Warm Audio Bus Comp - ratio backwards?

famous beagle

famous beagle

Well-known member
Hey y'all,

So I got this unit, and I'm wondering if anyone else has one and can chime in on this. I'm really loving the sound of it, but the ratio knob seems to be backwards on my unit!

If I set everything up so that I'm getting about 2 dB of GR on the 1.5 ratio setting (1.5:1), you would think that turning up the ratio knob to 10:1 (and touching nothing else) would result in much more GR and, therefore, a decrease in volume ... yes?

However, on my unit, it's exactly the opposite. I get the most GR (on the VU meter) and the quietest signal at 1.5 and the least GR (on the VU meter) and the loudest signal at 10!

I don't have a video of it yet, but I found one in which the same thing happens as mine!

Warm Audio Bus Comp 2 Channel VCA Bus Compressor Review - YouTube

If you watch at around 1:10, he moves from the ratio from either 1.5 or 2 (can't see exactly) up to 10, and you can see the VU meter showing less GR, and you can hear the sound get louder --- both exactly the opposite of what you'd expect to happen when you increase the ratio.

Right? Am I taking crazy pills?

There's no auto makeup gain on this unit (as far as I know), and he doesn't adjust the makeup gain at all.

So ... what the heck gives?

I sent an email to Warm about this, but I think I'll have to wait until after the weekend to hear back. So I thought I'd check on here to see if anyone could tell me something I've missed.

Thanks!
 
As you change the ratio, you are also changing the attack, release and knee settings.

In this case, the higher the ratio, the harder the knee. So that means that on the higher ratios, it is just knocking off the transient peaks. On lower ratios, it is riding the levels.
 
Farview said:
As you change the ratio, you are also changing the attack, release and knee settings.

In this case, the higher the ratio, the harder the knee. So that means that on the higher ratios, it is just knocking off the transient peaks. On lower ratios, it is riding the levels.
Click to expand...

Ok, thanks for the explanation, but I'm still not 100% clear on how a 1.5:1 ratio can mean more GR than 10:1 when the threshold hasn't changed.

I thought I was going crazy, so I loaded up six different EQ plugs on a channel and the did the same experiment. I got a few dB of GR going with a lower ratio and then only moved the ratio control without touching anything else, and in every one of them, the GR increased as I did it (which is what I would expect).
 
With a softer knee, the compression actually starts happening gradually at a threshold lower than the one you set. The threshold point you set is the point where it is actually compressing at 1.5:1. As the signal approaches the threshold, it is getting compressed at a gradually increasing ratio until it hits the threshold.

With a hard knee, there is no compression until the signal hits the threshold.
 
Thanks. I'd known about knee before, but I guess I didn't realize how drastic of an effect it can have. I guess I'd never paid close enough attention to it before and A/B'ed so many different units.

I'm still curious why the six different other compressors I loaded up didn't act this way. I guess they all had either a fixed knee or one that you can adjust?

The Bus Comp does not have a knee setting.
 
In trying to research about the Bus Comp's knee, I found this, which is about the SSL Bus Compressor (the unit the WA Bus Comp is supposedly cloning):

====================
KNEE
In order to fully grasp how this compressor works, we need to discuss the concept of internal settings, which regulate how quickly compression actually takes to occur.

We call this the “knee”.

On a graph, this would look like an actual knee, hence the fairly obvious term. “Hard knee” compression is immediate upon crossing the threshold. By contrast, “Soft knee” is a more gradual ramp up.

Internal settings adjust the SSL Native Bus Compressor, so it ends up having a softer knee the harder you hit it.

Even a ratio of 10:1 doesn’t end up completely squashing your mix.

Mixes retain their dynamics, as opposed to something that sound “flattened” and on the verge of “brick wall limiting.”
====================

This kind of confused me more, because it seemed to contradict what you were saying (that the knee got harder with a higher ratio).

