• We would like to remind our members that this is a privately owned, run and supported forum. You are here at the invitation and discretion of the owners. As such, rules and standards of conduct will be applied that help keep this forum functioning as the owners desire. These include, but are not limited to, removing content and even access to the forum.

    Please give yourself a refresher on the forum rules you agreed to follow when you signed up.

1:1 - Let`s talk A/B testing actual Amps to their Axe-Fx II Counterparts

guitardoc

Experienced
I compared to a 'modern' AC30TB with blue bulldogs. Top boost channel. Voicing selector on the back set to 'modern'. Hi cut set to zero, bass and treble at noon, gain at noon, master all the way up. Earplugs in (it's one of the loudest amps I've ever come across). Tele with cool rails pickups (the green one with the beautiful birds eye neck and the battered body - you might remember... ;-)) with high end summer cable 0,75 mm^2 into the amp.

Vs.

AC30TB block bypass double clicked and nothing changed into matrix into the same blue bulldogs. Same guitar, same guitar cable, summer cable 0,75 mm^2 to matrix and to speakers.

Matrix bridged and full throttle and clockwise with the Axe's volume taper until the same volume was reached (measured with iPhone dB meter) which was nearly all the way. Couldn't achieve the same volume with matrix unbridged (maybe possible with upping the volume in the amp block which I didn't touch).

The voicing is nearly spot on (the model is really excellent) but that was to be expected with the quality of the AxeFx and the speakers as the single most sound shaping components being exactly the same. Differences are presumably due to the different amp that was used for the modelling process. Only snag was that strange artificial high frequency hiss mentioned earlier and the significantly earlier dying of the sound.

FW12.03beta5.
 

Morphosis

Fractal Fanatic
So the "feel" thing i mentioned before is NO ISSUE for you in your A/B comparison scenario?!

I actually try to measure the frequency response from the Matrix, the Class H Rocktronic and the PowerAmp section of the studiotone (cab load & dummy load) and will do the tonematch as mentioned before and do the comparisons again ... will update when i got results...
 

guitardoc

Experienced
The feeling is a bit off too, that's true, but only slightly. When it comes to dynamics and the connected feeling my real AC30TB is somewhere in the middle of the AC30TB model and the Liverpool Trainwreck model (with delivers a different voicing of course) in the Axe.
 

Kriig

Fractal Fanatic
What I realised is that apart from the strange artificial hiss the AC30TB model produces (which I mentioned in the other thread) the sound from

Guitar->AxeFx->Matrix->AC30TB combo speakers

seems to die significantly earlier than from

Guitar->AC30TB combo.

I don't know why that is though. May it be possible that the signal coming from AxeFx or Matrix is not as 'hot' or long lasting as from the actual amp?

Btw, is there a difference in controlling the volume from the AxeFx output or the matrix input? Usually I use bridged mode and open up the matrix full throttle and adjust per AxeFx output taper. Tried it the other way round and didn't notice any which doesn't necessarily mean that there is none. Any experiences?
Is your gate off? The real amp has no gate...
 

PacoCasanovas

Fractal Fanatic
So he compared the [AC30TB Sim]+[Matrix Amplifier] vs. [actual AC30TB] ... both into the identical cab.

In my understanding this should result in a nearly "spot on" comparison. If this is not happen i can imagine different reasons:

- Maybe AC30`s are so different from each other like Plexis. Perhaps Cliff`s model based on a different type as the one from guitardoc?

source: AMP (block): list - Axe-Fx II Wiki

- The speaker used when modeling was another type as guitardocs used - This would require small adjustments on the speaker tab page, IMHO. That said, i don`t think those will be made the "night and day" differences, at least not those, guitardoc mentioned.
Dear Markus

