Beta Test: DISTORT - audibility of distortions

[Hayabusa]

This happens with even harmonics in the mix. Turn off even harmonics, leaving odd ones in and the distortion becomes symmetric around the 0 point:

View attachment 42842

Ok clear, so you could shift the curve back to 0,0 when using even harmonics, wouldn't that be more the reality in an amp?

Just curious: at what kind of distortions levels does the transfer-function visibly differ from a strait line?
Would maybe be nice to measure the transfer function of an amp to characterize its distortion....

 

[pkane]

Ok clear, so you could shift the curve back to 0,0 when using even harmonics, wouldn't that be more the reality in an amp?

Just curious: at what kind of distortions levels does the transfer-function visibly differ from a strait line?
Would maybe be nice to measure the transfer function of an amp to characterize its distortion....

You can easily see the level of amp non-linear distortion by measuring the harmonic content.

Anything less than about 0.1% THD will remain very close to a straight line (-50dB = 0.3%):

Getting to THD of 1% and above, it becomes more obviously curved:

 

[Alex-D]

That should be stereo. If you select L+R you'll have a two channel to one channel mix-down. Let me see why it's not writing a stereo file. You're doing good on finding bugs! Please keep going!

Processed file is saved correctly as stereo but reference file is saved as mono for some reason.
Same file, same settings.

Version 1.15

 

[pkane]

Processed file is saved correctly as stereo but reference file is saved as mono for some reason.
Same file, same settings.

Version 1.15

Aha! I'm not sure why saving reference file is even enabled, to be honest, other than as a left over from the other software (DeltaWave) this code was based on Do you see any reason to keep it around?

 

[Alex-D]

It converts file to the same format just in case it affects anything downstream.
Also in case you want compare L+R files instead of stereo for some reason reference file makes sense, because you can't compare that to the original.

 

[pkane]

It converts file to the same format just in case it affects anything downstream.
Also in case you want compare L+R files instead of stereo for some reason reference file makes sense, because you can't compare that to the original.

I don't think it should affect anything downstream, but I'll double-check. Thanks again for testing!

 

[pkane]

I don't think it should affect anything downstream, but I'll double-check. Thanks again for testing!

I confirmed that there is no downstream effect from the reference file being treated as mono -- this only happens when the file is written out using Save Reference menu option. This will be fixed in the next update.

 

[pkane]

I forgot to post an update here

DISTORT v1.0.16 is now available with the following changes:

• Added an option to hide THD legend on the spectrum plot
• Added an option to hide the transfer function plot
• Fixed a rare out-of-bounds condition when using jitter settings
• Save Reference file menu now saves stereo when two channel option is selected
• Added an option to dither the result to a specified number of bits

 

[Arpiben]

@pkane

Santa Paul came sooner with lots of gifts this year!
Many Thanks.

BTW what kind of dither is applied by Distort: 2 bits Triangle/Gaussian Power Density Function ?

 

[Pluto]

Hmmm - for the first time ever, attempting to update one of your programs has elicited this…

What is your suggested approach so as to avoid loss of existing settings?

Thanks.

 

[pkane]

Hmmm - for the first time ever, attempting to update one of your programs has elicited this…

View attachment 43332

What is your suggested approach so as to avoid loss of existing settings?

Thanks.

You must have version .16 already installed. Uninstall from Windows control panel, then install the new one. Settings will not be removed.

 

[pkane]

@pkane

Santa Paul came sooner with lots of gifts this year!
Many Thanks.

BTW what kind of dither is applied by Distort: 2 bits Triangle/Gaussian Power Density Function ?

Dither is TPDF.

 

[Pluto]

You must have version .16 already installed

Doh – true enough. I did indeed download the latest version before you announced it and I took the announcement to be another (newer) version. BUT…

…where has the bit reduction control gone?

 

[pkane]

Doh – true enough. I did indeed download the latest version before you announced it and I took the announcement to be another (newer) version. BUT…

View attachment 43383
…where has the bit reduction control gone?

Probably best to delete the installer ZIP file you downloaded and try to force a re-download. Looks like you still have the older version which was updated later. The date on the DistortSetup.exe file should be Dec 21, 2019. I verified that this is the version on the website and that it does indeed have the bits/dither setting.

 

[pkane]

Just in time for the new year! Distort version 1.0.17 is now available.

Changes in this version include:

• Added Bits tab to select dither type (none, TPDF, and Shaped) and to select the number of it bits to dither
• Added configurable Dynamic Range Compression distortion to Custom Transfer Function settings
• Added Noise Floor setting (no signal) to the available Test Signals
• Added Blue (+3dB/Oct) and Violet (+6dB/Oct) noise floor settings
• Minor bug fixes

Bits tab currently allows the choice of three dither types, none, TPDF, and Shaped. Bits setting allows the selection of how many bits to dither:

Dither is applied after all the other distortions to produce the desired # of bits of output. More settings will be added to this tab in the near future!

Dynamic Range Compression lets you experiment with loudness/compression effects. The setting is available under the Custom Transfer Function option on the Harmonics tab:

Compression ratio defines the amount of compression and threshold defines the level above which the compressor is applied. Turn on the Autoscale option to apply make-up gain to match the RMS level of the original waveform.

Example of a simple sine wave going through the compressor:

 

[pkane]

A couple of new features in v1.0.18:

• A number of distortion presets added to simulate various tube gear through a non-linear transfer function (model C is not currently working, and is turned off)
• Added a new display tab to show both, the original test signal waveform and its distorted version

Various presets can be found under the Custom Transfer function window:

Waveforms display is on a new tab on the main screen:

Another example, 1k sine wave with large random jitter:

 

[magicscreen]

I have added -120 dB white noise level to a wav file. But I can hear this noise playing this file. Is it possible or something is wrong here?

 

[pkane]

I have added -120 dB white noise level to a wav file. But I can hear this noise playing this file. Is it possible or something is wrong here?

Anything's possible Are you applying any other distortions or just the -120dB white noise? Are you playing the file directly in DISTORT or in another player? What is the volume level set to if you are playing in DISTORT? What's the original signal recorded in that file and what's the maximum level? If the original file has any signal reaching 0dBFS, adding white noise can cause some clipping that will be audible.

 

[Pluto]

I suggest a tab with a "system diagram" drawn on it to help the user keep track of what comes in what order. In due course you could add "enable" and "disable" to the various "modules" in the system!

 

[pkane]

I suggest a tab with a "system diagram" drawn on it to help the user keep track of what comes in what order. In due course you could add "enable" and "disable" to the various "modules" in the system!

Fully agree. I'm not 100% happy with the UI design, at the moment. It's not obvious what's turned on and what's turned off, especially if you are on a different tab. I'll need to rethink the overall layout and make things more obvious and apparent and easier to control all from one place.