point in saying that there's dedicated hardware (and a dedicated OS) coming with hardware samplers. So, at least some sort of quality is guaranteed.
But then, with software becoming more and more stable (unless when fooling around with some beta versions, I can't remember the last serious crash I had), the dedicated OS point becomes rather moot. Not sure about the hardware side of things, but these days good converters aren't all that expensive either, and you may have them already as well.

I do however agree on the update issue. There's no warranty the software sampler you buy today will still be running on tomorrows computers - and who wants to keep a stock of aging computers just to be backwards compatible? In case you're using a laptop, it might be less of an issue, as you could still use it as a standalone sampler, but dedicating a desktop with additional keyboard, mouse and monitor seems a bit tedious.

Then there's the live playing issue. I still have a tough time taking my laptop with me to a gig (just recently did so for the first time, permanently being in fear of something happening to it).

However, as far as the "software samplers don't actually sample" issue goes, I couldn't give less of a damn. Ok, it's nice to have some live sampling available, but these things are managed way more easily in dedciated hosts (such as Live) anyways. As far as plain sampling goes, I can't see a reason why recording something in whatever host and then instantly dragging the recorded file into your softsampler of choice should be so cumbersome. The benefits of working on a large screen, using my preferred host's (or audio editor's) interface to edit my samples weigh out everything by a fair margin. At least IMO.
The storage and total recall issue is another thing software wins hands down.

What's left is the actual quality of your sampling endeavours.
It's true that software samplers suffered of bad interpolation quality for quite a long time. Doing a patch that transposed/pitched well with just one single sampled note (or, say, one per octave) simply has been a nightmare in most softsamplers - until now. With computers getting faster and faster, you will most likely find high quality interpolation modes on most softsamplers by now, which don't cost your CPU an arm and a leg anymore.
The same might go for filters. The first software samplers (possible exception: Logic's EXS) featured horribly sounding filters. In addition, envelopes weren't what you'd exactly call "snappy". But then again: Software has went quite a mile during the last, say, 1-3 years. IMO these days you can find quite some decent software filters.

I have to admit that, others than borrowing an aging S1100 for a while, my hardware sampler experiences are close to non-existant, so I'm probably not the person to do any judgements on the matter, but apart from the backwards and OS compatibility issue (which indeed, let's face it, sucks big time - I'm just experiencing it, having switched from a PC to a Mac) and the issue of using things live, I can't see much of a reason to go hardware these days anymore.
For some other synthesis methods, hardware may still outperform software (hence the number of "what VSTi will sound like a Virus?" threads), but for sampling stuff, with the requirement of editing files and storing them somewhere, a well-equipped computer and an up to date softsampler are winning the race hands down. At least IMO. Your mileage may vary though...

[grillo]

I'd say go the software route, mostly for the workflow bonuses, and i mean multiple instances, easily edit almost anything, everything saved in the project, and that's a really big plus.

Just the idea of booting up my old emu sampler and having to load all the programs for the current project makes me cringe now.

[ckett]

Hollowsun

Thanks for your reply!

Sample interpolation will be very important to what I am doing. I have an Akai S6000 on the way this week and will be doing a sampler showdown between it and some software samplers. The Akai S5000/S6000 series is really the only hardware sampler I would consider. They have THE best interface of any hardware sampler and with the USB connection one can drag and drop samples back and forth with no problems. Almost like a software sampler.

I may be just a bit software/computer paranoid but there is nothing worse than to debug a computer software/hardware problem while trying to be creative. I am fully comfortable with building my own PCs and all. It just seems like having these nice dedicated hardware boxes doing one thing and integrating into a DAW with a simple Midi and audio cable seems much more reliable. No integration issues when upgrading your DAW OS or sequencer.

Well, I will see once I spend some time with both platforms.

On the discussion about whether samplers sample or not seems a bit goofy. I find using an audio editor on a DAW is far better to prep samples for the sampler. Never intended to sample directly into the sampler. So programs like Kontakt that don't "sample" isn't really that important of a factor.

Keep the input coming if you have more to share!

Thanks,

ckett

[kuniklo]

On the discussion about whether samplers sample or not seems a bit goofy. I find using an audio editor on a DAW is far better to prep samples for the sampler. Never intended to sample directly into the sampler. So programs like Kontakt that don't "sample" isn't really that important of a factor.

For example:

On the E-Mu samplers you can pop into sample mode, then sequentially sample a source into the sampler, just by hitting notes in series, and it will automatically map them to a range of keys, trim them, normalize them, and auto-loop them. In Kontakt the same procedure is a huge PITA.

On either the Akai or E-Mu you can take a sample, resample it in realtime while tweaking effects etc, trivially assign these to new keys, and have 10 variations of a sound mapped out in 30 seconds. On the PC this is a whole drag & drop ordeal with file management etc.

On the E-Mu you can convolve two arbitrary samples in-place with the touch of a button. Mix and match a couple of these and you've got a whole arsenal of crazy percussion effects ready to go.

If you haven't actually tried this workflow it's hard to appreciate that it's different than software but spend some time with it and you'll see. A huge part of FSOL and Autechre's sound comes from this kind of hands-on creative sample mangling and there's a reason they both still use hardware samplers to do it.

[Radek]

The Akai Z4 is kind of cheap now and it has 512MB capability. You might try to buy Yamaha A5000 as it has some interesting capabilities (every sample is a synth on its own) but can handle no more than 128MB of memory. Then there is FantomXR is both a "rompler" and sampler and can hold around 1GB of samples (both rom and ram). The MotifES isn't worth for its sampling capabilities unless you want all other its functions (as only keyboard versions can use ram memory). I don't know much about EMU samplers.

There is also a I/O perfomance and filesystem to consider with hardware samplers. They generally don't load samples too fast and sample management might not be exactly convenient. It's a pity those unit weren't really updates to more todays specs...