Voxengo needs a freeware compressor.

[Kingston]

the MjCompressor

Imho it has similar prob as blockfish, namely it doesn't like the -first- transient well. Not so good for drums etc.

depending on compressor topologies this is quite the universal problem. You know the tricks they used to do back in the day? They would add a hold control in the compressor, let the band play their loudest note before recording, and press hold. Then they would finally release the hold *just* before the band begins to play. Voila! No initial transient overshoot!

It's a tad easier these days as one can insert a dummy note right when the silence ends, and edit it out later on.

[bool]

:::: They would add a hold control in the compressor, let the band play their loudest note before recording, and press hold.

I know that. I'm not so young bassist.
Other approach is to limit/kill overshoots (peakstop) to get -more present- signal (not presence), more rms/peak, and compress and fiddle later. Oldschool but when going directly to hd (I know that was used with 16bit hd recorders), doing it with a limiter or drive conv. gets similar result. Digital may be less -audible-. But I wouldn't do it with a audio card convos. Imo this is stuff for guys who know their hardware well enough and know what is good with their setup. Out of my league and price range as well, but I've been around.

[brok landers]

alexey ... after this thread you _must_ know, why a lot of users need a limiter/saturator in the gainstage after the compressor ...
i really would like to have a statement on this ...
what do you think?

Actually, I have "add saturator as standard option to most plug-ins" line in my "to do" list. However, what makes me think twice is the necessity to add 2 knobs and 1 on/off switch to support this feature. This looks like a bit of a burden to me. What do you think?

alexey, thanks for seriously considering!
well, i would in_any_ case implement an on/off switch.
that way you are
1. compatible to previous versions (default state off)
2. if one still doesn't want to use it (there also might be no use for this in particular cases), he can freely choose to not use it.

also you should think about the saturation itself.
it should (imo) have some important properties:
- brick wall (no 0db passing)
- when saturating, one should not detect the saturation as a distorted sound (important to check the balance of odd/even harmonics) ...
as i described before i often have the problem that all plugins, that are labeled as "saturator"
actually _audibly_ sound like a distortion way too early ... i explained the behave of the gainstage on an analog console. there i can crank up the signal up to 6db over 0db, the signal transients get shaped, but it doesn't sound distorted an this stage ... the transient (i.e. from a 909bd) more shifted into another, richer frequency spectrum, ore at least the harmonics that are added, are only affecting the transients (i.e. the click of the 909bd), giving it a slightly richer bandwith "around" the original frequency of the "click".
so the whole signal is louder, the click is more "defined", but the body "sounds" mostly unaffected by the saturation ...
i played around with a multiband compressor where one could shift the phase of one band on a certain frequency while compressing ... that also reminded me to this behaviour i meant ... as by doing this you actually cancel out some harmonics, and boost others ... i just don't know of _which_ frequency must be affected ... but by saturating a signal (fi i'm correct, not sure on this) you actually are able to shift certain frequencies phase a bit, and on top the sat algo can be designed to only add a defined set of harmonics ...
imo _that_ (these two things, phase of certain frequencies and the combination of the harmonic sets) makes the pleasant saturation, which actually is not felt like distotion ...
i know, i went a bit on ice with this post, as it is very esotheric, but i don't know how to describe it in a better way, as my english again f**ks me ...

however, one thing i'd add in addition:
a final bipolar levelslider (in0,1db steps adjustable).
the reason for this is simple, but i'll explain it in depth:
imagin you have a snare.
it needs more snappyness, bit more punch.
it is allready leveled perfectly to 0db.
what i do is the following:
i comress the snare, leaving the attack open a bit, so that the important transients that create that snappyness slip through the compressor completely unaffected. after i while of trying out i found the perfect setting.
but the snare is now less loud in rms, as what i did was simply putting an level-envelope on the snare, which loweres the"body" (the part after the transients) of the signal.
now the snare sounds the way i want, but it still hit's 0db, i cannot adjust it any louder without peaking.
so what i do is, i switch on the brickwall-limiter/saturator.
then i raise the output of the compressor, pushing the compressed signal against the brickwall.
i know that i kill a certain ammout of transients again, but there's still enough snappyness, and now i'm equal loud or even louder (rms level, of course i'm not mathematiclly louder in peak, as i still don't overshoot) than the uncompressed signal.
now, if i'm even louder than the unprocessed signal, i can get fooled by the loudness, thinking it sounds better.
now, to avoid fooling my ears, i simply use the levelslider at the very end of the signal chain (_after_ the limiter/saturation stage) and turn it down to the point i have the same _audible_ level compared to the unprocessed signal.
this way i am actually hearing what the compressor does to the snare _besides_ making it louder.
i can now hear how the compressor is actually working.
after checking this in depth, i boost the signal up to 0db again with is level-slider, if needed.
now i have the complete fader range of the mixer, and no matter where i place it, i don't have to worry about the signal again, the relation of the transients to the body stays the same, and if i raise the fader up max it will not overshoot, so i am save finally.

