problems compressing OT libraries in Kontakt 4.1

[clif321]

I am getting error messages when trying to compress my OT libraries in Kontakt 4.1 using the lossless compression feature. The messages says "PANIC: could not write one or more samples! Probably the originals were removed?" In fact the uncompressed original samples are all there and the instruments work (and batch resave) fine with them. The samples just won't compress. Any idea why? I've used this Kontakt feature with other third party libraries without difficulty so I don't think the problem is with Kontakt or my setup more generally (Win7 x64).

Thanks in advance for any help.

[Gregjazz]

I've heard of a few libraries that have difficulty compressing, but that's strange--I'll check it on my system to see if I can get the same error. It might just be a bug with the current version of K4.1, so I'll try it on the K4.2 beta as well. It's still a pretty new feature, so I wouldn't be surprised if there are a few bugs to be ironed out.

[clif321]

Any update on this? Thanks.

[Gregjazz]

Which library/libraries in particular are not compressing right? Sorry, I forgot to ask.

[clif321]

None of the four CoreBass libraries would compress. By contrast, the Iconic Bass library compressed fine. The EEG library was mixed. The rhythm instruments compressed fine. Most of the sample subdirectories for the lead instruments also compressed fine, but several would not compress, which caused Kontakt not to copy over the lead instrument nki files into the target directory for the compressed instruments. However, I was able to construct usable mostly-compressed lead instruments by copying the nki files over manually along with the subdirectories containing the uncompressed .wav versions of the samples that would not compress and then batch resaving the nki files (after checking the box telling Kontakt to ignore sample format differences). The EEG sample subdirectories that would not compress were:

-- lead/miscellaneous(RAM)/resonance
-- lead/miscellaneous(RAM)/xEffects2
-- lead/miscellaneous(RAM)/xTransitions/UP 1
-- lead/Pre-pick(RAM)/single prepick
-- lead/Releases(DFD)/Open release [all six string subdirectories]
-- lead/Sustains(DFD)/xSoft sustain [all six string subdirectories]

A couple of other sample subdirectories also did not compress, e.g., lead/Pre-pick(RAM)/extras, but I think that was because the lead instruments do not actually use those samples (so far as I can tell, the Kontakt compression feature only tries to compress the samples actually referenced in an nki file, and leaves other samples alone).

Thanks again for your help.

[damstraversaz]

I had the same error message for CoreBass Pear with 4.1

Damien

[standalone]

Core Bass Cherry cannot be compressed here.

[Gregjazz]

I got the same result (an error) when trying to compress OTS libraries in the current version of Kontakt 4.1.3. However, it works fine using the latest beta, so it seems like that bug got ironed out, and you'll soon be able to use NCW compression with OTS libraries (I am going to be using it in future releases as well).

[damstraversaz]

great news !
thanks for checking this

Damien

[clif321]

Yes. Thanks. I've found one additional point to note and perhaps investigate, though. After reading your report I installed the Kontakt 4.2 beta and tried again compressing the four CoreBass libraries. The operation converted all the .wav files to .ncw format without any "panic" message about supposedly missing files and also created a corresponding new nki file instead of aborting that step with an error message in the log saying it couldn't save the new nki because of a "storage i/o error" (as happened under Kontakt 4.1). So far so good -- at least NI has fixed the problem of the operation aborting with an error message and failing to produce a usable instrument. But when I checked the size of the newly created instrument libraries I found that Kontakt hadn't really compressed them even though it changed the sample file format. The Corebass Cherry libraries were essentially the same size as before, and the Corebass Pear library was only marginally smaller, nowhere close to the 40-50% compression I've seen with most other libraries. So there still seems to be something about the samples that makes them not want to compress in the NI format. I'd be interested to know whether others get similar results.

[Gregjazz]

Yeah I was surprised--I wonder what the conditions of the compression are. In some libraries I've definitely got 40-50% before, but apparently it depends on some factors...