Stereo Tool
https://forums.stereotool.com/

Low latency secondary input (microphone)
https://forums.stereotool.com/viewtopic.php?t=853		 Page 41 of 75
Author:  hvz [ Tue Feb 23, 2010 1:45 am ]
Post subject:  Re: Low latency secondary input (microphone)
I just discovered that, if I use smaller latency settings, the AGC works remarkably well with aggressive settings. So apparently it suffices to use smaller portions of sound when aggressive settings are used - it's not necessary to change the volume at the exact sample where the spike occurs. I'll probably have an updated (finished) AGC filter by tomorrow.
Author:  Bojcha [ Tue Feb 23, 2010 11:19 pm ]
Post subject:  Re: Low latency secondary input (microphone)
hi Hans..
I notice that when Bandpass is at non-phase linear Bass is better/stronger...
Author:  hvz [ Tue Feb 23, 2010 11:21 pm ]
Post subject:  Re: Low latency secondary input (microphone)
Updated version available!

Changes to AGC:
- Now the AGC responds more at the moment a peak occurs, instead of before it occurs. This means that at more aggressive settings it sounds much better (at relaxed settings there's little difference)
- Added a 2nd step, corresponding to "Remove remaining peaks" (currently not controlled by that setting) - if there are very loud peaks in a song (for example at the start of a track), those peaks are now also lowered. I'm going to add an extra slider to control this 2nd step.

Note: The 2nd step may almost completely undo the improvements of the 1st step - that will be fixed when I've made the 2nd step configurable with a slider.

Links:
Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... _BETA3.exe
Stand alone version: http://www.stereotool.com/download/ster ... _BETA3.exe
VST version: http://www.stereotool.com/download/vst_ ... _BETA3.dll
Author:  hvz [ Tue Feb 23, 2010 11:24 pm ]
Post subject:  Re: Low latency secondary input (microphone)
Quote:
hi Hans..
I notice that when Bandpass is at non-phase linear Bass is better/stronger...
True - I also posted that a while ago.

Caused by 2 things:
1. Due to the better filtering, low bass frequencies are stronger
2. Because the filtering is not phase linear, low frequencies are pushed forward in time. This means that a single bass kick is spread in time, and Loudness has more possiblities to boost the volume of different frequencies separately.
Author:  Bojcha [ Tue Feb 23, 2010 11:53 pm ]
Post subject:  Re: Low latency secondary input (microphone)
Hans, if you have time look at this song:
http://www26.zippyshare.com/v/7301700/file.html

it's about bass sounds...
Here i can clearly hear that Stereo tool can't boost Bass sounds (specialy at biginning of song)

i'll test new prelimiter...
Author:  hvz [ Wed Feb 24, 2010 12:08 am ]
Post subject:  Re: Low latency secondary input (microphone)
Quote:
Hans, if you have time look at this song:
http://www26.zippyshare.com/v/7301700/file.html

it's about bass sounds...
Here i can clearly hear that Stereo tool can't boost Bass sounds (specialy at biginning of song)

i'll test new prelimiter...
I do clearly hear a difference when I increase "bass boost" - but I suppose others boost it further. The bass boost thingy is next on my todo list...
Author:  hvz [ Wed Feb 24, 2010 2:32 am ]
Post subject:  Re: Low latency secondary input (microphone)
A few remarks:

1. For "difficult sounds" (female voices etc.) the new AGC doesn't drop the volume as much as the old Pre Limiter did. This puts some extra stress on the Loudness filter. So far I've mostly noticed this for female voices combined with deep bass sounds - which now leads to some distortion-like effects in the female voices.
2. The overall output level of the AGC seems to be slightly higher (I'll measure it later to be sure).
3. There are now more situations with relatively low levels of high frequencies with a lot of lows (before the highs were often reduced a bit less than the lows).

To accommodate for this, and especially for some (mild) distortion caused by deep bass sounds, the following helps:
- Reduce the AGC output volume (left top slider) by a few %.
- In the Multiband filter, lower "Max punch" from +75% to ??? (I've only tested it very briefly, 35% sounds much cleaner, but higher values might also work well. The problem with this is that reducing Max punch also slightly reduces the bass level)

Since I'm still making changes I'm not going to spend a lot more time on this for now (I'm still changing things, especially working on deep bass boost might change things a lot, so it's not very useful to tweak everything now), but for audio quality tests lowering 'Max punch' might help.


