Stereo Tool

I'm addressing my concerns by moving the high-pass filter up to 45 Hz, maybe higher depending on how things test out. Just doing this has cut 2 dB off of the RMS power, which I can now use elsewhere. It would be nice to have a steeper high-pass though and/or changes to bass boost / deep bass boost in loudness (ability to configure a non-zero start point for bass boost / user defined frequencies for deep bass boost).

Edit:

Nope, that doesn't work good enough. When I played Wacka Flocka Flame - "No Hands", the sub-bass wasn't there anymore.

I need a truly steep high-pass. This discussion has come up before with the idea of stacking a high-pass or having a "double-strong" high-pass. Can that be had with this product, or do I need to look for additional processing tools?

The filtering happens *before* any processing. So that could very well be; the AGC uses the filtered signal for its calculations.

The non-phase linear filters are pretty steep, BUT other processing steps can introduce a DC offset which means low frequencies. So I guess that is what's happening. (If you turn everything else off, you should see a very steep cutoff). The earlier discussion that you mention was about the phase linear version of the filter which is indeed very flat.

If I turn everything else off, then what's the point? That is testing that individual piece in isolation, not as how it integrates with the other components.

If my car is out of gas and has flat tires, if I only inflate the tires, I still don't have any fuel, and if I only add fuel, I still have flat tires. Either way, unless something else happens, I'm not going anywhere. Similarly, if I turn off everything else, I can, in theory, have a steeper filter, but I can't have the other stuff, so...meh...

I'll see if I can work around this with a 30 Hz band that's heavily clipped. Wish I didn't have to try this though...

Especially the Loudness filter *has* to introduce lower freqs sometimes because it leads to a better audio quality. At least make sure that you turn the option there off... (Bottom of the Loudness panel, something with allowing DC offset).

It actually says:

Forcibly remove DC component caused by loudness (reduces quality!)

Based on your thinking it said the reverse (to allow it), I would guess I'd want to try enabling it.

Edit: Yeah, so that didn't do what I wanted. Didn't really appear to do much of anything.

I think a fundamental misunderstanding exists. I think you don't understand what I want.

There is absolutely no need for me to have a lot of content in the 0-35 (or 0-40) range. Your high-pass, once everything else is factored in, seems to give about -20 to -30 dB of cut, gradually, getting to that level of cut at 20 Hz when set at 35. What I'm looking for is something "wild and crazy", such as -40 to -60 dB of cut, with a sharp drop (high corner, if I'm reading and understanding correctly). Even that isn't as steep as the low-pass, which seems to be -60 to -90. The 25 Hz area has as much energy with the current high-pass design as some of the midrange. That means that in the total mix, the inaudible, and unplayable (unless you have really professional grade equipment) content is robbing power that could be used elsewhere in the mix. To combat the effect, I have to raise the mids and highs louder than what they normally would be to survive through the mix, which leads to pushing a higher RMS level for the entire mix. It's either that, or sacrificing the 40-80 Hz region, which destroys modern dance music / RnB / hiphop.

Now, I can try to find an additional plugin, but handling it that way means the mix produced by your product will need tinkering and will not sound the same unless someone also uses the additional plugin.

Finally, a great frustration to me is that you are asking questions about my settings inside your product, when you could visit the preset forum, download the preset, and use it for yourself. I think if you actually took a look, it might get us on the same page, or at least in the same chapter.

@Brian: You mean the Hypersonic V7 Beta 5? I've just tested it (both as it is available on the site and with non-phase linear filtering) and - at least for the tracks that I tested with - those low frequencies are really a lot lower than the mid frequencies. The filter is indeed less steep than I thought and nowhere near as steep as the lowpass filter*.

Having said this, I *think* that these numbers hardly have a noticeable effect on the RMS level. (Please tell me the name of a track where it does to convince me!). I used CoolEdit to filter out all frequencies ABOVE 5 Hz below the filtering frequency, and there really wasn't much left.

* actually it is much steeper, the problem is that I have to measure the steepness in absolute Hz).

Did you test with a heavy sub-bass track?

Try Wacka Flocka Flame - "No Hands"
I'm not sure I understand what you did. You used CoolEdit to filter out all frequencies ABOVE 5 Hz below the filtering frequency. I had erred and left the filter at 32 instead of 35, so are you saying that you left in the frequencies from 0 to 27Hz (or 0 to 30, if you adjusted for 35)? If yes, then again this is going to be highly track-dependent, specifically tracks with a lot of 20-35 Hz content (like "No Hands").

For the moment, I'm going to be rebooting. I've tried 3 times to process that track, and when viewed with CoolEdit, the low-pass is not even functioning. I've seen the low-pass work, so I'm hoping it is some transient thing with CoolEdit.

I have tried multiple times processing "No Hands", and there is no low-pass....

Edit: Problem found with the low-pass. If the MP3 sample rate is 48k, the Nullsoft DiskWriter output has to be matched to 48k output for the low-pass to show up in the WAV file.

The high-pass issues still persist.

@ Brian- I tested track. If i understand You want that bass at ~30Hz to be full there, but almost nothing under that (?)
That's just not possible
Here what i have with that track when lowest subz kicks - http://i.imgur.com/Lm88J.png - no phase linear 30Hz!

Difference between no phase linear and and phase linear filter are:
- Phase linear filter is placed AFTER Multiband
- no phase liear HPF is placed BEFORE AGC
- no phase linear filter is much more steep! but since it's before AGC, filtered sounds gets amplified by AGC and MB

Now you may ask, what would be if no_phase_linear HPF is after MB.
We will never know becaue that will never going to happen.

You partially understand, and yes, it is possible. See, that specific track and others like it are a torture test of their own. Unless you've got professional-grade equipment, you simply cannot accurately reproduce that spike at 32. Even then, you can't actually "hear" it. That is a frequency that is more felt than it is heard.

I'm thinking that I'd really like a HPF at 40. I'd like it to have a very steep rolloff, as in more than 36 dB/octave, probably double that at 72, with a sharp corner. This would leave 35-40 mostly untouched, then "turn the corner" and be strongly attenuated.

There is a filter type that could do this, known as a sinc, or "brick-wall". Normally used on the high end of the spectrum, it can be used on the low end too. That does have ringing impacts, but, at least on the high end, I'm reading something about "Faulkner Averaging".

Alternatively, something I mentioned some time ago in the Wish List area was a band-stop / notch filter. If I could define a notch filter from 28-35 Hz, it would address the bulk of the low frequency energy that my system cannot play.

Another idea would be to modify all kinds of things and allow users to enter their own frequencies, ratios, and Q levels.