Soft Clipper Central Station?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Soft Clippers versus expectations. They're not magic.

Update: Please see Sreten's informative post (#6) as soon as possible.
If I understood that post correctly, the layman's (my) expectations of a soft clipper is that which provides tube amp like softer clipping and headroom benefits; but, unfortunately, that isn't what typical soft clipper circuits do. Actually, except for being quieter hard clippers, real typical soft clipper circuits are not a desirable audio effect.

Therefore, please bear in mind that the goal of this thread is Non-Typical soft clipper circuits that are actually on-demand halfbreed compressors (for attempting some tube amp features and headroom). Thanks AndrewT, for mentioning Bob Cordell's Klever Klipper at post #3. As Mr. Cordell's creative spelling indicates (engineer humor there), Klever Klipper does compress-before-clip, which is prettier than a regular soft clipper.

Goal:
I have hopes that multiple slightly ineffective aka non-clipping, elementary soft clipper circuits can be stacked to do compress-before-clip (a Klever Klip) without op-amps or optocouplers.

See also Esperado's post #30.
More information at post #34.

The soft clipper is faster and costs less than a typical compressor; however there's a treble difference if the two are compared.
Some examples. . .

Here's Elvee's 4N25 Soft clip + Limiter for Circlophone. (a protector, not an audio effect)
What is shown is a protector with soft clip pattern; however, the detector shown on the far right is generally applicable, can use zeners as low as 5.6v and could be re-deployed to drive a compressor (not shown).
259412d1326104273-my-little-cheap-circlophone-circloclipd.gif

.

Here (below) is my diode based ltp soft clip (partially a compressor), inspired by Ken.
The feedback resistor, R17 is unaltered
The feedback-shunt resistor, R16 is replaced with a voltage divider or trimmer
The soft clip (partially a compressor) is connected from in- to in+ of an ltp
The circuit also fluctuates compensations thereby altering harmonics.
A trimmer at R16a, R16b can be dialed in so that compressor occurs while the amp is beginning to clip; but, if you push it harder, soft clip occurs.
280276d1336021613-my-little-cheap-circlophone-softclip.gif

.

Here (below) is my overload stopper. No matter if that is called a hard or soft clip, that protector is quite useless as an audio effect. However, it shows that a voltage detector at input of power amp, doesn't remove headroom (in normal conditions).
An externally hosted image should be here but it was not working when we last tested it.

Reduces the big loud bang of awaking the computer from standby power saver.
Maybe that circuit can be modified in order to answer this question:
Esperado said:
Please, i'm looking for a nice and simple idea to clip an input signal, and to can tune-it just under the clipping point of any amp. I would be very pleased if you have some idea better than my adjustable zeners's ones in the input line.
Here goes with sneaking up on the signal (a prettier effect). The high precision square corners of Zeners may sound like a square wave; so, instead of that distortion, an LED knee voltage is more gentle, and more current reliant, which gives softer corners for a prettier soft clipper (prettier = more like a compressor).
313472d1353670235-soft-clipper-central-station-thumpcushion2.gif

Variable resistor settings, lowest current on left side, higher current on right side.
Diode options:
1N34A, ECG109, SK3090, 1N60, BAT86, BAT85, 1N5711, fast/soft silicon diodes, and LED.
Tested: 1N34A, BAT86, LED's aren't as harsh as regular diodes.
 
Last edited:
I understand why we can use a clipper to protect an amp.
But why to use a "soft"-one ?
If it used to reduce disagreeable distortions effects during high levels's party's, why not to use a limiter, instead, with less distortion when after the limit ? Because of the less complicated and expensive circuit ?
 
I don't have a limiter.
I suspect most listeners don't have a limiter.
A bit difficult to add a limiter and set it up when we don't have any.

I tell a lie, I do have a DCX 24/96 and it has limiters of some type inside.

Could Baker Clamps be termed a soft clipper?

Cordell shows a soft clipper, he called it Klever Klipper.
 
Last edited:
Hi,

A thread that starts with a load of misinformation about "palatable clipping"
and ignores the general principles of the way dynamic range management
is usually done in subs, AV kit and PC speakers is somewhat doomed.

(Also the way its done for film, TV and radio, as well as music mastering).

