Quality of line out in load boxes (iron man mini 2 & suhr reactive load)

[Wonks]

...and if he increased the sample rate from 44.1 to 96 he'd reduce latency by a useful amount, too.

Presumably though the processor would need to be capable of dealing with the extra information? Fine if it can, but counterproductive if it can't as you'd need a bigger buffer to reduce the processor load and you're back to where you started. But the processor ten has to deal with all the info at the higher sample rate, so that's more than doubling the audio processing load of the processor.

[Hugh Robjohns]

True.... but worth doing if the processor has the capacity and the software is capable. It would probably improve the FX quality too. Certainly worth investigating, I'd say, as the converters probably account for a significant chunk of the overall latency, and running at 96k would more than halve their contribution.

[Wonks]

I don’t know all the the details but the processor is already overclocked to get it to do all that’s required, so there doesn’t appear to be a lot of spare processing overhead, unless he’s got it very wrong.

[James Perrett]

When reading the first part I was wondering at the madness of hacking a usb keyboard when the Pi offers plentiful GPIO. Glad he realised the error of his ways in part two.

...and if he increased the sample rate from 44.1 to 96 he'd reduce latency by a useful amount, too.

I must admit, having helped my lad with some of his Raspberry Pi projects, I read that and thought "yet another person who doesn't really know what he is doing". Using a USB keyboard and USB audio interface seems crazy when you have I2S and I2C pins available on the Pi. I know that there are tiny audio DAC boards available and I would assume that a quick search would turn up small audio ADC boards as well. That would be much more efficient compared to going through the USB interface.

And why not have some kind of display on there? They're not hard to implement. The next step would then be to make it fully controllable from the front panel with maybe a rotary encoder or two. And how about a variable pedal so that you can do wah or volume?

I don't think we've tried doing audio on a Pi Zero but a Pi 4 is perfectly capable of doing multitrack audio with some reasonably processor hungry effects.

[ajay_m]

And there is now a pi 5 with significantly better performance as well. I do like the idea of what was done here but would probably make some changes to the design for sure. But if you could build a decent pedal board for well under £200 I would have thought that might be quite attractive given the cost of commercial options. Hmmm.... Is a screen a good idea though or would doing things like the seqtrak where your phone, tablet or pc gives you a UI but the device itself doesn't have one makes sense. Just a display so you can see patch number for example? Or a small oled or TFT screen to show key info on the device?

[Wonks]

Commercial pedalboards with displays for £200 already exist.

I picked up one of these recently. https://www.andertons.co.uk/valeton-gp- ... processor/

The full-fat version is only £70 more and has more selector switches and a pedal. https://www.andertons.co.uk/valeton-gp- ... processor/

And both have things like three selectable input impedances (10k, 1 Meg and 5 meg).

Latency is apparently in the 5ms area, so not outstanding, but OK as long as you aren't using other digital effects or mixers in your chain.

[Martin Walker]

Commercial pedalboards with displays for £200 already exist.

I picked up one of these recently. https://www.andertons.co.uk/valeton-gp- ... processor/

The full-fat version is only £70 more and has more selector switches and a pedal. https://www.andertons.co.uk/valeton-gp- ... processor/

And thankfully evern though the prices are low, they're available from a reputable outlet rather than Temu

[johnwow13]

Hello! I want to ask something about TK2 Ironman.
I wanted to get rid cabinets and I bought TK Ironman 2 not for the attunate, but to use it on my tube amp as a loadbox and take the sound from it's analog cab simulator straight to PA. I can say that im not very happy with the result, because the sound is very bassy (from the speakers of PA - JBL PRX 815). I use TK2 on my 2 amps (DSL40 CR and PRS 15 MT) and on both, even if i have bass on 0 at the settings , the sound is still bassy and i need to cut bass also from ther mixer. Does anyone that have TK2 has the same problem using the cab similation straight to PA?

[Wonks]

I've never used the Tone King Iron Man , but have you ever used a modelling amp unit like a Helix?

I find that out of the box those are very bass heavy through almost all FRFR cabs, and usually need to really cut down on the low end in the output EQ section. I have a Fender Tone Master Pro as my modelling box and although the 1x12" cab models are fine, the 4x12" cabs have a very unnatural and excessive low end (before applying a lot of EQ).

So maybe it's just something about how manufacturers perceive the sound and fail to compensate for bigger FRFR cabs having a greater bass extension than the cabs or monitors they test them on so don't notice the bassiness. In theory a standard guitar shouldn't produce anything below 82Hz, but it feels like there's often a lot of stuff below that.

