Market Research: A amp modelling product

[tomographs]

Hi guys,



I'm a final-year Product Design Engineering student at Loughborough University, UK, and I'm undertaking some market research for my major project. I will be designing and building a new amp modelling product for guitarists, and would like to ask everybody a few questions.

With your answers I hope to piece together a detailed and concise snapshot of the thoughts of my target consumers of existing products on the market, and see whether there is indeed a need that has not yet been fulfilled.



There may be more questions in the future, but at this early stage I have just 10. You can answer them in as much or as little detail as you like. I'll be posting this on a few different forums, since I imagine needs will vary from group to group.



When answering the questions, please consider all devices that you are familiar with, both analogue and digital, that offer an amp-modelling solution (even if it is perhaps not it's primary purpose). Examples include: Consider POD, Sansamp, Axe-FX, Tonelab, PC/Mac Plugins, BOSS Multi-FX, etc.


1. In what areas do you believe amp modelling surpasses real tube amps? This might be in sound, convenience, versatility, I/O options, price, and so on. If there are features specific to a single unit only (e.g. An Axe FX or POD feature) then please state that. Also, if some of these advantages apply specifically to the kinda of venues or types of music you play, please comment on that also.

2. Likewise, in what areas to you believe tube amps are still superior?

3. In your experience it is possible to achieve an modelled sound that is indistinguishable from the original amp? (Of course there will be differences, but are you able to perceive them?) If only a certain product, or a number of products is able to achieve this, please state which.
(Please do not confuse this with the quality of the sound - please just comment on whether you can tell the difference).

4. Regardless of how similar it is to the original amp, can you achieve a sound using modelling that you enjoy as much as a sound shaped by a real tube amp? Again, if only certain products achieve this, please state which.

5. To what extent do you believe that DSP modelling devices sound 'digital'. Can you describe the sound, or tell-tale signs? If this only applies to some products and not to others, then please state which apply.

6. To what extent do you find the number of options available to the user overwhelming? Might this prevent you from purchasing a particular product? If so, please state which products you find overwhelming. (e.g. number of amp models and effects, I/O options, buttons, flashing lights, or other features)

7. Do you believe that price of amp modelling devices is restrictive for most? Which products specifically are pricing themselves out of the market?

8. As a performing musician, are you willing to sacrifice a degree of sound quality for convenience and simplicity? Or will you not compromise on sound, regardless of how much gear you need to lug about on public transport, or in your car?

9. Clearly the more you pay, the higher-spec hardware you get, resulting (usually) in better models and more complex algorithms in the case of DSP devices, or more complex circuitry in the case of analogue products. Which product do you believe offers the cheapest professional-quality solution, that achieves an acceptable sound?
To rephrase that, where is the price threshold, where modelling becomes acceptable for professional use?

10. In your rehearsals or live performances, do you regularly switch between clean and overdrive sounds within the same song? If the answer is yes, are floor-based modelling systems with built-in foot switches more attractive to you, or do you prefer a rack-mountable/other system, with an optional foot controller? Are you aware of any advantages or disadvantages with either design?


Thanks very much for your answers, and I'll keep you posted as the project progresses.

Tom

[bobhill]

No thanks. I'll stick with my amps.

[Enzo]

Taking this at face value, these are my personal opinions.  Others may disagree.

1.  Loaded question.   How about in what way do they surpass solid state amps?   In my view modellors do not surpass real amps.   Modellers make tones that sound kinda like this or that amp.  Fender-ISH, Marshall-ISH, etc.   What are they good for?  You can have all in one box amp that sounds more or less similar to a variety of things.  Maybe the easy way to cover tunes.   Like so many things, they do better when not listened to too closely.   I wish I had a dollar for every time I heard "I like my POD, but it sucks for live."

2.  Loaded question.  Not everyone prefers tubes, especially here on this solid state forum.  I like tube amps.  I like the dynamics of them and the way they work. (the performance, not the technology).  It won't sound or play like my Fender Twin Reverb.  It might be close enough for drunken night club dancers, but I'll know while I am playing.

3.  Possible?  Maybe.  Has anyone done it?  No.   There is and always will be the solid state versus tube arguments.   Someone will come up with some involved circuit with 15 op amps and a bunch of other stuff and claim it exactly makes the curves of a 12AX7... or whatever.   Well even if that were true, and we ignored other things like compression, then at best you would have some super convoluted circuit with hundreds of op amps, just so we can say we are not using a tube.   When someone finally comes up with a DSP that REALLY emulated my Fender, then I will be impressed.