Assuming circuits are a fix to a problem and describing them in those terms
doesn't cover what they really do and don't do, and thus will mislead a lot.

rgds, sreten.
 
A thread that starts with a load of misinformation about "palatable clipping"...
In a way, i was uncomfortable with this "palatable" as well as "soft". But did we have to discuss about producing an acceptable (intelligibility) movie sound track for both TV (listened at low level in noisy ambiance) and movie theater (with enough dynamic effects) ?
About what kind of peaks we can limit without we can notice the loss of "ease" to increase a subjective level in a CD ? It would take hundred of thread's pages and should be untitled with "limiter or compressor" in the subject.

So i believe the first poster limit its goal to speaker reproduction, and that its attempt is, not touching the gain curve, to reduce the subjective effects of rare clipping when playing a program at the limit of amp's headroom ?
While my own interest is just clipping without signal deterioration under the limit.
 
Last edited:
Hi, A thread that starts with a load of misinformation about "palatable clipping" and ignores the general principles of the way dynamic range management is usually done in subs, AV kit and PC speakers is somewhat doomed.
(Also the way its done for film, TV and radio, as well as music mastering).
Assuming circuits are a fix to a problem and describing them in those terms
doesn't cover what they really do and don't do, and thus will mislead a lot.
rgds, sreten.

Feel free to write a nice short concise intro for what soft clippers do and don't do, and I'll remove my intro and paste your quote at the top of post 1.

Right now I'll go attempt a re-edit to something less picturesque, but if that won't do, simply tell me something that will do. :)
 
. . . to reduce the subjective effects of rare clipping when playing a program at the limit of amp's headroom? While my own interest is just clipping without signal deterioration under the limit.
I had 4 thoughts about this:
A clip (of short duration) happens more often than expected at normal playback volume. References:
-Sound Impairment Monitor (SIM) - Is This The Answer?
-ESP SIM (Sound Impairment Monitor)
I'm trying to say that you don't want to activate a hard clipper several times per minute during normal volume playback.
If signal encounters a hard clipper, the result is always increased noise; however, a soft clipper is less noisy.
A visual indicator could explain more than anything I could type.

". . . without signal deterioration under the limit." probably favors a soft clipper instead of a hard clipper. You can use both simultaneously, if the hard clipper is set to a higher switch-on voltage, to allow the soft clipper room to work. For example a standard overload blocker (hard clipper) installed on an amplifier that is also equipped with Bob Cordell's Klever Klipper (soft clipper).

Esperado said:
. . .why not to use a limiter, instead, with less distortion when after the limit ? Because of the less complicated and expensive circuit ?
A compressor/limiter usually has a capacitive delay and sometimes misses the leading edge of signal but almost always gets the trailing edge just fine. In my experiences, it is nice to use both the compressor/limiter and soft clipper--The compressor/limiter's treble may be prettier but the soft clipper is faster. They can work together nicely.
 
Last edited:
Expired patent:
Patent US4318053 - Amplifier system with automatic distortion control - Google Patents
"charges a capacitor proportionate to the percent clipping of the output of the amplifier"
I'm not sure what you call that circuit, but in awaiting a cap to charge, the device will sound just like a compressor/limiter.

Here's a television compressor/limiter: http://www.amazon.com/Audiovox-Terk-Automatic-Volume-Controller/product-reviews/B00008VWOJ
That one is very aggressive so there's sometimes a pop or pulse when it engages. To decrease the disturbances, one would set the compressor/limiter to be less effective and then the difference between modified versus unmodified signal is decreased.

P.S.
I believe that a soft clipper is faster than either of those, no pop, no pulsing, and that a soft clipper has a treble effect that sounds somewhat like tube/valve amp clipping. So, that's different, isn't it?
 
Last edited:
I believe that a soft clipper is faster
There is no question of speed, here.
A compressor or limiter can be faster than anything else, because it can even reduce level before the peak happens. Using DSPs and delays.
A compressor change the average gain of an amplifier, it reduce the volume, like you could do it with a volume control. Its action is supposed to be linear from 0 to max of a signal, reducing the overall gain factor.. It does not clip any signal. You have access to the detection speed, the speed of recovery, the level threshold and the compression ratio. The limiter is just a compressor with a ratio > 10/1.

