Starting playback from loop offsets in DFD mode

In DFD mode it's not possible to specify a sample offset at which to start playback. This restriction is however unnecessarily limiting. Kontakt already keeps preload buffers corresponding to the start offset of loops, so in addition to offset 0 it should also be possible to start playback from any such loop offset. This would enable scripts to skip the attack of notes which is a very important component in scripts for legato/portamento, polyphony-saving techniques for crossfading between velocity layers, etc. Currently this is only possible in Sampler mode which has much higher RAM requirements.

I propose that KSP is slightly extended to allow for this:

• Eight builtin-variables are introduced - $LOOP1, $LOOP2, ..., $LOOP8 (having values -1 through -8).
• In DFD mode the play_note function will accept any of these variables as its sample-offset parameter and the playback will start from the corresponding loop offset (or zero in case the loop isn't active).

 

Agree with Kotori!

Thanks!

 

As the one who started this current discussion I of course strongly wished if you could implement this feature soon. I managed to get nicer string legato sounds than the recorded legato transitions by VSL, but it is quite unusable since the RAM requirements suddenly go way over the top.

The legato script is THE killer feature of Kontakt for orchestral composers. This missing bit would make it perfect.

 

Thanks for the support guys. I really hope NI will implement this because I think very little work would be required whereas the gain would be significant. For example, imagine loading a script on any sustained looped instrument, specify the number of velocity layers and get instant mod-wheel controlled velocity layer crossfading - at a lower polyphony than traditional crossfading patches.

 

It's pretty amazing that such an important feature isn't implemented yet.

>> I propose that KSP is slightly extended to allow for this[..]

With Independence I can set ANY offset in streaming mode and without a glitch.

If NI can't implement this, it may be time for them to start a donuts business.

 

Maybe for efficiency, ksp could have a function to let k2 know the offset range the script may use.
For modifiers, k2 already has this info.

But please do something as this is a must have for all wind, string (included guitars and basses) instruments, and many others.

It's also a quick way to have some random alternates for samples with strong attacks.

Even if we have to raise a little the dfd settings it's still better than being limited to a small quartet.

Marc

 

My personal situation is such that I can usually use Sampler mode but, I can certainly sympathize with the need to use DFD mode and it's a shame that at least some limited form of sample start offset can't be made to function in DFD mode.

The offset times required to get past the attack portion are usually not too large. So, maybe one way to do this would be for NI to allow offsets up to say 50% of what is in RAM. When an offset in excess of that is asked for, just limit it to that max. Then the user could adjust the DFD settings so that the prefetch was big enough to handle the max offset needed. The RAM needed will probably increase somewhat but not nearly as much as in Sampler mode where the whole sample has to be loaded in RAM.

Anyway, I cast my vote that something be done to provide at least some limited form of sample offset capability in the DFD mode.

Bob

 

Sorry, Bob, to disagree here. My attack offsets are usually some hundred milliseconds. The strings have almost a second attack offset. In that area is usually also the loop point, so still that would be the best idea, I think.

 

Kotori's original feature request has been submitted into the (long ) feature request list.

 

or, if it's possible, you could just load a larger amount of each sample in RAM when using DFD and sample offset.
maybe a feature in the script that will allow the user to load a predefined amount of the sample length in RAM, as needed by the particular application...

 

Or maybe a slider in the source section for a max sample start offset, so we can use normal modulation as well.
And eventually a script function to query this value.
Then just increase pre-load by this amount.

This would also make it easier, even in sampler mode when using modulation -> sample start, as of now, I still didn't find how to set up the slider so that I have a 0 to 100 msec offset with a 0-127 CC range

 

I guess K2.2 didn't bring any news to this, right?

 

No, and I'm very disappointed.
I need it on almost every instrument that I load. Either with a random generator, or connected to a foot pedal or with some scripts.

Well, at least I have Independence for that, as most of my patches are huge and sampler mode is not an option.

Too bad not many people understand the potential of such a feature

 

Any chance of this being implemented correctly in the next update to K3?

Justin M