4.  Enjoy?  Sure.  I usually enjoy the tone I get from an amp.   I like prime rib beef and baby back pork ribs.  That doesn;t mean I don't also enjoy a hamburger.

5.  As a player, I don't look at DSP as sounding digital, though on some occasions some things sound a bit "processed."  Can't articulate what that means really.  I look at DSP as either providing sounds I like or not, digital, schmidgital.    SOme poor (mainly much older) low resolution effects tend to tail off clumsily.

6.  I think many of us, if not most, when faced with the multi-efect boxes or multi-modelling amps, are not overwhelmed, if anything we are underwhelmed.  That DOD multipedal with all the settings and others like it?  AFter not long you usually find maybe three or four genuinely useful patches.   As to SELLING IT, whel the inexperienced are often really attracted to the eye candy of a million settings and a whole bunch of features.

7.  I have no idea what anything costs, I just fix the stuff and play it.   My impression is that the modelling amps tend to be entry level stuff like Line6.  Most are not aimed at the serious pro.

8.  There are the "I am a serious musician and I will not compromise on tone" types, and there are the guys who realize that people come to the bar to get drunk and get laid, not to critically listen to the band and comment on how the guitar tone was not exactly right on the Santana tune.

9.  I don't think the cost of the DSP silicon is the major cost factor.  Implementation is everything.

10.  Trick question?  Sounds like either way you have a foot controller.

[bobhill]

I don't mean to sound curt with my above response, so I will try to explain my feelings about modeling amps and pedals.

Like Enzo, Roly, Juan  and a few others on this board, I am an older player (Medicare next month ) with going on 50 years of playing with just about any type of rig I have ever dreamed of playing, from small amps up to Marshall Superleads' back in the 70's, but have pretty much stuck with Fender amps the last couple of decades ever since my hearing could no longer take the SPL from the bands I played in back then. Think of 110+ db stage levels four or more times a week for several years. This was before hearing protection became common, and like a lot of players from that time period I do have tinnitus as a result.

My current collection has three Fender SS amps and a pair of old Blackface Fender Bandmasters. I have quite a few pedals in my collection (20+), the majority that I have built myself. I use these to modify the guitar signal to produce whatever tones I am looking for out of the amps, but will usually use no more than three to five of them at any one time, something you will find with researching various pro rigs is pretty common. Unless you are Frank Marino, that is. With my rigs, I can produce the same sounds whether I am using SS or tubes for the amplification. It really doesn't matter to me.

I have tried out several of the Guitar processors that are on the market, and in my opinion (and that is my opinion, does not apply to anyone else) all of them suffer from a common fault. They all try to do everything and as a result they only achieve a certain level of mediocrity with nothing really standing out. I would much rather have them try to do just a few things and do those really well. Speaker and cab sims are really the weakest thing that modeling amps and processors try to do and fail to achieve. They may sound OK at low levels, but that does not scale at all well to gig levels. Think of it like a home stereo with the loudness button engaged at low levels to compensate for low base and high end. Works well at low levels but sounds like crap when the volume is pushed. Psycho acoustics comes into play whenever an instrument amp is pushed to the level where you are feeling the sound as well as hearing it.

I think that with the exception of AX-Fx, you will find that modeling amps and pedals are confined pretty much to the low end of the market, no matter how much power they push through however many speakers, and the resale of these amps kind of show that.

I do wish you luck with your endeavor, but expect to spend hundreds, if not thousands of hours and the money that entails in the programming of the dsp's and processors needed for the project. Then figure just what will be required to break into a market already dominated by the Line 6's, Boss, Digitech, Crate and others to get a small piece of that market. Admittedly, the beginner end of the market is the largest section of that market, but expect to have to produce a product that appeals to that segment of the market. And be prepared to make that product cheaply as a result for those beginners, as unless you are able to come up with something really outstanding, it will be hard to even get them into the local gig playing market. Forget the pros, it is hard enough to get them to use SS amps without modeling.

My suggestion would be to make your project as something that can be added to existing equipment. Looking at things from the maintenance side, there is nothing worse than a broken amp with a chip that was discontinued last year. All that can be done then is salvage the speakers, if they are worth salvaging, and the rest of it gets pitched into the landfill. That does not make the customer very eager to get your replacement product. Use common off the shelf components and keep as much of the programming off chip if possible, that way allows future improvements without having to pitch the previous generation. Look at the plug-ins available for DAWs for recording as an example.

