Impressions here:
Sounds really good, better than free ones (Classic Compressor etc), clear sound, can compete with Sinuwebs and even Golden Compressor soundwise - but it is not that versatile.
To sell this, you will need a good Gui, something as elegant as the Golden Compressor one becaouse the GUI colours sound perception a lot - good luck !
As i understand it, a real sidechain function isnt implemented, is this true ? Perhaps you could add this feature - is i seldom seen in the VST world but useful for some applications.
Richard
NEW! ChunkWare Vanilla Compressor - FREE ALPHA - please try
-
- KVRist
- 181 posts since 22 Aug, 2003 from Olympus Mons
-
Robert Randolph Robert Randolph https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=7328 - KVRAF
- 2226 posts since 25 May, 2003 from Saint Petersburg, Florida
Thats my main problem with it. I reach for a compressor knowing exactly what Im after... when I cant get that even after a few minutes of fiddling it gets deleted and sworn at.mauseoleum wrote:a) threshold/ratio/makeup controls interact in most illogical manner. this should be adressed.
-
- KVRist
- Topic Starter
- 415 posts since 8 Dec, 2003 from NYC (born & raised in Philly)
Hello all. first of all, thank you SO much for all the feedback. i (erm... we?) am so pleased that you have taken the time to give the Vanilla Comp a go.
i have read all of the comments, and i see that there are many features to implement in the future, and perhaps some bugs to address now.
SC filter: i will DEFINITELY make that go down to 50 Hz for you guys. i really hope that is low enough for you. please bear in mind: since the slope is so gentle (Q=0.707), there isn't much stop-band attenuation around the cuttoff. i could make it a little steeper, but that would result in envelope ripple, and IM distortion (i.e. it wouldn't be as smooth). anyways, 50 Hz is no problem.
Auto-makeup gain: i am a little confused as to the irregularities that some have experienced. the threshold and ratio don't interact with each other, but they do influence the auto makeup gain. the formula is simple: the auto makeup is equal to the inverse of the maximum gain reduction of an input at 0dB. basically, the auto gain makes the output peak at 0dB (not compensating for attack/release), regardless of the input.
i think that what people are reacting to is probably the fact that the auto gain does not make up for the ACTUAL gain reduction, but rather the MAX POSSIBLE gain reduction. therefore, the output with auto makeup will usually be a lot louder than the input. (but isn't louder a good thing?)
and i should note: the Vanilla auto makeup functions EXACTLY the same as the Golden Compressor's auto makeup.
also note: the Vanilla Comp is an RMS compressor--it uses short (1ms) RMS averaging of the sidechain signal. (i aplogize for not making this known in the description.) as such, it will react slower to peaks than a true rectified peak compressor. BUT, this is a trade-off for smooth compression, and you can always just "dig in deeper" to get that "squashed" sound. would a "Peak/RMS" switch be a good addition? it is pretty simple to implement.
if that isn't the problem, then it could just be a GUI problem, in which case i am terribly sorry about the whole GUI-less thing. that is my #1 priority after the bugs are sorted out.
Sidechain input: love the idea. unfortunately, it is trickier than it sounds to code it. i will do some research...
Releas times in ms: no problem. you got it!
GUI ideas: keep 'em coming! i will definitely have a gain reduction meter, LED or "virtual LED ladder." i probably won't do a transfer function graph, as i'm saving that for another project.
i'll try to get another build up on the site today or tomorrow. hmmm... well, probably tomorrow, as it's saturday and tonight i might be at the bottom of a bottle.
anyway, thanks again for the feedback. keep it coming, and keep an eye on the site and this thread for updates.
i have read all of the comments, and i see that there are many features to implement in the future, and perhaps some bugs to address now.
SC filter: i will DEFINITELY make that go down to 50 Hz for you guys. i really hope that is low enough for you. please bear in mind: since the slope is so gentle (Q=0.707), there isn't much stop-band attenuation around the cuttoff. i could make it a little steeper, but that would result in envelope ripple, and IM distortion (i.e. it wouldn't be as smooth). anyways, 50 Hz is no problem.
