Jump to content

SmartAmpPro


marlowg01

Recommended Posts

32 minutes ago, Esteban Villanova said:

There’s another version called SmartGuitarAmp which is way easier to install. If you’re not trying to create your own profiles I think that’s the better option.

The regular installation for just using the plugin doesn't require you to install the Python dependencies. It's just a VST3 file you drop at your plugins location.

Link to comment
Share on other sites

7 hours ago, Bruno de Souza Lino said:

The regular installation for just using the plugin doesn't require you to install the Python dependencies. It's just a VST3 file you drop at your plugins location.

Oh, ok. I think the Linux version needs to be compiled. How is the cpu usage? I still haven’t tested it yet, but I read it’s high.

Link to comment
Share on other sites

1 hour ago, cclarry said:

I certainly wasn't impressed...but that's just me....

So what impresses you?  I've actually wanted to ask you that for a while.  I'm guessing UAD or Kemper, but even more curious about pure software solutions worth raving about.  I enjoy the SmartAmpPro, Nembrini, Neural Granophyre, and surprisingly, Toneforge's Jason Richardson plugins.

Link to comment
Share on other sites

36 minutes ago, marlowg01 said:

So what impresses you?  I've actually wanted to ask you that for a while.  I'm guessing UAD or Kemper, but even more curious about pure software solutions worth raving about.  I enjoy the SmartAmpPro, Nembrini, Neural Granophyre, and surprisingly, Toneforge's Jason Richardson plugins.

UAD Amp Sims or "ok" IMO!  S-Gear is impressive, so I'm really anxious to see where version 3 lands.  (and when!)
The new Amplitube 5 is a huge improvement over A4, which to me was really "flat" sounding (no/very little life).  NeuralDSP
is impressive for mix ready tones, and I have the Jason Richardson, but haven't really sat down with it yet. Even  the new 1.3
iteration of TH-U has taken a big step forward.  This was kind of "eh".  Will it work?  Probably.  Is it "wow", not really.
Still sounds "Fizzy and thin" to my ears...but, as I said, that's just me.  I only went through the presets.  Cleans were decent.
Guitar tone is REALLY subjective...so that's a big part of it.  The only ones that made me go WOW right out of the gate
were S-Gear and NeuralDSP.

  • Like 1
Link to comment
Share on other sites

2 hours ago, cclarry said:

UAD Amp Sims or "ok" IMO!  S-Gear is impressive, so I'm really anxious to see where version 3 lands.  (and when!)
The new Amplitube 5 is a huge improvement over A4, which to me was really "flat" sounding (no/very little life).  NeuralDSP
is impressive for mix ready tones, and I have the Jason Richardson, but haven't really sat down with it yet. Even  the new 1.3
iteration of TH-U has taken a big step forward.  This was kind of "eh".  Will it work?  Probably.  Is it "wow", not really.
Still sounds "Fizzy and thin" to my ears...but, as I said, that's just me.  I only went through the presets.  Cleans were decent.
Guitar tone is REALLY subjective...so that's a big part of it.  The only ones that made me go WOW right out of the gate
were S-Gear and NeuralDSP.

So, the thing you're stressing the most about is impulse responses? Cause that's 99% of the "impressive" amp sims nowadays. There's also the hidden plugin oversampling you can't turn off, which gets rid of aliasing artifacts and such.

Link to comment
Share on other sites

Thanks for sharing!  I feel the same way with amplitube and THU, but didn't really give S-Gear much attention.  I forgot to mention Revalver 4, Vandal, and Mercuriall, which I also respect.  And I also agree with Bruno de Souza Lino, that the cab part is crucial.