I wish you luck in breaking into a rather small vertical market (figure less then 1% of the consuming population) that is becoming dominated by low end gear from China. You are entering the right end of the market, though. The only way to really make money in the music business is to sell to musicians, not playing music.

[Roly]

No interest.  "Do you want a real metal one, or a plastic imitation?"

After tech-ing and playing for the best part of 50 years I use a homebrew valve amp with occasional use of a minimal rack mounted analogue Fx unit (Ibanez UE400) for guitar, and a homebrew solid state for keyboards, simply because they have to do very different jobs.  I am quite capable of designing,  programming, and building my own digital Fx if the mood should take me, but it never has.

As a tech and performing musician I am horrified by the utter unrepairability of modern digital Fx, and their poor ergonomic design for stage use.  It is not a good look to see a very good musician made to look like a fool, doing a Riverdance, simply because of a cheap crappy power connector on their newly acquired wunderbox when it suddenly reverts to default in the middle of a lead break.  And it's often a case of "500 channels and nothing to watch" - about three useful patches hiding in a sea of selections.

I'm reminded of stories of Leo Fender leaping up on stage in the middle of a number and fiddling with the controls on some development amp he got the musio to test, and wonder if the creators of some of these modern units have ever even been to a gig, much less have a good understanding of the conditions under which they will be used.

It's a bit moot having 256 Fx selections if the rich tonality of passive guitar pickup resonances are first clobbered by a 100k amp input impedance.  And if you think I'm cork-sniffing, just try playing a $100 Strat copy into 100k then into a 2.2Meg FET buffer.

1 Megohm


100k

Ref: http://www.ozvalveamps.org/pickups.htm

As a keys player I have a lot of MIDI patches at my disposal, but the real problem is finding the right one for the song, and it's unusual for me to use more than four or six voices over an entire gig.

{@Enzo "3"  :lmao:  "6 & 7"   :dbtu:  post #3 yep, yep, yes, yep, yep, uh huh ... }

Frankly, like Enzo, I've taken your post at face value, but IMO it's right on the brink of being spam.    Quid pro quo - you want from us, but what do you offer in return?

[tomographs]

Thanks for your in-depth responses guys, it's very much appreciated.

I wrote the questions making the assumption that tube amplifiers were largely considered to be the benchmark of a quality guitar sound, and that other amp designs essentially emulate the tube sound and response. If you think that's a false assumption than that's fine, and useful info for me. I certainly overlooked the fundamental differences in opinion amongst different forum communities - apologies for that.

Once again this is very useful info and I'll keep you all posted as the project progresses.

If anybody else has any opinions, I'd love to to hear them - thanks!

[gbono]

I have a few questions for tomographs:

How are you impleneting the DSP processing - by any chance will you use an FPGA (i.e. Xilinx, Altera, etc)?

There is always a discussion regarding the electronics, either analog - tube vs solid state or digital - processing techniques. Musical instrument amplifiers are usually connected to loudspeakers. Will your design take this into account? Extra points if you even try to deal with the subjective nature of this issue - or maybe it is a marketing/brand feature ;-)

[Roly]

I wrote the questions making the assumption that tube amplifiers were largely considered to be the benchmark of a quality guitar sound, and that other amp designs essentially emulate the tube sound and response.

{at the risk of starting a s**tstorm...}

I think that is a valid assumption, it's just that it might not actually be true.

Like some of the others here I actually lived and worked through the middle of the valves vs transistors wars, and tried to remain Swiss.  I work on both, use both on stage, and on average, specifically for tenor electric guitar, valve amps seems to have an edge, but I'm dubious about valves for bass guitar, and wouldn't use them for anything else (e.g. like vocal PA).

Many early transistor amps were utter abominations but that has changed a lot in both directions; some modern valve amps sound pretty ordinary, and some modern solid-state ones sound great (particularly following the widespread adoption of "Valvestate" or current feedback topology which gives a higher output impedance).

A very important point that was ignored during the hot days of the V vs T wars was that while transistors could be superior using a Hi-Fi yardstick, tenor guitar amps are normally operated under overdrive conditions, something Hi-Fi-ists avoid like the plague, and above output stage clipping solid state amps generally sound pretty awful.  I attribute this to two main differences, the output transformer, and the very different levels of negative feedback; so in clip otherwise "identical" valve and transistor amps behave very differently.  These days it's not uncommon to see valve amps that don't use any NFB at all, while in solid state designs we tend to see signal clipping of more or less complexity applied upstream somewhere in the preamp so the actual s.s. output stage never does.