Ultimately, it is what it is, and you either complain to Tone King and ask them if there's anything they can do about it (maybe a component substitution makes them sound bassier than they should be), or else you decide to do something with the signal before it hits the PA cab and either use an in-line EQ or maybe a high pass filter module to reduce the level of bass.

[ore_terra]

Hi

There’s a lot of literature about load boxes impedance curves, and how the accuracy of these compared to a real speaker affects the low frequencies response. The Ironman fits in the not-so-accurate box . Not bad when just attenuating as the amp still “sees” the real speaker, but as 100% load boxes there is a difference.

This guy made an extensive comparison of load boxes including the explanation of why they don’t sound the same:

https://m.youtube.com/playlist?list=PLW ... b59VIY17ha

[ef37a]

Hi

There’s a lot of literature about load boxes impedance curves, and how the accuracy of these compared to a real speaker affects the low frequencies response. The Ironman fits in the not-so-accurate box . Not bad when just attenuating as the amp still “sees” the real speaker, but as 100% load boxes there is a difference.

This guy made an extensive comparison of load boxes including the explanation of why they don’t sound the same:

https://m.youtube.com/playlist?list=PLW ... b59VIY17ha

That video bothers me, maybe I am missing something and Wonks can help?
The output resistance of a valve guitar amp will not be close to zero, more likely 10 Ohms as a minimum. If you therefore measure the frequency response across a reactive load (which by definition has a resistance which changes with frequency) I do not see how you can ever get a flat response?

Dave.

[Wonks]

I haven't watched it all but I'm not sure exactly what he's doing with FabFilter there. The thick white line doesn't seem to correspond with the shape of the signal beneath it.

I don't know if he's using a recorded white noise source or using fresh white noise each time. With a recorded signal you can simply level match, invert one signal and sum the difference. That will show the difference more clearly.

I'm also wary of the fast update time on the real-time display. Unless the sampling period for the frequency analysis is pretty long, the low frequencies aren't displayed correctly. And you need to average the frequencies over at least 20 seconds to get a good full frequency picture as any noise source is producing random frequencies that only give you a proper straight line when averaged over a relatively long period.

I'll need to watch it again to see if I understand just what's going on.

And, yes Dave, I'd expect a natural variation in output levels. What was the amp setting. Clean or distorted? Same level each time, tone knobs not adjusted since the initial test? Was he going in via the FX return to bypass the preamp? And what was the signal like when using a real speaker?You've got a resonant peak in impedance that can sit between maybe 40Hz and 120Hz depending on the speaker. If you model your system on a speaker with a low resonant peak, you might not bother with getting the impedance down below that peak as they are frequencies a guitar won't produce (just guessing here as it may be easier just to mimic a speaker coil).

There are also phase issues which could make a summed signal reduce in level, especially around the resonant peak area.

As I said, haven't watched it all so might have missed important bits.

[Wonks]

I've watched the start of the testing in the video several times, and I have no clue what he's actually comparing. I'm not sure either where he's taking the signal. Is it from the emulated output. I don't even know if there's any compensation for the latency delay.

Anyone else have a clue?

[ore_terra]

check 2:40 and 3:35. his amp has an actual line out post power amp. the comparison is therefore before and after the load box.

flat line in fabfilter pro Q3 would mean both signals are similar and load box line out is "transparent"

[Wonks]

So it's the amp signal going into the load box and the amp signal coming out? A view of the actual DI connection arrangement wouldn't have gone amiss there.

It's then going to depend on the circuitry in the load box.

Again, a comparison with an actual speaker connected would have been more telling as to the expected shape of the output. The variable impedance of a speaker should not give a flat frequency response, so anything very flat indicates a pure resistive load.

So it may well be that anything that gives a flat response after the load box isn't a good 'reactive' load box but is mainly (or purely) resistive. So you really need to see what the profile should be from a couple of speakers and compare the shape of those curves to those of the loadbox outputs to the speaker.

[Wonks]

Thinking some more, the load box needs to present a near constant impedance to the valve amp, so the load box input circuitry is very likely to be buffered from the output circuitry. Otherwise as more attenuation is applied, the valve amp would see different impedances, which wouldn't be good for the output transformer if the impedance went too high.

So I'd definitely expect to see some difference between input and output.

It would be nice to see how steady the input signal level stayed as the load box attenuation went from minimum to maximum. A big variation would show it wouldn't be a good choice for the longevity of your amp.

[ore_terra]

So it's the amp signal going into the load box and the amp signal coming out? A view of the actual DI connection arrangement wouldn't have gone amiss there.

you have the view in 3:25

amp signal coming out of the load box (to the speaker).

in parallel, line out from the amp DI'd to the intereface.

