[Strangelove424]

If you roll off highs with EQ, roll off a little sub-bass as well, add white noise around the sound floor, a hint of inter-modulation distortion, you should have a fairly good idea of what exaggerated tube sound is. Not all tube amps were built to sound the same, but when people go looking for "tube sound", you can safely assume they aren't looking for tubes that measure/sound like an SS amp (which are quite expensive, and uneconomical compared to SS), they want the full on distortion experience. Getting back to the idea of using a software plug-in or filter, tube sound could be mimicked by introducing an artificial blend of distortion through sound processing. That's beside the ideological argument against distortion, the "transparent from the studio" unflinching accuracy routine, which each listener will have to debate and decide for themselves. Distortion can indeed sound pleasing, and a simple software solution could help any given listener decide whether the addition of an actual tube amp, and the distortion artifacts it would bring to their music, is welcome in their collection or not.   

 

[RoughSleeper]

I have tried a few "tube emulator" plugins and my problem with them is that all of them seem to be compressors in nature so whenever you turn up the dials that are meant to give off the tube sound the sound often distorts in an undesirable way when the music you are listening to is close to hitting the "brick wall". Because you are already compressing tracks that are most likely over-compressed to begin with.

I don't know if tubes work the same way exactely or what I said was completely true but those are my impressions of the tube plugins I have tried so far.

 

[Strangelove424]

Quote:

I have tried a few "tube emulator" plugins and my problem with them is that all of them seem to be compressors in nature so whenever you turn up the dials that are meant to give off the tube sound the sound often distorts in an undesirable way when the music you are listening to is close to hitting the "brick wall". Because you are already compressing tracks that are most likely over-compressed to begin with.

I don't know if tubes work the same way exactely or what I said was completely true but those are my impressions of the tube plugins I have tried so far.

 
Tube distortion has a few hallmark traits: slightly higher noise floor, less extension, and second order harmonic distortion that you can kinda sorta mimic with intermodulated distortion though they're not quite the same thing. I'm not sure the emulator you are using, but using multiple filters to achieve those goals, the data loss from processing becomes more apparent, especially if the files being filtered have low bit rates and/or heavy dynamic compression. The data will fall apart on you, like you were pulling on taffy that was too thin. Also, since you're still listening to an SS amp you won't get the soft-clipping benefits of a tube amp, so when all that processing leads to clipping of the signal (a "brick wall" of sorts), the clipping will still sound abrupt and harsh like a SS amp typically does. A tube amp clips more gracefully, but you really shouldn't be clipping your source at all to begin with. Ultimatley the best way to get natural tube distortion is via the electrons in the circuit, not the data in the source material.
 
That said, even if the emulator has its limitations, it's still useful. If you listen to a tube emulator and think "wow, that sounds like total garbage through and through" it might save you the cost of experimenting with a tube amp. On the other hand, if you enjoy the emulator (disregarding the limitations of its digital processing) and find yourself tempted to crank the emulated tube dial more, you may want to explore the tube amp experience a bit further, keeping in mind the digital "brick walls" aren't there on a real tube circuit. While a tube amp won't make compressed pop music sound any less compressed (claims that they do otherwise are audiophile wives tales) it certainly doesn't add any extra compression of its own. As has been mentioned, there are also tube buffers which would allow you to experience a more authentic tube distortion without the cost of an entire amp. There are times when a tube amp circuit is actually beneficial for more than just distortion, such as when a transformerless tube amp is outputting high impedance signals to high impedance headphones, but if you just want the tube distortion there are cheaper alternatives. Don't expect night/day differences either. The differences between certain tube and SS amps, in my experience, are very slight.    

