I made a new test to find out at which levels of the incoming signal which step of the sampled cab volume gets played back. The Dynamic template at 6 numbers of 1dB steps. The loudest sweep that ran through the gear peaked at around -1.5 dB, so the last sweep peaked at around -7.5 dB. - Repeats, Space, Cut= 1 - Algorithm: Enhanced - 64 bit - Mono Stone enabled with phase locked 5 Kernels
To see if the dynamics really worked I used a trick. For the first and loudest measurement I have chosen a different microphone position that sounded very different to the normal position. For the 5 remaining sweeps/measurements the mic was on the normal position. (I changed that on the fly)
Then I opened the program in Nebula and ran a signal through it. I changed the volume of that signal from very low to very loud and wanted to hear when the tone would change to the different mic position in the loudest sweep. But it didn't happen! From -40 to +20 dB the sound was always the same. I really don't understand why the loudest step wasn't played back on any level of the incoming signal.
When I made that test with the quietest sweep sounding different I also couldn't hear any difference in sound in the whole volume range of the signal that ran through Nebula.
Maybe you can help or send me a NAT session that will work for sampling a mic'ed guitar cabinet powered by a tube amp. Or tell me what might have been wrong in my session.
Hello again. Some of the problems got solved and I got it to move into the loudest step. BUT its only possible with ridiculous amounts of input signal. It needed atleast +5 dB (!) to get the loud step "triggered". I tried everything in NAT it seems. But I can't make the loudest sweep louder than 0 dBfs obviously. Maybe Enrique, you can help me again how to set down the loudest step kinda. Maybe its possible to have Nebula set up the way to shift dynamic steps in playback? Thanks a lot in advance!
i've been kind of thinking that it would be nice if, in NAT, we had control over the lengths of kernels 6-10 separately, as well as microkernel switches for each. also i'm wondering why kernels over 10 have to be microkernels. it isn't ever possible to get decent results with standard? -tim
I made an own test with white noise for myself. The loudest sampling step sounds a lot brighter than the others. You will hear it. And I found out that it is reached when I put in +7dB of white noise. That is of course a too high volume, even Nebula isn't distorting with so much input level.
Long story short, I would like to reach the loudest NAT sampling steps with lower input level on the Nebula playback side. Ideally around -10 dB Thanks so much for spending your time!
you can always adjust the padin and padout parameters to achieve that. open your program, click 'edit' then go to the 'glob' section. if you want it to distort at lower inputs, you would increase the 'padin', and then you might decrease padout to compensate.
best thing would be to open it with something like vstanalyser, to make sure that the end result is close to 0db. then click 'save' to save your changes.