IKM SampleTank 3 (ST3) Now Available

[neverenoughfunk]

just spoke to Sweetwater... they say Sept 15 for the box version...

Wow! They keep pushing it further back. Then again, maybe that's good news, as Komplete 10 should be announced by then. If it looks that much more enticing, we can cancel the yet unshipped preorder at Sweetwater and jump on K10.

i like your thinking!!!

[Rajah]

Groove three has come out with a viddy in case anyone is interested.

Thanks, got it in my mailbox today... Looks like it's chuck full of lessons to learn from...

[dmbaer]

I have read through the documentation carefully several times and have a few questions. I apologize if these have been answered previously in the 170 pages of this thread, but I don’t have a spare week to read through it start to finish.

I infer from the documentation that there is a one-instrument-to-one-sample-set relationship, given that there is no place on the instrument edit page to specify a set of sample files. Now, suppose one wants to take an existing instrument from the factory content and make a custom version with different effect settings, etc. When you do a save-as, does that duplicate the sample content or is it smart enough to keep that down to a single set of samples?

ST 2.5 allowed the user to load sample sets with velocity layers. There’s no mention of that capability in the ST 3 doc. Has this feature been removed?

Does the user have any capability in doing round-robin instruments from user-supplied samples? What about setting up multiple-articulation instruments?

How does ST3 decide on key ranges for user-supplied sample sets? Is there any way for the user to control this?

Is there any way for the user to extend the key range higher or lower for a factory-supplied instrument?

Thanks in advance for any insights.

[mhog]

(1) it's smart enough.

For the rest:

(2) "removed" (as far as I know). It works only if you import ST2 "legacy instruments". If you import your samples directly (having them different velocities) ST3 does not recognize them. No way, as far as I know.
(3) no capability in round-robin and/or multiple-articulation instrument.
(4) no way to control key ranges for user-supplied sample sets.
(5) no way for the user to extend the key range.

(but I hope I may be proved wrong )

[blueman]

I like as low as i can when playing my electronic drums. I used to use 32 samples with my RME PCIe.

It's generally accepted that latency under 10ms is considered realtime. You're probably not gaining a thing by working at 32 samples. When you're working at 44.1, 128 samples is already under 10ms latency.

Fwiw, I don't want IK Multimedia to waste time on making this work at latencies that most people are not using (i.e. < 128). I took the latency-obsessed trip for a while; it was a huge waste of time and energy for me--spending all my efforts tuning my system instead of enjoying music 128 is the sweet spot

It's fine that certain folks like to tweak their systems down to the lowest latency possible; does this represent the majority of users? Doubtful! That said, yes, IKM should state it up front.

Milliseconds (ms) are said to be roughly equal to feet. If 10 feet of distance is not acceptable then how are musicians able to perform live? Do you really need to be 2-4 feet from the sound source to perform? Sure, it's nice but I would place ultra-low latency in the luxury (not needed) category.

Consider this statement, found here:

"Actually, research sponsored by the National Science Foundation, through the Stanford University Department of Music, has shown that performers in an ensemble have no problem synchronizing with each other while experiencing latencies as high as 40 ms and even greater. In fact, latencies in the 10 ms to 20 ms range actually have a stabilizing affect on tempo and are thought to be preferred over zero latency"

[Bombadil]

With ST3, 128 is iffy for me, and depends on how many parts are loaded. OTOH, 128 manages other CPU intensive instruments quite well. I am using a Firewire Duet, which doesn't seem to like the Convo-room effect.

[MickGael]

Fwiw, I don't want IK Multimedia to waste time on making this work at latencies that most people are not using (i.e. < 128). I took the latency-obsessed trip for a while; it was a huge waste of time and energy for me--spending all my efforts tuning my system instead of enjoying music 128 is the sweet spot

The lowest possible latency is not "a trip" for me: it is a necessity when tracking. I primarily record live instruments with microphones, and an RME buffer size of 128 typically introduces an unusable amount of latency. Period. Why is this so hard to understand? And why do people insist on questioning this?