Then I found this on a forum, where someone was basically echoing my concerns. And someone explained it as follows:

=====================
"As others noted this is normal and to be expected from an SSL comp. The threshold and knee response changes with the ratios. An 1176 does this as well. It actually makes sense, if you think about it. With a 2:1 ratio, you aim at glueing the whole mix together slightly, right? While with 10:1 your goal is to reign in large spikes. So with a low ratio you're ideally compressing all the time and with a higher ratio you're ideally compressing mostly peaks."
=====================

So I guess it does all have to do with under-the-hood changes as you said, but he said that the threshold is affected as well. This makes a bit more sense to me, because that was the one thing I couldn't wrap my head around.

Anyway, thanks for the help!
 
I think you can kinda look at like...at lower ratios, it's a smooth, gradual compression ramp-up, so the gain reduction is obvious early...and at the high ratios, it becomes more of a limiter, as the knee becomes steeper, so it takes more signal to reach it before it grabs the top of the signal, and in that case the GR is smaller.

How do you like it overall?
I guess it's based on the SSL bus comp.

I picked up the Audioscape buss comp a few months back, which is also in that same camp, and I think it's about the same price as the Warm...though it's still sitting in the box because my new studio project has just dragged out, and I may not get a chance to really hear it for awhile...but the Audioscape was getting real good reviews so I rolled the dice.

BUSS COMPRESSOR — Vintage Pro Audio Equipment | AudioScape Engineering Co.

The Warm Audio unit has been getting pretty good reviews too.
 
miroslav said:
How do you like it overall?
I guess it's based on the SSL bus comp.

I picked up the Audioscape buss comp a few months back, which is also in that same camp
Click to expand...

I really like it from what I've heard so far, but I haven't used it very much yet. I acquired a bunch of new gear over the past several months, and so I just finished rewiring my patch bays. I'll be getting plenty of use out of it in the next few months for sure.

I am having one issue with it, though, which apparently has to do with my patch bay(s). It has this "engage transformers" button, which passes the signal through another opamp stage with CineMag transformers. It's basically just an extra tonal option. When I engage that button on my unit, I get a bit level jump --- about + 6 dB of gain. Someone from Warm told me it sounded like a grounding issue in my patch bay.

I took it out of the bay and tested it in a different room, and sure enough, the problem went away. So it's definitely something to do with the bay.

I saw another comment on YouTube, though --- actually in the video I posted above --- that addressed this same issue. The same thing happens in the video with the transformers button. The comment said that he'd had the same issue, and Warm told him it was likely because he was mixing balanced and unbalanced cables. I'm definitely doing that in my patch bays --- I'm using snakes that I had, and none of them are balanced --- but:

A) Some of the gear I have is unbalanced only, and
B) This is the first (and only thus far) anomaly I've ever experienced with my patch bays in this way.

So ... I don't even know if replacing all my cables with balanced ones would make a difference (since some of my gear isn't balanced). And I don't think it's worth it to take on all that extra expense --- i.e., buying 4 or 5 balanced 8-channel snakes --- in the hopes of possibly correcting an issue that I might not use that often (like I said, it's a tonal option) and that I could simply deal with by attenuating the level by 6 dB.

I don't make any long cable runs in my studio at all. And whenever I do, I'm always using balanced cables (running mic cables to a different room or something like that).

If the balanced cable issue didn't solve it, then I'm guessing it's most likely a grounding issue. And ... holy crap, I've never been more confused than for the 30 minutes I was researching grounding with patch bays!

So ... at this point, I'm tempted to let sleeping dogs lie.
 
famous beagle said:
...Warm told him it was likely because he was mixing balanced and unbalanced cables. I'm definitely doing that in my patch bays --- I'm using snakes that I had, and none of them are balanced --- but:

A) Some of the gear I have is unbalanced only, and
B) This is the first (and only thus far) anomaly I've ever experienced with my patch bays in this way.
Click to expand...

It's not a grounding issue as in AC electrical grounding issue...it's a signal grounding issue that you are creating with the unbalanced snakes.

Mmm...is the patch bay balanced capable...is it a TRS bay?

It's easier/better to add unbalanced gear into a balanced patchbay setup...then adding balanced gearing into an unbalanced setup.
With the unbalanced cables...you're basically shorting the ring to the sleeve...the Pin2/Cold to ground...of your balanced gear.
It's a good way to end up with all kind of noise issues, not just the +6 level thing...especially when you then patch in more unbalanced and balanced gear into the same signal chain.