 

[jaddie]

Tube distortion has a few hallmark traits: slightly higher noise floor, less extension, and second order harmonic distortion that you can kinda sorta mimic with intermodulated distortion though they're not quite the same thing. I'm not sure the emulator you are using, but using multiple filters to achieve those goals, the data loss from processing becomes more apparent, especially if the files being filtered have low bit rates and/or heavy dynamic compression. The data will fall apart on you, like you were pulling on taffy that was too thin. Also, since you're still listening to an SS amp you won't get the soft-clipping benefits of a tube amp, so when all that processing leads to clipping of the signal (a "brick wall" of sorts), the clipping will still sound abrupt and harsh like a SS amp typically does. A tube amp clips more gracefully, but you really shouldn't be clipping your source at all to begin with. Ultimatley the best way to get natural tube distortion is via the electrons in the circuit, not the data in the source material.

That said, even if the emulator has its limitations, it's still useful. If you listen to a tube emulator and think "wow, that sounds like total garbage through and through" it might save you the cost of experimenting with a tube amp. On the other hand, if you enjoy the emulator (disregarding the limitations of its digital processing) and find yourself tempted to crank the emulated tube dial more, you may want to explore the tube amp experience a bit further, keeping in mind the digital "brick walls" aren't there on a real tube circuit. While a tube amp won't make compressed pop music sound any less compressed (claims that they do otherwise are audiophile wives tales) it certainly doesn't add any extra compression of its own. As has been mentioned, there are also tube buffers which would allow you to experience a more authentic tube distortion without the cost of an entire amp. There are times when a tube amp circuit is actually beneficial for more than just distortion, such as when a transformerless tube amp is outputting high impedance signals to high impedance headphones, but if you just want the tube distortion there are cheaper alternatives. Don't expect night/day differences either. The differences between certain tube and SS amps, in my experience, are very slight.    

Huh. Well, there's a lot to disagree with here (data loss in DSP, second-order harmonic distortion simulated with IMD...no point in going on), but I will agree that listening to a few different tube-sim plugs is worth doing, if nothing else to reassure yourself that you ain't missing much.

My viewpoint, as someone that grew up with tubes, experienced early bad SS, then really good SS, studied David Manley's book (worth it, if you can find it), is: a bad amp is a bad amp, tubes or not, and a good amp comes as close as possible to a straight wire with gain. Changes in sonics can be inserted, if desired, but shouldn't be a forced mask over the original. It's just a philosophical difference. I like my signal processing and my amplification on separate plates, that way if I want the special sauce, I can have it, but if the chef already did a great job, why would I presume to re-season, especially with something that adds a flavor that was never intended to be there in the first place? Distortion, harmonic or non, is never an improvement over its absence.

 

[Strangelove424]

Quote:

Huh. Well, there's a lot to disagree with here (data loss in DSP, second-order harmonic distortion simulated with IMD...no point in going on), but I will agree that listening to a few different tube-sim plugs is worth doing, if nothing else to reassure yourself that you ain't missing much.

My viewpoint, as someone that grew up with tubes, experienced early bad SS, then really good SS, studied David Manley's book (worth it, if you can find it), is: a bad amp is a bad amp, tubes or not, and a good amp comes as close as possible to a straight wire with gain. Changes in sonics can be inserted, if desired, but shouldn't be a forced mask over the original. It's just a philosophical difference. I like my signal processing and my amplification on separate plates, that way if I want the special sauce, I can have it, but if the chef already did a great job, why would I presume to re-season, especially with something that adds a flavor that was never intended to be there in the first place? Distortion, harmonic or non, is never an improvement over its absence.

 
 
I leave philosophy out of my music enjoyment, and personally switch willy nilly between amps and EQ settings. I like variety, and I like the special sauce sometimes. Don’t confuse science with ideologies or philosophies. Science studies phenomenon to understand it better, and understanding the phenomenon of audio doesn’t necessarily translate into a philosophical conclusion about the ideal way in which it must be enjoyed. If a listener likes tube distortion, or is curious about tube distortion, and he or she wants to understand what tube distortion is, and different routes to getting it, it’s a pragmatic subject, not a philosophical one. Not every single discussion needs to become an ideological battle that invades on a listener’s personal listening preferences with external dictates and prohibitions.       