There have been occasions when I have had to crossover and use my valve amp for synth and s.s. amp for guitar, and while the synth/valve combination is just not very good, particularly with the latter there is an overload "wall" that you quickly learn to avoid at all cost.

So, if anything, there now seems to be an overlap between valve and s.s. tonality, and the assumption may be a bit dated these days.  It also has to be recognised that the V vs T arguments are not entirely rational, and that having a valve amp is a bit of a fashion statement.  The bottom line for me is that it's all very well having X, Y or Z gear, but what does it actually end up sounding like?  Cognitive dissonance, not actually hearing what you are sounding like, is sadly a very real affliction in some guitaring circles.

{grabs helmet and dives into fallout shelter.}

[Enzo]

Sure, there are a lot of people who want a tube amp because they feel they are SUPPOSED to want that.   Just as it is common to hear from young players who ask something like, "I just ordered my Peavey 5150.  What mods should I do to it.  I hear you should make the bias hotter.  What different tubes should I put in it."   They have jheard a bunch of things they SHOULD want to do, without really knowing what those things mean or even what the amp will sound like when it gets delivered.

I don't care if an amp is a wind-up or steam powered, whatever the technology behind it, there will be a wide range of performance and quality.  Solid stateness or tubeness is merely one factor.   One cannot make a blanket statement based upon it.    ANy more than "seafood" makes a definitive statement about food - that could mean anything from lobster or grilled sea scallops to cheap generic fishsticks.

[QReuCk]

Thanks for your in-depth responses guys, it's very much appreciated.

I wrote the questions making the assumption that tube amplifiers were largely considered to be the benchmark of a quality guitar sound, and that other amp designs essentially emulate the tube sound and response. If you think that's a false assumption than that's fine, and useful info for me. I certainly overlooked the fundamental differences in opinion amongst different forum communities - apologies for that.

Once again this is very useful info and I'll keep you all posted as the project progresses.

If anybody else has any opinions, I'd love to to hear them - thanks!

I'll voice a quite different opinion than Roly here: This asumption might be quite valid as far as marketability is concerned, but to my hears and without the notable exception of Kemper, SS gear that as been succesfull almost never is a true model of a specific valve amp. They sure try to reproduce some generally well accepted part of the behaviour of tube amps, but the good ones generally can't be looked at tries to exactly match one specific amp tone. They usually provide through various knobs and option a range of tones from witch some can approximate roughly what can be heard with various tube amps.

Not sure if I explain correctly, but if you take some of the well accepted and commercially available analog SS amps (think transtube, Tech21, etc...) the starting point is often to try to emulate the behaviour of tube amps, but admittedly or not without totally negating good aspects of SS gear (volume scalability, transparency against what instrument you plug into them, etc...).

Usually entry level DSP based products have a problem with transparency. I'm not an expert, but that would be logical for me that either the software (if not enough work is put in it) or the hardware (if undersized for the level of performance optimization vs complexity the software has) needs heavy filtering of the input signal to work properly, meaning, that all the subtle harmonics resulting from the delicate combination of sensible playing touch, strings, pickups might be lost in the process of feeding the software with something manageable for the processor.

As far as ergonomics is concerned, with the footswitchable 3 channels (highly configurable with independant character, drive, level and EQ controls), reverb and master boost of my trademark120 , the reactions of it to guitar volume pots level change and playing variation + a pre EQ stombox, I already have a lot more than what I do really need. Add to this the joker of a footswitchable stereo effect loop, and the fact you can get those things pretty cheap used thanks to the tube fashion, and you jump quite quick on the conclusion a DSP modeller would have to be darn good and inexpensive to convince me.

There are success stories in the DSP area though, but I fear this market is either limitted to semi-professional market (think Axe-FX, Kemper, Eleven Rack, etc...) or to entry level products (think Line6, Roland/Boss, Fender Mustang line, Vypyr from Peavey, Vox's take on that...).
For some reason, the presence of some glowing glass in the unit seams to help sell some of them at a bit higher price.