  • Like 1
Link to comment
Share on other sites

If we talk about cab sims, check this free pack from Jason Freeman here https://www.thegearpage.net/board/index.php?threads/my-framus-2x12-vintage-30-g12h-75-creamback-g12h-150-redback-g12h-30-anniversay-irs.2220503/

His impulses are IMO a bit darker overall  but I really like some of them (even more than from OH)

Edited by filo
  • Like 2
Link to comment
Share on other sites

4 hours ago, marlowg01 said:

Thanks for sharing!  I feel the same way with amplitube and THU, but didn't really give S-Gear much attention.  I forgot to mention Revalver 4, Vandal, and Mercuriall, which I also respect.  And I also agree with Bruno de Souza Lino, that the cab part is crucial.

I have all of those also, and Presonus Ampire, and a bunch of Free ones.  If you want to do the work you can get most of them 
to sound decent.  The issue becomes "how much time do I want to spend to get where I want to be", and when "Time is Money"
Less is MORE!

  • Like 1
Link to comment
Share on other sites

One plugin I'm starting to use quite a bit is AXP's SoftAmp PSA, which emulates the Sansamp PSA-1 rack preamp with some extra features. One recommendation is downloading the PSA-1 manual and studying it a bit. The controls are not something you'll learn by just tweaking, but you can get a lot of tones from this plugin. I also suggest using the cabinet simulation part of it as a global EQ and pairing it with a better cabinet simulator (I use Torpedo's Wall of Sound).

With good impulse responses you can even ignore amp altogether and just use drive pedals and such like amps.

Link to comment
Share on other sites

6 hours ago, Monomox said:

50% CPU use, which doesn't sound so bad based on what you guys are saying (I have a Ryzen 5 2600).

The CPU Passmark of your Ryzen 5 2600 is 13216.  I have Intel Core i7-4790K (passmark 8000) and what I remember the Neural Nolly needs about 7-8% of my CPU.

Are you really sure about mentioned 50% CPU use for one instance of SmartAmpPro?

Link to comment
Share on other sites

1 hour ago, filo said:

The CPU Passmark of your Ryzen 5 2600 is 13216.  I have Intel Core i7-4790K (passmark 8000) and what I remember the Neural Nolly needs about 7-8% of my CPU.

Are you really sure about mentioned 50% CPU use for one instance of SmartAmpPro?

My guess is that all Neural DSP plugins are optimized for Intel CPUs. I was starting to suspect why they were so slow to respond on my system, despite a good portion of the tones being reproducible in other plugins that used a fraction of ethe CPU.

Link to comment
Share on other sites

  • 2 weeks later...

Hey! Developer here, thanks for all the feedback and models!  High cpu usage is currently the trade-off for the flexibility of copying pedal/amps.  There's alot of number crunching behind the scenes that allow for the same model structure to produce a wide range of tones (as opposed to optimizing for one specific sound). From what I've heard from the Quad Cortex, my SmartAmp plugins are not quite there in terms of accuracy to the target hardware, but hoping to improve it.  Also, the SmartAmpPro is intended to be used with reverb or IR's, so if you're running it by itself it will sound dry.

Here's a model I made from recordings of the Gojira plugin; which definitely gave the capture some difficulty. The result is a pretty good sounding distortion model, but not 1:1 with the Gojira sound. Enjoy!

gojira_test_model.json

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

1 hour ago, Keith Bloemer said:

High cpu usage is currently the trade-off for the flexibility of copying pedal/amps.

But is that high CPU load necessary if you're not profiling the stuff?

Or better yet.

How much of that high CPU usage actually translates into audible differences?

Edited by Bruno de Souza Lino
Link to comment
Share on other sites

For this plugin, yes it’s necessary. It only runs on profiles, even the five included tones. The profiling works by training an artificial intelligence model, which basically amounts to adjusting a bunch of number values (you can see the values  if you open the .json files in a text editor). The values are adjusted during the profiling until they can make the input signal get as close as it can to your target signal. The resulting “trained” model is the loadable .json file.

I expect that as this tech develops it will get faster and more accurate, but as with any modeling it is still an approximation. I could make the model size smaller, which would reduce cpu usage, but it would sound worse. 

  • Like 1
Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...