St. Rock AMPERIUM LIVE

Funny you say that, since I got Kristian Kohle's Rainbows and Chainsaws IR pack the other day as it was on sale (less than 20 € for close to 30 (hopefully great) IRs is a no-brainer). I've already uploaded them to my Amperium but I have yet to try them. But I did try the two IRs from that pack that were available in the demo/trial version, and they sounded pretty good with almost no tweaking...which Bogren IRs' have you tried, and which ones do you recommend, Exluthier? Do they sound good with "clean" amps/amp channels?
 
just a question regarding the NNM Hub - there are profiles that say Tonex2NNM - what does this mean? are they Tones profiles that have been converted to NNM? or a profile of a profile?
I like the idea of the Amperium with the modelling and profiling but there isn't as much user base say compared to Tonex or NAM - but if it could play either profiles of these formats.
 
Yeah, considering NAM is open source, what's the chances of Amperium Live being able to load NAM profiles?
I`d say format of NNM is also fully open, because training is performed using not encrypted python scripts, you can check details on github.
Only profile-loading part (Amperium Live and Amperium NNM plugin) is closed source.
But someone can load NNM file and use, for example, RTNeural, it will be not as efficient as Amperium NNM, but can work.

Standard NAM (as it exists now) uses very computationally heavy neural network.
Lite variants will require retraining. So it`s easy just resample NAM test signal and reamp to 44.1k and train NNM instead.
But, formally, we can support NAM in LSTM form (using old training script) if training is performed using 44.1k signal pair. But no one will train such networks.

Neural networks should be used on training sample rate only or realtime resampling is required. But quality realtime conversion from 44.1k to 48k and vice versa will add significant latency. NAM using 48k and Amperium Live using 44.1k.
 
Last edited:
I`d say format of NNM is also fully open, because training is performed using not encrypted python scripts, you can check details on github.
Only profile-loading part (Amperium Live and Amperium NNM plugin) is closed source.
But someone can load NNM file and use, for example, RTNeural, it will be not as efficient as Amperium NNM, but can work.

Standard NAM (as it exists now) uses very computationally heavy neural network.
Lite variants will require retraining. So it`s easy just resample NAM test signal and reamp to 44.1k and train NNM instead.
But, formally, we can support NAM in LSTM form (using old training script) if training is performed using 44.1k signal pair. But no one will train such networks.
I think Steve from NAM said he was looking at resampling but not sure if that means it can be trained a lower sample rate.

So what does the Tonez2NNM mean from the NNM hub?
 
Funny you say that, since I got Kristian Kohle's Rainbows and Chainsaws IR pack the other day as it was on sale (less than 20 € for close to 30 (hopefully great) IRs is a no-brainer). I've already uploaded them to my Amperium but I have yet to try them. But I did try the two IRs from that pack that were available in the demo/trial version, and they sounded pretty good with almost no tweaking...which Bogren IRs' have you tried, and which ones do you recommend, Exluthier? Do they sound good with "clean" amps/amp channels?
I have the low tuned pack, I'm playing a seven string tuned down a whole step and I like the Bogren IRs better than the OH low tuned ones I have. The Bogrens seem more mix ready. So of the pack the two I like the best was one called Bogrenbalanced or something like that, and another that has Swede in it but I can't remember the exact name. That's the one I'm going to send you the FOH when I gig.
 
I have the Bogren Rhythm pack and they are quite nice, even though the style of Bogren of more on the metal side - they do work across genres
 
Thanks for the info, guys! I'll do some experimenting of my own with the pack that I have and I'll let you know about the results.
 
hey guys, is it possible to reamp with this? i see that usb audio has been removed in a previous version of the beta - just want to confirm/see if it can or can't be done and if it can how?
 
Do you have the Amperium and the Shiva?
I have played thru the real Shiva, but I do not own the Amperium right now. I want a modeler that can nail down the Bogner tones well because that's all I use to record most of the time and I would like to have an actual modeler that can replicate the Bogner tone live at a gig. I saw other companies that have Ecstasy plugins (Mercuriall obviously) and they did a great job, I hope the Amperium can deliver that tone
 
I have played thru the real Shiva, but I do not own the Amperium right now. I want a modeler that can nail down the Bogner tones well because that's all I use to record most of the time and I would like to have an actual modeler that can replicate the Bogner tone live at a gig. I saw other companies that have Ecstasy plugins (Mercuriall obviously) and they did a great job, I hope the Amperium can deliver that tone
We have a few revisions of extasy.
Shiva looks like jcm800 mod. We would to add it. But how you will test it? -)
If you were an Amperium user and have a Shiva, then we would be able to add this amp for test and compare..
But for now we have more interesting tasks -)
Thx for understanding.
 
Last edited:
Back
Top