Do you think modelers will get there in the next 10 years?

  • Thread starter Thread starter Tone Monster
  • Start date Start date
BatmansRigTalk said:
The main barrier to replicating an amplifier by a computer was in realtime processing of all those components without lag. When processing power got good enough it was then up to sound engineers to program systems to deep fake the real deal.
Click to expand...

Digital aliasing is still a problem to varying degrees with every commercially available modeler/profiler.
The processing hardware required to do an optimum job is simply not cost effective for any production unit.

Doesn't matter how good the sound engineer/programmer if he's working with limitations in processing
due to costs.
 
[ Donnie B. ] said:
Digital aliasing is still a problem to varying degrees with every commercially available modeler/profiler.
The processing hardware required to do an optimum job is simply not cost effective for any production unit.

Doesn't matter how good the sound engineer/programmer if he's working with limitations in processing
due to costs.
Click to expand...
Sorry, but can you show me where you are getting this information from?

Many years ago nearly all instances of this were a bug or using 96K sample rates where a Kemper was part of a larger complex DAW system which was solved by reducing the sample rates because 96K is totally placebo and all you need is 44.1

Today the least spec digital units can accommodate 6 to 8 blocks of anything (without more than two amps) not including noise gate or global EQ settings which are two more extra. You can easily have a Wah - Compressor - Distortion - Filter - Amp - IR - Delay - Reverb setup that isn't going to produce artifacts.

More importantly, what happened with Kemper was that the programmers figured out ways to process guitar sounds with algorithms that didn't demand supercomputer processors. We see this happening with game programmers for console systems after a few years. The graphics get better not because the console is better but because new algorithms (new math) do things better and faster.

Even the analog world does this. For example, Marshall has managed to get both the sound of a 100W Plexi and JCM800 into their 20W studio series. You can't even tell the difference when recorded and the only way to know the difference is with volume when those 100W amps opened up fully. The 100W are basically louder.
 
Indeed, aliasing is an artefact, but not an arty fact when it comes to guitar->amp->cab processing.

I know of no situation in which frequencies approaching 20kHz are desirable, necessary or even present to an audible extent when cabs are mic'd, and for those who don't know, it's the brick-wall filtering beyond 20kHz that causes frequencies to fold back down into the audible spectrum as aliasing.

IOW, whichever frequencies are present beyond 20kHz, not only would they be super-low-level, but once reflected back down into the audio spectrum, would likely be around the -80dB -> -100dB area, more-likely at the lower end of the range.

So, if you can hear it whilst playing, monitoring or just listening to a guitar part at -10dB or even higher, you have Superman Ears™.

All this said, Christoph Kemper has said that the algorithm for simulating the tubes runs at a greater than 700kHz sampling rate. That should be case closed.
 
BatmansRigTalk said:
Sorry, but can you show me where you are getting this information from?
Click to expand...

C.C. explained it as a current limitation of the AxeFX system in a PM to me about 2 years ago.
I had called it digital 'smear' and he was the one that explained it as aliasing and about the cost factor.
I figured he's the expert.
 
Spaceboy said:
holy shit y’all got weird in here
Click to expand...
That happened the moment people started to claim magical thinking over some mystery physics we don't understand while engineering amp circuits that they claim can't be modelled today by more sophisticated and more complex computers and even more sophisticated algorithms. Apparently the superhuman of hearing among us are now down to talking about software bugs or recording in ridiculous sample rates as the major differences between profilers and the real deal.

I'd actually still love to watch just one single youtube video of the golden hearing person who apparently exists and can tell the difference between real deal or profiler in double blind tests. The one who sends back their profiler because it just don't sound the same. 😆😆😆

I'm calling people out on this one as I learn more each time they get called out like they have been for the past decade.

I think it would be a hoot to take the most entrenched anti-profiler person around who says their rig can't be duplicated and then do it and run 20 comparisons to eliminate random chance guessing correctly and watch their faces after getting the results.
 