I was refering to the default settings which were made to reach similar results over a FRFR system vs. the real thing miced. Maybe you have to tweak your settings to get "closer", since every amp acts differently (what you also wrote above....) in reality.
PS: My ears were unable to conduct a comparison test at very high levels. This reminds me on my test drive over an old Laney 100W amp from a friend - I've recorded the test, you hear me saying "great, it's still clean and punchy..." but you hear the recording tells a different story - the sound was already crunchy. So on the AxeFx you hear that sound on lower volumes - and you always think it sounds different, but it doesn't sound that much different.
The thoughest part is, that most people were unable to adjust their poweramps and AxeFx to each other. Remember my test of the GM50 module - when the Sensitivity knob is on full, the maximum input signal shouldn't be higher than -3dBu level before the amp start to clip - we don't want solid state clipping - that's why we need to know the exact signal level. The AxeFx itself punches out approx. +18dBu signal before the output clip led goes on.
Another thing goes for the speaker to amp adaption - if there are two speaker in serial you get 8 or 16ohms total depends on the speaker. A tube amp has an output transformer which does an impedance conversation of the output impedance of your tube amp into the speaker load. If there are 16Ohms used - you have a 16ohms tap on your output transformer . On a solid state amplifier there is no impedance conversation needed - but somehow your output power would be limited by a higher load - 16Ohms (two 8Ohms in series) makes approx. 125W per Channel on a GT1000FX. When wiring both speakers in parallel, you get a total load of 4Ohms - and you got 500W power from your Matrix solid state amp.....just in case you might wonder why ;)

Since you need more headroom to compare a solid state amp to a tube amp, there is much more power needed to reach the same loudness! A tube amp compresses the signal peaks, which makes the average to peak ratio higher (it's called "Crest Factor") which also results in higher perceived loudness. On a solid state amp we want to avoid clipping so - we set our maximum level to the maximum of possible output power that could be done by the soild state amp - there is no peak signal compression on solid state amps just as tube amp will do it....this is the job for the axefx.

Cheers
Paco
 

Morphosis

Fractal Fanatic
Allright, back to OP ... before Tonematching the Koch Studiotone into the Axe-Fx to get a better comparison base i tried to measure the different PowerAmps. What I`ve found out is to me very intereseting and hopefully help my understanding, what`s going on:

Testify scenario: Capturing the frequency response of the PowerAmps: For that i use the IR capture tool from the Axe-Fx:

[Axe-Fx Sine-Sweep OUT2L] into [PowerAmp] into [Behringer GI-100 w/o Spkr Sim] into [Guitar Cabinet / 8 Ohm]. The Behringer give also the [Line Out Signal] -> [Axe-Fx IN2L]

So, all the Amps will see the load of the real Cabinet!

MATRIX GT1000FX:


This is what i call a flat response, no?! Fine.

ROCKTRONIC PC750 (Class H / ring core transformer / rated at 2x250W / 8Ohm / to me subjective louder than the Matrix) (But also their damn 3 fans: loud as hell and the thing is damn heavy ;( )


Wow, this is damn the same frequency response as the Matrix, no?! Also very flat. Fine.

And now comes the Picture that make me very astonished:

Koch Studiotone Power Amplifer (FX return) WITH DUMMY LOAD (!), NOT the real Speaker (using it`s Line Out)


What the heck? This is also damn flat, no? This is the Power 2x EL84 section of the Koch Studiotone (seeing the inbuilt dummy load!).

And now the Koch`s Studiotone WITH the same Cab load:

Koch Studiotone Power Amplifer (FX return) WITH Guitar Cab as load:


So, this shows a frequency response which looks like the ones on the speaker tab page on the Axe-Fx, right?

So, now i have following questions:
- Am i right, that the last plot shows me the voltage feedback respectively this thing, we can adjust on the speaker tab page? If yes: Should i try to reproduce this curve to the speaker tabs page, when comparing the actual amp vs. Axe-Fx II Sim+Solid State Amp? So, first adjust this curve to the speaker tab page and than tonematch the sim to the Koch to get as near as possible? (Also DAMP to zero / Hi Cut to zero as Cliff state...)

- What does the Koch`s Plot WITH dummy load will say about it`s PowerAmp? Is it a super clean, flat Tube PowerAmp in theory? In theory, because of damping issue the flat response is not usable at the moment, you connect the Amp to a speaker...

- Would other Guitar Tube-PowerAmps react similiar flat, when connected to a dummy load?

Next step: Get a Axe-Fx Sim similiar to the Koch (Perhaps a DC30 or AC30 with Marshall Tonestack? Speaker Tab settings like the last plot of the Studiotone PowerAmp, Damping+Hi-Cut to zero and than tonematch the whole thing to the actual amp...)

If this will work, i should get a very, very similiar feel and tone experience when comparing the Axe+SS Amp vs. the Koch at the same guitar speaker, no?!
 

barhrecords

Axe-Master
IIRC, the Amp block speaker page is the impedance curve not a frequency amplitude curve.

