So How good is amp modeling these days?

[sk8centilli]

This model kicks ass.

 

[BryanMichael]

That's YOUR opinion as the end-user, and I much agree with you. However, when a product claims to re-create XYZ amp, you cannot be surprised when people place a huge amount of emphasis on whether or not the product meets those claims, right?

 

 

Yes, but as has been pointed out over and over again, someone can say "That doesn't sound anything like a Marshall JCM 800!!!" (Meaning that, in their minds it should sound like (insert artist here that uses said amp) )when really, on any given day taking any JCM800 "off the rack" so to speak would result in a tone that is exceptionally close to, if not exactly like that model. I also think that a LOT of it has to do with the sound you get playing from a REAL AMP when you are in the room with it. Just as people are always going on and on about Leslie speakers and how no effect can reproduce the sound of a REAL Leslie, blah, blah, blah- they are right- nothing can reproduce the PHYSICAL response you have to a real amp- but as far as RECORDED tones? You are going to have a hell of a time pulling out the "real vs. fake" from someone who knows their stuff in setting up a modeler.

[Arr0wHead]

 

Yes, but as has been pointed out over and over again,

 

 

I'm trying to follow you, really, but I think you're trying to use my own quotes to argue against me? I'm the one "pointing out over and over again" these things you reference.

 

And for response, as I also keep saying over and over, response CAN be modeled, and it IS in the newest crop of modelers. The more processing (computer or hardware unit like an axe-fx) the easier it is. Earlier modelers lacked "feel" (a word blues guys use to describe dynamic response) because it was too much processing needed to accurately create the modeled sound AND get those dynamics. That was the whole upgrade from the POD XT/X3 to the HD series, they went back and rebuilt their models to include this more complex algorithm, then put enough processing power and memory into the unit to compensate. It's also why you get 16 models instead of the previous 80+, because these new more complex models take up more power and RAM.

[Eddie]

That's just outstanding!!! All of the cable snobs should be forced to watch this all the way through

 

Since I posted the video originally, I'd like to address this. Cables do make a difference.

I recorded a comparison a while back, with a video showing a frequency analyzer while the audio was playing. It was clear from listening that the sounds were different, and the analyzer was a good tool to look at the frequency content.

 

It's quite easy to do a valid test for this purpose. Record a loop, and record it to a DAW using a different cable each time. That way the only variable is the cable. Ideally all cables should be of the same length, but even approximations (e.g., 18 ft vs. 20 ft.) should be "good enough" for such a comparison.

[Arr0wHead]

 

Since I posted the video originally, I'd like to address this. Cables do make a difference. I recorded a comparison a while back, with a video showing a frequency analyzer while the audio was playing. It was clear from listening that the sounds were different, and the analyzer was a good tool to look at the frequency content. It's quite easy to do a valid test for this purpose. Record a loop, and record it to a DAW using a different cable each time. That way the only variable is the cable. Ideally all cables should be of the same length, but even approximations (e.g., 18 ft vs. 20 ft.) should be "good enough" for such a comparison.

 

 

The cables they reference in the video are NOT instrument cables. He's talking about people claiming that top dollar POWER cables affect sound. As Ethan says in the video, it may be true. It may not. But either way that fancy shmancy cord is still plugged into the wall feeding off hundreds of feet of standard wiring anyway, so how could it work.

 

Also, for the cable snobs, it MAY make a difference, but it is by FAR not the weak link in a recording situation. As Ethan himself says, the room itself and it's acoustical treatment are more important that every single other stage in your recording chain. Of course, grain of salt, this is coming from a guy who happens to run an acoustical treatment company. The guy who runs monster cable may instead tell us that the cables are WAAAAY more important.

 

And THANK YOU for posting that video, I have it bookmarked for future reference and have sent it to several of my audio enthusiast buddies. I learned a lot from it.

