Stereo Tool

Hey everyone,

I visited Leif today, we talked a lot about things, checked the declipper and then we also did some A/B comparisons. Unfortunately he (Omnia 9) clearly won all of those, because it sounded more consistent and the highs were much brighter. So I couldn't stand that and I quickly (15 minutes or so!) hacked some new clipper control code according to the redesign I talked about earlier... And guess what: Now suddenly the 2 clippers are nearly identical (of course mine isn't composite - yet). The transients sound good now (more punchy), and the output looks just as tight as the one from the Omnia 9... That's something new for Stereo Tool! . So expect some big improvements coming soon!

Ahm ..
Leif was needed to get that out?


Phase-Linear HPF: You just need to look at one of my last posts and read what i get from tests! Also post from Eldorado (Audition picture)

Basicly, You just need to move it before AGC and make it 2 times stronger (two phase-linear in raw) or, if any better idea for same result.
I think that's about 10 minutes of job (sry if i am wrong in this). Checkbox "before AGC" another checkbox "2x"
I think there will be no artifacts, or maybe just small one around loudness.

No, I already had that change planned, but I wanted to fix my declipper first. But after hearing the difference between the Omnia 9 and my own clipper I just couldn't stand not knowing whether my idea would work. Well, now I know: It does. It's definitely not ready to post yet (among others, I have to do something with all the protection steps such as the voice vibrations filter, and the CPU load is insanely high because it's a very crappy implementation). But at least I know what's wrong now. And better: When this is fixed I expect to be able to create a composite clipper within a few days... (I might be wrong though: Leif said that there are some caveats. And of course he's not telling me which they are ).

I know - that fix would require increasing the latency though (if you use 2 Stereo Tools, you also get twice the latency!) Which is of course no problem for many users, so it should at the very least be an option.

About moving the highpass filter: The issue here is that the highpass filter causes these annoying artifacts

I've heard artifacts with the test file from Bojcha: http://www63.zippyshare.com/v/14583087/file.html play this sample in a loop.

The artifacts are around 400 Hz (between 300 and 400) when i check it with SpectraLAB.
You can see the spectrum moving up and down in a strange way. You can hear it if you play the track.

But now my question, why are these artifacts only present in the following combinations?
HPF with 2/3-Band AGC ON
HPF with multiband ON
HPF with Loudness ON

Why are there no artifacts when i only use HPF ON ?
Why are there no artifacts with HPF and 1-Band AGC ON ?

My conclusion about this is that there's a bug in stereo tool that's causing this artifacts.
But i don't see a bugfix for this added to the todo list.

Is it not a bug then?
Why is this then only happening with this combinations?
And why don't i hear this artifacts when i use 2 Stereo tools for only HPF (because then i don't have the artifacts and i have strong filtering) and a 3rd stereo tool for the processing, so that the HPF is also before the AGC ?

All above is of course about phase linear.

De-Clipper is done (?) Will public 6.20 version be released soon ?

De-Clipper is done...?

I hate declipping!! i want clipping !!! ( mpx ) hahahhaaha

Hey guys, I found a nifty way to decrease cracks and pops if you have a lot of things running on ST and especially if you have a PCI sound card. If you can get into your BIOS settings and find an option called PCI latency timer change it to something higher. This is usually set to Auto, or something low (in my case 32) I had two other options 64 and 128 so I set it to 128. I did some reading into it while I was messing around with other things in the BIOS (RAM timings) and apparently it controls the timing of information being sent to different devices on the PCI bus. Usually there is only one thing on that bus and that's the Audio device, especially true if you have a PCI audio card. You can also have other things on that bridge like a WIFI card or a LAN card, or even a VERY old graphics card. So try it out if you want to cause after doing that I'm not getting any pops or crackles from ST, and I have the Natural Dynamics on, Over Samplig on and the latency is set at 4096. I get pops and crackles when I start up programs like Photoshop, Sony Vegas or Chrome with about 30 tabs open but so far nothing, not a single pop. Thought I'd share my finding.

Though be careful cause if you have other things on the PCI bus their latency will increase, I'm not sure if it'd be dramatic or not but the jump was high in my case, 32 > 128. But I don't have anything else on my PCI bridge, so everything is running fine.

Also Hans, you said the last beta you turned natural dynamics off, unless you go into the ST file and manually mess with the settings. So what if I have a ST file from a previous beta where ND was there, and saved those settings? Using that with the new-er non ND beta would still have ND on since the settings were saved right?

_________________
<-- I like this guy!

Yeah it would - at least for now...

But that doesn't mean it's the only thing on your PCI Bus. Case in point, my box only has a WiFi card plugged into a PCI slot, but here's what's actually on my PCI bus:
http://i.imgur.com/yLAfj.png
And half of those things have other devices connected to them.

Of course, with PCI Latency Tool
http://downloads.guru3d.com/download.php?det=951
you can adjust the latency of each device independently, and if you run it at startup (with the settings set to apply at startup of the app too) then you'll get persistence.

I haven't researched if this work with Windows 7 yet. But it should still work great with 2K/XP.

Whoops, wasn't thinking when I wrote that. Thank you!

_________________
<-- I like this guy!