BatmansRigTalk said:
That happened the moment people started to claim magical thinking over some mystery physics we don't understand while engineering amp circuits that can't be modelled today by more sophisticated and more complex computers and even more sophisticated algorithms. Apparently the superhuman of hearing among us are now down to talking about software bugs or recording in ridiculous sample rates as the major differences between profilers and the real deal.

I'd actually still love to watch just one single youtube video of the golden hearing person who apparently exists and can tell the difference between real deal or profiler in double blind tests. The one who sends back their profiler because it just don't sound the same. 😆😆😆

I'm calling people out on this one as I learn more than when they get called out last time like they have been for the past decade. I think it would be a hoot to take the most entrenched anti-profiler person around who says their rig can't be duplicated and then do it and run 20 comparisons to eliminate random chance guessing correctly and watch their faces after getting the results.
Click to expand...
I have difficulties with recording since they are all compressed, but I’m willing to bet my discerning ears can tell the difference in the room while playing through different tube and digital amps.

Hence my comment earlier. I’m thinking a digital preamp with a monster tube power amp. That would likely meld the best of both worlds.... both in the room and recording.
 
psychodave said:
I’m willing to bet my discerning ears can tell the difference in the room while playing through different tube and digital amps.
Click to expand...

Me too. But I guess we just don't get the math.
Same with the dozens of paid engineers working for all of the modeling companies for the past decade.
Bunch of dullards.

Controlled A-B testing results are not the be all end all when it comes to this stuff.
Tons of data on why this is especially true with audio. This is a good one.

 
Last edited:
psychodave said:
I’m thinking a digital preamp with a monster tube power amp. That would likely meld the best of both worlds.... both in the room and recording.
Click to expand...
I have the AFXIII. I've tried it with a massive tube power amp (Mesa M180), but I prefer it with a massive solid state power amp (Crest CA-9; thing feels heavier than the damn Mesa).

I sounds very good either way, though. The Crest has a bit more slam and is faster feeling than the Mesa.
 
cardinal said:
I have the AFXIII. I've tried it with a massive tube power amp (Mesa M180), but I prefer it with a massive solid state power amp (Crest CA-9; thing feels heavier than the damn Mesa).

I sounds very good either way, though. The Crest has a bit more slam and is faster feeling than the Mesa.
Click to expand...
I’ve not heard of Crest. How would it compare to an HH V800 solid state power amp? The HH is also crazy heavy, more so than I think I remember my Strategy 400 being
 
braintheory said:
I’ve not heard of Crest. How would it compare to an HH V800 solid state power amp? The HH is also crazy heavy, more so than I think I remember my Strategy 400 being
Click to expand...

Crest made PA amps back in the day. A lot of bass players adopted them into their rigs (like Rex Brown), and I've been running one with a bass rig for a bit but moved it over to the AFX, which is doing bass duty now too.

Something about the CA9's feedback design makes it really great for instrument playback.
 
cardinal said:
Crest made PA amps back in the day. A lot of bass players adopted them into their rigs (like Rex Brown), and I've been running one with a bass rig for a bit but moved it over to the AFX, which is doing bass duty now too.

Something about the CA9's feedback design makes it really great for instrument playback.
Click to expand...
Sounds cool, I’ll have to check one out. I’ve already been using their toothpaste for years
 
Not here to start an argument but only as observing from the other side of modelers and real amplifiers:
I actually love getting ready for a gig with 2-4x12 cabs,amp heads, and its accessories. Totally love it.I allow said time for setup,again,totally diggin load in,hook up,etc.It doesn't bother me one bit.And if I or the other guitar needs help loading in his rig,we all help each other.no biggie.
Oth,I have friends that come in to sound a chk at 845 and throw down thier (modeler of choice in front of thier mic stand ) hoping they'll be good to go by 900 start time.thats just not for me.
Oth,also...tube amps have come a long way too.Major amp companies are putting out new models all the time,they're extremely reliable, sound killer,and some even have cab Sim and ir loading, if that's your thing.
My tube supply is enough to last me well into 2099,so I'm good there.
My fear is that someone that grows up in the current modeling boom,only playing the modelers, when they do plug into a finely tuned jcm800 ,or soldano slo,or mesa mkiic+ into a nice greenback loaded 4x12 might not even like it all cause its so immediate, incredible ( insert ur own awe description here) that they simply can't get on with it so to speak.I hope I'm wrong
 
