I do not understand what this means; can someone clarify it for me?Contextual Info wrote:Attack
Attack defines the attack time, that is how quickly the level detector increases the measured input level. When the input peak level is higher than the current level measured by the detector, the detector moves into the attack mode, in which the level is increased depending on the input signal. The higher the input signal, or the shorter the attack time, the faster the measured level rises. Once the measured level exceeds the Threshold then the dynamics processing (compression, limiting, gating) will start.
Attack shape
Attack shape controls the shape of the attack stage. The shape mainly affects the ratio between pumping and distortion, which simply cannot be avoided. Note that the attack time parameter is quite dependent upon the mode, so you may expect differences in the actual attack time between different modes of the Attack shape.
Slow modes usually produces more pumping, but less distortion. Conversely fast modes reduce pumping, but cause more distortion. The type of the distortion is different between modes. You may actually profit from the distortion caused by some modes as the generated higher harmonics may enhance the audio. The default fast mode provides a good compromise between distortion and pumping.
There are also 2 custom modes available. With these modes you can actually draw the shape. Note that what you draw is NOT what you get. The custom shape graph converts the difference between the input level and the drawn envelope value into the speed of level detection.
For example, if you set the default graph to show 100% across the X axis, then the results will be similar to the Slow mode. If you then move the point on the right upwards to say 400%, it will mean that, if there is a big jump in the audio, the envelope will initially follow it 400% faster than it normally would. The closer the envelope is to the current audio level, the lower the speed-up. Similarly, if you take the point on the left and move it downwards to 0%, it will slow down the envelope as it approaches the audio level.
Q: MDynamics / Attack Shape?
-
- KVRAF
- Topic Starter
- 10310 posts since 2 Sep, 2003 from Surrey, UK
I was looking at this, after reading another thread.
-
- KVRAF
- Topic Starter
- 10310 posts since 2 Sep, 2003 from Surrey, UK
Bump ... anyone?
I've been doing some investigation and experiments but cannot make sense of the results.
I've been doing some investigation and experiments but cannot make sense of the results.
-
MeldaProduction MeldaProduction https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=176122
- KVRAF
- 14019 posts since 15 Mar, 2008 from Czech republic
Well, it controls the attack based on actual level... How to get a grasp of it, only experiments, trust me, it's really hard to "imagine" this . Just make test signal from say a sine wave modulated using a rectangle and check what it does - it will essentially control the attack/release curves.
-
- KVRAF
- Topic Starter
- 10310 posts since 2 Sep, 2003 from Surrey, UK
Yep, I tried a few experiments and could not understand the results I was getting, so I thought that I was doing it wrong.
(a) "the difference between the input level and the drawn envelope value" -
what drawn envelope value is used to calculate the difference?
(b) does the word "envelope" used in the Info refer only to the drawn envelope shape in the graph or does it sometimes refer to the Attack (or Release) shape?
(c) in the envelope graph, what does the X-axis represent? The input level? The difference between the input level and the "measured input level"? Something-else?
The underlined bits are those that I do not get.There are also 2 custom modes available. With these modes you can actually draw the shape. Note that what you draw is NOT what you get. The custom shape graph converts the difference between the input level and the drawn envelope value into the speed of level detection.
For example, if you set the default graph to show 100% across the X axis, then the results will be similar to the Slow mode. If you then move the point on the right upwards to say 400%, it will mean that, if there is a big jump in the audio, the envelope will initially follow it 400% faster than it normally would. The closer the envelope is to the current audio level, the lower the speed-up. Similarly, if you take the point on the left and move it downwards to 0%, it will slow down the envelope as it approaches the audio level.
(a) "the difference between the input level and the drawn envelope value" -
what drawn envelope value is used to calculate the difference?
(b) does the word "envelope" used in the Info refer only to the drawn envelope shape in the graph or does it sometimes refer to the Attack (or Release) shape?
(c) in the envelope graph, what does the X-axis represent? The input level? The difference between the input level and the "measured input level"? Something-else?
-
MeldaProduction MeldaProduction https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=176122
- KVRAF
- 14019 posts since 15 Mar, 2008 from Czech republic
The envelope is actually the graph... that's a bit of a chaos here, because originally the graphs were developed for MCompressor's custom shapes, where it is processing envelopes... ugh... I thought all the instances have been removed, apparently not .
Anyways the difference between the input level - ha! Apparently I remember it wrong. Anyways, my current guess , is that there's about infinite possible ways to implement a level detector, and this one is trying to be as generalized as possible using a single graph, which some here call arcane, but it's actually a brutal simplification . Anyways what it probably does is that it takes the difference between current level detector value and the input level on the X axis and transforms it into speed of the detector. Now that's reaaaaaly hard to understand, that's why I asked about trial-and-error approach. Anyways just a hint, since I don't really have a time to write an essay about level followers , it's called follower, because it literally follows the input level. The detector takes the input level (peak, rms...) and updates the internal detector level, which is what is displayed by the vertical line jumping in the processing shape graph. And these attack/release curves basically define how quickly it will react depending on the difference - sort of how far behind it is.
Anyways the difference between the input level - ha! Apparently I remember it wrong. Anyways, my current guess , is that there's about infinite possible ways to implement a level detector, and this one is trying to be as generalized as possible using a single graph, which some here call arcane, but it's actually a brutal simplification . Anyways what it probably does is that it takes the difference between current level detector value and the input level on the X axis and transforms it into speed of the detector. Now that's reaaaaaly hard to understand, that's why I asked about trial-and-error approach. Anyways just a hint, since I don't really have a time to write an essay about level followers , it's called follower, because it literally follows the input level. The detector takes the input level (peak, rms...) and updates the internal detector level, which is what is displayed by the vertical line jumping in the processing shape graph. And these attack/release curves basically define how quickly it will react depending on the difference - sort of how far behind it is.
-
- KVRAF
- Topic Starter
- 10310 posts since 2 Sep, 2003 from Surrey, UK
Thank you No need for an essay.
After reading through the above (twice!) I get it. Or at least I have formed a coherent mental image of how it works. And, believe it or not, it fits in with the theory I had earlier.
I'll have a go at revising the Info, if that's OK.
After reading through the above (twice!) I get it. Or at least I have formed a coherent mental image of how it works. And, believe it or not, it fits in with the theory I had earlier.
I'll have a go at revising the Info, if that's OK.
-
MeldaProduction MeldaProduction https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=176122
- KVRAF
- 14019 posts since 15 Mar, 2008 from Czech republic
Hehe then you are definitely smarter than 99% of the population, because understanding that gibberish I created must be pretty brutal
-
- KVRAF
- Topic Starter
- 10310 posts since 2 Sep, 2003 from Surrey, UK