Stereo Tool

V2 RMS 7

viewtopic.php?p=14393#p14393
Here we go !
A family of 9 Presets with different RMS (dB) output target.
For example:
RMS-2 (extreme loudness similar to Web radio Loud)
RMS-5 (loudness similar to Bojcha's Addiction v11)
RMS-7 (similar to modern CD masters)

As mentioned in the other thread, the difference in RMS level explains why the mids in this were wearing me out. I see another slight change in the -6 version (the reduction in the 2.8k band) that helps, but to let you know, the -6 still sounds mid-heavy to me, but it doesn't wear me out like the previous IO preset did...

More than 50 downloads until now.

I need more feedback (besides Brian).
Do others agree with Brian that these presets are possibly too mid-heavy ?

More opinions would really help to improve things.
I don`t know who I can trust more - Brian or myself ?

There's only one possible answer to that question....

Re-visiting this, listen to Van Halen's Hot For Teacher and I'll Wait and hear the absolutely muffled percussion. The -7 version (bundled with 6.2x). David Lee Roth's vocals also have no upper end to speak of. Are you that averse to highs, or is your hearing super-sensitive (or mine totally degraded)?

Edit: Now I'm completely understanding your singleband limiter comment in the other thread. You're having to do that because your multiband mix is super-duper, way, way, way over-the-top too hot. Unrestrained by that, if you just turn Final Limiter on, you're solid bright red.

You have to tone down multiband. You're pumping way too much RMS through there.

I've looked at this more now that I've gotten sleep. The main contributer to the overpowered multiband is the EQ. That, in turn however, is fueled by the soft limit and up/down speeds.

I had a preliminary modification when I posted early this morning, but I felt it was too high-rich for your taste. It was also made before I started researching the singleband limiter / hot multiband issue. I do not have time until later tonight to do anything further, and possibly not until tomorrow night, but here is the very quick modification, which again, I feel is too bright with highs (possibly only the high midrange). You'll note that the percussion is markedly improved.

Oh, and yes, that Scorpions track sounds "ok" with your original settings. The reason why is because it is not as sonically "busy" as the Van Halen tracks, particularly with respect to the percussion. Because of this, your multiband limiting is not squashing all that much with Still Loving You. Some other tracks I've listened to are ok as well.

This really dovetails into what Bojcha (I think) was requesting about the ability to choose limit or compress on a per-band basis.

Retesting DR / RMS of Still Loving You

Bundled BASE 7 preset:

DR: 6.9 / 7.0
RMS: -9.6 / -9.7

The quick mod I did:

DR: 8.0 / 8.2
RMS: -10.6 / -10.8

My current HypersonicV6_Web:

DR: 8.6 / 8.6
RMS: -10.9 / -11.0

Much improved with only the quick mod and there's still much more to change, so it's only going to get better. Due to less loudness, I should be able to get your DR / RMS better than mine, with still mostly following your sound profile. You can blunt the highs if your ears are now bothered by adjusting the de-esser.

Hawkeye T.'s BASE RMS 7 V2

Not a complete package for different RMS levels like V1.
This is only RMS 7 V2.
Meant as a replacement of the built-in user preset in Stereo Tool 6.2x versions.


changelog:

A) I am too busy to write it.
or
B) I am too lazy to write it.

C) It is more fun for you to find out what I do wrong now (maybe worse than in V1 - if that might be possible ?) !

(Little hint: I even restrict hihats and cymbals more in V2, but overall it sounds now cleaner than V1, because I have changed BB and MB bass and mid settings too)

It is especially meant to be used via (big) speakers and high volume !
If you prefer headphones and/or low or medium listening volume then don`t be surprised that you possibly still won`t like it.

(note: I had first uploaded this preset with 2048 latency and one person downloaded it. I have now replaced it with the correct 4096 latency setting version for best sound quality)

So, you went even further with stiffling highs in favor of the fundamental male vocal frequency range (100-350Hz, bass, baritones, and low tenor), neglecting all of the harmonics, and neglecting to understand that vocals actually exist across nearly all of the frequency spectrum (100ish to way on up to 16000 or so...the upper end is called "breath" / "air")?