[Bombadil]

Fwiw, I don't want IK Multimedia to waste time on making this work at latencies that most people are not using (i.e. < 128). I took the latency-obsessed trip for a while; it was a huge waste of time and energy for me--spending all my efforts tuning my system instead of enjoying music 128 is the sweet spot

The lowest possible latency is not "a trip" for me: it is a necessity when tracking. I primarily record live instruments with microphones, and an RME buffer size of 128 typically introduces an unusable amount of latency. Period. Why is this so hard to understand? And why do people insist on questioning this?

I would track my instruments at lower latency, and if there's anything that doesn't like that, I'd freeze the tracks. I track vocals and guitars @ 64.

[MickGael]

I would track my instruments at lower latency, and if there's anything that doesn't like that, I'd freeze the tracks. I track vocals and guitars @ 64.

I can work around the edges (even if it means more work for me) but will say it again:

ST3 is the only - the only - plugin I have that crashes at the lowest latency WITH NO INSTRUMENT LOADED AT ALL. Surely, that is something IK has an interest in fixing?

Hell, even Diva (which beats the hell out of my CPU) does not does that.

[blueman]

Fwiw, I don't want IK Multimedia to waste time on making this work at latencies that most people are not using (i.e. < 128). I took the latency-obsessed trip for a while; it was a huge waste of time and energy for me--spending all my efforts tuning my system instead of enjoying music 128 is the sweet spot

The lowest possible latency is not "a trip" for me: it is a necessity when tracking. I primarily record live instruments with microphones, and an RME buffer size of 128 typically introduces an unusable amount of latency. Period. Why is this so hard to understand? And why do people insist on questioning this?

I can only speak for myself I didn't find it worth the effort at all. I would rather make performance adjustments, and focus on other things. I find the kind of tweaking that's required for getting < 128 laborious and tedious with only marginal returns, for the effort involved. Of course, ymmv; at least, you will find plenty of people on this forum to agree with you.

[Bombadil]

I would track my instruments at lower latency, and if there's anything that doesn't like that, I'd freeze the tracks. I track vocals and guitars @ 64.

I can work around the edges (even if it means more work for me) but will say it again:

ST3 is the only - the only - plugin I have that crashes at the lowest latency WITH NO INSTRUMENT LOADED AT ALL. Surely, that is something IK has an interest in fixing?

Hell, even Diva (which beats the hell out of my CPU) does not does that.

No argument from me there. My system doesn't particularly love ST3, either. Until it does function properly within my system, it isn't going to play any part in my recordings. As is, it is a buzz-killer.

[blueman]

ST3 is the only - the only - plugin I have that crashes at the lowest latency WITH NO INSTRUMENT LOADED AT ALL.

Well, you do have a point there; however, I think you should challenge this notion that you need to be at the lowest latency. I cannot be there to see what you are experiencing in your environment. Perhaps there are other things adding latency?

For both timing and frequency balance, 10-15ms is the sweet spot for total latency--university research and science support this claim. I do think there is a strong mythological (psychological) affect around this topic that gets into people's heads. It got into mine, and that is what I meant by the word "trip"

[BBFG#]

Do we have the demo to try yet?

[gmontano]

Even though I also ride the edge of low latency for vdrum tracking. I wouldn't even care if st 3 officially didn't support the low latency, but it hard crashes cubase and the standalone when the sound card is at set low. Debate what you want about the needing for low latency, but ST 3 shouldn't crash your daw. Playing fast 16th or 32nd not patterns on the hihat benefit from low latencies. Agreed for playing a keyboard or guitar sim, 64-96 or even 128 may be ok for me. Not so with with my drums. The issue is when tracking drums, I can't do it with a ST3 scratch track. I can with Halion or Kontakt or anything else that exists. If the nay sayers or IK (not that IK has said this) think hard crashing is completely acceptable, then there really isn't much else to say.

[Uncle E]

"Actually, research sponsored by the National Science Foundation, through the Stanford University Department of Music, has shown that performers in an ensemble have no problem synchronizing with each other while experiencing latencies as high as 40 ms and even greater. In fact, latencies in the 10 ms to 20 ms range actually have a stabilizing affect on tempo and are thought to be preferred over zero latency"

It's an interesting point. Certainly, playing live with an ensemble is not the same as monitoring over headphones in this regard.