[Kaz Kylheku]

I think modeling is fantastic. It lets people get into guitar and make some nice sounds who will never play outside of their bedrooms, and who could never put together a decent tone using a normal way even if they spent their life savings on real non-emulated gear. Emulation is probably very useful to the pro. You can practice with headphones, do some convincing sounding direct recording, or even play live by going into a PA instead of lugging an amp.  Not all music needs the best possible guitar tone. Just like not all music needs the best possible sax or strings, and synthesized sounds are passable. The pro can set his ego aside and realize that sometimes his contribution is just a small part of a larger tapestry.

I think modeling can do a good job of things like speaker cabinets and room reverberation. However, the distortion isn't there. At least not the heavy distortion. The mild crunches and such, perhaps.

The problem is that if your digital sampling process introduces X decibels of digital noise, and then you simulate 80 decibels of gain (or whatever), then you have X + 80 decibels of digital noise. The quantization noise and jitter artifacts are amplified too by the same amount of gain.

These guitar boxes do not have the quality A/D conversion to begin with. Nothing like expensive audio interfaces used in professional studios.

I think it will be a long time before it makes sense to do anything but generate distortion with analog circuits and sample it afterward. At least for situations when music is being made in which the guitar is important, rather than just some embellishment to add a rock and roll flavor to a passage.

[lossfizzle]

A little late to the party, but hopefully this will still help you in your project. I think there are still some big gaps in the modeling-gear world and hopefully you'll steer your project in a direction that may help fill some of those gaps.

I'm new to the forum, but I'm going to take Enzo's cautionary approach here and suggest that you substitute "classic SS / tube amps" for "tube amps." I see you've already taken that into account. For the most part, at least when considering a good basic clean tone, I feel that good vintage solid-state amps are much closer to the performance behavior of conventional tube guitar amps than modeling amps (though, of course, you can't expect any solid-state amp to produce pleasing overdrive when the power section is hit hard), and for the most part I'm primarily using good vintage solid-state amps at this point in my playing "career." So all my answers pertaining to your "modeling vs. tube amp" questions are really responses for "modeling vs. vintage-style / purely analog SS amps."

1. The "sonic value" factor of modeling is certainly what has made it such a big deal in the last 15 years. There are certain types of tones-- particularly the super-distorted metal sort of tone in which playing dynamics are much less of a factor than they would be in lower-gain genres-- at which modeling devices excel dollar-for-dollar against the usual tube or even SS suspects that normally populate the boutique end of the equipment spectrum. The sheer number of tones to play with, along with the added possibility of DSP effects, makes them a great / somewhat pedagogically useful value proposition for less experienced players and less critical situations.

2. The biggest problems I have with most amp modeling algorithms is their lack of dynamic response and the palpable latency that occurs even on a faster, low-latency platform. I don't hear a lot of players critiquing the latter, but when I play through most modeling devices I can feel a clear "disconnect" with my guitar that doesn't occur when I'm plugged straight into a good analog pedal and amp, and I believe this disconnect to be related to inevitable DSP-related downtime, no matter how short. Most modeling algorithms do a very poor job with things like mild, dynamic overdrive that are easily attainable with even basic solid-state analog electronics, and I find that their clean tones are not much better, lacking the warmth and sparkle that any good tube / SS amp will effortlessly produce.

3. No. Further, I think it's easier for a picky modeling user to be satisfied when they are told that their model is only a "holistic" genre of tone (the Yamaha approach) vs. a specific amp (nearly everyone else's approach).

4. There are certain VST plugins (Studio Devil, some patches on Amplitube, the LePou plugins) that can generate nice-sounding results *on "tape"* with the right cab emulation selected. I also like the sound of the old Yamaha DG series amps in actual performance, and the Vox Valvetronix amps are passable if much less so, but unless I am playing death metal that day, I would generally still prefer to play through a good SS head with a couple good dirtboxes in front.

5. This is, of course, a classic "argument" not only in the guitar world but in recording and hi-fi as well. The comparatively useless adjectives you often see bandied about in regards to "digital-sounding" devices: "cold," "icepicky," "brittle," "one-dimensional," "uninspiring," "sterile." To try and describe these things more quasi-empirically, particularly in relation to various not-so-great-sounding guitar effects (and modelers) I've heard, they nearly all have to do with added (sometimes undesirable) clarity in the highs and the aforementioned missing dynamic distortion / compression characteristics. And I've heard many of these adjectives leveled somewhat unfairly at analog SS guitar amps as well. Amp modeling in general is an attempt to mimic the distortion and compression behaviors of valves, but at this stage in the game it's only gotten to a certain point, and the difference in the "grey area" of just-breaking-up is still so vast that I believe most players can still hear the difference.