Fundamentally, I think the sound you're wanting is something that is not going to be found being made by audiophiles. You seem to like very, very, very prominent male vocals, with specific focus on the fundamental frequencies. What you also seem to be looking for is a Dolby B and/or Dolby C noise reduction, because of your maximum setting of Noise Gate. Dolby B/C is simply not needed for music today, because it is an analog technology. The technology itself is great and is still of great use for analog recordings. The problem is though, analog recordings are becoming obsolete, as digital recordings are now the norm, thus the technology of Dolby Noise Reduction is being made obsolete by transitioning to digital recordings, which do not have the tape hiss that the Dolby string of technologies (A, B, C, HX, R, S, SR) was designed to help alleviate.

Going back to the male vocal range, the problem is that the funamental male vocal frequencies are at the same frequencies of the first few order harmonics of the kick drum, toms, snares, and the bass guitar. Your attempt at highly boosting the 100-350 range causes boosting of the high RMS power of the kick drum and toms (primarily) and the bass guitar, thus causing the the rest of the percussion to be significantly muted, as well as a muting of vocal harmonics.

I have a similar problem with HypersonicV6, except it is in the female vocal range. It's not always evident. A track that exhibits this is Kelis - Brave, yet Diana Ross - Upside Down, I can't pick out the problem. I think I need changes to the way multiband works in order to address that issue, as I've tried many, many, many things. The male vocal range is something completely controllable though, if one chooses to.

Anyway:

- Hot for Teacher sounds even worse, as David Lee Roth has the capability of reaching high baritone / mid tenor, which you're clamping.

- You're still pouring through a massive amount of RMS through Multiband.

- Comparing V1 to V2 via using the disk writer functionality of Winamp and evaluating the output via TT DR Offline shows you have decreased Dynamic Range of Scorpions "Still Loving You" from 6.9 / 7.0 to 6.4 / 6.5

To reiterate, the previous bundled Base 7 preset had:

Dynamic Range (L/R): 6.9 / 7.0
RMS (L/R): -9.6 / -9.7

This V2 has:

Dynamic Range (L/R): 6.4 / 6.5
RMS (L/R): -9.0 / -9.2

So, what was accomplished with my feedback, as well as my effort, was that:

1) You made it louder.
2) You made it have less Dynamic Range.

Revisiting the RMS through multiband, if I get rid of your singleband compressor, TT-DR shows for Still Loving You:

DR: 1.9 / 2.0
RMS: -4.1 / -4.2

Do you understand how insanely loud that is? One of the loudest recordings ever made, Iggy Pop's 1997 remix of "Raw Power" averages RMS values in that range. All things considered, Still Loving You is a quiet track, so RMS values in the -4 range is no small feat given the track. When I use TT-DR on the unprocessed track I have, RMS is -16 / -16.4 , and you're making it up to -4!

By comparision, HypersonicV6_Web produces RMS of -10.4 / -10.5, when I remove the final limiter and adjust to 0.08 dB of loudness. This means that your multiband RMS level is just over a full 6 decibels louder than mine. 3 dB is considered to be doubling, so you've doubled, and then doubled again, or 4 times louder.

Is this where you say, in a George W. Bush tone, "mission accomplished"?

I spent literally hours of time gradually working on getting the RMS power more in line so that clamping via singleband wasn't needed. I gave up when I realized that it was going to take a major investment, but not until after making it to where I had moved to .65 final limiting verses your .45 singleband compressor / limiter setting. Attached is that effort. Bear in mind that HypersonicV6 is using .80 final limiting volume drop, so I was getting it closer and closer to the same RMS being sent through multiband.

Do you have additional processing going on either before or after StereoTool that could be changing the coloring of the audio? The only things I have post-StereoTool are:

-3 dB attenuation via Creative Labs' EQ (EQ otherwise flat)
Creative Labs' Crystalizer set at the default 50%
1 dB bass boost from 0-60 Hz via the THX control area of the card's drivers.

I use Winamp, with no other DSP plugins running at all, so nothing before ST, and only those things after it.

The attenuation doesn't happen for writing to disk, since it is a function of the sound driver. The issues I hear happen with Crystalizer on or off, and bass boost on or off.

Like I said in the other thread, the mix being too hot is a function of the mix software and hardware, not the output device. Yes, full-range speakers would give a different tone for all kinds of reasons (room acoustics, baffling of the enclosure, possibly better crossovers, etc, etc, etc...), but it doesn't change the mix being too hot (too high RMS).

The use of the singleband compressor / limiter as a volume control to tame the high RMS is the cause of a lot of the issues. That's not the real purpose of that. It should be used as a burst (peak) control, not as a continual restrictor plate.