 

[jaddie]

Quote:

 
 
I leave philosophy out of my music enjoyment, and personally switch willy nilly between amps and EQ settings. I like variety, and I like the special sauce sometimes. Don’t confuse science with ideologies or philosophies. Science studies phenomenon to understand it better, and understanding the phenomenon of audio doesn’t necessarily translate into a philosophical conclusion about the ideal way in which it must be enjoyed. If a listener likes tube distortion, or is curious about tube distortion, and he or she wants to understand what tube distortion is, and different routes to getting it, it’s a pragmatic subject, not a philosophical one. Not every single discussion needs to become an ideological battle that invades on a listener’s personal listening preferences with external dictates and prohibitions.       

Well that's a mighty philosophical post for someone who leaves philosophy out of music enjoyment.  
 
Here's the problem.  The "tube" experience is only partially about what they do to the music.  The experience includes lots of other stimuli, visual (no tube amp ever has it's tubes not in full view, and what tube amp owner would hide it in a closet?) and tactile (you do feel the heat), that are highly influential.  Why do you think the tube sound is often called "warm"?  But, if you don't know you're listening to glowing glass bottles, the real tube sound alone isn't all that compelling.  That's why I suggested in another post that a tube-sim plug would not be representative of the experience, and I'd add that it probably wouldn't succeed.  To completely expose yourself to the tube experience, you have to have real, hot, glowing tubes, and know that's what's responsible for the sound you hear.  
 
There's a reason why, though the tube sound is not difficult or expensive to exactly duplicate, it's just not done.  The one manufacturer that tried it only met with limited success.  And today with a DSP here and a DSP there, running a really accurate tube-sim would be no great shakes at all.  But, do you fine the "tube" effect on AVRs?  Nope.  How about in high-end pre-pros?  Nope.  How about a really good totally analog simulation in a high end SS amp?  Nope.  It's not because it can't be done, or it's too expensive, or some other reference to practicality, it's because people don't want just the tube "sound".  That's not what's fun.  It's the entire visual, tactile, and aural experience. I know, I've owned tube amps, I've felt the heat, and the love. But take away the reinforcement of knowing you've got red hot filaments and hundreds of volts on the plates, take away the sparkling glass surrounded by glistening industrial design, and just leave the tube-processed audio, and you've missed the point.  Sure, you'll find an occasional listener who might like just the sound, but I seriously doubt adding distortion of any kind would win in a blind preference test.  "If you roll off highs with EQ, roll off a little sub-bass as well, add white noise around the sound floor, a hint of inter-modulation distortion, you should have a fairly good idea of what exaggerated tube sound is."  That quote, yours, pretty much says it all.  None of those modifiers get you closer to hearing the original music, and if taken far enough, will stand prominently in the way of the suspension of disbelief.  Not one advancement in sound recording and reproduction technology in 5 decades has moved toward deliberately making response non-flat, or adding distortion or noise.   In fact, technology has prominently moved in the other direction.
 
There are some who might say that the tube-sim, DSP or analog, isn't accurate enough, even when it's dead on.  What they are really saying is that without the sensory support of the physical presence of the tube amp, it's just not all that fun.  And I would agree.
 
What's particularly telling about tube-sims is where the find their market: music production, where tube amps become a part of the creative palette.  A market where you'll find the few devices that were designed to deliberately add distortion and create non-flat response.   
 
I would encourage anyone to find a couple of tube-sim plugs and go nuts, and I hardly call that invading a listener's personal preference with external dictates and prohibitions.  In fact, it's a fantastic idea, and it certainly has the economic upper hand over the alternative.  

 

[esldude]

Something that is revealing I have done is load a tube amp with power resistors, tap and reduced it so it basically has unity gain and feed it to a high quality SS amp.  You pretty well get all the effects of tube sound.  As some FR effects are related to output impedance, loading with power resistors isn't quite exact. But other things like mild compression, low order distortion and such are there.  If you wished you could even had some inductors or caps to simulate your loudspeaker load.  Have done this with ultralinear tube amps, triode in push-pull and SET amps.
 