I had at one time a good amount of unbalanced gear along with balanced...but when I wired my TRS patchbays, I wired them all for balanced use, and for the unbalanced gear, I simply took both signal wires in the snake cable going to the unbalanced, and wired them together to the Tip/Hot....so unbalanced gear only saw a Tip/Sleeve connection...and the balanced saw the TRS. When connecting a balanced piece to an unbalanced...there was no shorting of the balanced Ring to Sleeve...and the Tip signal from the unbalanced simply went to the Tip of the balanced, because the patch cables were TRS...so the Ring/Sleeve remain separate...with simply no signal going to the Ring of the balanced gear, and the unbalanced gear connects to balanced and vice-versa without issue.

So a TRS bay, with TRS snakes and TRS patch cables will make it easier to have a mix of balanced and unbalanced gear...without running into issues.

You can run with it as you have it...but it's a sleeping dog that has many teeth. ;)

At this point, I have maybe 3 pieces of older, unbalanced gear that I still use...and I've gone pretty much all balanced.
 
I have one balanced patchbay (Behringer PX3000) and two unbalanced ones (a Tascam and a Hosa).

So, regarding this:
but when I wired my TRS patchbays, I wired them all for balanced use, and for the unbalanced gear, I simply took both signal wires in the snake cable going to the unbalanced, and wired them together to the Tip/Hot....so unbalanced gear only saw a Tip/Sleeve connection...and the balanced saw the TRS

To make sure I understand correctly ... this is what I would do:

1. Get an 8-channel (let's say) TRS snake.
2. Leave all the ends that are going to the balanced patchbay alone.
3. On the other end, for any cables that are going to unbalanced gear, I would do this:
A. Unscrew the plug to get access to the cable connections.
B. Unsolder the ring cable connection and solder it to the tip connection.
C. Seal it back up and plug this new cable into the unbalanced gear.

Is that right? Or would I need to make this modification at both ends of the snake?

Assuming the above is correct (and please correct me if it's not), then if I replace my two unbalanced PBs with balanced ones and then use all balanced cables --- taking into account the above-mentioned caveat for interfacing with balanced gear --- I shouldn't have any issues?
 
The snake is wired unbalanced at both the bay and the gear end for the unbalanced gear.
So at the balanced TRS bay, you would wire both signal leads to the Tip/Hot, and the shield/ground to the Sleeve connections and nothing for the Ring...and at the gear end, you would wire the two leads to the 1/4" Tip and the shield/ground wire to the sleeve.

The balanced stuff would be just straight forward...TRS at the bay to TRS or Pin 2/Pin 3/Pin1 (Hot/Cold/Ground) respectively.
Then you just use TRS patch cables for interconnects between the balanced and unbalanced gear.

When going from balanced to balanced or unbalanced to unbalanced...it's all straightforward....when you patch a balanced to unbalanced...the TRS patch cable will see the expected TRS connection at the bay, and on the unbalanced side, the TRS cable will simply not see anything for the Ring, only the Tip and Sleeve, because both signal leads for unbalanced were wired to the Tip...and yes, that technically unbalances you balanced gear, but you don't get the Ring shorting to ground like you have now with the unbalanced cables.
Do you need to wire both leads to the Tip for the unbalanced...?...No, you could simply cut one lead off, or whatever, but I find that in case you ever decide to put a balanced piece of gear on those bay points down the road in place of the unbalanced gear...you'll already have both signal leads ready to go, if you cut one off, or don't strip it and just tuck it back or something, you'll have to do it then, which might be more harder at the bay end, getting in there, stripping back, etc...so I just wire both to the Tip...no other benefit, and certainly nothing negative in doing it. :)

Will that solve all your problems...? :D
Well...I can't swear to it...but that's how I've wired my bays and gear, and I do have a bunch of rack gear, and TBH, I've had almost no issues ever. No loops, no hum or buzz...nothing. I think there was one piece of gear that I had to individually sort out for some hum...and all it took was grounding its case to the next piece of gear's case in the rack.
It's hard to guarantee everything, because you never know how they wired things for each piece of gear...what did they do with power ground vs audio ground...etc...but generally speaking, you should avoid the most issues, and I would certainly tailor my snakes and bays for balanced use, especially if you're expanding your gear, since most of it will be balanced these days, or maybe they offer both...but there's not too much being made 100% unbalanced for audio use (not getting into home stereo systems and all that).

