Modeller "Dynamic Range" ... how meaningful is it for Guitar Modellers ?

[BenIfin]

Hey all !

This has always had me a bit confused ..... just did a quick look and found some results:-

=> Kemper 105db
=> Helix Floor 123db
=> Opus Pedal 101db
=> FM3 114db
=> Fender TM Pro 117db
=> NUX Amp Academy 110db
=> Boss GT1000 Couldn't Find
=> QC Couldn't Find

What exactly does it mean in real-world terms for guitar playing ?

Does higher always = better for guitar playing ?

Will a unit with a 101db input range always be less dynamics and have less feel-range for guitar playing than one with a 123db range ?

What's the dynamic range of a guitar when played through a high quality "real" Tube Amp ?

I have always *assumed* that electric guitar always had a [relatively] fairly narrow-limited dynamic range ? compared to say something like an acoustic drum kit or a brass wind instrument ?

Ben

 

[Ed DeGenaro]

Hey all !

This has always had me a bit confused ..... just did a quick look and found some results:-

=> Kemper 105db
=> Helix Floor 123db
=> Opus Pedal 101db
=> FM3 114db
=> Fender TM Pro 117db
=> NUX Amp Academy 110db
=> Boss GT1000 Couldn't Find
=> QC Couldn't Find

What exactly does it mean in real-world terms for guitar playing ?

Does higher always = better for guitar playing ?

Will a unit with a 101db input range always be less dynamics and have less feel-range for guitar playing than one with a 123db range ?

What's the dynamic range of a guitar when played through a high quality "real" Tube Amp ?

I have always *assumed* that electric guitar always had a [relatively] fairly narrow-limited dynamic range ? compared to say something like an acoustic drum kit or a brass wind instrument ?

Ben

Yeah not what it is…
Dynamic range is the distance from lowest possible sound before self noise sets in I.e. noise floor and the loudest possible sound produced.
Think about this way the DNR of 24 bit audio 144dB not like any gear get to 144dB.

The human ear does about 30 to 120 dB …that’s 90dB DNR.

What the numbers tell you is part of the story, as if you have signal to noise ratio which will be lower than DNR the difference is the headroom before digital clipping.

 

[Sascha Franck]

I think one of the main benefits of greater dynamic ranges is simply that you can be much more generous regarding input levels. Pretty much similar to 16bit vs. 24bit recordings (96 vs. 144 dB), so you can always leave quite some headroom.
In the land of modeler input I guess that this might translate to more realistic input behaviour as you could already sufficiently use the entire dynamic range of whatever guitars without ever running into the risk to clip your inputs when, say, using a boost or a guitar with pretty hot pickups (or even a combination of that).

 

[MirrorProfiles]

What's the dynamic range of a guitar when played through a high quality "real" Tube Amp ?

I have always *assumed* that electric guitar always had a [relatively] fairly narrow-limited dynamic range ? compared to say something like an acoustic drum kit or a brass wind instrument ?

Yep.

A real guitar amp has inherently more noise because valves, high voltages, transformers etc al pose issues that modellers don’t have to worry about.

There’s also the noise that gets into the pickups, even with good shielding and grounding etc, the level of the guitar signal is relatively weak and prone to background noise seeping in. The earlier in the chain the noise is, the more it’s amplified.

Same goes for pedals, there is limitations on how low the noise will get.

Lower is more betterer but you’ll likely never find yourself in a situation where you can notice or that it matters

 

[laxu]

Yep.

A real guitar amp has inherently more noise because valves, high voltages, transformers etc al pose issues that modellers don’t have to worry about.

To add to this, the noise is somehow also part of the sound. Whenever I flip my single coil guitars to a dual mic setting where they are out of phase and the noise goes away, it somehow sounds "wrong". Of course it doesn't mean I like having noise, but it becomes part of "the sound."

The lack of noise has long been a telltale sign in "modeler vs real amp" clips too.

 

[MirrorProfiles]

To add to this, the noise is somehow also part of the sound. Whenever I flip my single coil guitars to a dual mic setting where they are out of phase and the noise goes away, it somehow sounds "wrong". Of course it doesn't mean I like having noise, but it becomes part of "the sound."

The lack of noise has long been a telltale sign in "modeler vs real amp" clips too.

Yep, absolutely.

Very true for plugin emulations for all types where they add noise otherwise when people do blind tests or comparisons they’d say they don’t sound right.

Weirdly for effects like reverb/delay/chorus/flange etc where there is feedback loops, the noise really changes the overall sound a lot and is an essential part of the charm of old gear. In tape machines it unintentionally covers up other noises.

 

[Sascha Franck]

To add to this, the noise is somehow also part of the sound.

True, noise is good at "masking" dynamics. The higher the noisefloor, the less (at least perceived) dynamics.
There's been some experiments they made in Japan (I think) in those big ass bureaus/offices, where you had multiple folks hammering good old analog type writers. They were then playing (or rather "injecting") some white or pink noise through multiple speakers, which would in fact raise the noisefloor and hence lower the perceived dynamics of those hammered type writes. IIRC, even if it seemingly was successful, they got rid of it, though, because in the end people were exposed to more noise, not less - which was raising the stress level more than the typewriting.

Fwiw, personally I still prefer noiseless pickups, including pseudo singlecoils. It might be a pretty different story in case the noise level was always the same, but as it's pretty much completely unpredictable, I rather don't deal with it at all.

 

[Cirrus]

In the real world, if I'm gain staging a mic to peak at -10db I feel pretty lucky if the rig has a noise floor around -50dB

 

[jay mitchell]

There's been some experiments they made in Japan (I think) in those big ass bureaus/offices, where you had multiple folks hammering good old analog type writers. They were then playing (or rather "injecting") some white or pink noise through multiple speakers, which would in fact raise the noisefloor and hence lower the perceived dynamics of those hammered type writes.