[ Donnie B. ] said:
Me too. But I guess we just don't get the math.
Same with the dozens of paid engineers working for all of the modeling companies for the past decade.
Bunch of dullards.

Controlled A-B testing results are not the be all end all when it comes to this stuff.
Tons of data on why this is especially true with audio. This is a good one.

Click to expand...


You don't have to get the math. Plenty of us can't even see what the proprietary algorithms are because they are the companies secrets. We just know its algorithms and engineers have experience with others that are freely available.

As for the video, that's why you do it with a larger sample of people. Bigger population samples. Which is what has gone on for the past decade. He is just talking about himself alone. In proper experiments, you re-run it with multiple people to reduce confounding variables (what he is talking about). He is an audio engineer. I would actually pay more attention to what physics scientists have to say about controlled tests as that is their field. It involves statistics and null hypothesis testing.
 
psychodave said:
I have difficulties with recording since they are all compressed, but I’m willing to bet my discerning ears can tell the difference in the room while playing through different tube and digital amps.

Hence my comment earlier. I’m thinking a digital preamp with a monster tube power amp. That would likely meld the best of both worlds.... both in the room and recording.
Click to expand...
You can get power amps that aren't tubes that do the same thing as tube power amps. A lot of loadboxes even do it. Real amp plus loadbox with IRs. Line out so you just take the amps pre-amp section to the loadbox for power tube and IR emulation.

I would love to see videos of these people who can tell these differences.
 
My issue with modelers is how complicated they can become. I don't get very geeked out about configuring them. Plenty of real amp models out there today, still in production or otherwise relatively easy to grab if I'm chasing an old familiar tone, and I know with a few knob turns I'll have good tones and can get down to the business of playing.

I'm not opposed to carrying less heavy gear, or not having to worry about tube maintenance, and all the other benefits the modelers can bring. But every time I go down the modeler route I spend tons of time dialing it in, convince myself I'm there, and then start second guessing if it's really as good as the real thing.

When I hear the results some people get (both recorded and live), it does amaze me how far things have come. But again, simple just works for me better. When I use a real amp, I don't second guess myself if it's "real" enough.
 
braintheory said:
I’ve not heard of Crest. How would it compare to an HH V800 solid state power amp? The HH is also crazy heavy, more so than I think I remember my Strategy 400 being
Click to expand...
I have an HH V800 as well. It sounds glorious, but quite often I like the added sound of the Strategy 400. Keep in mind my S400 has added depth pots and can rattle your teeth out if needed. 🤣
 
psychodave said:
I have an HH V800 as well. It sounds glorious, but quite often I like the added sound of the Strategy 400. Keep in mind my S400 has added depth pots and can rattle your teeth out if needed. 🤣
Click to expand...
Can a depth pot be added to the V800 or something that would give the same effect?

I liked the Strategy 400, but my coliseum sounded better to me when used as just a poweramp and once I engaged the GEQ it was game over and rattled everything like you mentioned. I’d think even with a depth pot the strategy wouldn’t do it for me next to my coli with the geq, but who knows, maybe I’d be surprised. I haven’t even tried my blueface as a poweramp and can only imagine what would happen there because even my coli has nowhere near the bottom end of that. The Wizard and CCV also had some interesting sounds as poweramps. I haven’t played with that kinda setup in a few years now. I should give it another go
 
Last edited:
You must log in or register to reply here.
Back
Top