6. I'm a tweaker, a guy who will occasionally get down to nuts and bolts in order to get a particular sound I can't get otherwise, so a well-designed product with a lot of depth doesn't scare me. If anything, I find that a big opportunity is being missed in terms of truly *new* sonic options. The modeling products out there, particularly effects floorboards, are more bent on mimicing "classic" devices with plug-and-play ease of use emphasized front and center over really wild new sounds and tweakability. I'm disappointed that nothing in the DSP world has come along which can baffle heads like the simple misuse of the LFO on an 80s Korg SDD-3000. This all said, when it comes to *live* use / performance and particularly my amp, I do value immediately accessible fixes and simplicity. I do not want to have to dig into a menu and hold down a value key to access the same behavior as the "bass" knob on a classic amp. An optimally designed digital device would let me dig into things to my heart's content on my own time, but keep it simple and fast to fix things that matter - volume, gain settings, basic EQ voicing, maybe the amp model itself - when my brain and hands are busy trying to make a performance happen.

7. Aside from some very expensive brand-name VST-only options (like what Waves used to offer), crazily overpriced devices like the Axe-FX, and perhaps to a much lesser extent the Yamaha THR series (a little tool I'd love to have but do not find worth the $300 asking price), modeling is about as proletariat as gear comes in the current market. This is as it should be. CPU time has never been cheaper, and when you're shoving nothing more than a basic modern solid-state power section, a processing chip and a speaker into an MDF box, there's no reason this stuff needs to cost very much money.

8. I think nearly all musicians are willing to compromise in certain situations. For instance, modeling is a compromise that actually makes more sense in many home recording situations vs. the headaches of cleanly recording a tube amp at whatever crazy required volume in a typical residential room. There are others who would feel differently and would never dream of using anything short of their nicest amp in the studio. When it comes to live situations, I want the gear to be the cheapest and simplest rig I can get away with. I have had way too much beer thrown at me on stage by overenthusiastic patrons to present them with a valuable, temperamental, fragile piece of equipment. (And that's where my love affair with old solid state amps began.) I think the flowchart of gear value / complexity is hard to sum up and will vary greatly from player to player, but I strongly suspect that most of us do have situations where equipment compromise makes a lot of sense. It's not like most guitarists can afford to be full-time divas.

9. The nod would have to go here to VST plugins; there are some really nice ones out there for guitar modeling which have zero cost to the end user, and they can definitely produce acceptable results when under the thumb of a patient and knowledgable user. Latency / interface issues aside, there's nothing that would stop you from using this stuff live, and I've seen more than a few bands that are obviously using their laptops as guitar preamps. Calculating the cost becomes trickier in this new world - do you include the cost of the laptop, especially when a) everyone has one and b) we're in an era when you can buy a VST-capable netbook for a hundred bucks? Oh yeah, and now there are also mobile iOS and, shortly, Android solutions that can achieve similar results for guitarists. In any event, your stated notion that quality goes up with cost has been totally thrown out the window in the last 15 years, and anyone designing a product to compete in this marketplace will have to deal with the reality of so many insanely cheap viable options / alternatives for those who can't justify, say, an Axe-FX.

10. I would definitely, without question, want the expensive guts of the box off the floor. I want the controls where I don't have to bend over and squint at them to adjust them in a live context. I don't want the CPU (or signal path) liable to disruption or destruction via my clumsy feet or those of my bandmates. I want to be able to replace the footswitch as easily / cheaply as possible if it gets broken and I want the show to be able to go on if it *does* get broken. 

[Frank]

I'll give it my thoughts as well. I haven't read the other replies (yet) BTW. My experience is mainly based on the excellent Vox AD120VTH top, that I have applied to several different cabs, plus the occasional checking out of other amps in shops.

1. In what areas do you believe amp modelling surpasses real tube amps?

The short answer is: It doesn't. At least not on the parameters in which tube amps excel. Modelling is handy if you are recording, and it brings a lot of versatility to the table. "Proper" amps, tube or SS, have one type of sound, while a modelling amp will have many.

2. Likewise, in what areas to you believe tube amps are still superior?

Tube and SS amps will last forever, as they are made of discrete components and are repairable. Modelling amps are built with custom chips that will eventually die. When this happens you may as well throw out the amp. That is the plain truth!

