New Friedman IR-X preamp

[Jarick]

Did some measurements of the IR-X and can confirm that there is a low pass filter on all outputs in all configurations.

The three way switch seems to affect the SLOPE of the low pass filter, not the frequency cutoff:
Middle ~ 27 db / octave
Left ~ 22 db / octave
Right ~ 18 db / octave

Rolloff seems to happen right around 3.5 kHz, so not sure what the cutoff frequency technically would be.

I did measure a similar rolloff with the following configurations:
Balanced out with IR bypassed
Balanced out with a null (flat) IR
Send out (no power amp sim)

Unfortunately I don't have an earlier version of the software to roll back and see if an earlier version also had a roll off.

I also did one quick measurement of the Axe FX 3. Same guitar and cable and all that. No IR or high cut. There is a roll off but it's much milder.

Here's a visual comparison, with some heavy frequency smoothing for clarity.

Blue = IR-X cut left
Red = IR-X cut middle
Green = IR-X cut right
Purple = Axe 3 no cut

Didn't include the other outputs as they all had the same overall EQ shape.

 

[Shredder777]

The results don't seem to make sense. Can you double check?

Been staring at it for 20 minutes and the groupings don't make sense.

Send should have less cut that the direct modes. I cant see a pattern besides there is a software bug.

 

[KingsXJJ]

The results don't seem to make sense. Can you double check?

Been staring at it for 20 minutes and the groupings don't make sense.

Send should have less cut that the direct modes. I cant see a pattern besides there is a software bug.

I sure would like to know asap if a firmware will address this. And if it was just a bug or intended. Does any one know which version introduced this and if previous versions are available? I’ll look around a bit…

 

[Shredder777]

I sure would like to know asap if a firmware will address this. And if it was just a bug or intended. Does any one know which version introduced this and if previous versions are available? I’ll look around a bit…

I'd like to know too. It looks that the cab sim is a boost at 100Hz and a low pass.

He didn't plot any of the send curves, so unknown if it would also have the 100Hz boost, in which case that would suggest cab sim is active even on the send. That would be bad.

 

[KingsXJJ]

Did some measurements of the IR-X and can confirm that there is a low pass filter on all outputs in all configurations.

The three way switch seems to affect the SLOPE of the low pass filter, not the frequency cutoff:
Middle ~ 27 db / octave
Left ~ 22 db / octave
Right ~ 18 db / octave

Rolloff seems to happen right around 3.5 kHz, so not sure what the cutoff frequency technically would be.

I did measure a similar rolloff with the following configurations:
Balanced out with IR bypassed
Balanced out with a null (flat) IR
Send out (no power amp sim)

Unfortunately I don't have an earlier version of the software to roll back and see if an earlier version also had a roll off.

I also did one quick measurement of the Axe FX 3. Same guitar and cable and all that. No IR or high cut. There is a roll off but it's much milder.

Here's a visual comparison, with some heavy frequency smoothing for clarity.

Blue = IR-X cut left
Red = IR-X cut middle
Green = IR-X cut right
Purple = Axe 3 no cut

Didn't include the other outputs as they all had the same overall EQ shape.

View attachment 34891

I’ll take the Axe’s high frequency output and the IR-X’s low end… please.

 

[GuitarBilly74]

There's definitely no cab simulation in the Send.
I've a/b the IR-D into my Dirty Shirley fx return vs the Dirty Shirley own preamp and it's pretty much identical.

 

[Shredder777]

There's definitely no cab simulation in the Send.
I've a/b the IR-D into my Dirty Shirley fx return vs the Dirty Shirley own preamp and it's pretty much identical.

I belive you, but why would toggling the cut switch affect the send in any way?

 

[Jarick]

There’s no cab on any measurement. I loaded a null cab aka no impulse. But that generally matched ir bypass outside volume difference.

Low end bump doesn’t matter. Not measuring the low frequency shape. Look at the high frequency rolloff slope.

 

[Shredder777]

There’s no cab on any measurement. I loaded a null cab aka no impulse. But that generally matched ir bypass outside volume difference.

Low end bump doesn’t matter. Not measuring the low frequency shape. Look at the high frequency rolloff slope.

What I'm asking is why the lopass software toggle woiuld have any affect on the send, which is supposed to be a direct preamp output?

In your chart, the lopass filter affects the send output in a similar way as it does the DI output.

 

[Jarick]

What I'm asking is why the lopass software toggle woiuld have any affect on the send, which is supposed to be a direct preamp output?

In your chart, the lopass filter affects the send output in a similar way as it does the DI output.

It just plain does. The low cut apparently is not on the cab, it's on the preamp itself.

Here's a fresh set of measurements.

Signal chain is Telecaster > IR-X input > IR-X output > UAD Apollo input > Room EQ Wizard

I used the RTA feature of REW to record PEAK frequency response, then smoothed the response.