While related to amplitude / freq I don't think the speaker impedance curve is an exact analog of the frequency amplitude.

What application are you using to plot the IR .syx files?
 

FractalAudio

Administrator
Fractal Audio Systems
Moderator
What the graphs say:
1. The built-in dummy load is a resistive load. A "good" dummy load is reactive and will attempt to simulate the speaker impedance. The schematic verifies this. The dummy load is simply a 25W, 8-ohm resistor.
2. The last graph doesn't show you "voltage feedback". Your amp has no feedback. The last graph shows you that the power amp (because it has no feedback) is following the speaker impedance. You want to try to match this graph using the SPKR page in the amp block.
3. The speaker has a resonance of about 12 dB at 120 Hz or so. So set the SPKR page in the Axe-Fx to 120 Hz, Q = 2.0 and Res to 5.0.
4. The HF response is pushed out a bit. That indicates the speaker has a low voice coil inductance. I'd start with HF Freq at around 1500 Hz.

Now for my "harsh reality" advice. Stop trying to match the sound of your amp. Pick a model that sounds good TO YOU and play. The Axe-Fx II modeling is the most accurate available. Agonizing over minutiae is time that could be spent making music.
 

FractalAudio

Administrator
Fractal Audio Systems
Moderator
IIRC, the Amp block speaker page is the impedance curve not a frequency amplitude curve.

While related to amplitude / freq I don't think the speaker impedance curve is an exact analog of the frequency amplitude.

What application are you using to plot the IR .syx files?
The speaker page is the impedance vs. frequency. For a guitar amp with no negative feedback the voltage frequency response of the power amp will very closely match this since the power amp is basically a current source (V = I * Z). There will a slight reduction in the peaks as the output impedance isn't infinite but it is very high and will therefore be very close.
 

barhrecords

Axe-Master
The speaker page is the impedance vs. frequency. For a guitar amp with no negative feedback the voltage frequency response of the power amp will very closely match this since the power amp is basically a current source (V = I * Z). There will a slight reduction in the peaks as the output impedance isn't infinite but it is very high and will therefore be very close.
Is there an easy way to plot this for amps with negative feedback?
 

Morphosis

Fractal Fanatic
Dear Paco,

i here (and heard ;) ) what you`re saying. And you`re for sure right on everything. But most things you pointed out are totally considered in my AB testify:

So on the AxeFx you hear that sound on lower volumes - and you always think it sounds different, but it doesn't sound that much different.
No. All comparisons are made at low volume. Axe and Koch (or what Power Amp else) - always through the same cabinet under really low levels.... so i have to impose this argument.


The thoughest part is, that most people were unable to adjust their poweramps and AxeFx to each other. Remember my test of the GM50 module - when the Sensitivity knob is on full, the maximum input signal shouldn't be higher than -3dBu level before the amp start to clip - we don't want solid state clipping - that's why we need to know the exact signal level. The AxeFx itself punches out approx. +18dBu signal before the output clip led goes on.
I heard you my friend! But as i said: The tests are so low volumed, that i never have to raise the PowerAmps attentuator knobs over you 70% Paradigm, also i am really FAR away to blast the output of the Axe with the frontside levels fully to the +18dBu. Most of the time it`s somewhere at 9 o`clock...

Another thing goes for the speaker to amp adaption - if there are two speaker in serial you get 8 or 16ohms total depends on the speaker. A tube amp has an output transformer which does an impedance conversation of the output impedance of your tube amp into the speaker load. If there are 16Ohms used - you have a 16ohms tap on your output transformer . On a solid state amplifier there is no impedance conversation needed - but somehow your output power would be limited by a higher load - 16Ohms (two 8Ohms in series) makes approx. 125W per Channel on a GT1000FX. When wiring both speakers in parallel, you get a total load of 4Ohms - and you got 500W power from your Matrix solid state amp.....just in case you might wonder why ;)
Yeah, i know that. Because of that reason i let see the Power Amps always the same one speaker with 8 Ohm. (Because of that reason i build me that Cabinet i CAN (but did not here in these tests... ) drive at 2x 8Ohm (stereo or dual mono) or 4 Ohm mono (speakers in parallel). Most of 2x12" are 16Ohm speaker to come to 8Ohms. The intention using 8Ohm speakers to get them to 4 Ohm was made in consideration to using it with SS-Amps ... but as i said: Here only: One speaker / 8 Ohms to ALL Test candidates!

I was refering to the default settings which were made to reach similar results over a FRFR system vs. the real thing miced. Maybe you have to tweak your settings to get "closer", since every amp acts differently (what you also wrote above....) in reality.
Again: In these tests, this thread ... i am not talking to FRFR, not comparing FRFR to anything else and i have pointed out my view, how i understand what Cliff said. But I am totally with you when it comes to "you have to adjust the PowerAmp Sim to match to the used real PowerAmp". Every PowerAmp is more or less different - so the simulation must be tweaked to balance this issue.

The Best!
Markus
 

PacoCasanovas

Fractal Fanatic
Matrix bridged and full throttle and clockwise with the Axe's volume taper until the same volume was reached (measured with iPhone dB meter) which was nearly all the way. Couldn't achieve the same volume with matrix unbridged (maybe possible with upping the volume in the amp block which I didn't touch).

The voicing is nearly spot on (the model is really excellent) but that was to be expected with the quality of the AxeFx and the speakers as the single most sound shaping components being exactly the same. Differences are presumably due to the different amp that was used for the modelling process. Only snag was that strange artificial high frequency hiss mentioned earlier and the significantly earlier dying of the sound.

FW12.03beta5.
Dear Markus

Sorry, I mostly refered to the post of guitardoc above ;)

Cheers
Paco
 

Morphosis

Fractal Fanatic
What the graphs say:
1. The built-in dummy load is a resistive load. A "good" dummy load is reactive and will attempt to simulate the speaker impedance. The schematic verifies this. The dummy load is simply a 25W, 8-ohm resistor.
2. The last graph doesn't show you "voltage feedback". Your amp has no feedback. The last graph shows you that the power amp (because it has no feedback) is following the speaker impedance. You want to try to match this graph using the SPKR page in the amp block.
3. The speaker has a resonance of about 12 dB at 120 Hz or so. So set the SPKR page in the Axe-Fx to 120 Hz, Q = 2.0 and Res to 5.0.
4. The HF response is pushed out a bit. That indicates the speaker has a low voice coil inductance. I'd start with HF Freq at around 1500 Hz.
This is (hopefully) very helpful for me to learn and understand. Thank you for clarification!

Now for my "harsh reality" advice. Stop trying to match the sound of your amp. Pick a model that sounds good TO YOU and play. The Axe-Fx II modeling is the most accurate available. Agonizing over minutiae is time that could be spent making music.
It may appear, that i want to match my actual amp, but this is really not my priory intention. There are dozens of Amp-Sims in the Axe i love much more than my actual Amp. It`s just, that i`ve missed something when playing them thru my guitar cabinet and i just will to find out for myself what`s and where the reason is maybe located for that. So matching the Koch should be only be something like "a bridge" to find out what i can do, to get more completely satisfying results out of the Axe+SS Amp+Guitar Cabinet.

To clarify: I won`t proof that the Axe-Fx is not spot on. I can`t. In fact i am confident that it is! Because of that i assume i am wrong somehow. In Amp-selection, adjustment, PowerAmp decisions or anywhere else. So i will locate that, find out a recalibrate to best satisfaction.

That`s all. :star: really ...
 

Capricorn

Member
Slightly off topic, but this thread got me thinking...If I ever wanted to use the Axe through a tube amp + cab with PA sims off, should I flatten the impedence curve in the SPEAKER tab?
 

FractalAudio

Administrator
Fractal Audio Systems
Moderator
Slightly off topic, but this thread got me thinking...If I ever wanted to use the Axe through a tube amp + cab with PA sims off, should I flatten the impedence curve in the SPEAKER tab?
Unnecessary as the SPRK tab only affects the PA sims.
 

MikeyB59

Power User
The Axe-Fx II modeling is the most accurate available. Agonizing over minutiae is time that could be spent making music.
I enjoy reading about people analyzing the various differences between the Axe and xyz, but mostly I roll with what Cliff said above. I find so many great sounds that I don't find time to agonize over what I don't find.

I'm glad some people focus on all the details though as I think it makes it better for all of us in the long run.
 
Top Bottom