New Headrush pedalboard: Headrush Prime

[Sascha Franck]

I made a quick size comparison on Pedalplayground of Headrush pedalboard in the back and then Helix LT, Boss GX-100 and Stomp XL in the front.

The GX-100 comes at a most excellent form factor. Too bad that, in case I went Boss (which I may in fact do one day), I'd want some features exclusive to the GT-1000 (which fortunately is at least just somewhat deeper but not wider).

 

[Evilution]

The GX-100 comes at a most excellent form factor. Too bad that, in case I went Boss (which I may in fact do one day), I'd want some features exclusive to the GT-1000 (which fortunately is at least just somewhat deeper but not wider).

Ok, what features is it that you want that the GT-1000 has but not the GX-100 ?

 

[Sascha Franck]

Ok, what features is it that you want that the GT-1000 has but not the GX-100 ?

Mainly global blocks and the more complete output section (I like to have my global EQ on my monitoring path only, that's not possible with the GX - ok, maybe with some trickery, but the GT is ready to roll from the start), then the 6 encoders and 2 FX loops. I also seem to remember there's some more FX and amp blocks, but that wouldn't be as important. If the GX had global blocks, I'd possibly think about it again, though, because that's really the most important feature for me. But then, it hasn't - and I don't think Boss will add that.

 

[the swede]

Except it sounds really bad. If it had the oversampling update, I would be ll over it.

I need to determine if i can hear the difference. Cant say for sure 100% heard a difference with my stomp after that update. But at least I felt better

 

[JCW308]

I need to determine if i can hear the difference. Cant say for sure 100% heard a difference with my stomp after that update. But at least I felt better

No squirrely bits / fizz anymore after the update.

 

[Jim Roseberry]

Measured latency of the Headrush Prime (using an Impulse response).
Audio interface was loop-back tested before... to make sure the recording process wasn't adding/subtracting samples (latency).


Rig with just Amp/Cab models is 5ms

Rig with EQ>Drive>Amp/Cab models is also 5ms (a couple additional samples vs just the Amp/Cab)

Rig with Amp Clone is 14ms (see below for clarification)

Rig with Pedal Clone > Amp Clone is 26ms (see below for clarification)

Side by side look

I created the Amp Clone using one of my favorite Helix BE-100 patches.
This was done to keep things quick/easy.
Helix's round-trip latency is ~2ms.
The Pedal Clone was done using a real Tube Screamer.

Don't know about anyone else, but 12-14ms is too much for my liking.
When running a pair of Clones (Pedal/Amp) in series, 26ms (for me) makes it unusable.

The Cab "model" was the CT 4x12 Bogie V30 IR

 

[Sascha Franck]

Don't know about anyone else, but 12-14ms is too much for my liking.

Complete no-go. I would sort of get along using monitors but it'd drive me mad using headphones.
Thanks for measuring!

 

[Bob Zaod]

Measured latency of the Headrush Prime (using an Impulse response).
Audio interface was loop-back tested before... to make sure the recording process wasn't adding/subtracting samples (latency).


Rig with just Amp/Cab models is 5ms
View attachment 5782


Rig with EQ>Drive>Amp/Cab models is also 5ms (a couple additional samples vs just the Amp/Cab)
View attachment 5783


Rig with Amp Clone is 14ms (see below for clarification)
View attachment 5784


Rig with Pedal Clone > Amp Clone is 26ms (see below for clarification)
View attachment 5785


Side by side look
View attachment 5786


I created the Amp Clone using one of my favorite Helix BE-100 patches.
This was done to keep things quick/easy.
Helix's round-trip latency is ~2ms.
The Pedal Clone was done using a real Tube Screamer.

Don't know about anyone else, but 12-14ms is too much for my liking.
When running a pair of Clones (Pedal/Amp) in series, 26ms (for me) makes it unusable.

The Cab "model" was the CT 4x12 Bogie V30 IR

I start to feel it around 6-7 and anything over 10 for me is a no go.

 

[Evilution]

Measured latency of the Headrush Prime (using an Impulse response).
Audio interface was loop-back tested before... to make sure the recording process wasn't adding/subtracting samples (latency).


Rig with just Amp/Cab models is 5ms
View attachment 5782


Rig with EQ>Drive>Amp/Cab models is also 5ms (a couple additional samples vs just the Amp/Cab)
View attachment 5783


Rig with Amp Clone is 14ms (see below for clarification)
View attachment 5784


Rig with Pedal Clone > Amp Clone is 26ms (see below for clarification)
View attachment 5785


Side by side look
View attachment 5786


I created the Amp Clone using one of my favorite Helix BE-100 patches.
This was done to keep things quick/easy.
Helix's round-trip latency is ~2ms.
The Pedal Clone was done using a real Tube Screamer.

