Stereo Tool

That's very fair having other important things for stations to do first to get out of the way! It does really seem progress is being made with the software on the forums but yeah, most people don't notice any actual changes using the software.

I'd say in my experience so far, the current multiband AGC isn't best for replicating the sound of music transparently, but it's excellent for processing my mic (and probably other single instruments). I don't own a BIMP license yet, but a multiband aspect to its AGC seems like something that it could really benefit from.

Re: AGC

I have ver. 9.92 and in my specific use, I think the multiband AGC is excellent. Please do not remove controls to simplify it. Instead have different levels of access. Level A could have a three-control set-up that produces great results for many users. Level D (for example) could have a multitude of controls.

I really like the AGC in 9.92 because I can set it up to not make gain changes unless they are necessary. The platform window action is superb. All the adjustments make this possible. User ability to limit window attack is brilliant design work, please do not remove these adjustments. In my application, most of the time the AGC meters suggest the AGC is stuck, maybe broken. Then it makes a smooth fast change gain, according to the parameters I set. It is beautiful thing to see, and not hear. The big advantage of this is because the AGC is not moving up and down a lot, I can be more aggressive in the following multiband.

Everything in above paragraph has been known and used for over 50 years. Remember the Audimax? Please don't remove user ability to extreme tweak it in the modern DSP era, which I see as a major benefit of Stereo Tool.

Regarding the AGC not sounding good on music, unless there is a "circuit" problem in the coding of the gain reduction, I don't see how that is possible. My guess is subjective dissatisfaction with AGC sound quality is unhappiness with single band control, or with parameters of multiband control. If a user is unhappy with the sound of multiband AGC, they should consider their crossover frequencies and ratios, GR coupling between bands and overall depth of GR. Noting GR coupling can be asymmetrical, and time based in "multiple dimensions" such as (but not limited to) (1) the amount of cross coupling, (2) the direction of the cross coupling, (3) the extent cross coupling can increase or decrease gain compared to in band GR, and (4) if cross coupling has a time constant on it. This means a large number of adjustments. It's not supposed to be easy. If users are interested in these things, they need the huge amount of adjustment controls available in some way, like "Level X" access mode.

Re: GUI- Would you consider a visual depiction of time-based GR settings, like the parametric EQ display? Y-axis could be GR voltage and X-axis could be time, with test input signal profile superimposed in another color. User would select background test input signal from a common list or maybe draw one, then adjust GR parameters for the desired action for the test signal. Additional levels of control could possibly be added in a fashion, such as limiter GR at a higher threshold, etc. I think this would make setting up a desired profile visual and fast.

For example, user might want to delay onset of GR 2 milliseconds, then slope up to needed GR over a period of 5 milliseconds, and be able to see GR hold, RMS/peak detector anomalies, and the release profile right there on the screen. Slope could be linear, exponential, etc. and visual on the screen. For me X-Y visual is a better workflow than interpreting control descriptions and listening to reference material over and over. Doing that requires taking Stereo Tool off-air, which puts a back-up processor on air, OR having a duplicate Stereo Tool offline to try things out on.

In broadcasting we will be listening to program material over and over, but by then the train has left the station. With the ability to set up GR time/amplitude parameters with a visual X-Y, we can set up many things in advance with efficiency. Users could set up a dozen attack-release profiles with X-Y display, save each one as a new modified preset of the existing on-air preset, and then A-B them on the air when programming comes along that demonstrates the need for hearing another setting.

In the task of developing user-friendly GUI that can achieve results quickly, AND can produce nuanced complex adjustments, I suggest looking at mixing consoles used for sound reinforcement by companies like DiGiCo, Yamaha, SSL, Avid, etc. This industry has developed GUIs that keep nearly every adjustment yet make it manageable by humans who may be very busy or stressed out.

btw- with all those XLR connectors on the back, does STXtreme have a way for the user to bridge and listen to a multitude of test points within the processor signal flow, without interrupting on-air audio? Is that a headphone jack and does it have a volume control and an easy way to switch the headphone feed to A-B between two test points? You'll probably need a gain trim to enable a smooth A-B. This dovetails into an earlier comment- will you consider an internal gain structure/headroom dashboard? Does the processor have excellent internal audio processing bit depth resolution regardless of input/output level settings of each processing block? Gain structure dashboard would ease concern.

Finally- consider adding band input level controls in multiband mode. Plus and minus 8 dB adjustment trim should be fine.

Thank you for developing the processor. There was obviously a lot of thought and skill put into it.

Hello. I suggest that you insert the latest version of the X-Mas Sensation preset into the Stereo tool. viewtopic.php?f=10&t=5396 For me, this is the best Stereo tool preset so far.

STXtreme multi-instance with Nielsen PPM watermarking is available now.

I've tested the New BETA031 for:

"Check startup pre-emphasis issue mentioned here: viewtopic.php?f=14&t=32702&p=73104#p73104 - solved; value was not updated without GUI updates after the sound card logic had executed."

But it still happens, Starting my Automation: Pre-Emphasis is on on the de-emphased Webstream, only after opening the GUI the sound is correct.

I have just tested it again, and it works fine here (and before it didn't). Am I missing some step?
- FM on with pre-emphases, MicroMPX on,
- If you switch MicroMPX off, the pre-emphasis is there on the output to the host application, it disappears as soon as you enable MicroMPX
- With these settings, restart with no GUI

I don't get pre-emphasized audio if I do this. Tested with BETA034 but that shouldn't matter.

So, I've started messing with using 2 instances of ST at once: one for Music/Spotify alone using VAC 2, and another for everything else using VAC 1 (real-time stuff like games). The chain is Music player > VAC2 > Music ST > VAC1 > Main ST > output. I'm curious if there's a way to detect silence so the Music instance of ST stops using any resources while it's not being fed anything above a certain threshold. I know I can just close the music instance to recover that 5%, but I am supremely lazy.

I remember testing ST Enterprise and being woken up by your voice a couple of times since it speaks out every 12 hours regardless of whether it's processing silence. That made me think of how BreakawayOne doesn't nag you to register when using Trial features unless you're processing stuff. The clock only seems to accrue time while it's being used. The reminder that's supposed to sound every 3 hours isn't incredibly intrusive even though it's programmed to potentially be much more common than the 12-hour reminder of ST Enterprise.

Just some thoughts. Thanks!

Greetings on this thread also and many Kudoz for this magic software!

So, I tried to use the latest 10.02 Beta (and 10.0.1 earlier) and unfortunately the output of the sound was filled with crackling noises. My setup currently is this :

Jazzler --> VAC Input --> StereoTool --> Stream with the build-in engine or with RadioCaster.

At first I got scared and rolled back - most settings seemed to have been loaded properly which meant the preset was ok. Today, I tried again and I still got the crackling sound.
Finally I spotted that the culprit (in my case) was that for some reason ST had selected for my Input Kernel Streaming/ VAC - B Output . Theoretically Kernel/Wasapi/Mme should all be working correctly, but only MME seems to be working.

Red for Crackling, Green for sweet sweet sound!

_________________
Keeping The Rave Spirit Alive
https://raveunited.dance/

Hi dR_wH0,

So did the old version sound fine with Kernel Streaming? The selection should be the same. What changed is that we moved to a new I/O framework.