This has been understood and used for several decades. There is a field of audio/acoustic design known as sound masking. In open-plan office spaces, the sound from people speaking in nearby cubicles can be distracting and privacy-invasive. Tailored noise at an appropriate level can mitigate this. See Sound Masking 101.

 

[FractalAudio]

I think one of the main benefits of greater dynamic ranges is simply that you can be much more generous regarding input levels. Pretty much similar to 16bit vs. 24bit recordings (96 vs. 144 dB), so you can always leave quite some headroom.
In the land of modeler input I guess that this might translate to more realistic input behaviour as you could already sufficiently use the entire dynamic range of whatever guitars without ever running into the risk to clip your inputs when, say, using a boost or a guitar with pretty hot pickups (or even a combination of that).

No. Dynamic range doesn't set the maximum input level. Maximum Input Level sets the maximum input level.

Dynamic range in a modeler determines how much noise you will hear when using amp models with moderate to high amounts of gain. For example, if the DR is 100dB and you've set the input gain such that you just barely hit full-scale then the noise from the conversion will be -100dB re. FS. If you now apply, say, 60dB of gain the noise will be -40 dB re. FS.

Now, the quoted DR for a product is not a standardized specification. For our products we use the quoted DR of the A/D converter. However, we use multiple channels of the converter so the effective DR is around 132dB for the instrument input. I'm guessing the Helix is using effective DR since it uses the same converters as our products. The other inputs on an FM3/9 are 114dB.

 

[Sascha Franck]

Dynamic range doesn't set the maximum input level.

Hm, I did not say that. A language thing?

 

[yeky83]

Hm, I did not say that. A language thing?

I think one of the main benefits of greater dynamic ranges is simply that you can be much more generous regarding input levels.

 

[Sascha Franck]

@yeky83 - but that hasn't got anything to do with "setting" the input level.

 

[Cirrus]

Yeah I think all Sascha is saying is that lots of dynamic range means you've got a much wider useable range of input below the typical levels we'd normally shoot for - playing really softly or accidentally messed up gain staging and getting peaks at -45dB? Want to boost it right up in a mix? No problem. Less so if you're using say an old 12 or 13 bit digital unit.

 

[Ed DeGenaro]

Yeah I think all Sascha is saying is that lots of dynamic range means you've got a much wider useable range of input below the typical levels we'd normally shoot for - playing really softly or accidentally messed up gain staging and getting peaks at -45dB? Want to boost it right up in a mix? No problem. Less so if you're using say an old 12 or 13 bit digital unit.

Still not how it works…multiple guys explained what it is, yet folks gleefully continue to believe what they think “dynamic” implies.

 

[DLC86]

No. Dynamic range doesn't set the maximum input level. Maximum Input Level sets the maximum input level.

Dynamic range in a modeler determines how much noise you will hear when using amp models with moderate to high amounts of gain. For example, if the DR is 100dB and you've set the input gain such that you just barely hit full-scale then the noise from the conversion will be -100dB re. FS. If you now apply, say, 60dB of gain the noise will be -40 dB re. FS.

Now, the quoted DR for a product is not a standardized specification. For our products we use the quoted DR of the A/D converter. However, we use multiple channels of the converter so the effective DR is around 132dB for the instrument input. I'm guessing the Helix is using effective DR since it uses the same converters as our products. The other inputs on an FM3/9 are 114dB.

But does it really matter when the noise introduced by the guitar (EMI, hum, etc.) is much higher than that? I guess it might in some particular cases, e.g. when rolling off ther guitar volume...

 

[Cirrus]

Still not how it works…multiple guys explained what it is, yet folks gleefully continue to believe what they think “dynamic” implies.

Getting into arguments about digital theory on forums is pretty much the most boring thing imaginable, I just came in to reassure @Sascha Franck that not everyone is willfully misinterpreting what he's saying in order to tell him he's wrong. I won't be back in this thread.

 

[jay mitchell]

The phrase "dynamic range," like so many well-defined engineering/scientific concepts, is so poorly understood and widely abused as to have minimal meaning to a consumer interested in making a purchase decision.

The actual definition of dynamic range is the difference (in decibels) at a device's output between the noise floor and the strongest signal the device can produce. The conditions of measurement must be specified, e.g., input terminated with x ohms resistance, system voltage gain of y dB, measurement bandwidth of z decades, etc. Best-case, dynamic range characterizes the performance of a single device or subsystem. When multiple devices are cascaded, the total DR cannot exceed that of the component with the lowest DR. Optimal gain structure will create the result that the output noise level of the system is equal to the noise at its input multiplied by the total voltage gain in the system.

Here's a very important fact: guitars have dynamic ranges that are substantially less than that of a halfway-decent (or better) signal processor. IOW, best case, the limiting factor in the DR of a guitar/modeler is the instrument, not the processing. It's certainly possible to reduce DR even further than that of the guitar, but you'll never make it any better.

Another important fact: nonlinear processes that involve gain reduction - compression, limiting, etc. - reduce dynamic range by the amount of the reduction.

 

[OneEng]

According to Christof Kemper, the new Player has 127db dynamic range and is the same as the main profiler, rack, and stage units.

I did find quite a bit of older stuff on the web citing >108db for the KPA. Perhaps they have cleaned up their noise floor to raise the numbers?

 

[FractalAudio]

But does it really matter when the noise introduced by the guitar (EMI, hum, etc.) is much higher than that? I guess it might in some particular cases, e.g. when rolling off ther guitar volume...

If you roll the volume all the way off then you'll hear the noise. The greater the DR, the less audible noise.