Don't know about anyone else, but 12-14ms is too much for my liking.
When running a pair of Clones (Pedal/Amp) in series, 26ms (for me) makes it unusable.

5ms is the same as the Gigboard according to Leo Gibsons tests and should be just fine but over 10ms seems like it could start to become too much, 26 seems ridiculous. Also consider if you use a digital wireless as well or even some digital pedal in the loop, even with the lower 5ms to start with can start to build up in that case. BTW Boss was the clear winner in those latency tests with 0,7ms for those that might not know it.

 

[JiveTurkey]

It was nice while it lasted, Headrush Prime.

 

[2112]

HR/FM9/QC/TX/Stomp Size comparison

 

[Lysander]

Rig with just Amp/Cab models is 5ms
Rig with EQ>Drive>Amp/Cab models is also 5ms (a couple additional samples vs just the Amp/Cab)
Rig with Amp Clone is 14ms (see below for clarification)
Rig with Pedal Clone > Amp Clone is 26ms (see below for clarification)

Ouch. That's way too high, IMHO - even for just the EQ/drive/amp/cab rig.

If i recall correctly, Leo Gibson recently measured the latency of a HX Stomp with a ToneX pedal in the fx loop at ~6ms. And that involves 6 discrete A/D-D/A conversions.

 

[the-trooper]

Rig with Pedal Clone > Amp Clone is 26ms

 

[Evilution]

Ouch. That's way too high, IMHO - even for just the EQ/drive/amp/cab rig.

If i recall correctly, Leo Gibson recently measured the latency of a HX Stomp with a ToneX pedal in the fx loop at ~6ms. And that involves 6 discrete A/D-D/A conversions.

The AD/DA conversion hardly takes any time though, its the processing (effects, routing etc) inside the pedal that takes the time and adds latency.

 

[Jim Roseberry]

Ouch. That's way too high, IMHO - even for just the EQ/drive/amp/cab rig.

If i recall correctly, Leo Gibson recently measured the latency of a HX Stomp with a ToneX pedal in the fx loop at ~6ms. And that involves 6 discrete A/D-D/A conversions.

Yeah, Ideally I'd liked the model based RTL to be ~2-3ms.
That would put it on-par with top tier competitors.

If you add wireless, FOH digital mixer, digital processing on powered-speakers, etc... latency can start piling up.

 

[JCW308]

Ouch, hope they have a fix in place soon! Sticking with my QC for now I guess.

 

[Sascha Franck]

Also consider if you use a digital wireless as well or even some digital pedal in the loop, even with the lower 5ms to start with can start to build up in that case.

Even without a digital pedal, any used loop will double the "baseline" latency (which would be the device's latency without any blocks loaded).

After all, the HR boards seem to be pretty weak on the CPU side of things because otherwise they'd likely be able to do more things within a set/fixed (and smaller!) buffersize, just as many plugins in a DAW on a modern computer would do. Set your buffersize to 32/64/128/whatever samples and that's it. Usually only very CPU hungry or "lookahead" plugins need additional buffers to get their processing done.

 

[Sascha Franck]

The AD/DA conversion hardly takes any time though, its the processing (effects, routing etc) inside the pedal that takes the time and adds latency.

Yes and no. See my last post. Inside a DAW (and after all, a modeler is nothing else but that) you usually set your buffersize and in many cases plugins are able to do their job within that set buffersize, not adding any additional latency. This is also true for most modelers (so far at least), unless you were adding some really CPU intensive blocks.
As an example: If I recall correctly, there's no difference in latency when you have just an amp&cab block in a HX series modeler compared to slapping a whole truckload of (active) drives in. It's only rather special FX (such as the new-ish polyphonic pitch affairs) that there's additional latency introduced.

 

[Sascha Franck]

Fwiw, you can usually "buy" the processing juice for additional blocks by either having a more powerful CPU or by raising latency. And all too apparently, the HR boards have to make use of the latter.

 

[Evilution]

Yes and no. See my last post. Inside a DAW (and after all, a modeler is nothing else but that) you usually set your buffersize and in many cases plugins are able to do their job within that set buffersize, not adding any additional latency. This is also true for most modelers (so far at least), unless you were adding some really CPU intensive blocks.
As an example: If I recall correctly, there's no difference in latency when you have just an amp&cab block in a HX series modeler compared to slapping a whole truckload of (active) drives in. It's only rather special FX (such as the new-ish polyphonic pitch affairs) that there's additional latency introduced.

What you are talking about isnt AD/DA conversion though, you are talking about the internal processing, AD/DA conversion both in a DAW and in a modeler takes place in dedicated hardware components which are unaffected by the amount of processing that is done in the computer. The AD/DA conversion hardly adds any latency.