Presswerk patch/emulation area + inits

[3ee]

OK, here's what I have so far: https://app.box.com/s/qf8jtkidi5yqjafn1q6e

I really needed to bring in some release adapt since the release curves seem different and with about that setting it feels to recover just about right. It may surely need a bit more of testing and fine-tuning...

Threshold is set to 0 so that it catches one's eye that it needs adjustment.

added: btw, I left the analog option off.. since it tends to have a very significant character that derails from the emulation (maybe in the future we'll have several "analog" flavors to choose from or something)

[3ee]

btw, just for reference, I tested Presswerk with the VST plug analyzer the other day found out that if the knee is @ 25% then it's about 5db transition, @ 50% it's 10 db and so on... so a max value of 20 db soft knee is possible with the max value.

[woodsdenis]

https://www.dropbox.com/s/lpcc81ff2kgii ... k.mp3?dl=0


Hope that works, A quick drum loop 2bars bypassed and the 2 engaged. Great stuff 3ee. Listen to how it just catches the loose snare and tightens it up

Sorry doesn't seem to play but you can download , any hints on how to get this to play in the browser ? 320 mp3

[woodsdenis]

Double post

[3ee]

^^ Well, thanks to OneRoomStudios also for pointing out some useful tips! Yeah, in your example, the drum loop receives some of that extra attack..

Well, will test the 160 patch tomorrow for some more finetuning, in the mean time, waiting for a confirmation if it's in the ballpark or what?

moving on, the Teletronix LA2A is next...

[OneRoomStudios]

That preset does sound nice, but it sounds much "grabbier" than any 160 VU I've worked with. I think part of the issue is the attack time, and part is the negative delay in the side chain.

I think 7ms is far too fast for the attack. If you check the original manual (http://mixonline.com/online_extras/dbx_160.pdf), on page 14, you'll see that the dynamic attack time is about 15ms for changes of 10dB above the threshold, and speeds up as the signal pushes further past the threshold. For general compression, the signal will almost always be within 10dB of the threshold - and the attack time will almost always be 15ms. You could make a second initial preset for "160 VU Limiting" or something like that with a faster attack time, a slower release, and more distortion, but for compression, 15ms is far more realistic.

The negative delay in the side chain makes the plugin behave in a way that the original unit never could. In fact, theoretically there should be a minuscule positive delay - the time it takes for electrons to pass through the unit. In practice though, this is probably less than a hundredth of a millisecond.

Here's my stab at it. It's not any better, just a different take.

https://drive.google.com/file/d/0B4-39s ... sp=sharing

[3ee]

^ Sounding great! Will test it tomorrow again.. in the mean time, what RMS window should I use for the LA-2A? I have it currently set to 15ms but I'm guessing maximum again? FB mode, right? I have about 8 db of softknee and non-linear slightly to the left. Big release amount about 400 ms with about 64% release adapt etc, again RMS value is something I have no idea where to begin for a fairly precise start.

About the sidechain delay. Theoretically you seem to be right, but in the digital, things already need to have a delay/latency to work so yeah, I guess it's best to leave the sidechain delay as close to 0 for now.
The reason I tend to give it negative values is: 1st of all it seems to eliminate that clicky/un-natural attack associated with digital compressors (especially when using at extreme values) and 2nd; for compensation regarding the attack shape, so in practice I assume it can help achieve more pleasant/natural sound in the digital. If not, the saturation will help get more natural results, I'm guessing that NI specifically added high amounts of saturation precisely for that reason = to sound good in digital.

Try an experiment and tell me what you think: try to get a slightly artificial sounding attack while slamming, then try to get rid of it by setting the sidech delay to negative a bit and compensate with the attack value.

[3ee]

here's the LA-2A in the works: https://app.box.com/s/lhtjw7geqxeftqrsufyt (updated1)
Still need to make it more laggy and less pumpy and shape the attack.

[aaron aardvark]

ORS,
I haven't tried your DBX160 patch yet, but what would be a good application for that? My primary interest in compressors are vocals first, drums second (usually toontrack stuff). I own a dbx 163x compressor, though it's seen better days, I don't use it anymore. Just using the compressor that came with Cubase LE4 presently. I have used compressors for about 20 years, but I have a lot to learn. I have barely tried Presswerks so far, don't really have the hang of it yet.

[OneRoomStudios]

3ee:
The LA-2A is a pure peak limiter, so I would have the RMS window at nearly zero. Yes, it is FB. The real trick with the LA-2A is emulating the non-linearity in the T4 cell. I'd have to sit down with the transfer curves and try to mimic them in Presswerk, which I won't have time to do until later this evening. Getting the saturation right is going to be tricky too (I wish you could do pre and post).

It would probably be a lot easier for you if you stuck to the same family of compressors for now - since you've got the DBX 160 VU working well, try some other VCA plugins like the SSL buss compressor, SSL G-series channel compressor, API 2500, etc. That way you can start in the right ballpark and tweak. Then move on to FET, then Opto, then vari-mu, then zener, then PWM, etc.

aaron:
I tried to add a little of that into the preset description. Generally, the DBX 160 VU is great on drum buss, electric guitar, and other percussive sources. It can be great on staccato vocals too (I know a successful hip-hop engineer that uses them on vocals all the time). Anything where you want to exaggerate the transients and bring out the low-level detail.

Compression really is more of an art than a science, especially since people seem to use it at least as often as an effect as they do as a dynamics tool. It's a big topic, but maybe as we go through different presets we can discuss what they would be good for.

One last thing - if you use my DBX 160 VU preset, you should only be adjusting the threshold, ratio, and output (that's all the original had).

[3ee]

^ I will update my LA-2A patch with peak response asap and put it aside to focus on the SSL bus compressor. (RMS value for this? )

About extra saturation for the LA-2A... I've noticed that the soft clipper gives a high amount of saturation/distortion. So, we can maybe use the saturation in pre mode and engage the soft clipper so it would saturate more if pushed for the "post" mode.

[3ee]

Updated LA-2A patch same link: https://app.box.com/s/lhtjw7geqxeftqrsufyt

[OneRoomStudios]

I'll check out that preset when I'm back in front of my system.

In the mean time, here are some things to think about for the SSL buss compressor.

You should probably go with the SSL 4000 G series "Quad Compressor." Some people like the 9000, but for the "classic" SSL sound, the 4000 G is the way to go.

The real trick with the Quad Compressor is that the knee changes significantly as you increase the ratio. I would suggest making two presets - one for "glue" with a very soft knee and a 1.5:1 ratio, and one for "smash" with a hard knee and an 8:1 or 10:1 ratio. I'll add more details later.

[3ee]

According to this thread: https://www.gearslutz.com/board/so-much ... rward.html

...the SSL G4000 bus comp is a hybrid between FF and FB ... so I guess it would probably be a good idea to set it to 'INT' in Presswerk

added: OK, good idea to make 2 patches for it.

[woodsdenis]

This thread really should be a sticky. Will try the la2 later. Starting points that mimic existing compressor designs is really helpful.