[thunder100]

 

However, keep in mind that all the plugins I listed DO have demo versions you can download. I encourage people to try them out! Just remember how important speakers are - don't compare your crappy laptop speakers or $30 ipod buds to your guitar amp, no matter HOW good the models it will NEVER be the same. Then again, your AMP won't sound the same through laptop speakers or these ipod buds EITHER.

 

 

Thats a very valid point as stated earlier.Run through same speaker differences diminish to 30%,run through same poweramp then differences are down to 10%

 

As AXE owner and as well as a full POD (Farm 2 with all paid) and Jamvox -->YES Harware is the difference

 

But the cheaper ones are much more user friendly then the AXE-->but there you have one of the greatest communities to help you out

 

Roland

[thunder100]

This model kicks ass.

 

Hmm no-->because if so his hair should blow forward

 

Roland

[Eddie]

I'd like to see someone do the following: (although, since I'm telling you up front it wouldn't work now anyway) Do a blind test, comparing a modeler to the same style amp - as an example lets say a JCM800. However, do NOT use a modeler. Just use two different JCM800 heads through two of the same cabs, with the same type of mic in the same position on each. Then watch people go on and on about which one is better, more realistic, etc...... ...I think the "myths in audio" video someone posted earlier is an amazing and educational thing, probably the single most valuable post in this entire thread...

 

I think I posted something similar to what you propose. At an amp shootout we had a few models that were repeated. In particular I recall 5 or 6 vintage Vox amps, and at least 3 vintage Gibson amps. I don't recall the models, but these were "identical" amps. They all sounded quite different. Some sounded quite mediocre. Compared to modeling, I would say that 1 or 2 Voxes sounded better than my zenTera, but the zenTera sounds much better than those other Vox amps. Makes me think about the source used for some modeling. In the zenTera's case, the amps came from Thoms Blug's own amp collection.

 

I'm glad you watched the video. IIRC, Ethan himself posted it a while back and there was some interesting dialogue about it.

 

The only point where I disagree with you is regarding feel. A lot of modelers I've used still don't respond well to my touch. With some amps, as well as some pedals, the resulting sound varies depending on my picking. I still have not found that in the latest modelers I've used (e.g., POD Farm 2, TH2). I believe that both appeared in 2010.

 

When it comes to digital modeling, I think that the first modeler was Line6 Amp Farm in the late 90s. Different companies have used different approaches to modeling since then. Using samples has been mentioned for years, with the problem that a sample was static. You could get maybe the nicest sampled Marshall sound, but it was just one sound. It did not take into account changes in volume or voltage or picking attack. Now it may be more usable with multiple samples, like they use in drum modeling software. Instead of having one snare sound, now we have a range of sounds from soft to hard hits, in various places of the same snare drum, with different mics in place.

 

For guitars, the latest I've seen is Kemper. Here they talk about their approach:

http://kemper-amps.com/faq.htm

It sounds like Roland's COSM concept, but IMO Roland never really got a great sound, at least up to VG-88 which is the last one I used. The sounds were good and usable, but not great. They also talk about taking a snapshot of an amps sound, but with greater flexibility, ability to get multiple snapshots, modify the snapshot, etc.

 

I like what I've heard from them. It's interesting that in the past some companies have developed modeling from an engineering standpoint, and they've failed. Hughes & Kettner was one of them, so starting with the zenTera it was a collaboration of engineering and an actual guitar player.

 

A drummer gave us Marshall amps, and now a keyboard player gives us Kemper amps.

[Matt Gnarly]

 

Yes. And that 5% feel you're missing is a direct factor of price/processing power. Which is why people say the Axe FX nails it 100% - it's got the hardware (and the PRICE TAG) to get that huge amount of detail and response out of it.

 

I'm not sure if I'd agree with the people who say Axe Fx nails it 100%, from the clips I've heard it may come close to 90%, and that's the top of the line modeller available.