It's then going to depend on the circuitry in the load box.

yes! that's what the guy's testing.

Again, a comparison with an actual speaker connected would have been more telling as to the expected shape of the output. The variable impedance of a speaker should not give a flat frequency response, so anything very flat indicates a pure resistive load.

So it may well be that anything that gives a flat response after the load box isn't a good 'reactive' load box but is mainly (or purely) resistive. So you really need to see what the profile should be from a couple of speakers and compare the shape of those curves to those of the loadbox outputs to the speaker.

watch the video... in this part he's not looking at the load box impedance curve just the "transparency" of the line out of the device.

flat or not flat is because he's using amp's line out as reference, and then measuring the line out of the load box against it. flat means they're the same; not that the impedance curve is flat (he's not measuring impedance curve here)

[Wonks]

I still want to see exactly where he's connecting to. If he's measuring the emulated speaker output of the loadbox, then there's no way it's ever going to be flat as speaker responses aren't flat. So I'm not sure what the actual point of the video is. I see he's got a load of other vids on the subject, but hey seem to be more ones where he knows what he's doing and talking about but isn't explaining fully to the viewers.

[ore_terra]

it's it not what he is measuring, mate. it is what he is comparing

he's taking the amp line out as reference, and then stacking the load box line out in fabfilter. Flat line only means both measurements are equal. but we are not seeing any of the actual responses.

he does measure responses in further videos of the series.

[Guitarking]

Update:
I've have figured out a way to have a real speaker 'float' inside my iso cab.
The baffle is basically suspended with strong rubber straps which are attached to a frame I made from wood. I will add a picture.

https://imgur.com/a/LDGsb3I

The bass transmission is much less now!
I will not buy a reactive load, this is great this way I think. No noise, but still the impedance curve of a real speaker!

I have a question though: does taking a speaker out of a cab change it's impedance curve? Or is this to be neglected?

Tnx!

[ef37a]

"I have a question though: does taking a speaker out of a cab change it's impedance curve? Or is this to be neglected?"

Yes! Well it makes a profound difference for "proper" cabinets, i.e. "hi fi" which guitar cabs decidedly are not! Still, the two major difference are that a totally sealed cab (most 4X12s) will have a single bass resonance. That peak is partly due to the natural cone resonance of the speaker(s) and partly due to the 'spring/mass' quality of the air in the cab. A reflex, ported cab will have two such resonaces, one the box+cone and another from the port.
In free air there will only be the cone resonance and that will not be quite right because there is no cab. A single 12 in an open backed cab, a combo is essentially that, will I suppose approximate to free air conditions?

Do keep in mind Wonk's warning regarding compromised power handling? Both from a bass capability and thermal. You do NOT want the speaker on a valve amp to burn out O/C at full chat!

Dave.

[Guitarking]

Yes!
I have my loudest amp (Reinhardt jkc33 33w) on powerscaling.
I have a 65w speaker in the isocab.
I have also put a fan in the isocab for some airflow.
Hope this will be safe!
Thanks!

[level-eye]

I'm late to this party so my comments won't necessarily be relevant, and are inevitably subjective and according to my own tastes.

I've been using a Suhr r/l box with built-in IR's for some time (but not used any from other manufacturers). The amp is a 50w PRS Archon, and I also have a PRS single speaker cabinet.

When using the r/l box (into a Focusrite Clarett+ 8pre) I haven't discerned any detrimental HF roll-off or noticeable impact on the overall tone. It's surprisingly close to what I get when using the speaker cabinet. However, this could be subject to the type of guitar/amp combination in front. My EBMM JP15 has powerful humbuckers whose character I would describe as dark, woody, with a pronounced midrange, and the sound of the amp is very much the same. Not a bright sounding combination then, so maybe this is a case of a guitar and amp combination that is well-suited to this load box (and my ears ). Dialling in different IR's changes tonal characteristics to varying degrees, yet the overall character of the guitar/amp is retained.

I've also noted that a compressor in the signal chain, even with moderate settings, helps to tame wayward transients that still manage to creep through despite proper gain-staging and allowing some headroom. Bearing in mind the use of compression will also change tonal characteristics, to one degree or another! I've also experienced no grounding or noise floor issues, which I was concerned about before purchasing the unit.

Prior to the PRS amp I was using a Marshall JVM205C (with an obviously different sound palette), and my experiences were much the same.

I don't own a Strat or Tele, but maybe these guitars with single coils might induce the r/l box to exhibit a more pronounced difference, e.g. significant HF roll-off. Either way, there are lots of variables to consider, but judging by how established most of these type of units now are, most will do the job, subject to personal expectations.

John