Stereo Tool

I'll listen again after some sleep, but I have already been working on a RC2 that has a little more bass to it. If you're talking about the sub-bass region though (30-60, maybe as high as 80), it's very difficult to handle with the current peak-based multiband.

@DJ_DOGGY: I'll have a listen later today, but the bass is not locked to stay below the mids - in fact the opposite is true, the mids are not allowed to drop below the bass.

So it must be not a bass drop but a total volume drop. Where my guess is that if you switch to a singleband AGC, it will be worse.

Yes, extra filters, more CPU. Allowing users to defeat some of these more CPU intensive (or perhaps handle it via an 'overall quality/cpu slider like BBP) is a good idea to allow usage on older cpu's or in applications where quality trade offs are acceptable. It's a challenge to meet multiple markets/applications. In FM broadcasting, I personally find the notable recent beta improvements in the loudness clipper to be worth every ounce of extra CPU usage. It's all subjective I guess. A new mid level I7 pc as a dedicated single or dual instance ST pc handles it reasonably comfortably with ST settings optimised for higher quality. Nothing beats driving along dial surfing, where your station stands out as the cleanest and most consistent for a given loudness level.

This latest beta on FM notches ST up another rung in my book. It's great. Hole punching on FM - now largely gone. We have a number of <100w FM isolated sites in our network, where the FM tx is the most expensive box at the site, so even buying a dedicated entry level hardware processor for each site is largely out of the question from a viability point of view. That's what initially started my search for a good viable pc based solution. I didn't expect that search to end with a product that competes ever increasingly well (and in some respects can surpass) the popular mid/high end hardware boxes on the market. Fantastic work Hans. I really can't wait to see where you take this over the next year, and can (somewhat selfishly!) hope it remains within our reach for our smaller FM sites. You deserve great success with this. Cheers AJ

sounds really good and sweet

_________________
by GAP
"Less is More" (Bob Katz)

Yes it goes worse .That is clear .
I say that Bass not locked to stay below the mids but opposite - bass is locked to go above(not below) mids . In other words max bass level is mids level .But in this song we need bass to go up more (in the mentioned parts) .You will find for what I am talking about

If there is not a problem for one set of users, you should not force the "fix" onto them if it is otherwise avoidable.

You have explicitly stated that the issues being discussed here were only when you started using pre-emphasis. That does not concern non-broadcast users. Why cannot the code be structured just like the check box in AGC to "prepare for BS-412 if enabled"? You check to see if FM or pre-emph is enabled by placing an If statement around the extra filtering and directing the code down the appropriate path.

If this code remains in place for non-broadcasting users and stays at the same performance level, then beta015 is as far as I will be able to go with this software on my current system. Forcing a doubling of CPU load on me to fix something that doesn't pertain to me is quite off-putting.

It is either a total volume level issue driven by your settings, or the hole-puncher, since you're using FM and pre-emphasis.

Not sure how you listened to HypersonicV7_RC1, but at worst I only hear perhaps a slight drop in relative bass level. This is most likely not driven by the AGC itself, but by Loudness attempting to make room for more mids and highs. If I toggle between processing and bypass all, the ratio of bass with processing on seems to be either almost the same or the same.

With RC2, so far I had already lowered the downspeed of the 240, 480, and 960 bands, giving a slightly increased presence of that area and bass harmonics.

Overall, I think this is a settings issue or a broadcaster issue that does not pertain to those of us who are not broadcasting.

I'm curious about one aspect of the behavior of the AGC. I play a track using the FM TRANSMITTER SOFT EUROPE preset that has a fairly consistent level that always sits comfortably within the range of the AGC (i.e. the AGC meters spend most of the track hovering around mid scale). If I then increase the pre-amp gain slowly by say 10db, the MPX output loudness level of the processor notably increases for the remainder of the track, by at least 1db or so.

If I take an CRL SGC800 AGC/leveler operating in its comfort zone and gently increase its input level by 10db, it makes minimal difference to the units output level. Likewise, on an Orban 1100 (operating about midscale on its AGC window) slowly increasing the AGC drive by say 10 db, you can't hear/see a notable increase in the output loudness of the processor. It remains fairly consistent. This suggests to me the AGC in those units are working as a true leveler.

If a studio operator drives their mixer output levels too high, as some announcers do, while I expect the processor to react by increasing quiet parts of tracks on air (more so than if driven at the correct level), I don't expect to increase the overall loudness of the processor on air. But ST seems to do that.

Is there a reasoning that the ST AGC does not seem to take a more consistent leveler approach? Cheers AJ

...and again, this seems to be a FM broadcaster issue.

I'm not trying to be rude, but I am trying to protect more changes from being foisted upon people who don't need the change.

I understand Brian, no problem. Consistency of loudness and AGC Leveling action is of interest, to a slightly lesser degree perhaps, for those internet streaming as well. We internet stream with a different processing setup to to our FM processing, and that's not an uncommon approach for those that want to maximise quality on both platforms. *If* there were to be any such changes to the leveler in this respect, I can't see it (audio wise) being detrimental to home users, and I suspect **although I don't know** that any such change in behavior would not require additional CPU intensive steps - it's just a different way of doing things. As I said in my earlier post, I'm all for being able to disable certain cpu intensive features, or alternatively, have different 'overall quality/strictness' settings that scale from low cpu, small speaker home audio use, right through to the best fm mpx signal you can get out on an i7. Hans seems to have accommodated those differing needs in the non-beta releases - allowing for the continuing evolution of the pc, and that we have to retire our older pc's eventually!