[Eddie]

The cables they reference in the video are NOT instrument cables. He's talking about people claiming that top dollar POWER cables affect sound. As Ethan says in the video, it may be true. It may not. But either way that fancy shmancy cord is still plugged into the wall feeding off hundreds of feet of standard wiring anyway, so how could it work. Also, for the cable snobs, it MAY make a difference, but it is by FAR not the weak link in a recording situation. As Ethan himself says, the room itself and it's acoustical treatment are more important that every single other stage in your recording chain. Of course, grain of salt, this is coming from a guy who happens to run an acoustical treatment company. The guy who runs monster cable may instead tell us that the cables are WAAAAY more important. And THANK YOU for posting that video, I have it bookmarked for future reference and have sent it to several of my audio enthusiast buddies. I learned a lot from it.

 

You are correct, I just wanted to clarify that cables do make a difference since many (most?) people won't take the time to actually look at the actual video. Glad you liked it

In the context of the video I totally agree. In the ampfest I mentioned, one cable that people did not want to use with vintage amps and guitars was GeorgeL because it sounded a bit harsh. This coincides with the test I did. The GeorgeL cables are not harsh, but retain high frequencies much better than other cables. Many guitarists at the ampfest preferred cables that were working as a hi-cut filter (with that specific gear).

[Arr0wHead]

 

I'm not sure if I'd agree with the people who say Axe Fx nails it 100%, from the clips I've heard it may come close to 90%, and that's the top of the line modeller available.

 

 

You can HEAR response and "feel"?

 

I bow down to you sir, you have the most amazing ears on the planet.

 

 

In a less sarcastic response, - you take my words too far out of context.

[Arr0wHead]

 

The only point where I disagree with you is regarding feel. A lot of modelers I've used still don't respond well to my touch. With some amps, as well as some pedals, the resulting sound varies depending on my picking. I still have not found that in the latest modelers I've used (e.g., POD Farm 2, TH2). I believe that both appeared in 2010.

 

 

Pod farm 2 is still first generation models. EVERY line 6 product, all the way up to the HD, uses the exact same modeling. They added new amps and effects over time, but it's still the same algorithm's and technology as the original amp farm sounds. For the record, Amp Farm was early, but I don't know it that or the original combo amp came first. (I forget the amp model, but someone here mentioned it earlier in the thread). The HD series is the first and ONLY update to the modeling technique used by line 6. This includes everything, including POD farm, Gearbox, the Toneports, and all their respective VST plugins.

 

However, TH2 IS brand spanking new modeling. This is one I would recommend as an example of everything GOOD about modern modeling. If you're not happy about the feel of those models, and assuming you are using a good set of speakers and a box powerful enough to run it, then perhaps your argument about the feel is legitimate. Honestly, at this point I see ZERO difference between the response of a proper amplifier and the Overloud plugin. Latency can change that as well, but again that brings us back to processing power.

 

So I ask, what was it about the responsiveness and dynamics of the TH2 that you found different than a real amp?

[SnowViking]

They might be able to model the sound of a JCM 800 / vintage Marshall but dammit they can't model the feeling in the guts when it's turned up. There nothing else like it.

 

Other than that I think modeling amps are cool - even if they really give too many sound options. I can live with clean or overdriven plus some spring reverb thanks.

[DaleH]

Read the manual.

[BryanMichael]

I'm trying to follow you, really, but I think you're trying to use my own quotes to argue against me? I'm the one "pointing out over and over again" these things you reference. And for response, as I also keep saying over and over, response CAN be modeled, and it IS in the newest crop of modelers. The more processing (computer or hardware unit like an axe-fx) the easier it is. Earlier modelers lacked "feel" (a word blues guys use to describe dynamic response) because it was too much processing needed to accurately create the modeled sound AND get those dynamics. That was the whole upgrade from the POD XT/X3 to the HD series, they went back and rebuilt their models to include this more complex algorithm, then put enough processing power and memory into the unit to compensate. It's also why you get 16 models instead of the previous 80+, because these new more complex models take up more power and RAM.