Now in principle good plug ins could do all of this.  I don't know of those commonly available that do all of it. I have suggested to those that like Single Ended tube sound, they could use a 3 watt SET, feed it to a SS amp and not be so limited in speaker selection.  Most of course don't believe that.  But it is a good way to get that sound if you like it without going to horn speakers.
 
My personal preference in tube colored sound is push-pull triodes.  SET's are a bit too colored for my taste.  PP Triodes seem to fool me into thinking there is more there without bringing too much notice when listening to music. 

 

[penmarker]

Hey guys.
 
It's possible to emulate any sound characteristic if you have ideal equipments. In the case of tube sounds using computer plugins, I'm sure it is very possible.
Guitarists who use computers as their interface would know there are emulators that can emulate +6000 types of guitar amps and distortion pedals and etc almost perfectly. It's not a matter of "can or cannot?" it's a matter of how good the developer can create the plugin. Very much achievable, only limited to human abilities.
 
Thanks.

 

[cupofwrathcom]

According to this, Bob Carver (someone who actually knows about the differences in amps) says that you get a lot of the tube sound by increasing the impedance of the amp.
http://audiokarma.org/forums/index.php?threads/1-5-ohm-resistor-in-series-with-speaker.250403/
He said to place a resistor in series between the amp and speaker. This gives more a of disconnect between the driver cones and the amp which would make them less damped. I think that this would produce a more lively sound.

But the answer can't be that simple. It also means that tube pre amps aren't giving the real tube sound.

 

[pinnahertz]

According to this, Bob Carver (someone who actually knows about the differences in amps) says that you get a lot of the tube sound by increasing the impedance of the amp.
http://audiokarma.org/forums/index.php?threads/1-5-ohm-resistor-in-series-with-speaker.250403/
He said to place a resistor in series between the amp and speaker. This gives more a of disconnect between the driver cones and the amp which would make them less damped.

Remember, he said that the resistor and turning up the bass would get you to 80% of the tube sound, and that was with a power amp. In his initial infamous demo where he matched a SS amp to a reference tube amp it took quite a bit of experimenting and testing to get a good null with the tube amp. The resistor creates a variance in frequency response because the impedance curve of the speaker load along with the resistor creates a frequency dependant voltage divider.

I think that this would produce a more lively sound.

...OK...

But the answer can't be that simple.

It's not, there is a bit more to it. You cannot generate distortion and noise with a resistor. Then there's the huge psychological factor that with a tube amp you see these big glowing glass bottles, and can feel the heat coming off them. The is no question they will sound warm. But most of it is frequency response, and you can do a lot of that just by raising the amps output impedance with a resistor.

It also means that tube pre amps aren't giving the real tube sound.

Interesting conclusion. There is likely less tube sound, or a somewhat different sound with a tube preamp, but the much higher operating impedances and less linear characteristics, plus higher noise, do present a tube sound even with a preamp. It won't do the same thing as a power amp, and even tube power amps will interact differently with different speaker loads.

What has always amazed me is that the "tube sound" is though of as better. Having lived through the tube/ss transition, there was a time when some early SS amps did sound much worse than tube amps of the day. Then we went past that, and things got way better. Yes, it's different, but using a tube amp as an equalizer just to change the FR of the system never seemed a good idea to me. There also have been a few tube amps designed not to sound like tube amps. I'm not sure what the point of that was, but it has been done.

 

[cupofwrathcom]

It's not, there is a bit more to it. You cannot generate distortion and noise with a resistor. Then there's the huge psychological factor that with a tube amp you see these big glowing glass bottles, and can feel the heat coming off them. The is no question they will sound warm. But most of it is frequency response, and you can do a lot of that just by raising the amps output impedance with a resistor.

