Reply to thread

Message

<blockquote data-quote="Rich Frembes" data-source="post: 68525" data-attributes="member: 432">Re: Limiting Within Line ArraysI couldn&#039;t agree with you more Ivan. Measure a few different limiters and you are likely to be horrified.&nbsp; I think limiter definitions vary even more then parametric filters and their ilk do.&nbsp; It&#039;s especially frustrating as a loudspeaker manufacturer, as the second most popular request we receive from customers is &quot;can you recommend limiters for processor/amp X that will protect my system?&quot;&nbsp; The short answer is nearly always &quot;perhaps&quot;.The process to measure and match parametric filters is fairly straightforward.&nbsp; It is a lot of work to support processor du jour, but it&#039;s doable.&nbsp; Measuring and accurately characterizing every limiter topology on the market is another matter altogether.&nbsp; Food for thought: I surveyed approximately twenty processors last year while preparing for an AES presentation. There are a vast assortment of ways to describe each limiter parameter. To wit:Threshold: dBu, dBFS, VU, volts, watts, Off/-3/-6/12 dB&nbsp;&nbsp; Note that dBu and dBFS are often simply labeled &quot;dB&quot;, so you have to decipher the GUI or look at a help file to learn what they are.

Attack:&nbsp; Milliseconds, Seconds, ms/dB, ms/20 dB, Slow/Medium/Fast

Hold:&nbsp; Milliseconds

Release:&nbsp; Milliseconds, Seconds, dB/seconds, ms/dB, ms/20 dB, Slow/Medium/Fast, multiplier of Attack Time, e.g. 4x or 16x
There are also &quot;bonus&quot; parameters found on some limiters:Ratio:&nbsp; Values typically 1.2:1 to INF:1

Over Easy:&nbsp; Integer values from 0 to 10

Overshoot:&nbsp; Values typically 2 dB to 6 dB

PeakStop:&nbsp; Values either On/Off or range 0 dB to 6 dB

Soft Knee:&nbsp; Values typically On/Off

Knee:&nbsp; Soft, Medium, Hard

Corner:&nbsp; Values typically 0 dB to -100 dB
&quot;rms&quot; and &quot;peak&quot; limiter detectors (the detector circuit, NOT the numeric values found in the GUI) are not strictly defined.&nbsp; For instance, a &quot;peak&quot; detector may be just that, or it may in fact be an rms detector with 3 dB added.&nbsp; If we as an industry cannot come to an agreement on this most basic of limiter functions, how on Earth can we agree what all the other parameters listed above mean?In short, as Ivan said you have to MEASURE.&nbsp; And you have to measure A LOT of parameters, many of which are interdependent. I don&#039;t see how we can realistically talk about how different limiters &quot;sound&quot; when we generally don&#039;t even know how they work!</blockquote>

[QUOTE="Rich Frembes, post: 68525, member: 432"] Re: Limiting Within Line Arrays I couldn't agree with you more Ivan. Measure a few different limiters and you are likely to be horrified. I think limiter definitions vary even more then parametric filters and their ilk do. It's especially frustrating as a loudspeaker manufacturer, as the second most popular request we receive from customers is "can you recommend limiters for processor/amp X that will protect my system?" The short answer is nearly always "perhaps". The process to measure and match parametric filters is fairly straightforward. It is a lot of work to support processor du jour, but it's doable. Measuring and accurately characterizing every limiter topology on the market is another matter altogether. Food for thought: I surveyed approximately twenty processors last year while preparing for an AES presentation. There are a vast assortment of ways to describe each limiter parameter. To wit:[INDENT][B]Threshold: [/B]dBu, dBFS, VU, volts, watts, Off/-3/-6/12 dB Note that [I]dBu [/I]and [I]dBFS[/I] are often simply labeled "dB", so you have to decipher the GUI or look at a help file to learn what they are. [B]Attack: [/B]Milliseconds, Seconds, ms/dB, ms/20 dB, Slow/Medium/Fast [B]Hold: [/B]Milliseconds [B]Release: [/B]Milliseconds, Seconds, dB/seconds, ms/dB, ms/20 dB, Slow/Medium/Fast, multiplier of Attack Time, e.g. 4x or 16x [/INDENT] There are also "bonus" parameters found on [I]some [/I]limiters:[INDENT][B]Ratio: [/B]Values typically 1.2:1 to INF:1 [B]Over Easy: [/B]Integer values from 0 to 10 [B]Overshoot: [/B]Values typically 2 dB to 6 dB [B]PeakStop: [/B]Values [I]either [/I]On/Off or range 0 dB to 6 dB [B]Soft Knee: [/B]Values typically On/Off [B]Knee: [/B]Soft, Medium, Hard [B]Corner: [/B]Values typically 0 dB to -100 dB [/INDENT] "rms" and "peak" limiter detectors (the detector circuit, NOT the numeric values found in the GUI) are not strictly defined. For instance, a "peak" detector may be just that, or it may in fact be an rms detector with 3 dB added. If we as an industry cannot come to an agreement on this most basic of limiter functions, how on Earth can we agree what all the other parameters listed above mean? In short, as Ivan said you have to MEASURE. And you have to measure A LOT of parameters, many of which are interdependent. I don't see how we can realistically talk about how different limiters "sound" when we generally don't even know how they work! [/QUOTE]

Verification