I didn't think I was arguing against you, just making and clarifying points. That said, as a POD HD500 owner, I still use a Tube Driver and analog distortion before the modeler because the models don't quite feel or sound right to me - that and all of the {censored}ty corners they cut on the HD500, as well as the bugs, like the sound just disappearing and having to toggle between patches to get it to come back, running out of DSP processing power when you try to stack too many effects (more of an engineering concept that shouldn't be advertised the way it is...), and a few other things, but overall the amp models sound great- definitely good enough to record with, but I wouldn't rely on it for live use until they fix the bugs. It would be nice to have an ON/OFF switch on the unit as well

[Jarick]

 

They might be able to model the sound of a JCM 800 / vintage Marshall but dammit they can't model the feeling in the guts when it's turned up. There nothing else like it. Other than that I think modeling amps are cool - even if they really give too many sound options. I can live with clean or overdriven plus some spring reverb thanks.

 

 

Not that I disagree, but isn't that just the reproduction of the low end at high volumes? In other words, why can't a modeler cranked way up through either an amp and speaker or a P.A. system shake your guts?

[Eddie]

Pod farm 2 is still first generation models. EVERY line 6 product, all the way up to the HD, uses the exact same modeling. They added new amps and effects over time, but it's still the same algorithm's and technology as the original amp farm sounds. For the record, Amp Farm was early, but I don't know it that or the original combo amp came first. (I forget the amp model, but someone here mentioned it earlier in the thread). The HD series is the first and ONLY update to the modeling technique used by line 6. This includes everything, including POD farm, Gearbox, the Toneports, and all their respective VST plugins. However, TH2 IS brand spanking new modeling. This is one I would recommend as an example of everything GOOD about modern modeling. If you're not happy about the feel of those models, and assuming you are using a good set of speakers and a box powerful enough to run it, then perhaps your argument about the feel is legitimate. Honestly, at this point I see ZERO difference between the response of a proper amplifier and the Overloud plugin. Latency can change that as well, but again that brings us back to processing power. So I ask, what was it about the responsiveness and dynamics of the TH2 that you found different than a real amp?

That explains why I've NEVER liked Line6's sounds. I have not tried HD though.
With TH2 I'm not getting a variation in sound as a result of varying my pick attack to the degree I'd like, for example. I don't know if this is due to the cabinets I can use, since most of TH2's cabinets do not work on my Mac. I first reported this back in December, and a few days ago read about another person having the exact same issue. This happens when using VST, AU, or standalone mode.
It's hard to evaluate just based on a sound sample, but here it is:
http://soundclick.com/share.cfm?id=10250570

A 0:00-0:25
B 0:27-0:57
C 0:58-1:23
D 1:24-1:50

I minimized the volume variation just a little, but it's still present. The idea is to focus on the change in sound (like a real amp) more than on the mere change in volume. One problem with audio is that, at least at first, everything seems to be better when its louder.
Which one (or ones) seems to vary mainly the volume?
Which one (or ones) seems to add a more complex tone (overtones), and not merely volume?

To me, 2 sounded richer. One felt more responsive, and also is one of the rich sounding.
I spent probably 1 second selecting the amps, and did not spend any time at all setting up eq., drive, volume, or any additional processing. Only at the end I tried to get them to sound at approximately the same levels.

[cratz2]

I liked ganier portion of C the least... esp starting around the 1:07 point. I didn't enjoy that at all. I probably liked D the second least, but I liked the cleaner part of C more than the cleaner part of D.

 

Between A and B... they are pretty close. I think I like the cleaner portion of A more than the cleaner portion of B but B just seems to have more overall gain/saturation/whatever than A so I can't really compare them.

 

Overall, I'd say I liked A the best, then B, then D followed by C.

