Create New Project
Create 1 kHz square wave
Watch artefacts (vertical lines) in the spectrum, that DO NOT belong there !!!
→ This programm is completely broken. Bugs after bugs after bugs…
It seems like the very basic functions of the programm are messed up in so many different parts, that make everything else, that relies on their functionality, will - of course - also be messed up even further!
How can this pass a quality control at all? If there is any…
Confirmed.
Same result here, when using the exact same process. Initially, it appeared to be visual only, but it’s definitely audible.
Here is one of the transients, cranked up 50dB with the bulk of the tone removed:
I’ve noticed mystery transients in other processes as well. Hopefully the next version will address this issue once and for all.
As a workaround, set the “Start Phase” to a low value, i.e. 1 or 5.
It only seems to happen with square waves.
Nope!
It also happens with Sawtooth
This is a simple no go. I don’t care about workarounds, instead expect every function/module/method 100% exactly to work as it is supposed to do and without any odd sideeffects, caused by sloppy programming. Not more, not less! That’s what I paid for.
When it comes to a signal generator of this simplicity: That is basic stuff even for a beginner.
And btw, are those weak lines in the background aliasing artifacts?
Yes, you’re right … also confirmed. For comparison, I generated the same tones with WaveLab and loaded them into SL12 (just to rule out a possible display issue) and they are clean.
The generated waveform is definitely corrupted. If I generate the same square wave in Sound Forge, layer both on top of each other, and invert one of them, they do not cancel each other out. The square wave generated in SpectraLayers deviates irregularly by exactly one sample. And each beginning and end of a deviation causes a click.
Square waves have harmonics.
That’s not what the signal generator produces… it produces distorted crap.
It seems, that at the fundamental algorithmic level of SL there is something wrong. It might be just a glitch, some index in a loop or something like it, but it distributes all over other functions of the program.
For the signal generator, it might be better to produce the waveforms in the pure time domain and than transform them into the spetral domain afterwards to not introduce unnecessary errors. It seems, that everything here is calculated already in the spectral domain, which in the case of simple signal generation is not appropriate, imho.
Looking at the posted images i would say those are harmonics.
Wich are part of why Square waves sound how they do.
But proceed gentlemen.
I’m always eager to learn when there’s something to learn.
As a side note, i wouldn’t use SL to generate waveforms but that’s just me. I’m picky against audio software where level measurement occupies less than 0.7% of the workspace.
If a program offers a function, than the function should work 100% as it is supposed to be. Otherwise don’t offer the functionality. Not more, not less.
The function generator comes in handy for many use cases. Not everyone uses SL for some artistic poetry!
Replying to your first sentence: that is a Square Wave. Square Waves have harmonics.
Replying to your second sentence: i don’t deal well with arrogance and condescendence and you’re being both.
That is stupid because even if you were right, it would be the wrong approach to write things that i will read inside my home.
Remember that the utmost respect is needed when you talk with people you’ve never met because your words and attitudes (the arrogance and condescendence) resonate in places that you don’t know.
Behave. Be polite. Be nice.
We’re talking about a f@cking Square Wave not a firing squad.
A square wave is a theoretical (mathematical) construct.
It isn’t possible in nature.
And it produces unlimited harmonics as you see in your screenshot.
So that is what you should expect to see.
I guess the vertical lines come from a period not aligned to the zero crossing point.
Did anyone read what I wrote>
Those lines are not harmonics.
Well, I did. Maybe some others don’t understand the point of this thread…
It’s just another case, where the emotion of arts and the reason of science clash together…
Who cares?
Unfortunately the forum does not allow the upload of WAV files (!), so I can’t upload the tones created in WaveLab for comparison..
As further proof, the spectra of the SL and WL generated tones differ … and for that I used Sound Forge as the “honest broker”.
Just to be absolutely clear, the vertical lines in @Sunnyman 's screenshot from SL in the OP are not harmonics and should not be there.
Quite likely, but that’s what is at issue here, the vertical lines, not the horizontal ones (which are harmonics, of course). Furthermore, if the “Staring Phase” is changed a little, the vertical lines are diminished. Either way, the generator in SL is faulty and it is a bug.
Exactly.
You were too emotional and “lost it”. No science in what you have posted before.
You must be a poet.
See?
It’s bad idea to antagonize someone on the internet.
Make better posts next time.
___
As for the OP, I was referring to the horizontal lines, of course.
They’re harmonics and are a natural part of a square wave.
The vertical lines (wich i ignored) were later mentioned (and appropriately so) by @st10ss
They’re just SL doing what it’s good at.
Actually, the vertical lines were mentioned right at the beginning and they are the key to understand this thread.
You’re right, sir!
