Transient behaviour when picking the first note

of course we are, I'm still trying to reproduce this.

could you guys that are experiencing this pm me, so you can mail me the rigs that behave that way?

thanks.

Quote from DonPetersen

of course we are, I'm still trying to reproduce this.

could you guys that are experiencing this pm me, so you can mail me the rigs that behave that way?

thanks.

PM'd

Thanks!

We are taking issues as seriously as they are, and this is not really serious.
The Profiler is out now for almost two years. The dynamic behaviour was not changed during that time. This overshoot issue was only reported once before, without even the complaint about a tonal flaw.

What you see is a bit of power sagging. It creates an overshoot of about 6 dB. There is no risk of coloring the audio due to clipping, because high transients are smoothly catched by the soft clipping algorithm of the outputs. Have a closer look at the recordes transients to ensure yourself that the signal is not cut. The red lights on the level led's do not indicate a digital clipping, but a soft and tube-like compression on the input or output.

The Pick parameter and the Power Sagging parameter will produce and emphasize similar transients, btw.

I made a quick test and disabled this part of the dynamic engine of the profiler. The result was a *visible* suppression of those peaks, but no change in sound color. The only audible change was a slight reduction of the perceived energy of this first transient, maybe one or two dB - similar to changing Pick to - 0.5 or so.

Any sound that you perceive "unnatural" would sound the same, since this overshoot does not yield to a coloration of the sound.

If these transients are recorded, wouldn't they be there when we play live?
Regardless of what they sound like, they overshoot the volume of the signal by far - so running them into a PA, for example, might clip the input, while the signal is still relatively quiet.
It gets worse when you deal with hardware compressors in a studio.

If we could lower the power sagging parameter maybe we could deal with it, but at 0.0 the transients are still very much there.

I was able to reproduce the transient this time (sort of) with a pick scrape. It's still quieter than what the signal strength is when playing regularly, and regular playing doesn't produce spikes for me. This was with Bombed! and through s/pdif. Just resting the pick on the strings didn't really produce anything.

First the pick scrape and then a palm muted single note on the E-string.
[Blocked Image: http://i.imgur.com/yXF8bMU.jpg]

Since I cropped the timeline out of the pic, the length of the first hit is 0.4s.

Edit: I found that you can easily reproduce the transient by bashing the strings perpendicular with the edge of the pick. But I'm pretty sure that's just characteristic to that type of "technique".

Quote from Quitty

If these transients are recorded, wouldn't they be there when we play live?
Regardless of what they sound like, they overshoot the volume of the signal by far - so running them into a PA, for example, might clip the input, while the signal is still relatively quiet.
It gets worse when you deal with hardware compressors in a studio.

If we could lower the power sagging parameter maybe we could deal with it, but at 0.0 the transients are still very much there.

Sure they could clip the input of a PA. But this would happen even more with Pick or Powersagging turned up, by nature.
Clean sounds have even much higher transients.
When clean sounds have the same energy as distorted sounds (thus Clean Sense adjusted correctly), you will achieve a transient of up to 15 dB more than the "quieter" average signal or distorted signal.
This has to be taken into consideration, when leveling the input of a mixing desk.
Best is to level it with clean sounds due to this reason.

The overshoot we discuss here is 6 dB (only) above average. But much more headroom should be given to cover all dynamics of the Profiler.

Quote from laardi

Maybe "unnatural" and "clipping" are not good terms as such, but there is still an audible "snap" every time one hits the strings for the first time:

Nakedzen: it is so weird that I can't avoid the spikes and you can't produce them, no matter how hard you try

Ninja edit'd

Edit: I found that you can easily reproduce the transient by bashing the strings perpendicular with the edge of the pick. But I'm pretty sure that's just characteristic to that type of "technique".

[Blocked Image: http://i.imgur.com/Rg8qUG2.jpg]

The transients produced this way are still quieter than my regular overall level when playing normally.

But I think it's good that I'm not getting clipping, since this narrows possibilities down as to what is the cause.

Edit: I did another test by playing regular palm mutes and pick scrapes and the input and output led's stayed green on the scrapes and went to orange on regular playing so I'm not getting internal clipping on the scrapes either based on that. Hmmm

Quote from ckemper

Sure they could clip the input of a PA. But this would happen even more with Pick or Powersagging turned up, by nature.
Clean sounds have even much higher transients.
When clean sounds have the same energy as distorted sounds (thus Clean Sense adjusted correctly), you will achieve a transient of up to 15 dB more than the "quieter" average signal or distorted signal.
This has to be taken into consideration, when leveling the input of a mixing desk.
Best is to level it with clean sounds due to this reason.

The overshoot we discuss here is 6 dB (only) above average. But much more headroom should be given to cover all dynamics of the Profiler.