My point is not that there aren't frequency response changes and distortion in tubes. We get that. We also get that they look nice. But these factors are relatively unimportant to the tube effect, at least according to Carver. In fact a number of sources say that any distortion, even second order sounds lousy.

The main sonic benefit of tubes is that they sound more open and less "blanketed" than solid state. The higher output impedance could affect this.

How? You take a typical full range driver with a QES of .25, and put that in a big horn, and its going to be overdamped. You raise the output impedance of the amp and its no longer overdamped. This would make a noticeable difference, especially at lower volumes. However some speakers might no benefit much and others might sound worse.

The difference is physical. This article absolutely nails it. http://education.lenardaudio.com/en/14_valve_amps_7.html

It all depends on the properties of the speaker and amplifier etc. Connecting a tube amp to a boom box with a system QTC of 1.25 would sound very mushy I'd imagine.

 

[pinnahertz]

My point is not that there aren't frequency response changes and distortion in tubes. We get that. We also get that they look nice. But these factors are relatively unimportant to the tube effect, at least according to Carver. In fact a number of sources say that any distortion, even second order sounds lousy.

The main sonic benefit of tubes is that they sound more open and less "blanketed" than solid state. The higher output impedance could affect this.

How? You take a typical full range driver with a QES of .25, and put that in a big horn, and its going to be overdamped. You raise the output impedance of the amp and its no longer overdamped. This would make a noticeable difference, especially at lower volumes. However some speakers might no benefit much and others might sound worse.

There is no case where underdamping would "improve" the response of a speaker! In fact, that would compeltely ruin transient response. The high SS damping factors that people assume are also completely wrong becuase they would apply only if a SS amp were directly connected to the speaker with nothing inbetween, no wire, no crossover, etc...

The difference is physical. This article absolutely nails it. http://education.lenardaudio.com/en/14_valve_amps_7.html

The article contains a number of inaccuracies. Here are just two:
"A valve amp does not require negative feedback to enable it to function."
You can design a tube amp with no negative feedback, but there are other trade-offs. Most use some form of negative feedback.
"The primary purpose of Negative feedback is to quieten the amp and provide some damping to the speaker."
No, the primary purpose of negative feedback is to reduce distortion.

If the guy doesn't understand that, and that's just in one paragraph, the rest of the analysis falls apart.

It all depends on the properties of the speaker and amplifier etc. Connecting a tube amp to a boom box with a system QTC of 1.25 would sound very mushy I'd imagine.

Don't know, but likely.

The point of all of this is, nothing about a tube amp cannot be simulated in software.

 

[castleofargh]

My point is not that there aren't frequency response changes and distortion in tubes. We get that. We also get that they look nice. But these factors are relatively unimportant to the tube effect, at least according to Carver. In fact a number of sources say that any distortion, even second order sounds lousy.

The main sonic benefit of tubes is that they sound more open and less "blanketed" than solid state. The higher output impedance could affect this.

How? You take a typical full range driver with a QES of .25, and put that in a big horn, and its going to be overdamped. You raise the output impedance of the amp and its no longer overdamped. This would make a noticeable difference, especially at lower volumes. However some speakers might no benefit much and others might sound worse.

The difference is physical. This article absolutely nails it. http://education.lenardaudio.com/en/14_valve_amps_7.html

It all depends on the properties of the speaker and amplifier etc. Connecting a tube amp to a boom box with a system QTC of 1.25 would sound very mushy I'd imagine.