[cratz2]

I didn't think I was arguing against you, just making and clarifying points. That said, as a POD HD500 owner, I still use a Tube Driver and analog distortion before the modeler because the models don't quite feel or sound right to me - that and all of the {censored}ty corners they cut on the HD500, as well as the bugs, like the sound just disappearing and having to toggle between patches to get it to come back, running out of DSP processing power when you try to stack too many effects (more of an engineering concept that shouldn't be advertised the way it is...), and a few other things, but overall the amp models sound great- definitely good enough to record with, but I wouldn't rely on it for live use until they fix the bugs. It would be nice to have an ON/OFF switch on the unit as well

Interesting. I didn't get the HD500 for recording, just noodling around with higher gain tones than my tube amps realistically allow. I'm mostly an AC15/Tweed Deluxe/Blackface Twin kinda guy. I ran the 500 into the FX Return on a Peavey Prowler tube amp (so the only thing that is actually working on the amp is the tube power section) with an EVM 12L speaker in it.

I thought the 500 sounded great. And felt pretty responsive once it was up pretty loud. At first, I pretty much lived on the Park/Plexi model and grew to really love the Dr Z model as well. I also liked combining two amps basically to bring in a clean/solid feel that I thought the straight up gain amps lacked a bit. For example, I'd dial in the crunch I wanted with a Park model, then add the Dr Z pretty clean, but it didn't really sound like two amps. It sounded more like... a crunchy amp, but with the solidity and impact that a clean amp has.

I ended up returning it after 58 days. I was pretty damn impressed with the amp models and I guess the cabinet models which I almost always used. I liked how much of a difference the mic choice made, but I have no idea how realistic that portion was. I was pretty disappointed in pretty much all the effects, esp the dirt boxes and compressors. I played around with them pretty extensively and never found any of them or any settings I was truly happy with. I still have an original Vox Tonelab... the desktop model and I firmly believe it's easier to dial in a believable sound with that than the HD500, esp when using dirt boxes. Of course with the HD, there's really no reason to use the dirt box models. You can always have two presets where one just has more amp gain or you can put an actual dirt box in front of it as you suggested which is basically what I did, but it still didn't feel quite right.

I've only played through an AxeFX and two powered QSC speakers for about an hour... maybe a bit more. I'm not sure one is clearly better than the other... at least for my style/needs. The AxeFX certainly can sound more 80s or 90s processed (I mean that in the best possible way... not saying 'overly processed'... I just mean the AxeFX clearly sounds better with mid to high gain with compression, reverb, chorus and a couple delays. In that instance, it KILLS the HD in my opinion. But for just straight up plugged into an amp kinda things... Well... I spent 100 or more hours with the HD and 1 hour with the AxeFX, but I was very impressed with the HD.

[Eddie]

I liked ganier portion of C the least... esp starting around the 1:07 point. I didn't enjoy that at all. I probably liked D the second least, but I liked the cleaner part of C more than the cleaner part of D. Between A and B... they are pretty close. I think I like the cleaner portion of A more than the cleaner portion of B but B just seems to have more overall gain/saturation/whatever than A so I can't really compare them. Overall, I'd say I liked A the best, then B, then D followed by C.

 

I get what you're saying, but in this case don't focus that much on sound quality, but rather on the change in sound between the soft and hard parts.

That way the focus is more on the actual responsiveness to a players touch, which was the purpose for these files.

 

In terms of actual tone, I'm sure I could improve each one of the samples. I just plugged, selected a "Vox" model, and played. At the very least I usually adjust the eq. (bass, treble, mids, presence), and this time I did not even do that

 

Listen again with this in mind, and let me know if this new listening approach changes your opinion (or not). I'll post the source of each Vox along with my favorites later today.