Display More

It is manageable, true, but i still think is quite the problem.
When i play clean, i know i play clean. even if i'm not clipping a mixing desk, we play differently when playing clean or distorted amps.
Regardless, the overshoot occurs when lightly touching the strings - as if some upward expander is working, or that something triggers when a sound comes in and its activation induces popping.
So, at the moment, the profiler responds unlike a guitar amp would, and wasn't that the idea?..

Anyway, it would seem we have a solution -
according to CK, if we set the volume of the rig so that it spikes near the red (output wise), the output compressor should catch and squeeze the transient.
Can someone verify?
Still is weird that they're so gentle with your rig, nakedzen. They're as obvious as they get with mine.

Quote from Quitty

Still is weird that they're so gentle with your rig, nakedzen. They're as obvious as they get with mine.

Which leads me to believe something we haven't thought of yet is the cause, since logically thinking it can't be any of the things that my setup shares with yours (high definition parameter, the profile itself, high output pickups, clean/distortion sense settings, firmware).

I'm not sure if I missed if this has been tried yet, but have you tried deleting the rigs that are producing pops, redownloading them and do a fresh import of them? Just to rule out any possibility of corrupted rig files.

Quote from nakedzen

...
I'm not sure if I missed if this has been tried yet, but have you tried deleting the rigs that are producing pops, redownloading them and do a fresh import of them? Just to rule out any possibility of corrupted rig files.

Yup. Haven't tried resetting the machine, but that's next up.

Quote from Quitty

It is manageable, true, but i still think is quite the problem.
When i play clean, i know i play clean. even if i'm not clipping a mixing desk, we play differently when playing clean or distorted amps.
Regardless, the overshoot occurs when lightly touching the strings - as if some upward expander is working, or that something triggers when a sound comes in and its activation induces popping.
So, at the moment, the profiler responds unlike a guitar amp would, and wasn't that the idea?..

Anyway, it would seem we have a solution -
according to CK, if we set the volume of the rig so that it spikes near the red (output wise), the output compressor should catch and squeeze the transient.
Can someone verify?
Still is weird that they're so gentle with your rig, nakedzen. They're as obvious as they get with mine.

Display More

This is not unlike a tube amp. Tube amps with power sagging have a signal overshoot. That is actually the trick of power sagging.

You will hopefully leave 6 dB headroom on your mixing desk anyway when you record or play live. This is recommended anyway.

A regular compressor by nature would produce a similar initial transient, rather than catching that transient.
That's why there is an adjustable attack time on compressors. It controls the length of the transient.

Quote from laardi

So the cure for the "snap" sounds (associated with the output going red and the waveforms being unexpectedly spiky) is simply to lower the rig volume?

The other way around.
CK explains that the spikes are a product of the power-sagging of the original amp - the distorted signal from the preamp hits the poweramp, a bit goes through and then it gets compressed.
He also said that there's a soft-clipping compressor on the output of the profiler that will smooth these transients if you get it to work - meaning, set the rig volume as high as you can without clipping (make it peak in the red on palm mutes).

CKemper: would it be possible, then, to let us turn power-sagging down below the original amp's level? It would be extremely useful in a live setting, when more volume is required, or when we're stressed for headroom in the studio.

Had a couple more minutes of testing today.
No scope screenshots, as they seem to piss Mr. Kemper off

Tried a profile of a Mesa Mark IV from the rig exchange that i've been using for a while -
definition is pretty high, power sagging on 3 or so and the rest is at 0 - crazy transients.

• Raised the cab volume until the profiler spiked red on palm mutes (+4.0) - no improvement at all.
• Raised the cab volume to +12.0 so the profiler output stayed red - no clipping was heard and the transients were improved, but still very much there.
• Cab volume at +12.0 and power sagging to 0 - transients are almost gone. Pretty much perfect.
• Power sagging at 0 and cab volume at 0 - crazy transients.
• Power sagging to +10.0 - no difference.

in fact, it seems that changing the power sag did absolutely nothing either way, until the cab volume was crazy high.
Are we absolutely sure this is the cause?

The way I've always seen it is that the KPA is not subject to internal gain staging issues - so if you for instance added 64 db gain then subtracted 64 db later in the signal chain, there would be no real affect on dynamics or tone - no compression...no digital clipping. Only when you pummel the input or output do you get compression or clipping. It sounds like the Cab Volume +12 db is forcing the soft clipping...inside the chain?

Some more thoughs:

The so called spikes are not spikes, as there is no additional signal.
It is simply a transient.

The transient has no audible consequence. Otherwise it would have been revealed in thousands of profiling procedures and A/B comparisons.
We have not received a single message of those guys in two years.

The transient is omnipresent. It cannot be overlooked when leveling the input of a mixer, as it occurs with every hit of the string, after the strings were muted below the distortion level.

But if I got it right they seem to be produced by simply lining the pick on the string, without even picking them...

Quote from viabcroce

But if I got it right they seem to be produced by simply lining the pick on the string, without even picking them...

Exactly.
And the problem is that when it occurs, the level far exceeds that of the distortion level.