well that was a weird read. when Jesus and Karl Marx comes to tell us about how they like more bass no matter the consequences, so everything and everybody else is wrong and in a cult. amp manufacturers are wrong because they wished to get a more stable amp, speaker designers are wrong because they made speakers that would be flat and ring less with such a stable amp. and I like how flat response is always equal to lifeless all along his page.
and the funny part is that he mostly explains the technicalities of why it would be desirable to use low impedance amps and good damping factor.
then a strange statement about musicians who don't like SS guitar amps... hmmm ok so what? should I ask for an overdrive setting on my speaker/headphone amp? ^_^

anyway I'd like to point out that massive electrical overdamping doesn't mean that the speaker will behave that way. for starters as he mentions, we assume a fully resistive behavior and that's usually not the case. and then we have the actual mechanical behavior that the electrical damping will usually try to fight against. but anyway if some drivers end up better with a little less electrical damping, that can be measured. and if it's all about personal preference, then let people do what they like.
as for "the difference is physical", yes indeed, but those differences are mainly:
-some distortions, extra signal can be added at the digital level.
-some FR change, also can be done. if phase is a thing, that too can be done in the digital realm.
- as for the damping difference, it is one where the tube amp would usually have higher impedance resulting in decreased damping. so a signal that stabilizes slower and that's not really a problem to simulate with a more damped system. the other way around would be a different problem, if we had to make a low damping sound like a high one. because removing ringing isn't the same as adding some.

maybe I'm missing something, but right now I don't see why it shouldn't be possible to simulate most tube amp behaviors digitally. although I'm not really a fan so I never spent any time trying any VST of that kind, even for compressors if all they offer are some sort of tube effect I usually remove them right away. so I actually can't say how "good" those things can be. I'd imagine that the best tools are those for mastering engineers and are probably not oriented toward copying any specific consumer tube amp, but instead decompose each sort of effect for more versatility. at best I'd imagine they would try sometimes to mimic old vintage professional devices? but a plug in that copies the tube amp I just bought, I don't know of any.

 

[cupofwrathcom]

There is no case where underdamping would "improve" the response of a speaker! In fact, that would compeltely ruin transient response.

First its a matter of taste, a lot of people prefer over damped bass but some don't. Do some research on speaker qtc and speaker building. These are all well known principles. A qtc of .7 is considered ideal because the cone comes to rest faster than any other value, but it really depends on a lot of things. Also these concepts are not easy to apply to bass reflex and non sealed speakers, but transients still exist nonetheless.

The high SS damping factors that people assume are also completely wrong becuase they would apply only if a SS amp were directly connected to the speaker with nothing inbetween, no wire, no crossover, etc...

Huh? The tiny little impedance from proper wires has no noticeable effect. Generally crossovers are designed exactly not to add to the amps output impedance (remember zero output impedance is infinite damping).

The point of all of this is, nothing about a tube amp cannot be simulated in software.

No, the electro acoustic system of driver-amp-box and sometimes crossover will have unique properties for a given non-zero impedance. You can't change a physical value like damping through the signal itself. Perhaps you could mimic a certain sound with varying degrees of success.

 

[pinnahertz]

First its a matter of taste, a lot of people prefer over damped bass but some don't. Do some research on speaker qtc and speaker building. These are all well known principles. A qtc of .7 is considered ideal because the cone comes to rest faster than any other value, but it really depends on a lot of things. Also these concepts are not easy to apply to bass reflex and non sealed speakers, but transients still exist nonetheless.

So...generally...higher damping is desired. I agree.

Huh? The tiny little impedance from proper wires has no noticeable effect. Generally crossovers are designed exactly not to add to the amps output impedance (remember zero output impedance is infinite damping).

The series impedance of wire and crossover aren't tiny when compared to the output Z of a SS amp. Take that amp as some tiny fraction of an ohm, and add 20' of 16ga wire and who knows what all in a crossover that must work into an impedance of many, many ohms...you blow that amp damping factor away really quickly.

No, the electro acoustic system of driver-amp-box and sometimes crossover will have unique properties for a given non-zero impedance. You can't change a physical value like damping through the signal itself. Perhaps you could mimic a certain sound with varying degrees of success.

Whatever... You can model the entire mess. Start with the impulse response and convolve from there.

I really don't care what anyone thinks is the source of the "tube sound", or if it's better or worse in any specific case. All of it can (and is) modelled. Just fine.