Null Tests

[SillyOctpuss]

View attachment 41844
View attachment 41845
View attachment 41846

 

[Orvillain]

Anytime I think about grabbing a kemper out of fascination, I just have to remember @Orvillain posts

Buying a Kemper will take you so far back in time, you'll become a prisoner of the British Empire once again.

 

[Orvillain]

(that one was mine, not Chatters)

 

[ian_dissonance]

The only time I wanna see a null test is when IK says they’re doing something the best, lol.

 

[JiveTurkey]

Insert "if I was a Marshall person and a Marshall person only; I'd rock a powerhead through a 4x12 all day long" response here.

 

[Chocol8]

A null test is only as good as the signal you feed it. For me, the Kemper does great in many situations but falls apart in a few. I could make a sound comparison or null test that shows it is pretty good, or I could play with the guitar volume and use palm mutes for example to make it sound and test noticeably different.

In other words, he who designs the test can determine the results in advance.

 

[[Nathan]]

A null test is only as good as the signal you feed it. For me, the Kemper does great in many situations but falls apart in a few. I could make a sound comparison or null test that shows it is pretty good, or I could play with the guitar volume and use palm mutes for example to make it sound and test noticeably different.

In other words, he who designs the test can determine the results in advance.

If you’re doing null tests and you have to select a certain snippet to only show the best results then it’s flawed and you’re fooling yourself more than anything. I’d prefer to hear it in the extremes as well as the easy stuff and make an overall assessment.

In the past when I’ve done null tests with tonehub, NAM and tonex I can generally tell how well something is going to null based on how the profile plays in isolation. The null test typically backs up what I already felt was a good profile.

Anyway the point of any kind of test is not to rig the stage and omit details. If that’s what someone has to do on a kemper to show it in its best light then that’s not great.

 

[Chocol8]

Yeah, I was just trying to say other peoples null tests are worth less than what you paid for them.

 

[hunter]

I agree there is a Kemper Sound. There's also a Tonex Sound. The Kemper sits well in a mix though. With Tonex it's easy to have too much treble which can sound harsh/thin on a PA.
Actually, the product Christoph Kemper made before the Profiler, Access Virus Synthesizer also was known for warmth and smoothness. It definitely had a sound. Nord Lead was the first "virtual analog" one year before the Virus, and it was very snappy and in your face, The Virus was more pleasant. I feel the same signature in the Profiler and I like it.

 

[AFKAEjay(retired)]

I agree there is a Kemper Sound.

I don’t, Kemper would have been easily identifiable in blind tests…and I haven’t seen any where that was the case.
In TOP I posted a few clips in response to someone claiming the “kemper sound” was very clear. The request was to spot which one was the kemper. Offcourse no play from the claimer other then criticizing the clips

 

[Sascha Franck]

It’s not the whole story

Why exactly?

 

[EOengineer]

 

[Eagle]

That is pretty much the test I did.

 

[Sascha Franck]

So, @Eagle, would you finally elaborate why a null test is *not* the entire story? You keep coming up with that kinda stuff any never even just once managed to even remotely answer why a null test would not be the entire story.
And no, please don't tell us that playing an amp through a cab vs. a modeler through fullrange monitors is "the entire story" - because that's not what null tests are about at all.

 

[KingsXJJ]

What I’d like to know is how much of a difference in a null test is audible?

Has that been determined with solid data?

 

[KingsXJJ]

Come on oh long winded ones. Don’t be dicks about a simple question.

Please. Lots of stuff is hitting the market or is here. Ampero Ii, ToneX, GP-5, AA Stomp, etc. Give us/me the TDLR.

 

[KingsXJJ]

@Orvillain @2dor @MirrorProfiles @AlbertA et al. Please bring me up to speed.

 

[2dor]

@Orvillain @2dor @MirrorProfiles @AlbertA et al. Please bring me up to speed.

If you take a listen to Leo's vids for example, whatever you hear when the NULL test is done (when the signals have opposite phase) is what the profiling is missing. If the signals would be identical, a Null tests would rendered silence.

If the stuff you're hearing is a noticeable part of the frequency spectrum, then accuracy isn't as good and you're most likely going to be able to tell a difference.

If whatever you hear is just static etc then it means the profile is pretty spot on.
Just my 2c.

 

[AlbertA]

What I’d like to know is how much of a difference in a null test is audible?

Has that been determined with solid data?

I don't think there's studies in the context of amp modeling.

There's been some perceptual metrics developed in the context of perceptual codecs (like mp3, AAC) etc - like PEAC, PEMO-Q.

PEMO-Q for example will give you PSMt in a range of 0 to 1 (1=> identical, 0 => no similarity).

A potential "rule of thumb" could be (don't quote me on this);

• -50dB RMS in residuals of a null test + PSMt > .95 could be "virtually the same"
• -60dB RMS in residuals of a null test + PSMt > .97 could be "perceptually inaudible"

You would have to calibrate that against a cohort of people making subjective comparisons in a controlled test in this context though.

 

[AlbertA]

So, @Eagle, would you finally elaborate why a null test is *not* the entire story? You keep coming up with that kinda stuff any never even just once managed to even remotely answer why a null test would not be the entire story.
And no, please don't tell us that playing an amp through a cab vs. a modeler through fullrange monitors is "the entire story" - because that's not what null tests are about at all.

Because the residual of a null test only tells you if there is a difference, not if the difference is perceptually audible.