Where are you setting your input level?

yek said:
So the conclusion of this thread is:
Adjust Input A/D Sensitivity to where it "tickles the red", and then turn it down more?
Click to expand...
That works for some, sure. I already have my Input Sensitivity at 5%, and I'm still tickling the red. That is, assuming the activity I'm observing is "tickling." It's too vague of an instruction. I don't really want to turn it down further at the risk of affecting the tone of the signal chain. Doing the DI test, for me, would be a worthwhile sanity check.
 
yek said:
So the conclusion of this thread is:
Adjust Input A/D Sensitivity to where it "tickles the red", and then turn it down more?
Click to expand...

IMHO I would say the conclusion of the thread is: if you follow the instructions in the manual, you might be clipping the input converter. If you want to be sure you're not clipping when following those instructions, you must record and examine a DI.

Alternatively, the conclusion might be: it would be helpful if the Axe-FX had a clip light on the front panel.
 
Well, this is embarrassing… I have had the Axe Fx III since 2018 and only now realized that 50% is way too high for some of my guitars. The guitar that sounded best at 50% has been my Les Paul with Mojotone 59 clones. And they sounded like raging high output pickups. I turned it down to 15% and that has made a huge difference in the clarity of the tone. There isn’t all this extra noise now that I am not clipping the inputs. Just goes to show you can still learn things 5 years in…
 
Welp, I know what I'm doing tonight.

Edit: Cliff just shaking his head reading this entire thread.
Edit 2: Page 5 in the manual PDF
 
ak49climber said:
Well, this is embarrassing… I have had the Axe Fx III since 2018 and only now realized that 50% is way too high for some of my guitars. The guitar that sounded best at 50% has been my Les Paul with Mojotone 59 clones. And they sounded like raging high output pickups. I turned it down to 15% and that has made a huge difference in the clarity of the tone. There isn’t all this extra noise now that I am not clipping the inputs. Just goes to show you can still learn things 5 years in…
Click to expand...
did u confirm via DI review that u were clipping at 50%? (I still find it hard to believe people are hearing material differences in a/d convertible signals between 5% and higher + not clipping).
 
Haven't looked at the wave form but I've had to set mine to 5% for my 92 Ibanez RBM-1 with it's HB-2 wich is just a Ibanez Branded Dimarzio Megadrive. Only lights up the red real quick when really digging in and only on the Low E. If i set it higher it lights the red more often and longer.
 
sprint said:
did u confirm via DI review that u were clipping at 50%? (I still find it hard to believe people are hearing material differences in a/d convertible signals between 5% and higher + not clipping).
Click to expand...
Based on what the manual says, I feel the exact same way. This is why this whole thread baffles me a bit, because if the signal is auto compensated level wise, you should be getting the same signal going in all of the time. And if you somehow hit a db or two over clipping 10 out of 1000 strums when setting it correctly, that microsecond to millisecond transient clip should literally be inaudible.
 
sprint said:
did u confirm via DI review that u were clipping at 50%? (I still find it hard to believe people are hearing material differences in a/d convertible signals between 5% and higher + not clipping).
Click to expand...
The most likely answer is: they are clipping. As you say, without examining the DI, you get no indication that's happening, so people are unaware this clipping is occurring. I don't know if this is widespread, but that's the most likely explanation for the people who have reported it in this thread.
 
Oh good, a thread that validates my neuroticism about this!

Seems like most people with really high output pickups are setting it at 5% to avoid clipping the waveform and/or overloading the converters. I've been using a combination of low input sensitivity with my bridge pickup (PRS HFS) so low in the pickup ring that it's almost falling out. I guess I just need to try a lower number.

I feel like the lights on the front aren't that good of a gauge for this kind of thing. The waveform in the Looper block seems to show clipping pretty accurately. Debating making a feature request that indicates input clipping somewhere. Like maybe it could explicitly flash the words "INPUT CLIP" on some screen or something so there'd be no ambiguity. I dunno...people have been confused about input sensitivity for ages but most of the things I read or watch seem to create more questions than they answer.
 
yek said:
How can you know that there's no limiter at work? Can you tell from the peaks of the waveform whether it's clipping or the limiter kicking in?
Click to expand...

Yes the tip of the waveform peaks is hard clipped off like someone drove a lawn mower over it. A compressor or limiter would need nearly instantaneous attack and release times to recreate that effect on the signal. At that point it stops being a compressor or limiter and becomes a hard clipper.

