(UPDATED for new FW!) EVH amp vs Axe Fx, QC and TMP...

[MirrorProfiles]

why not just default the amp to a "close enough" SIC when not using the cab block, just like many other products that sound pretty accurate? Problem solved.

I think they already said that a generic SIC curve gets applied when using 3rd party IR’s. Its not clear what this curve is, or how close to any particular real curve it is.

Giving the ability to choose is ideal (and a good move) but I’m not sure why the current implementation sounds so much worse than other non-selectable options from other companies.

Using the EVH amps as an example, these models use different SIC and one is an EL34 and one is 6L6 Stealth. They don’t sound the same but they’re also as close as you’d expect those 2 amps to sound:

 

[Aleksi]

I think they already said that a generic SIC curve gets applied when using 3rd party IR’s. Its not clear what this curve is, or how close to any particular real curve it is.

Giving the ability to choose is ideal (and a good move) but I’m not sure why the current implementation sounds so much worse than other non-selectable options from other companies.

Using the EVH amps as an example, these models use different SIC and one is an EL34 and one is 6L6 Stealth. They don’t sound the same but they’re also as close as you’d expect those 2 amps to sound:

It seems very probable that the SIC is in the cab block and is based on the chosen cab. The generic 3rd party IR SIC curve is then exactly the same for all amps from JC120 to the Uberschall because it has no way of knowing what amp and cab is being used. So the SIC is probably pretty far from optimal for anything. If the SIC was in the amp, then the TMP should sound fine with a 3rd party IR, much like nearly all other modelers, as it could just use a SIC curve that's appropriate for the usual amp-cab pairings.

 

[Lysander]

Read this and it broke my heart a bit. @GuitarJon , you shouldn't be apologizing here.

I appreciate Fender people lurking around here and being open about updates but, make no mistake: this IS an issue, and no review/shootout should be tip-toing around it, just because a fix is promised for the future.

I particularly didn't enjoy the suggestion that the A/B comparison was "invalid" somehow. Jon A/B'd what comes in the box today with the real amp; that's as valid as it gets.

 

[GuitarJon]

Read this and it broke my heart a bit. @GuitarJon , you shouldn't be apologizing here.

View attachment 18844

I appreciate Fender people lurking around here and being open about updates but, make no mistake: this IS an issue, and no review/shootout should be tip-toing around it, just because a fix is promised for the future.

I particularly didn't enjoy the suggestion that the A/B comparison was "invalid" somehow. Jon A/B'd what comes in the box today with the real amp; that's as valid as it gets.

Thanks, I appreciate it and I do agree with you! I was feeling the "heat of the battle" and people were saying things like "they lost respect for me" etc. So I just wanted to address some things. Leaving the IC thing out of the video wasn't done on purpose because as I explained, I don't think it matters as much as some other people believe. I figured I'd apologize anyway. And I never set out to make people upset because of my videos, I want everybody to be happy ideally so I apologized for that too, I figured, why not? :) I did however not apologize for my video as a whole and I still stand behind my methods. People can keep complaining about it but no matter what you do to " improve upon the methods", the results will remain the same. The amp model just doesn't sound like an EVH whether it's a 50w EL34, LBX or 100w Stealth. I've played through various of the amps and aside from some quirks here and there like the blue channel of the 50w EL34, you can dial in the red channels to sound very similar. I don't care what anybody says. To what extent all of this matters is up to everybody to decide for themselves. Some people don't care how accurate the model is and that's 100% fine! Enjoy whatever you enjoy! I applaud that!

I also don't think that any IC or whatever is magically going to make the models THAT much more accurate. So people can keep going back to that but it doesn't really change my point. So again, I stand by my points, I want to be able to be pure and honest towards my viewers and it appears that all my viewers want that from me too. I don't want to hurt anybody's feelings though, that's never the goal, but as long as I remain honest I guess it just sometimes happens. In the end, I'm just a messenger and I shouldn't be blamed for inaccuracies in modeling. IC's, differences in exact amp itterations, tube types etc haven't stopped some other models from sounding damn close to my tube amps. I do believe that I kind of know what I'm doing, so I'm confident about my methods. Just sayin'

 

[GuitarJon]

I also find it funny that a lot of people like to say "all the amps on my channel sound the same" (which of course I don't agree with), but now all of the sudden there are HUUUUUGE differences between the various 5150IIIs, all similarly voiced amps.

 

[Los-D]

Yeah, I personally don’t buy the SIC argument. TMP looks like pretty good device but perhaps it’s modeling is just not as good as some of the top tier devices (at least on the high gain side). Nothing wrong with that at all since it has other features not seen on said devices. But this treatment of @GuitarJon is waaay out of line.

 

[JiveTurkey]

I also find it funny that a lot of people like to say "all the amps on my channel sound the same" (which of course I don't agree with), but now all of the sudden there are HUUUUUGE differences between the various 5150IIIs, all similarly voiced amps.

This is low key a HUUUUUGGGGGGGGEEEEE point to observe. "They (modelers) all can sound the same" has been the truth for your channel forever but now suddenly; the latest $16-1800 thing can't quite do it.

 

[2112]

Nice work as always Jon.

A video comparing different loads/impedance curves one one of your amps would make for an interesting video!

 

[LP59]

It seems very probable that the SIC is in the cab block and is based on the chosen cab. The generic 3rd party IR SIC curve is then exactly the same for all amps from JC120 to the Uberschall because it has no way of knowing what amp and cab is being used. So the SIC is probably pretty far from optimal for anything. If the SIC was in the amp, then the TMP should sound fine with a 3rd party IR, much like nearly all other modelers, as it could just use a SIC curve that's appropriate for the usual amp-cab pairings.

+1.

I don't understand what Fender is saying. They told Jon the problem is the SIC in the IR block, but there's no such thing as an SIC in an IR block. You can have EQ in the cab block, but not impedance, but that's a poor way to model an amp.

And I don't see how a selectable SIC will necessarily solve the problem anyway. If, as they claim, the problem is the SIC, what if they don't provide a curve that matches whatever gear Jon is using (load box or cabinet)?

In the end though, I think this going to come down to Fender saying "You don't have the exact amp we modeled".

 

[Aleksi]

I definitely get the feeling that the ARM stuff isn't all that for real time amp modeling compared to a dedicated DSP. It probably doesn't have enough juice to run modern amp modeling on a single core without compromises like 44.1khz and offloading parts of what should be in the amp block to the cab block. ARM seems to be fine for fx processors like the Eventide H90.

Source: I have a master's in internet speculation.

 

[jellodog]

I definitely get the feeling that the ARM stuff isn't all that for real time amp modeling compared to a dedicated DSP. It probably doesn't have enough juice to run modern amp modeling on a single core without compromises like 44.1khz and offloading parts of what should be in the amp block to the cab block.

I feel that isn't entirely fair for ARM as a processor architecture in all MI devices. There's a wide range of processing power available at different price points.

However, where I would be inclined to agree with you is that this particular processor in the TMP may well be under specified in terms of the per core performance (cost savings!) and yes, there likely isn't enough juice for modelling on a single core without some compromises, as you've said. An amp model must be run on a single core.

Having said that, there are always compromises made in all modellers. But in this case, the compromises may be far from ideal. We will know for sure within the next 6 months or so, I suppose.

 

[the-trooper]

You can have EQ in the cab block, but not impedance, but that's a poor way to model an amp.

I definitely get the feeling that the ARM stuff isn't all that for real time amp modeling compared to a dedicated DSP. It probably doesn't have enough juice to run modern amp modeling on a single core without compromises like 44.1khz and offloading parts of what should be in the amp block to the cab block.

Pretty much exactly what that one guy from Fractal said about it, day after he got it.

 

[Lysander]

I definitely get the feeling that the ARM stuff isn't all that for real time amp modeling compared to a dedicated DSP. It probably doesn't have enough juice to run modern amp modeling on a single core without compromises like 44.1khz and offloading parts of what should be in the amp block to the cab block. ARM seems to be fine for fx processors like the Eventide H90.

tl;dr: nah, not really. Modern ARM SoC's are plenty powerful to run digital models; it all comes down to the quality of the software implementation these days.

If i recall correctly, Headrush and Tonex also run on ARM.

 

[LP59]

tl;dr: nah, not really. Modern ARM SoC's are plenty powerful to run digital models; it all comes down to the quality of the software implementation these days.

If i recall correctly, Headrush and Tonex also run on ARM.

+1. They are running short on cpu power for some reason, otherwise it would be using 48kHz, but an ARM processor shouldn't be that reason.

 

[MirrorProfiles]

I was recommended to read the discord server for some SIC details (that I’ve actually been asking ever since the product was announced).

Seems that the cab block applies a general SIC to the built in cabs even if there is no amp in the chain.

Scratching my head at that one. Have asked for more of an explanation but it sounds like you can’t use the TMP cabs with external amps or modellers, unless you feed them a resistive load.

Definitely something that should be in the manual and explained to users because that is not how these things normally behave.

 

[timbuck3]

Yeah, I personally don’t buy the SIC argument. TMP looks like pretty good device but perhaps it’s modeling is just not as good as some of the top tier devices (at least on the high gain side). Nothing wrong with that at all since it has other features not seen on said devices. But this treatment of @GuitarJon is waaay out of line.

 

[ian_dissonance]

I was recommended to read the discord server for some SIC details (that I’ve actually been asking ever since the product was announced).

Seems that the cab block applies a general SIC to the built in cabs even if there is no amp in the chain.

Scratching my head at that one. Have asked for more of an explanation but it sounds like you can’t use the TMP cabs with external amps or modellers, unless you feed them a resistive load.

Definitely something that should be in the manual and explained to users because that is not how these things normally behave.

That sure sounds an awful lot like an eq ahead of the cab/IR and also seems to confirm the vehemently-denied suggestion from a certain amp modeler recently.

 

[JiveTurkey]

I was recommended to read the discord server for some SIC details (that I’ve actually been asking ever since the product was announced).

Seems that the cab block applies a general SIC to the built in cabs even if there is no amp in the chain.

Scratching my head at that one. Have asked for more of an explanation but it sounds like you can’t use the TMP cabs with external amps or modellers, unless you feed them a resistive load.

Definitely something that should be in the manual and explained to users because that is not how these things normally behave.

Dumb.

 

[MirrorProfiles]

Some of the questions I asked on how they handled this stuff were asked in October last year, and it’s still not really clear what exactly is going on.

That email I sent Softube in 2022 above was answered within about 2 hours and with no ambiguity. There’s nothing to gain by concealing how the power amp/cabs/IR’s behave, users need to know that to get the right results.

 

[LP59]

I was recommended to read the discord server for some SIC details (that I’ve actually been asking ever since the product was announced).

Seems that the cab block applies a general SIC to the built in cabs even if there is no amp in the chain.

Scratching my head at that one. Have asked for more of an explanation but it sounds like you can’t use the TMP cabs with external amps or modellers, unless you feed them a resistive load.

Definitely something that should be in the manual and explained to users because that is not how these things normally behave.

How is that an SIC? That sounds like an EQ in the IR/Cab block, which is not the same thing as impedance. They even call it a "filter" instead of impedance.