All that said...you will always have to deal with the difference in levels across the board, because most balanced is +4 and most unbalanced is -10....but that's a normal thing, unlike getting that +6 thing because your running unbalanced cable with balanced gear and shorting the Cold to ground...which is what I think your problem was/is.

There's also many ways to wire balanced gear...floating one end...sometimes connecting Pin1 to the sleeve, sometimes not...etc...etc.
Rane has all kinds of tech info on all that...which is where I did most of my reading back in the day before doing some experimentation, and then coming up with specific protocol that I always adhered to when wiring. Mind you, I also have a decent amount of AC power conditioning and I also did my best to make sure all power/ground was running back to a single point source...etc...but those things are to prevent other issues...not what you were experiencing.

Frankly...some of it does feel like voodoo...and you do occasionally have to perform an exorcism to get rid of some weird hum/noise issue. :p ;)
 
Great, thanks very much for the info; I really appreciate it. I suppose I may have been lucky thus far, because I've never experienced any hum/buzz or anything until now.

I suppose I'll give this a try. I've already ordered another balanced patch bay, so I'll just need one more. And I just ordered the first of several 8-channel TRS snakes.

And I just finished completely rewiring everything, too. Ha!

Oh .. I just thought of something interesting ...

When I was testing my Bus Comp in another room, this is what I did:

1. Ran an 1/8" stereo-to-dual RCA cable from my phone, which was playing music. I then used two RCA-to-1/4" adapters to get into the 1/4" inputs on the Bus Comp.

2. I ran two balanced patch cables from the 1/4" outputs on the Bus Comp to two balanced inputs on a small Behringer mixer.

3. I plugged phones into the Berhi mixer and monitored that way.

As I mentioned before, I did not have the increased gain transformer issue when I did this. However, I was clearly mixing unbalanced (signal coming from phone) and balanced (signal coming out of Bus Comp) signals here. Was it because of the way I was doing this --- i.e., an unbalanced input --- that prevented me from having the issue?
 
It's hard to say how the mixer treated all the connections...but I think it was mostly at the bay where things were messed up.
Sometimes things can be hooked up "wrong"...and it all works right...but in general, I think it's best to adopt a more accepted protocol, and then just stick to it, rather than rolling the dice with every hook-up...that way you don't have to wonder where in the chain might be the problem. Keeping things the same through out will make it easier to find any problems when they do pop-up.
 
miroslav said:
Keeping things the same through out will make it easier to find any problems when they do pop-up.
Click to expand...

This definitely rings true for me, and it's one of the main reasons I'd like to go ahead and give it a shot. It's all been pretty haphazard up until now, which does not make for easy trouble-shooting.
 
I have a lot to learn about compressors in audio, though I have used three or four different brands with different characteristics. Two compressors were built into mixers and sported what they called "single-knob compression" with not a lot of advice on how to set them. My DBX compressor has knobs for threshold, ratio, attack and release. I believe one of the buttons controls something about the knee. As you might suspect from my email name, four or five knobs gives me more with which I can play. So I just "keep on knobbin'" until I get the sound I want for the track I am recording at the moment, though I am also open to learning more about compressors in general and their best usage. I imagine that the "single knob compressor" is a big compromise to make things somewhat easier for a novice; and I'll readily admit to recording at least one track on which I got carried away with the knobs and recorded a track with pumping and other problems making it "sound like hell!" In the instructions with one mixer with the single knob compression, I got the impression that it was weighted somehow toward vocal miking, and the mixers I have do not have the single-knob compression on the line-input channels - only on the mike inputs, giving me a bit more credence to the idea that those controls benefit vocals but are not necessarily limited to vocal miking. Also, based on my experience with "bad knobbin'" don't let the simplicity of the single knob lull you into the thought that you can set it on any value and everything will be fine. I've been there and done that! The single knob is a "tool" in an audio toolbox - and one which I am still learning how to use to its best effect.
 