1679606224796.png
 
Last edited:
How well you can hear any differences between
deathbyguitar said:
I feel like the lights on the front aren't that good of a gauge for this kind of thing. The waveform in the Looper block seems to show clipping pretty accurately. Debating making a feature request that indicates input clipping somewhere. Like maybe it could explicitly flash the words "INPUT CLIP" on some screen or something so there'd be no ambiguity. I dunno...people have been confused about input sensitivity for ages but most of the things I read or watch seem to create more questions than they answer.
Click to expand...

Any kind of visual indication like an LED or screen message would be limited by our eyes/brains' response time to changing light. We can't really see a blink that's faster than around 50 Hz or so. Our persistence of vision blurs it all together. Our ears on the other hand are sensitive to events up to 15 to 20 kHz. As a result you can't have a visual indicator exactly follow the clipping of a waveform. It has to persist long enough to be seen. That means such an indicator can only tell you that you are clipping but not accurately when or how much. The transition into and out of clipping is often too fast unless it is significant.
 
Last edited:
mr_fender said:
How well you can hear any differences between


Any kind of visual indication like an LED or screen message would be limited by our eyes/brains' response time to changing light. We can't really see a blink that's faster than around 50 Hz or so. Our persistence of vision blurs it all together. Our ears on the other hand are sensitive to events up to 15 to 20 kHz. As a result you can't have a visual indicator exactly follow the clipping of a waveform. It has to persist long enough to be seen. That means such an indicator can only tell you that you are clipping but not accurately when or how much. The transition into and out of clipping is often too fast unless it is significant.
Click to expand...
True, but I don't think that's what he's suggesting. He's just suggesting a conventional overload indicator that signals a clipping event occurred and persists long enough for you to notice. That's generally how overload indicators work.

I've never understood the rationale for the input led meter scale. Seems like it would work better if it used the same scale as the output led's (red means -1 instead of -6). That would probably solve the problem posed in this thread. I don't know if that can be done with a firmware update though, which is why I've always suggested it as a Axe-FX IV wish instead of a Axe-FX III wish.

That would make the instructions easier to understand IMHO. Instead of "Tickling", which is rather vague and might lead to clipping, you would simply adjust the senstivity so the yellow light comes on, but not the red.
 
Last edited:
The guitars that were really clipping the inputs at 50% were EMGs and a set of Fluence Moderns. The red led on the input meter was staying lit up pretty solidly with all those guitars. I just didn't think much about it because "tickling" is a bit subjective. The red lights were coming on, good, enough input gain! But backing it off so those guitars are only just lighting up the red led for a moment made a pretty significant difference in how clear the tone sounded. I went back and forth a bunch to see if there was a difference and it was pretty noticeable.
 
ak49climber said:
I went back and forth a bunch to see if there was a difference and it was pretty noticeable.
Click to expand...
It's just your ears playing tricks on you! :mad:

Just kidding :D. I found the same here. Note separation seems better at lower settings. Maybe I was clipping?
 
mr_fender said:
Yes the tip of the waveform peaks is hard clipped off like someone drove a lawn mower over it. A compressor or limiter would need nearly instantaneous attack and release times to recreate that effect on the signal. At that point it stops being a compressor or limiter and becomes a hard clipper.

View attachment 118167
Click to expand...
Thanks for sharing the photo! I was wanting to see an example. In regards to this clipping, how often is the red light on the Axe Fx going off? Does it turn on at every one of these highlighted sections? That’s what I’m wondering about, because if it is, I would assume the inout sensitivity is set wrong. If it isn’t tickling the red or barely, but it’s still happening visually, then that would be very interesting.
 
The Red LEDs on the input meter light at -6 dB below clipping. They don't indicate clipping at all, only that the level has exceeded -6 dB. A quick flicker on hard hits is probably fine, but the only way to know for sure if you're clipping the input or not when that LED blinks is to look at the DI waveform. Some pickups and players are more dynamic than others, so how close you can ride that -6 dB level will depend on your pickups and playing.

As mentioned earlier, the looper block's waveform view can give you a quick and dirty, low resolution view of the input waveform if you put it right after the IN 1 block (or bypass any block between) and turn down the level from the IN 1 block so the waveform doesn't fill the whole display window. If that waveform has a very even peak level and looks like a freshly mowed lawn, you're almost certainly clipping the input. Inspecting the waveform in the DAW just gives you a lot more detail when zooming in on the waveform to look at individual peaks.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom