Stereo Tool

Was the Dutch Moose preset updated a little bit quieter than in 9.00?

Hi Hans,

Great work on the new builds, sounding sweet.

Just a few points I've come across:

• Running the latest VST beta 901-001 in Adobe Audition CS6 in the tabs can only see about 2/3 across the screen, unable to get to any adjustments/buttons further than that, for example in the Processing tab only as far as half way across the Multiband 1 button. Audition is 32-bit with the 32-bit VST.

• Processing a file in the VST has your jingle placed over it a couple of minutes in when using unlicensed AutoEQ/BassEQ however this doesn't happen when using the Winamp plugin. Everything is licensed except AutoEQ/BassEQ however this works fine in the winamp plugin without any jingles or beeps. If I switch off AutoEQ & BassEQ in the VST it still says unlicensed in use but so far has not placed a beep or jingle. I will probably buy the AutoEQ license when there is a full/final release, bugs fixed and I can justify the extra 150 euro, but so far I've spent 250 and I use this purely for personal use, re-mastering music for CDs/iPod so it sounds nice and crisp Being that I'm only processing a few files at a time it seems a bit odd that I can play the winamp plugin for hours with no interruption yet the VST won't allow me to process any audio without interruption.

• That being said, I have done an A/B comparison between the VST and the Winamp plugin, the VST has excellent peak control whereas the Winamp plugin has some overshoot, however on the flip side, the Winamp plugin sounds warmer, more open and natural whereas the VST sounds constrained and too 'digital'. Same preset settings, same source material. Clipper enabled, hard limit enabled, etc.

Hope this helps

It's 9.01 here (I just downloaded and installed it from the site). Might be a browser cache issue on your end? If you just downloaded the previous version...

Resizing doesn't seem to be really supported in VST plugins. And until 8.54 I didn't support it at all. I found a trick to do it, which appears to work (and I actually tested it in Reaper, because so far it's the only program I've found that supports resizing of VST windows). I do see some white flashing while making the window bigger, but when I stop, everything looks fine.

I'm not drawing anything white myself, that's probably Windows.... If it just flashes but looks fine when you stop resizing (as it is on my end), it's just how it works... it things are worse (you end up with a broken display), let me know - in that case I'll just have to completely disable resizing support again. Please send me a screenshot or if possible a video in that case!

It had a bit too much bass so we reduced it. So in total it will indeed be a bit quieter.

Hi,

1. Can you try if this was ok with version 8.54? Someone else reported this before I released version 9, and because of that I (thought I) removed all the changes w.r.t. Windows DPI sizes for VST plugins. This issue ONLY seems to affect Audition though. Long story here, but Windows DPI support was added in Windows 7, changed in Windows 8 and changed again in Windows 10, and every program has to indicate which things it supports or doesn't support. In this case, the problem might be caused by a conflict between how Audition handles it and how Stereo Tool handles it. I have restored the default settings (let Windows handle it, I don't resize myself), which *should* have fixed this issue - and it works fine in every other program I've tested, including older Audition versions. So right now I don't know if this is a bug in Audition or in my code. Could you do 2 things?
1. Test version 8.54
2. Make a screenshot of how the window looks and send it to me (or post it here). DO NOT resize the image, as I need to be able to see if Windows is resizing the image that I'm sending to it, and the Audition part of the window.

If you set your DPI size to 100% the issue should be gone - but then everything is probably way too small (I'm assuming you have a 4K screen?)

That's on purpose. VST plugins are often (and typically) used to process files, so I need to somehow block that for the unpaid version - the Winamp plugin is mostly used for long term listening, and it's not very suitable for file based processing; for one it doesn't support correction for processing delay, so if you use the DSP plugin to process files, you'll always get some silence (or worse, the end of the previous track) at the start and the end gets cut off. This isn't the case for the VST plugin. Hence the different way of handling unlicensed features.

Of course both can also be used for 24/7 processing, and in that case the VST nags are way over the top,. but it's not really possible to see which of the two is happening.

There could be all kinds of reasons for this, but it's almost certainly not Stereo Tool. People have reported things like this in the past, I've done comparisons multiple times where I wrote the output to file and subtracted it, and the only difference I ever saw was a rounding difference in the last bit (leading to white noise at -120 dB when using 16 bit audio).

How do you see that there are overshoots?

Possible causes:
- If you're actually running it in Winamp: The Windamp equalizer might be used. Or the volume slider in Winamp could be used, which causes audible artifacts if it's not set to 100% full!
- Source material: If it's not a lossless file, different decoders might sound different. Very different even in some cases.
- Sound card: I think Winamp by default uses MME, a VST program might use Wasapi, Kernel Streaming. If the sample rate doesn't match what's set in Windows, Windows will resample any MME audio and the resampler causes clearly audible artifacts, and overshoots. So if you're measuring Winamp's output you could very well see overshoots that are generated by the Windows resampler.
- Some programs purposely send a different signal to a Winamp/DSP plugin than to a VST plugin. There are many DSP plugins that only support 44.1/48 kHz and sound weird if you feed it something else, and some programs "protect" DSP plugins by resampling any audio to a standard sample rate. If they convery back afterwards, that can also lead to overshoots.
- Some programs oversample audio to overcome bad quality lowpass filters in sound cards. If they show the oversampled audio, you might again see overshoots, and on low quality sound cards, the sound might improve.

Note: There is a DSP plugin that lets you run VST plugins, so you can run the VST plugin in any program that supports DSP plugins. That's probably the safest way to compare. As soon as you use different programs to compare things, there are far too many variables.

Just taking note of some new "Factory Reset" defaults:

True Bass:
Old was Band 1 and 2 strengths = 1.25 and 1.00, new is 1.00 and 0.50


AutoEQ and BassEQ:
Old was Kick Speedup = 0.25, new is 1.00


Basstardizer:
Old was "Place in Chain" = "After Multibands", new is "Between Multibands"

Maybe one of the new Ryzen+, or Threadripper setups? Seems like they would be pretty good for compiling these days. Especially for the price vs Intel's options in the high core count market.

No, that would probably be a lot worse actually. For one, I'm using the Intel compiler, so I'm guessing that if anything, it's optimized for compiling on an Intel CPU. Beside that, the biggest issue - and I can work around that, it just takes some more effort to set things up correctly - is that most of the time, the compiler can only use 1 core. My system has 4, 8 if you count Hyperthreading, but despite the fact that it's set to use all of those, 90% of the build time it only uses one. It should be possible to fix that by running builds simultaneously, for one, the 32 and 64 bit builds don't interfere with each other at all so I can run those at the same time, which would speed things up a lot already. I will look into that soon (I need to update my build scripts). Similarly, I don't have to wait for the Linux builds to finish while doing the Windows builds - I am doing that now, until recently that didn't make much of a difference but the Linux builds are starting to get very slow as well. (The Pi and Mac builds are already happening simultaneously with the rest; the Pi build itself takes about 6 hours or so).

can someone update the documentation with the new functions how everything works and I can set that if I want to adjust my preset