Steenamaroo
...
Ok, so adjusting the release time shows a change in transition time without going below the threshold again, so
Post 22 is the one I think.
Post 22 is the one I think.
Bobbsy
Boring Old Git
Looking at those latest pics, I'm going to guess that BSG either has "look ahead" switched on in his DAW...or, if that's not a control, the compressor algorithm has a permanent "look ahead". The lack of any overshoots even at 0ms attack and release would tend to suggest this.
Steenamaroo
...
How do you mean lack of overshoots?
Surely the fact that the compression amount reduces over time on the lower level signals means it's overshooting?
Surely the fact that the compression amount reduces over time on the lower level signals means it's overshooting?
Bobbsy
Boring Old Git
Sorry, I should have specified "last" picture--the one that is just a flat-topped file which corresponds to 0/0ms attack and release.
Since, as per the explanation I got, a hardware compressor IS only reacting when levels get to the threshold level, at very least the first or last half cycle will have to exceed threshold to turn on the compression--then add the extra time to "throw the switch" and calculate the new level with regard to the ratio setting and there's always a spike or dip in hardware compressors.
However, also mentioned in that explanation I got, software compression in a DAW has the facility to "look ahead", find the level changes, then switch them in accurate to the sample. I got a similar result to an experiment in Audition--except that simply going to 0ms attack and release, there were still minor peaks and dips. However, going to 0ms and a small amount of look ahead, I got the same "flat top".
....which, it occurs to me, may bring us full circle. The addition of "look ahead" would, in effect, result in something pretty similar to a compressor reacting to "downward changes". It's not trully the downward change that's being reacted to--but, like a stage psychic, the ability to know where the threshold crossing is going to be a few samples early lets the compressor appear to be more clever than it is!
Since, as per the explanation I got, a hardware compressor IS only reacting when levels get to the threshold level, at very least the first or last half cycle will have to exceed threshold to turn on the compression--then add the extra time to "throw the switch" and calculate the new level with regard to the ratio setting and there's always a spike or dip in hardware compressors.
However, also mentioned in that explanation I got, software compression in a DAW has the facility to "look ahead", find the level changes, then switch them in accurate to the sample. I got a similar result to an experiment in Audition--except that simply going to 0ms attack and release, there were still minor peaks and dips. However, going to 0ms and a small amount of look ahead, I got the same "flat top".
....which, it occurs to me, may bring us full circle. The addition of "look ahead" would, in effect, result in something pretty similar to a compressor reacting to "downward changes". It's not trully the downward change that's being reacted to--but, like a stage psychic, the ability to know where the threshold crossing is going to be a few samples early lets the compressor appear to be more clever than it is!
Steenamaroo
...
You know what.
I don't care anymore, Sorry.
I don't care anymore, Sorry.
bouldersoundguy
Well-known member
The compressor I used has something like look-ahead. But look-ahead doesn't really bear on our question, it just allows the target gain reduction value to be calculated earlier, even ahead of the signal.
Attack and release do nothing until the target gain changes which in turn doesn't happen until the threshold is crossed.
Rane Note 155 has a block diagram that illustrates this nicely. First threshold is set and the signal measured for the amount over threshold, then the scaler calculates the target gain value, then that target gain value is modified by the attack/release filter. Attack and release don't "know" anything about threshold crossings, they only "know" what the target gain reduction is doing, and that's what they modify.
Attack and release do nothing until the target gain changes which in turn doesn't happen until the threshold is crossed.
Rane Note 155 has a block diagram that illustrates this nicely. First threshold is set and the signal measured for the amount over threshold, then the scaler calculates the target gain value, then that target gain value is modified by the attack/release filter. Attack and release don't "know" anything about threshold crossings, they only "know" what the target gain reduction is doing, and that's what they modify.
N
NYMorningstar
Recording Modus Operandi
You must spread some Reputation around before giving it to bouldersoundguy again. TY!
N
NYMorningstar
Recording Modus Operandi
I woke up this morning thinking of this question, lol. I'll take a stab at the answer - your sampling rate of 44100 shows a 4.41 kHz sine wave with good detail.
bouldersoundguy
Well-known member
It was so I could generate tone segments in multiples of 10ms and have the wave at the end of one segment match seamlessly with the beginning of the next. I listened to the clips repeatedly and it played more smoothly.
bouldersoundguy
Well-known member
Gate attack and release controls, on the other hand, do operate in response to the threshold. Expanders function more like compressors in reverse, but I don't recall using an expander with attack or release.
Similar threads
- Chrisulrich
- Mixing Techniques
