Updated Chris's Dynamic Compressor

http://pdf23ds.net/software/dynamic-compressor/

A new version of my dynamic compressor, 1.2.6, contains a new parameter, hardness, that replaces four others that were, apparently, hard to understand. It gives you the ability to get a more traditional, hard sound out of my compressor.

What did you take out?

Koz

The new GUI certainly looks less intimidating.

Using the default settings, this new version appears to produce more compression (the result is “louder”) than the previous version.

When tested on simple generated tones, the raised parts show noticeable ripple in amplitude (which did not happen with the older version) as illustrated here (new compressor at top).
tracks003.png
I’ve not looked at the code, but I assume that you have linked the 4 attack/decay width and exponent controls into the one “hardness” control. I think it would be useful to keep the “attack” and “decay” controls separate (possibly just labelling them “attack/decay time”) as it is quite often useful to have separate control over attack and decay (for example, when compressing drum samples it is common for the attack to be very much faster than the decay - more so than with other types of sound).

sounds like the old version was better
shouldn’t the user be able to set the loudness due to compression

i vote for more parameters never fewer
even if people are confused at first
they can just use the defaults
and twiddle what they understand
until they learn more and can take full advantage

To avoid any confusion, that is not what I am saying.

It is more usual to use something before offering feedback about it.
The amount of compression can be set using the “Compress Ratio” control.

That’s what I’m afraid of.

Is there a “Legacy” download for the old one?

Koz

OK, this is serious. I really need the public link to the old version. I can’t recommend the new one before trying it and I don’t have continuous time in the next couple of weeks to do that. The new one doesn’t sound the same, so I don’t know if I’m solving the right problems or not.

Thanks,

Koz

Sorry about that.
I was not replying to you.

I went by his message.
Where it says the sound is now “harder”.
Sorry if harder is not louder in your dictionary.
Many English speaking folks would take it that way.

is it just my eyes
or an artifact of my pc display
but i see a periodic spike in the new compression results
wider in the low volume sample and narrower in the high one

No problem.

There was an earlier discussion on the forum about the previous version of Chris’s Dynamic Compressor in which Chris pointed out that more aggressive (harder compression setting were possible with the effect.

The default settings in both the old and new versions are very gentle and produce a subtle, near transparent, compression of the dynamics. With the old version, it was largely not understood by users how to achieve more aggressive settings (other than increasing the compression ratio). Chris has addressed this issue by changing the interface and making it much easier to make the affect more aggressive/harder. The “hardness” setting appears to adjust the attack/decay parameters which were previously 4 “cryptic” parameters, but are now combined into one “hardness” parameter.

No it’s not your eyes - there was a periodic ripple. This was using pretty extreme settings, but the thing that is bugging me now is that I can’t reproduce the problem.
Also, the default settings only appear to be very marginally louder than the old version and the attack/decay slightly softer.

With the old version, users would generally only use the “Compress Ratio” setting and leave everything else at the defaults. With the new version you can do the same and it will produce very similar results.

If you want to produce the same effect as the old version on the default settings, increase the “hardness” of the new version to 0.52 (the default in the new version is a little bit “softer”.

Some settings that were favoured by many with the old version was to increase the Compress Ratio to 0.7
To achieve the same effect with the new version:
Increase the Compress Ratio to 0.7 and increase the Hardness to 0.76
The result will be virtually identical (the old version will be about 0.3 dB lower)

Thanks

I think you mean by harder then that the compression ratio is higher
but wouldn’t that also translate to louder – at least to some extent?
If not, then hardness is not at all clear. Unless it means a sharp knee instead of an easy over gradual knee?

I am not sure that I would have called it ripple, but maybe for digital signals that would qualify for the use of that term.

I agree that I would still prefer the old version.
Hide the other parameters until somebody clicks on them and until they do use the default values. Best of both worlds. Those who don’t understand can use the simple approach, those who want to tweak stuff can click through and diddle all the settings.

Anybody know of an APL to nyquist translator? This is getting interesting but I am not sure that I want to take time to learn nyquist.

No, the “Compress Ratio” slider does that.

The “Hardness” slider is a combined control for the attack width, attack exponent, release width and release exponent.
The reason that Chris has combined them into one “Hardness” slider is that no-one understood what he meant by “attack width, attack exponent, release width and release exponent”.

Basically, if you move the “hardness” slider to the left, you get a slower and more progressive (softer) attack and release. Move it to the right and you get a faster and more abrupt (harder) attack and release.

Personally, now that I’ve tried it a bit, I like the combination of the width and exponent controls. Chris seems to have engineered a nice balance between width and exponent. However I would (personally) prefer to have two controls, one for attack and another for release.

You can’t do that with Nyquist plug-ins. Settings are either visible or not visible. There is no mechanism in the interface for changing what is visible.

It is very similar to Lisp, or more precisely, to XLisp.
It’s a pretty easy language, though Chris has done some fairly trick stuff in this plug-in.
http://audacityteam.org/help/nyquist

Thanks

maybe folks who don’t want to learn what the parameters are should be using some other plug in for compression.

<<<maybe folks who don’t want to learn what the parameters are should be using some other plug in for compression.>>>

Maybe not. At this writing, Chris is the only compressor that can do what it does. Make the show louder and even out the volume variations with almost complete transparency. You can’t tell what it’s doing, it sounds terrific, and it just works.

Try and get that effect with any of the single element compression tools. When Chris was introduced, even with small problems here and there, he mopped the floor with the other tools. As I wrote then, “You have no idea what you have here.”

Koz

Maybe it’s time for a fork… lets have two versions of the plugin… call one Chris Compressor for Dummies and the other Chris Compressor for Nerds, or alike :slight_smile:

Update:
Regarding the “issue” of “ripple in amplitude” that I mentioned at the beginning of this topic.
I’ve still not been able to reproduce the problem and have no idea why it occured, so unless anyone else notices similar effects I think it is safe to assume that it was a freak occurrence and probably caused by some issue outside of the dynamic compressor plug-in. It’s just very strange that it occured at all and was 100% repeatable on my first day of testing.

Chris has clarified this question in his blog:

look in the source of the plugin for instructions on how to re-enable the advanced features. (You just comment out some lines and uncomment others.) The tutorial mentions this, so hopefully new users will be able to find out about them. I would definitely make it a button to switch, but I have very little control over the parameters screen. I basically just set the parameter name and ranges, and I can’t do anything else.

The hardness adjusts the two width settings in sync, on a x^2 scale. The exponents just stay at the defaults.

You can get to any old version by the following scheme:

http://pdf23ds.net/projects/compressor/compress-1.X.Y.ny > (e.g. compress-1.2.5.ny).

BTW, 1.2.6 doesn’t change the compressor algorithm. Results with the default settings should be identical.

I can confirm that when the settings are equivalent, the results in 1.2.5 and 1.2.6 do indeed appear to be identical. In the old version, I think people were dissuaded from experimenting with settings other than the compress ratio because of the complexity of the interface. Perhaps with the simplified GUI people will be encouraged to experiment more and thus get more out of it.