I have a question about this control I’ve never used so far; according to Cubase manual it “Sets the time the applied compression affects the signal after exceeding the threshold”; so, searching on the web I’ve also read the exact opposite, i.e. “hold” acts “after the signal falls back below the threshold”? I would like to know:
This is an interesting question. I tested it - and I got inconsistent results. Generally the Hold parameter seems to apply after the signal falls below the threshold level. But trying several values for Hold and several run throughs with the same value I get varying results.
That is quite alarming.
Here are my settings, where Hold will be tested with varying values:
I think that’s a question of topographics and their definition. IMO, both definitions aim at the same thing:
Is a rather unfortunate choice of words. I believe exceeding is meant as being beyond the compressor’s required trigger range (defined by the threshold) which means the signal itself is actually below the threshold in a topographical sense at the same time.
Basically, hold will prolong the compression affect to the signal although it doesn’t meet the set requirements (threshold) anymore. Afterwards, the release phase will take over. That’s where I heard strange jumps in level, not sure if this is intended behaviour. @Johnny_Moneto : Is it that what you described as “alarming”?
No. Alarming is that without changing any value I get a differing duration of compression in certain cases. In other words, the compressor does not always work in the same way and is thus unreliable.
This happened when I first used 4000ms for hold and then switched to 5000 ms for hold. A couple of runs with 5000ms then yielded different results.
I had different results too but I still doing test; I noticed that when I use Hold set to 1000ms compression lasts more, actually, compared to when using release set to 1000ms… And there’s also a pumping effect..
Not so clear, in my comprehension; to me exceeding sounds more like ‘over’ but I could be wrong, of course…
Here two screenshot of my settings test (obviously a picture is not dynamic so take them for what they are…)
That is indeed alarming. Did you try within a loop? I am sure you took that into consideration, however, 4 to 5 seconds is a long time so it has nothing to do with a looped test with too little time to reset, right?
This is difficult to compare: Attack & Release do not always represent what you might think they stand for. It also depends on the definition of the reference compression level which is used to calculate attack/release. It doesn’tnecessarily have to be reaching/leaving the 100% compression level.
I am with you - that’s what you’d think reading the manual. I tried to “think” like the Steinberg documentation team which sometimes describe things in a way you’d not expect using plain English/German. But you are right: Other times, they want to say exactly what you’d think they want to say because they turned the functionality of a known feature (like wet/dry mix) upside down.
In this case, it wouldn’t make sense and judging by the compressor’s behaviour my interpretation seems to be correct. But who knows, I have been wrong before…
BTW: I am not sure what to make out of your last screenshots @arfo62 …
I was doing tests, actually, trying to guess something more but I know compression is a tricky process to understand completely; I often read about different opinions about terminology (what’s attack, what’s release…). I’m not obsessed with definition but I would like to understand the tools I use the best I can…
It’s not a competition at all and this is what I generally like in this forum: we try to help one another if possible
Me too, for the truth… I hoped someone more expert than I am could ‘read’ something in them…