Note: It could be that this means that the Trumpet filter (which also handles female voices) isn't working properly. Will check that.
Author:  hvz [ Wed Feb 24, 2010 7:07 pm ]
Post subject:  Re: Low latency secondary input (microphone)
Bojcha: I just found a solution to get rid of the artifacts that I had before for Deep Bass Boost on higher frequencies. I'm now getting pretty good results for frequencies upto about 150 Hz. And I don't have to combine it with Loudness anymore, so I could even put it between AGC and multiband if it turns out that that has advantages (but I think putting it behind Multiband will give better results).


Edit: Hm, I see now why they put it before multiband... *going to do the same thing*
Author:  Bojcha [ Wed Feb 24, 2010 8:15 pm ]
Post subject:  Re: Low latency secondary input (microphone)
Quote:
Bojcha: I just found a solution to get rid of the artifacts that I had before for Deep Bass Boost on higher frequencies. I'm now getting pretty good results for frequencies upto about 150 Hz. And I don't have to combine it with Loudness anymore, so I could even put it between AGC and multiband if it turns out that that has advantages (but I think putting it behind Multiband will give better results).
I was thinking ..
Actually .. In current stereotool case, ANY changes between PreLimiter and Multiband will not be good at all. But who knows.
It's maybe possible but other things must be changed and i am not shure which thigs exactly.

In Breakaway and other, bass level is somehow connected with AGC and it have kind compressor to push it more, if i put any bass tone <200Hz, at any level... it will be boosted anyway to 100% level. In stereotool that is not case. StereoTool will rise it by PreLimiter, but there is multiband to set it down again.

So behind Multiband (in StereoTool case) is logical. but again Final Limiter and Loudness will suffer (?)
Author:  hvz [ Thu Feb 25, 2010 12:33 am ]
Post subject:  Re: Low latency secondary input (microphone)
Quote:
Quote:
Bojcha: I just found a solution to get rid of the artifacts that I had before for Deep Bass Boost on higher frequencies. I'm now getting pretty good results for frequencies upto about 150 Hz. And I don't have to combine it with Loudness anymore, so I could even put it between AGC and multiband if it turns out that that has advantages (but I think putting it behind Multiband will give better results).
I was thinking ..
Actually .. In current stereotool case, ANY changes between PreLimiter and Multiband will not be good at all. But who knows.
It's maybe possible but other things must be changed and i am not shure which thigs exactly.

In Breakaway and other, bass level is somehow connected with AGC and it have kind compressor to push it more, if i put any bass tone <200Hz, at any level... it will be boosted anyway to 100% level. In stereotool that is not case. StereoTool will rise it by PreLimiter, but there is multiband to set it down again.

So behind Multiband (in StereoTool case) is logical. but again Final Limiter and Loudness will suffer (?)
The maximum bass peak level isn't going up, so between AGC and Multiband could still work. (Mixing it with AGC would also cause the volume to drop a low when bass sounds are present, so I don't think that's a good idea). Among others, Multiband protects the Final Limiter and Loudness against too loud bass peaks, which would probably occur when I place it after Multiband.

In about an hour I'm going to upload my first bass boost version, which basically boosts bass sounds below 100 Hz, and sliding down to no boost at 300 Hz. So far I hear much louder bass sounds, but I'm also getting more distortion from the Loudness filter. So I think it will require some extra tweaking. And I'm probably overdoing it right now (a first measurement showed a bass increment of about 6 dB before Multiband, which is probably FAR too much).

Note: This will be a first test, only useful to check if the filter itself is doing what it should do. It cannot be configured in any way - actually it cannot even be turned off :( So I'm going to upload it as BETA3A, BETA3 will for now remain the best version for normal usage, BETA3A will be for testing Bass Boost only.


Edit: The new bass boost method that I'm using relies less on the input volume. If it's low, the bass level is just increased. If it's high, the shape of the bass is changed. If it's very high, the shape of the bass is still changed, but less strongly - but it's still changed. As long as the bass peak level isn't TOO far away from 100% it should work.
Page 41 of 75 All times are UTC+02:00
Powered by phpBB® Forum Software © phpBB Limited
https://www.phpbb.com/