Here you can see all three settings recorded with both the send (no power amp sim or IR) and with the output (power amp sim but IR bypassed). The low cut definitely impacts the output in a consistent manner.

You can also see the power amp adds a low frequency bump as well, but again that's not really what we're measuring here.

 

[Jarick]

And if you're curious, same settings on the output with the IR bypassed vs not. Of course adding an IR rolls off the high end further.

Now all that to say, it's not necessarily wrong that the high frequency is rolled off. It's possible that's what the amp is supposed to do with its tone stack. But it is odd that there's a switchable cut in the preamp, apparently.

I have a suspicion there may be a bug, but maybe that was there all along? Again would need to roll back the firmware to check.

 

[Shredder777]

Cool. Much better graph. This shows how it all works. Noir and send are same on top, both influenced by lopass. The noir has a different low end than the send.

Bottom and top both different from fractal. How similar does the irx sound to the fractal be100?

 

[Shredder777]

And if you're curious, same settings on the output with the IR bypassed vs not. Of course adding an IR rolls off the high end further.

Now all that to say, it's not necessarily wrong that the high frequency is rolled off. It's possible that's what the amp is supposed to do with its tone stack. But it is odd that there's a switchable cut in the preamp, apparently.

I have a suspicion there may be a bug, but maybe that was there all along? Again would need to roll back the firmware to check.

View attachment 34899

I would agree that selectable lopass on send seems wrong.

Maybe send df an email if youre so inclined? Or maybe i should but i didnt make the graphs.

I am hoping this can be corrected with eq.

Btw, what channel were you using?

 

[Jarick]

Stay tuned...going in for extra credit...

 

[GuitarBilly74]

I belive you, but why would toggling the cut switch affect the send in any way?

The low cut is before the loop. Deep, Presence and IR is after.

The Send goes to through the ADA converter because Friedman wants to make updates in the feature that will include basic effects etc. It's not an analog out, but there is no IR in it.

 

[Jarick]

Okay last one for the night.

Friedman IR-X - balanced output with IR bypass (power amp sim, no cab), channel 2, knobs at 5
Fractal Axe FX 3 - Friedman BE V1 amp model, no cab, knobs at 5
Quad Cortex - Friedman BE amp model, no cab, knobs at 5
HX Stomp - Friedman BE amp model, no cab, knobs at 5

Here I did less smoothing, honestly I could do practically no smoothing and you'd see the super steady rolloff.

Anyways, you can see the IR-X definitely rolls off much quicker than the digital modelers. The three modelers are really close in terms of frequency and slope roll off, so you likely wouldn't hear much difference there (not to say there wouldn't be audible differences elsewhere).

For S&G I extended up to 30 kHz. I'm not sure how accurate that is or isn't, but interesting that the Fractal has a very steep cutoff and the other two just do there own thing.

And again, this isn't saying IR-X is worse than the others...it just definitely has a more noticeable high cut. Some people might call that analog warmth? I dunno.

If someone wants to send me their Friedman BE-100 amp I'd be happy to run that through as well

Also, this is why I can't sell anything...for the twice a year I get extremely curious.

 

[Shredder777]

That explains why harmonics sound dead to me on top, but normal down low. I can say with all my belief this is not natural and the modelers are more correct.

I suppose i can use a shelving boost that is roughly inverse to bring back some top and bake it into the irs. Or i could just use the 3 ir slots for varying levels of top boost and use speaker irs in my mfx. That is what i will do. So the ir toggle would become a selectable brightness control. It needs it.

Thanks for your work. This tracks with what im hearing. Unknown why no one else has complained.

 

[Shredder777]

It really sounds like he took a PEQ with a fairly broad Q at 3K and boosted it +4-6, and took a low pass filter and started chopping off content above 6K.

Quoting myself from a couple weeks ago. My ears are good.

 

[GuitarBilly74]

Unknown why no one else has complained.

I didn't complain because it's not an issue to me. No dead harmonics or anything.

The pedal is just not what you're looking for.
Doesn't mean there's anything wrong with it. Works for me and many others I wouldn't replace it with any modeler I've tried so far, especially live.

 

[Shredder777]

I didn't complain because it's not an issue to me. No dead harmonics or anything.

The pedal is just not what you're looking for.
Doesn't mean there's anything wrong with it. Works for me and many others I wouldn't replace it with any modeler I've tried so far, especially live.

Abrupt chages to frequency response like in graph above always sound unnatural. When you remove high frequency content it affects the sound of notes lower in frequency band. (They are missing their harmonics).

The ir d may not be doing that. I didnt hear it in your videos.

The irx rhythm sound, feel and response was always more natural than digital modelers from when i first plugged in. I agree with you its better which is why i didnt give up. Now i should be able to fix the irx with shelving eq.

Thanks again to jarick for helping me visualize it.