3. In your experience it is possible to achieve an modelled sound that is indistinguishable from the original amp?

No. Is it the purpose? I think a modelling amp can get very close, but indestinguishable? No. But that said, a modelling amp can get plenty of very useful sounds. In my book there are three equally rated types of amps; valve, solid state and modelling. I proudly own and play all three types.

4. Regardless of how similar it is to the original amp, can you achieve a sound using modelling that you enjoy as much as a sound shaped by a real tube amp? Again, if only certain products achieve this, please state which.

Yes, but only with the Vox AD120VTH, and the secret here is, that the output stage really is a valve coupled with a transformer and speaker with a feedback loop to the valve. A very unique feature for the original Blue Vox Valvetronix series (AD60VT/VTX, AD120VT/VTX). When these Vox are turned up loud (!), they really push like a true valve amp. All other modelling amps I have tried lack that satisfying sound, including Johnson Millenium and Line 6. Pure modelling from line-out sounds way artificial.

5. To what extent do you believe that DSP modelling devices sound 'digital'. Can you describe the sound, or tell-tale signs? If this only applies to some products and not to others, then please state which apply.

All amp modelling plugins I have heard on for instance Garageband sound ridiculously fake. Amp models on my effects units also sound fake. The main problem is the lack of feedback from amp to instrument, but modelling also sounds like it has some unnatural high frequency content.
I tried to put the Vox through an on Defiant cab. The Defiant has a Midax horn that enhances high frequency. The Vox sounded horrible through this cab with weird high freq artifacts I have never heard before, and I am sure this gave the amp away as being a modelling amp. On a traditional 2x12" cab it sounds just right though. But the real difference is what happens when you push the amp. A pure modelling amp - without an output stage like the Blue Vox Valvetronix - will not feel right when pushed. The Vox provides that crucial feedback from amp to instrument.

6. To what extent do you find the number of options available to the user overwhelming? Might this prevent you from purchasing a particular product?

No. But I keep it simple anyway. In general I use only the same 4-5 amp models, and same similar presets, and then add what I need from time to time. Blackface Twin, Tweed, 60'ies Marshall and Boutique Clean. I don't even use the Vox models, as I have the real thing. Better approach, I think, to get familiar with a limited set of sounds before moving on to other sounds. Also, I don't rely on the built in effects, but always use my traditional set of stomp boxes.

7. Do you believe that price of amp modelling devices is restrictive for most? Which products specifically are pricing themselves out of the market?

No, not in general. Actually I think you get a lot of value for the money.

8. As a performing musician, are you willing to sacrifice a degree of sound quality for convenience and simplicity? Or will you not compromise on sound, regardless of how much gear you need to lug about on public transport, or in your car?

I could definitely live with the Vox AD120VTH as my main performance amp, but then I don't think that is a compromise. I would not use another modelling amp.

9. Clearly the more you pay, the higher-spec hardware you get, resulting (usually) in better models and more complex algorithms in the case of DSP devices, or more complex circuitry in the case of analogue products. Which product do you believe offers the cheapest professional-quality solution, that achieves an acceptable sound?
To rephrase that, where is the price threshold, where modelling becomes acceptable for professional use?

I disagreed with the "the more you pay..."-sentence. I believe that the Vox Valvetronix Blue series is still the best value for the money that ever was in the modelling market. Forget about later Valvetronix. It's cheap and it doesn't have the same output stage. The new Valvetronix Pro are said to have the same quality output stage of the original, using an EL84, but they are too expensive considering that the old Blue series is practically up for grabs. Paying more that the used price of the Blue series amps is a waste of money. I'd rather stock up a number of AD120VT's for the future.

10. In your rehearsals or live performances, do you regularly switch between clean and overdrive sounds within the same song? If the answer is yes, are floor-based modelling systems with built-in foot switches more attractive to you, or do you prefer a rack-mountable/other system, with an optional foot controller? Are you aware of any advantages or disadvantages with either design?

I prefer real stomp boxes, anytime. The only built in effects I really use are those that are commonly associated with the actual amps modelled, such as tremolo, reverb and sometimes fixed frequency wah as it sounds like the old SS Vox MRB effect. I would like a foot controller for those, though.
And then, I'm a real old school guitar player, who believes that the coolest way to switch between clean and overdrive is by using the guitar volume knob. This demands a lot from the amp (model). If the DSP sucks, it will not deliver the proper dynamic response to the volume knob setting.