Quote:
Originally Posted by
Insanelysane
➡️
The equal loudness was the reason I couldnt use the original HG2 in my workflow, it was always great for adding juice but there was no way to properly balance the levels with the tweaks, they really needed a link button that locked an output volume knob moving counter-clockwise to the Pentode and Triode knobs that we (lets face it) would crank clockwise.
Regarding the sound, in the blurb it does say the 4 filters have been reworked, TMT is probably much of a muchness to a lot, too subtle and only really worthwhile when stacking multiple instances of the plugin to get its mileage.
The M/S may be much more interesting for spatial fattening, the tweaking on that front is most welcome IMO.
With the original, you have to use the bypass button as you adjust the output until it's at the level you are after. It's just like the hardware. Obviously it's harder to do with the software and although I've never found it to be bothersome on any level, I know it's something a lot of people asked about and REALLY wanted for a feature. With that in mind, the ability to link the pentode or triode to the output with an inverse response has been added but it's far from perfect, which is why we were extremely hesitant to add it at all. I'll explain....
When you have a single gain stage with a linear input trim and linear output trim, it's extremely easy to simply lock the two together and have them move in opposite directions. This is how the density works and it works amazingly. The problem is that things change dramatically when you start attempting to do it inside the unit with individual controls.
The saturation, pentode, triode and output pots are not linear and actually have different curves. That means that if you divide the pot up into steps (for ease of explanation), one click up or down will produce a very different change in volume depending on where the control is positioned. So if you imagine that on the Pentode, one click up adds 3db of gain but one click down on the output knob (based on its current position) gives you 1db down, you can immediately see the issue. But that's just the beginning....
Now keeping in mind that the curves on each pot are different, imagine that the stages interact so not only does the position of the pentode knob change how much gain you get for each step, so does the position of the triode knob (it affects not just the triode gain but also the pentode gain)! Aaaaaaaand if that isn't enough, each tube has a saturation curve that isn't linear so how much gain you get out changes based on how hot the signal is and how hot the stage you are adjusting, AND the previous stages are!
So yeah, it's mind numbing and the best you can really do is to try and find a curve that works best at most settings, which is what we've done.
Since I've already written an (unbelievably boring?) essay, I'll touch on the other approach that a lot of people suggest. What about using RMS detection and matching?
As I'm sure most of us know, RMS is measuring the energy over a period of time. It's allowing us to get an idea of the average level, which corresponds fairly well to perceived level in a balanced spectrum. By its nature, this takes a period of time, or a sampling window that needs to be analyzed. The longer the window, the more accurate a reading we get in terms of average level. So if we have say a 3 second window, that would mean that every time we turn a knob, we have to wait 3 seconds. Obviously that is a non-starter.
So what if we make the window much shorter and look at only milliseconds, the delay would still be super annoying but maybe workable. The problem is that we no longer have an RMS detector. What we have is something that is measuring the instantaneous level or peak level and if we set the output to compensate for or ride that, we just created a compressor (or expander depending on what the rest of the unit is doing)! Obviously the output acting as a dynamics processor, mangling the output signal is a non-starter as well so we're trapped between these two unworkable solutions.
Sorry for the long response but with so many people asking about it, I wanted to explain exactly what the hurdles are.