This is the actual answer to the balanced<>unbalanced interconnect question: RANE Commercial - Knowledge Base
Unfortunately, it’s not necessarily simple, depends on the specifics of all of the gear involved, and actually indicates a different solution depending on whether that button is pressed or not. Sure, it’s user-unfriendly design on the part of WA, but we suffer for fashion. Or are expected to be able to call ourselves “engineers” with a straight face. ;)

What’s worse though is that even that table won’t always give us the best answers for a given situation. Maybe we don’t know as much as we’d like about the gear itself. Maybe it was built on full moon Friday the 13th. In those case you just try different options until the problem goes away or you learn to live with it.

But there’s really the three options, short the R to the S, don’t, or pseudo-balanced.


Edit - Honestly, I think the problem here is not so much that the transformer is too loud as that the non-transformer output is quieter than it should be. That is, the way you’re unbalancing the signal is wrong for the transformerless output.

The transformer only has two wires coming out and the only reference either has is to the other. The shield is literally not part of that circuit, so “shorting” those wires has no effect. At the other end, we’re still essentially reading the difference between the two wires and everything is fine. You might lift the shield at one end if there seems to be a grounding issue. Theoretically, lifting the R connection should leave the transformer path open and therefore silent, but IRL the results will be difficult to predict. Might be quieter, different frequency response, more noise....

The active output is actually two different circuits, each with two wires coming out, but one wire from each goes to a common point. Each signal wire is referred to the shield, and the shield is the “bottom wire” of each output. In this case, shorting R to S actually shorts the “bottom circuit” altogether. Technically that could cause damage, but protection is easy. More importantly, it means that bottom signal is not available at the other end to add to the difference between the two wires of the input. In this case, just ignoring (disconnecting) the R is less dangerous than shorting, but won’t be any louder. Usually, you can force this type of output to “float” the way the transformer does by disconnecting the ground connection and using the negative output as the reference at the other end. That is, connect R to S, but disconnect the actual shield in the cable at one end. This should match your levels closer by bringing the transformerless output up. Bonus that it should also work fine for the transformer.
 
Last edited:
ashcat_lt said:
This is the actual answer to the balanced<>unbalanced interconnect question: RANE Commercial - Knowledge Base


But there’s really the three options, short the R to the S, don’t, or pseudo-balanced.
Click to expand...

Mmm...I'm not seeing where shorting Ring to Sleeve is a recommended option...except in a single specific situation.

I guess the question is...are you talking about controlling hum/ground issues...or to solve the +6 dB issue with the Warm Audio gear?
Also, I would think that if using some of the options that unbalance a balanced piece of gear because you are connecting it to unbalanced gear, which then causes some issue with the balanced gear...then don't do it. :)

I've got a Joe Meek SC 2.2 stereo comp...and one time I simply tried to use it for a situation with unbalanced connections...and it totally whacked out the stereo output.
Thought I had an issue...called...they said, nope, it's designed to only work with 2-channel stereo, balanced connections...and what it was doing when tried something different, was basically normal.
So I didn't do that anymore. ;)

I still prefer to follow Rane's "best case" wiring...and use full balanced, though AFA connecting the Pin 1/Sleeve to chassis ground...I don't opt for trying that UNLESS there is an issue, because like they say, many manufacturers adopted the practice of doing that internally....and that was my point in earlier posts...that sometimes you don't know how they did signal ground and chassis ground in their design....so I start with the fully balanced cables, and pretty much using 95% balanced gear, and then any individual issues that come up, I deal with them. I also have a couple of boxes that are used to interface bal & unbal, and also to bump up/down -10/+4 connections when needed.
 
Last edited:
Oh god...I recall spending many days reading through a lot of Rane's tech notes about cable wiring, proper grounding...etc...etc...back in the late '80s/early '90s...and sometimes feeling like there was no obvious answer.
It can be quite overwhelming, and you're left with a LOT of options that you then have to decide which to use for your situation or issue...and in most cases that information can only be gleaned from trial-n-error using the clearly outlined options Rane has provided...one by one, starting with the best case, and kinda working your way down from there.

I still say always start with the best/proper way they outline...don't make assumptions that some of the other ways may work better for your situation until you have a reason to.
 
You must log in or register to reply here.
Back
Top