Relationship between sampling Impulse, Kernel length +RTE

Tips & tricks, working results, technical support
Post Reply
Gstell
User Level II
User Level II
Posts: 23
Joined: Sun Jul 24, 2011 5:32 am

Relationship between sampling Impulse, Kernel length +RTE

Post by Gstell » Sun Jul 14, 2013 8:14 am

Having read through the Acustica-audio Forum for a few weeks I am still not sure of following:

What are the relationships between:

NAT sampling impulse length
Nebula3 Program rate (RTE)
Nebula3 Kernel length

1. If program rate is tied to Kernel length does that mean they can't be adjusted independently or are they the same thing?

2. Does RTE in Kern page show us the Kernel length of a particular program?

3. In the Kern page of Nebula3 do TIMED/FREQD, clean, even and odd options allow one to adjust the processing algorithm regardless of Kernel length?

For example if a Kernel of a loaded program has a length of 40ms and I set the parameters in Nebula3 in the Kern page to Classic + normal selecting TIMED for Clean at 20ms, does that mean the Kernel is being truncated (I understand using splith that then from 20ms on the kernel will be processed with Freqd. But if freqd is set to only 5ms is the kernel being truncated)?
What if I set TIMED to 60ms for a Kernel length that is 40ms does that just waste CPU resources needlessly?

Thank you all for your patience

User avatar
enriquesilveti
Expert
Expert
Posts: 2858
Joined: Sun Mar 28, 2010 9:00 pm
Location: Lodi | Madrid | Buenos Aires
Contact:

Re: Relationship between sampling Impulse, Kernel length +RT

Post by enriquesilveti » Tue Jul 16, 2013 11:06 am

What are the relationships between:

NAT sampling impulse length
Nebula3 Program rate (RTE)
Nebula3 Kernel length
You setup NAT sampling kernels length and Nebula reads the kernels length.

About Nebula3 Program rate (RTE) read: http://www.acustica-audio.com/phpBB3/vi ... f=11&t=899 and http://www.acustica-audio.com/phpBB3/vi ... 39&t=25012
2. Does RTE in Kern page show us the Kernel length of a particular program?
No.
3. In the Kern page of Nebula3 do TIMED/FREQD, clean, even and odd options allow one to adjust the processing algorithm regardless of Kernel length?
RE: Yes it's the hibrid engine that mix direct and FFT convolution in real time.
For example if a Kernel of a loaded program has a length of 40ms and I set the parameters in Nebula3 in the Kern page to Classic + normal selecting TIMED for Clean at 20ms, does that mean the Kernel is being truncated (I understand using splith that then from 20ms on the kernel will be processed with Freqd. But if freqd is set to only 5ms is the kernel being truncated)?
What if I set TIMED to 60ms for a Kernel length that is 40ms does that just waste CPU resources needlessly?
I don't know I truncate this the correct word for this process. In a perfect world you should only use TIMED but in other to use less CPU you can use SPLITH and use both TIMED and FREQD, depending the emulation preset and the sound 10 to 50 millisecond should be enough to process the attack of the sound with TIMED.

http://www.acustica-audio.com/phpBB3/vi ... =11&t=1846
http://www.acustica-audio.com/phpBB3/vi ... =25&t=1859
Enrique Silveti.
Acustica Audio customer and technical support.

MBP 11.5 (i7-4870 | 16 GB | 512 SSD)
SP4 (i5-6300 | 8 GB | 256 SSD)
UFX | Lyra2 | USBPre2
VM U16 | VM Win10 CU | VM OSX 10.12
N4/NAT4 | SPX3 | RX6 | LN2C | Smaart8 | R5 | PT12 | PX10 | NIK5

Post Reply