i hope you understand what i mean, and why this is so necessary in digital ...
analog i don't really need this way of working with a limiter, i just raise the compressors output, driving the channel for the saturation.
but in digital 0db is the border when reaching the converter finally, no matter how many db i have downwards. thsi is why it is so necessary to keep track of the transients.
that is, if you actually like to have snappy sounds, otherwise you just close the attack, if you want it not snappy, no need to worry about a too less loud mix then...

however, if i was unclear at any point, just let me know, i try to explain it different then ...
if my english allowes it ....

btw, i'm sure that it will sound perfect if you implement it, as all your saturation stuff actually sounds_ very_ good ... its just a different type of saturation i am after ...
oh and of course, you are aware of most of my described, i'm sure ...
so, i don't wanna come off like a teacher or so ...
i just want to make sure that you actually know why _i_ find this so important, so i described my way of working, and what the desired result should be ...

again, thanks for listening!

[brok landers]

just as a sidenote:
mda dynamics is a very good freeware compressor

love it, absoluteley ...
i don't know how how often i wrote paul an email, praising him for that ...
he must think i'm gay or so ...

[brok landers]

There is a much better freeware compressor IMHO, the MjCompressor from smartelectronix too (by Magnus, nyquist fame).

It has no custom GUI nor gain reduction meters (which make it a bit tricky as it can be EXTREMELY transparent) but work with it a while and I guarante you fantastic results.

It's on my "best digital compressors top 10" list and this is counting all the hardware I have too!

Cheers!
bManic

yeah, thatone is also stunning ... it really adds the colour when you need it, and it can handle a lot of different type of signals ...
also, whe tweaked a bit harder, it reminds me to my old and beloved valley-people compressor ...
i should've not sell it ...
yeah ... now i know better ..
sorry, back on topic again ...

[brok landers]

but also don't forget the free VBaudio StripTool V1 - incredibly low on the cpu and very good sounding (but not many controls)

amen to that ... i wish i could afford the big one, as the free one is too limited to my needs ...

[brok landers]

but also don't forget the free VBaudio StripTool V1 - incredibly low on the cpu and very good sounding (but not many controls)

Yeah, that thing is great - really good top end on the EQ too. But the worst controls ever; up is down, left is right, etc. Could really use an update.

you can chage that behaviour to your likes ...
just right clock on the knob ... uncheck inverse mousemode. voila...

edit. sorry, someone already mentioned this ...
damn, you fast typers ...

[friteuse]

Hi brok,

just to help you out a bit

I understand your problem very well. I think, the thing you're describing with your snare example is easier to explain when looking at peak/rms ratio.

The original snare has a certain peak/rms ratio. By shaping it with the compressor in the way you mentioned, rms level decreases drastically, but the peak level keeps constant. So, your aim is it to reach or even top the original rms level and, in consequence, the original peak/rms ratio. This must be done with a limiter or a saturator after the compressor, within the track, and has nothing to do with the master fader or mastering itself (like others mentioned in this thread).

It's more the way to limit dynamics (ratios!) of each track to get a reasonable overall dynamic, which is easier to master (as the master compressor hasn't too much work with it). It has nothing to do with 'I wanna be louder', but with the fact, that to reach a *reasonable* loudness. You just do not have to quench the master compressor that much, what leads to a more open, transparent and pleasing sound. (Note, we are not talking about classic or acoustic jazz here!).

In analog world, this was 'autmatically' done by the console, as I saw in tests with an ADT console (just summing leaded to a quiet denser mix). In digital domain, you have to care about this yourself.

But the problem in digital domain is, that there are few good plugins that can handle this situation well. Either they shoot over, or they sound distorted, muddy or else. I recently buyed an analog compressor (a vintage radio broadcast one) that opened my eyes. It has a built in limiter, that perfectly prevents overshoots without distorting or influencing the sound in a negative way (in a certain range, of course). I didn't reach this behaviour with any of my plugins yet.

So, for an intuitive production workflow, it would be very great to have a good compressor with a good gain stage. imho, at least...

[brok landers]

thanks friteuse!
that is exactly what i wanted to describe ...
i just wanted to mention things that influence the desired result, besides what you explained, such as the balance of odd/even harmonics, and the behave of these in a dynamic way, etc ...
just to make sure that alexey doesn't just implement a simple brickwall (though i don't really think he does just that, simply because it wouldn't equal to his philosophy) ...
alexey, what do you say?
i'm very interessted in your opinion ...