Stereo Tool

From what I’ve already found, Orban’s LPF gets way too close to the pilot. What I personally would do is at 16.4k is start slowly sloping the audio down until 17.2k where I would leave a little bit of audio (similar to the Orban) at about or below -60 (dB? whatever MPXTool is using for measurement on the MPX spectrum analyzer) until 17.8k have it start to slope down again and then by 18k have the audio completely cut off.

17-18k is way too much for audio to be there. It might create conflicts with pilot and rds

What you see is in that LPF is just minimal CPU usage in use for LPF - minimum dB per octave, just enough to fully protect pilot. There is other reasons, but that's main one.
Also you can see that 'slope' starts from around 15kHz. Important is first few dB's, so if you compare it with LPF from ST @16kHz, wich have much steeper/narrow slope (more dB/oct), it's actually superior to that one in Orban.
Both are just fine tho, receivers will anyways filter that on their own.

_________________
control point
control point2

Some replies:

- The new compressor is almost finished, when that's the case we will also make the AGC use the same algorithm because it's vastly superior (and far less audible) than the current one. You don't even really need an AGC anymore with the new compressor, but I can imagine cases where you might want to combine an AGC with a lower-ratio much faster compressor to create a specific sound. So that's coming.

- Naturan Dynamics indeed needs a better dynamics detection, very few presets use it and with the new compressor it's even less needed. Having said that, on some material it can sound really good and add more dynamics, so we do need to look at it at some point. Adding detection should help a lot. But ND is also still based on the digital compressor code; it might be possible to update it to the new compressor design as well. But we have not looked into that at all.

- Lowpass filter: Any noise around the pilot (such as in the image that was posted) big reception issues in modern receivers, because many receivers measure the audio around the pilot to determine reception quality. If it's noisy they blend to mono, and we have had many reports from stations that switched from other processors to Stereo Tool that their stereo reception area got bigger, in some cases (where there's no interference) they had stereo reception up to 30 km/20 miles further from the transmitter after switching. The reason why we set the maximum to 16.4 kHz as the maximum is that anything above that will interfere with RDS, and that hurts RDS reception. RDS is modulated at 57 kHz +/- 2400 Hz, so it goes down to 54600 Hz. With a lowpass at 16.4 (and we only START to go down from that frequency), the L-R signal ends up at 38000 Hz + 16400 Hz = 54400. That leaves 200 Hz to go from full signal to below -90 dB.

Someone mentioned Omnia LPF, I don't know what they do differently? (The 9/9sg uses our filter, but the 11 doesn't).


(AutoPilot and ND dynamics detection added to todo list. We got more requests for AutoPilot recently so I guess people really want it, but I have seen receivers that won't go back to stereo once they have lost the pilot).