A clip is just an electric limitation of the max voltage of an electric signal signal. Due to rail voltage limitations, in an amp. A soft one is one witch has a transition curve near this point.
 
I dont see how a (perfect) clipper (hard or soft) can add noise to the signal. Noise is a non harmonic signal, random. A clipping add distortion. A soft clipper just reduce upper harmonics from it. jut just smooth the corners of a square wave.
Please pretend that I said distortion. :)
Apparently, I was using the word noise in the most common way according to Wikipedia: "In relation to sound, noise is not necessarily random. Sounds, particularly loud ones, that disturb people or make it difficult to hear wanted sounds, are noise. "

The article goes on to explain that it has different meaning to engineers, specifically random or hiss: Noise - Wikipedia, the free encyclopedia
I did not know that engineers used this word differently than common usage.
 
The compression ratio of a compressor determine the non linear gain factor of the action. You can set-it, for example, for each 6db increase of a signal will produce a 3db gain at the output. It has to be clear: it is the gain of the amp during all the duration of this signal (at least one sinusoid).

You can set delays for different feels. With a fast attack and a long recovery, you can reduce the volume of any loud signals.Volume decrease at the attack and stay at a lower level. After the average high level material of the signal has stopped, the gain will slowly increase.
On the contrary, using slow attack and fast recovery, you can let the peaks, and reduce volume after. Adding attack to a bass that way is a common practice in mixing processes.
 
Last edited:
Used in audio, noise is an added random signal, not in correlation with it.
Can be produced by different factors, like random movements of atoms in transistors or resistances due to thermal effects, random repartition of the elementary magnets on an magnetic tape, random roughness of the vinyl groove.
Distortion is a deformation of the signal itself.
When no signal, noise can continue, while no distortion can occur.
In certain circumstances, noise can be modulated by the signal, but, again, no harmonic correlation with the signal itself.
 
Last edited:
There is no question of speed, here.
A compressor or limiter can be faster than anything else, because it can even reduce level before the peak happens. Using DSPs and delays.
A compressor change the average gain of an amplifier, it reduce the volume, like you could do it with a volume control. Its action is supposed to be linear from 0 to max of a signal, reducing the overall gain factor.. It does not clip any signal. You have access to the detection speed, the speed of recovery, the level threshold and the compression ratio. The limiter is just a compressor with a ratio > 10/1.
Very true! More modern compressors do have DSP's that can match the timing up so there's no audible pulsing.

By your description, what you're wanting is a DSP based Line Level PA/Prosound type compressor. With those, its not necessary to clip anything in order to activate a detector. Instead, you program the level and the DSP regulates the output of this line level device. In that line level scheme, you don't have to reduce the output power of your amplifier by early engagement of compressor.
A clip is just an electric limitation of the max voltage of an electric signal signal. Due to rail voltage limitations, in an amp. A soft one is one witch has a transition curve near this point.
That's a really good definition. Thanks!!
 
About the Pipipedia " "In relation to sound, noise is not necessarily random. Sounds, particularly loud ones, that disturb people or make it difficult to hear wanted sounds, are noise. " i suppose their refer to the content of the audio material, not the quality of its reproduction.
It is obvious that any unwanted background behind a dialog, by example, like car ambiance, is called noise. Not at all the same thing, as a good amp with -110db of signal/noise ratio will reproduce perfectly this noisy recording.
 
About what i'm looking for, is a clipper able to clip a signal in a identical way that following amp can do.
Without adding noise or distortion in the non clipped signal.
The reason is that i use a comparator in a protection circuit, between the input and the output signal, to detect any DC, oscillation, cutoff, even noise or unwanted distortion added by the amp, that would reveal its dysfunction.

The problem is that amp clipping introduce such a difference between input and output, while no dysfunction of it.
So the idea was to clip the input signal, just before the max peak level of the amp. I will compare the two, clipped input signal with amplified not clipped signal of it at the output, whatever the input level.
Of course, because it is supposed to be a protection, even in high end amps, it is requisite that this clipping circuit will not deteriorate the quality of the signal in any way.
Such a perfect circuit will just remove little percent of the peak power capability of the amp.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.