Auto-makeup gain: i am a little confused as to the irregularities that some have experienced. the threshold and ratio don't interact with each other, but they do influence the auto makeup gain. the formula is simple: the auto makeup is equal to the inverse of the maximum gain reduction of an input at 0dB. basically, the auto gain makes the output peak at 0dB (not compensating for attack/release), regardless of the input.
i think that what people are reacting to is probably the fact that the auto gain does not make up for the ACTUAL gain reduction, but rather the MAX POSSIBLE gain reduction. therefore, the output with auto makeup will usually be a lot louder than the input. (but isn't louder a good thing?)
and i should note: the Vanilla auto makeup functions EXACTLY the same as the Golden Compressor's auto makeup.
also note: the Vanilla Comp is an RMS compressor--it uses short (1ms) RMS averaging of the sidechain signal. (i aplogize for not making this known in the description.) as such, it will react slower to peaks than a true rectified peak compressor. BUT, this is a trade-off for smooth compression, and you can always just "dig in deeper" to get that "squashed" sound. would a "Peak/RMS" switch be a good addition? it is pretty simple to implement.
if that isn't the problem, then it could just be a GUI problem, in which case i am terribly sorry about the whole GUI-less thing. that is my #1 priority after the bugs are sorted out.
Sidechain input: love the idea. unfortunately, it is trickier than it sounds to code it. i will do some research...
Releas times in ms: no problem. you got it!
GUI ideas: keep 'em coming! i will definitely have a gain reduction meter, LED or "virtual LED ladder." i probably won't do a transfer function graph, as i'm saving that for another project.
i'll try to get another build up on the site today or tomorrow. hmmm... well, probably tomorrow, as it's saturday and tonight i might be at the bottom of a bottle.
anyway, thanks again for the feedback. keep it coming, and keep an eye on the site and this thread for updates.-
- KVRist
- Topic Starter
- 415 posts since 8 Dec, 2003 from NYC (born & raised in Philly)
one more note: don't forget to play with the Release Shape setting. for those of you craving more "pumping and breathing", try out the pump setting. if you want a smooth, master buss compression, try ARC-2, which is compression dependent.
-
- KVRAF
- 1907 posts since 29 Oct, 2003
man, take your time ...
BUT since you have the sc and key-filters already implemented, why not throw in the L-key / R-signal sidechaining ? now that wouldn't take too long ?
or is it R-key / L-signal ?
phew ...
BUT since you have the sc and key-filters already implemented, why not throw in the L-key / R-signal sidechaining ? now that wouldn't take too long ?
or is it R-key / L-signal ?
phew ...
-
- KVRist
- 120 posts since 7 Sep, 2004
Hi citizenchunk,
I think you didn't understand the problem several people found with the auto gain. When I have a kick sample playing and lowering the threshold brings up the level more than 12 dB than before I did that I would call this a "problem".
Please check it with single drums. Keep up the good work, really like this compressor.
I think you didn't understand the problem several people found with the auto gain. When I have a kick sample playing and lowering the threshold brings up the level more than 12 dB than before I did that I would call this a "problem".
Please check it with single drums. Keep up the good work, really like this compressor.
-
- KVRist
- Topic Starter
- 415 posts since 8 Dec, 2003 from NYC (born & raised in Philly)
ok guys. i just threw up a NEW BUILD, with the following minor changes:
- filter frequency now ranges from 50 Hz to 5 kHz (i hope that's wide enough)
- release is displayed in ms (rather than sec)
i'll think about this L-key/R-signal (or whatever) thingy. it would be a little easier than implementing a real sidechain (with an autonomous sidechain input plugin). but it will take a little "brain time."
sorry about the confusion with the auto makeup gain. i see your problem with it. unfortunately, there's not much that can be done about its behavior. as i said, this is how other auto makeup gains work. (see Kjaerhus Golden Comp for example.) the difference might be in the fact that, since it is an RMS compressor, it might not be grabbing as much as other compressors, which means that the auto makeup has nothing to counteract it.
however, i do want to accomodate. i've been thinking about creating an auto makeup mode that will attempt to match the RMS (average) level of the input. this should be useful for A/B'ing.
or perhaps i should just have it off by default?
enjoy the new build, and keep the feedback comin'!
- filter frequency now ranges from 50 Hz to 5 kHz (i hope that's wide enough)
- release is displayed in ms (rather than sec)
i'll think about this L-key/R-signal (or whatever) thingy. it would be a little easier than implementing a real sidechain (with an autonomous sidechain input plugin). but it will take a little "brain time."
sorry about the confusion with the auto makeup gain. i see your problem with it. unfortunately, there's not much that can be done about its behavior. as i said, this is how other auto makeup gains work. (see Kjaerhus Golden Comp for example.) the difference might be in the fact that, since it is an RMS compressor, it might not be grabbing as much as other compressors, which means that the auto makeup has nothing to counteract it.
however, i do want to accomodate. i've been thinking about creating an auto makeup mode that will attempt to match the RMS (average) level of the input. this should be useful for A/B'ing.
or perhaps i should just have it off by default?
enjoy the new build, and keep the feedback comin'!
-
- KVRAF
- 1907 posts since 29 Oct, 2003
automakeup overshooted original signal (wholemix) for 8.5 dB, (threshold-12dB, softknee8db, ratio 5.25:1, attack5ms, arc2, sc hipass50hz)
now you don't want to scare ppl off with this monster
now you don't want to scare ppl off with this monster
-
- KVRist
- Topic Starter
- 415 posts since 8 Dec, 2003 from NYC (born & raised in Philly)
hmmm.... that seems odd... do you have any other plugs with auto makeup to compare this with? i mean, i tested it against others, and it seems to be working exactly like they do. the "x factor" in all this is the envelope detection, which will affect how quickly it can grab the peaks.mauseoleum wrote:automakeup overshooted original signal (wholemix) for 8.5 dB, (threshold-12dB, softknee8db, ratio 5.25:1, attack5ms, arc2, sc hipass50hz)
now you don't want to scare ppl off with this monster
regardless, i am about to upload a new build, in which the auto makeup defaults to OFF. perhaps that won't scare off first time users.
-
- KVRAF
- 1907 posts since 29 Oct, 2003
look citizenchunk, I don't want to sound like a hyper smartass, but there's got to be more than just turning a *scary* control off.
today I listened again to what this compi does (on a mix only) and it *has* something in it, especially those arc modes. and it seems to be simple, which is a big plus in my book, imo ppl who have to work fast don't look with positive emotions on crowded interfaces.
o.k. - what R.Randolph said may sound like a rant, BUT - hell I think other users won't give a f**k getting into details either if they turn a knob and meters go into red
so - I still think you'll have to figure out how to BALANCE interaction between threshold and ratio controls in automakeup mode - from the user pov.
soundwise, to me, it sounds nice as it is.
< zeitgeist >
" but does it alias ? "
< /zeitgeist >
today I listened again to what this compi does (on a mix only) and it *has* something in it, especially those arc modes. and it seems to be simple, which is a big plus in my book, imo ppl who have to work fast don't look with positive emotions on crowded interfaces.
o.k. - what R.Randolph said may sound like a rant, BUT - hell I think other users won't give a f**k getting into details either if they turn a knob and meters go into red
so - I still think you'll have to figure out how to BALANCE interaction between threshold and ratio controls in automakeup mode - from the user pov.
soundwise, to me, it sounds nice as it is.
< zeitgeist >
" but does it alias ? "
< /zeitgeist >
-
- KVRist
- Topic Starter
- 415 posts since 8 Dec, 2003 from NYC (born & raised in Philly)
i don't really understand, but i hope that this is a good thing.mauseoleum wrote:... today I listened again to what this compi does (on a mix only) and it *has* something in it, especially those arc modes. ...
are you saying that it has its own color? or are you saying that it "has something," like, "you've got moxy, kid!"? i really didn't intend this to be a "colorful" compressor. it definitely has its own sound when it's active, but with all controls at neutral, it is completely passive. (there's no valve modelling going on here. net yet...)
i completely agree. that is why i have tried to minimize the controls, while maintaining as much functionality and versatility as possible. i have a few more features planned for it, but i don't want it to get too bloated.mauseoleum wrote:... and it seems to be simple, which is a big plus in my book, imo ppl who have to work fast don't look with positive emotions on crowded interfaces.
who's this R. Randolph guy? and what has he been saying about me?mauseoleum wrote:... o.k. - what R.Randolph said may sound like a rant, BUT - hell I think other users won't give a f**k getting into details either if they turn a knob and meters go into red
[edit] oh yeah. that guy. i didn't take it as a rant--criticism is just feedback to me right now.thanks again for all the input. i'm glad that you like the sound--that's the best possible compliment it could get. now go, tell a friend! tell them how wonderful it sounds!...
-
- KVRist
- Topic Starter
- 415 posts since 8 Dec, 2003 from NYC (born & raised in Philly)
(quick, before my battery runs out...)
NEW BUILD - build 36, updated 09/26/04. (the file now has the build number.)
http://www.chunkware.com/vanilla/
new stuff:
- new "Detection" mode parameter, switches between peak/RMS detection.
- "Auto Makeup" is now "Max Makeup" to better define its effect
- RMS averaging time lengthened, for smoother envelope, less IM distortion and better bass response.
i hope that the new Peak mode will offer those who want "squash" or "crunch" some more options. if you really want to grab the transients, use peak mode. also, if you intend to use it as a maximizer, i.e. with Max Makeup enabled, you might want peak mode to help prevent overs. but remember: Max Makeup still does not take sidechain averaging into consideration, so watch your meters!
check it out! http://www.chunkware.com/vanilla/
NEW BUILD - build 36, updated 09/26/04. (the file now has the build number.)
http://www.chunkware.com/vanilla/
new stuff:
- new "Detection" mode parameter, switches between peak/RMS detection.
- "Auto Makeup" is now "Max Makeup" to better define its effect
- RMS averaging time lengthened, for smoother envelope, less IM distortion and better bass response.
i hope that the new Peak mode will offer those who want "squash" or "crunch" some more options. if you really want to grab the transients, use peak mode. also, if you intend to use it as a maximizer, i.e. with Max Makeup enabled, you might want peak mode to help prevent overs. but remember: Max Makeup still does not take sidechain averaging into consideration, so watch your meters!
check it out! http://www.chunkware.com/vanilla/
-
- KVRAF
- 11380 posts since 3 Feb, 2003 from Finland, Espoo
Hi,
I'm a compressor junkie and I just gave your compressor a quick spin. Verdict: Me like it. Quite a lot actually. I still can't put my finger on what it's that I like but it seems to be very smooth even at pretty agressive settings. So far I've only tried it on synthesized drumloops (trying some free drum synths at the same time ) but I'll do some more testing during the next few days. Keep up the good work!
Cheers!
bManic
I'm a compressor junkie and I just gave your compressor a quick spin. Verdict: Me like it. Quite a lot actually. I still can't put my finger on what it's that I like but it seems to be very smooth even at pretty agressive settings. So far I've only tried it on synthesized drumloops (trying some free drum synths at the same time
) but I'll do some more testing during the next few days. Keep up the good work!Cheers!
bManic
-
- KVRAF
- 1907 posts since 29 Oct, 2003
again wholemix test only
this rms tweak seems to add more *cohesive* feel to it. sounds quite stable and *holds* the mix nicely in place to my ears. i've managed to *find* a setting which sounds almost identical to unprocessed mix (no coloration or excessive loudness, no flapping) - except it's *more as it should be* or *more perfect*. can't really put this into a written statement. nice work.
this rms tweak seems to add more *cohesive* feel to it. sounds quite stable and *holds* the mix nicely in place to my ears. i've managed to *find* a setting which sounds almost identical to unprocessed mix (no coloration or excessive loudness, no flapping) - except it's *more as it should be* or *more perfect*. can't really put this into a written statement. nice work.
-
- KVRAF
- 11380 posts since 3 Feb, 2003 from Finland, Espoo
Wow! I'm quite impressed with this compressor. Nothing that I've thrown at it has caused any serious problems, in fact, it's stood up very well against the Sonalksis Compressor and Voxengo crunchessor. Your vanilla compressor is smoooooooooth, very smooth and hard to abuse in an un-musical way which is good. I've now squashed some more instruments and mixes.
I don't know how you do it but it behaves extremely nicely at less than 1ms attack time. It's very easy to do some serious low level compression with it, that is, low threshold around -30dB or so with a small ratio around 1:1.2 or less. Also, the peak setting with an agressive 1:8 or more ratio with high threshold around -10dB with fast attack and down to 5ms release causes very rarely bad artifacts (!!!), this is very nice indeed.
Feature request: Wet/Dry controll if possible for sample accurate parallel compression.
Cheers!
bManic
I don't know how you do it but it behaves extremely nicely at less than 1ms attack time. It's very easy to do some serious low level compression with it, that is, low threshold around -30dB or so with a small ratio around 1:1.2 or less. Also, the peak setting with an agressive 1:8 or more ratio with high threshold around -10dB with fast attack and down to 5ms release causes very rarely bad artifacts (!!!), this is very nice indeed.
Feature request: Wet/Dry controll if possible for sample accurate parallel compression.
Cheers!
bManic