[Arr0wHead]

Eddie: I admit, I'm currently too broke to afford the upgrade/purchase for TH2, but I do have TH1. I picked it over all the other demos (Amplitube, Guitar Rig, Waves GTR (they were giving out free 1 year licenses, don't know if it's still available), etc... because I felt it had exactly the response you aren't getting from TH2. Are you familiar with the TH1? I'm wondering if it's a bug with TH2, or if we just have wildly different expectations/picking styles. I REALLY hope it's the latter, because I've been creaming myself to get 2 since it came out. (although I am honestly happy with one, the sounds are amazing except the brutal high gain amps, and I have my floor pod and a billion other amp sims for that).

 

Cratz: I can't click back to last page as I type this. Was it you who said you experienced all kinds of bugs with the HD 500? I was just reading a thread over at SS.org where another user was saying the same thing. He also commented that he feels the quality of the models themselves has gotten WORSE with the last few updates they have released for the unit. Was this your experience as well? This is a bad development indeed, since I was considering TH2 or a POD HD as my next purchase.

[Arr0wHead]

Another tidbit:

It's not just about DIGITAL amp modeling. I just picked up a BOSS ST-2 Powerstack pedal. It really is like a stack crammed into a box. I love it. I run it direct into my interface, and then stick a cabinet impulse on there. HUGE sound all by itself.

[cratz2]

Cratz: I can't click back to last page as I type this. Was it you who said you experienced all kinds of bugs with the HD 500? I was just reading a thread over at SS.org where another user was saying the same thing. He also commented that he feels the quality of the models themselves has gotten WORSE with the last few updates they have released for the unit. Was this your experience as well? This is a bad development indeed, since I was considering TH2 or a POD HD as my next purchase.

 

No. No big bugs for me. I did originally get the HD400, then exchanged it for the 500 about two weeks later and I swear certain models sounded better on the 400 than on the 500 including the Park/Plexi. Maybe that was placebo effect, maybe not. My 500 had a later firmware and I also updated that to an even later firmware while I had it, but I tend to go with my hands and ears and my hands and ears said that MY 400 sounded better than MY 500.

 

[Arr0wHead]

That's what the guy on ss.org was saying, that after the firmware updates the models didn't sound as good. Hmm, this will definitely keep me from picking up one of these.

[cratz2]

I get what you're saying, but in this case don't focus that much on sound quality, but rather on the change in sound between the soft and hard parts. That way the focus is more on the actual responsiveness to a players touch, which was the purpose for these files. In terms of actual tone, I'm sure I could improve each one of the samples. I just plugged, selected a "Vox" model, and played. At the very least I usually adjust the eq. (bass, treble, mids, presence), and this time I did not even do that Listen again with this in mind, and let me know if this new listening approach changes your opinion (or not). I'll post the source of each Vox along with my favorites later today.

 

Ah... well, yeah... if your intent was to find which of those four had the most range from clean to gainy then I'd say #2 had the best range AND sounded the best. #3 and #4 had pretty decent range, but #3 sounded pretty terrible when full up, at least compared to the other selections. But sometimes those differences are relatively small when you only have access to one example, if you know what I mean.

 

For example, as I mentioned, I still have an original Tonelab and without a doubt, that's the modeler with which I have the most experience and the most actual hands on time. I've had a POD 1 for a very long time... Maybe 8 years, but other than a brief while, it was never my 'main rig' so to speak whereas the Tonelab very much was. Anyway, I had dialed in countless Vox presets... some clean, some with a tiny bit of breakup, some pretty much in JCM800 territory, but with the Vox flavor rather than Marshall. I have one preset that was literally the 4th preset I ever dialed in and I used it endlessly. Then I got a hair up my ass to make a bunch of presets and spent literally a week dialing in a bunch of tones, laboring over small details but flash forward a few months, pretty much any of the mid gain presets can be used, as long as you don't compare them to other, similar presets.

 

That's kind of confusing to word and I'm sure something was lost in the translation.

 

Basically, I have about 10 equally usable presets and they all sound great as long as I just plug in and turn on and go with any of them.

 

Probably similar to if I actually owned ten 1964-1966 AC15s. Each of them sounds great and works great, but they sound a bit different.