Cacophony Forums
Unofficial Mackie User Forums => DL1608/DL806/DL32R/ProDX Mixers => Topic started by: Harpman on July 18, 2013, 03:58:56 AM
-
You would think having compression on a vocal channel whould be used to eliminate feedback on a channel rather than cause it. I ran into a scenerio last Saturday where I isolated feedback coming from one of the vocal channels and the only way I was able to quickly eliminate it was to DISABLE compression on the vocal channel. Anyone care to explain that one? ???
-
You would think having compression on a vocal channel would be used to eliminate feedback on a channel rather than cause it.
Nobody I know thinks that - quite the opposite as you demonstrated.
-
I know, it was weird. I did my typical cutting surrounding frequencies in the 5kHz range to eliminate the mid-range FB and didn't do a thing. Disabled the compression and voila no more feedback. Goes against the "laws of sound" :lol:
-
I think you misunderstood what I said - compression DOES increase the likelihood of feedback.
-
+1 Compression does increase the chances of feedback.
That's why there are quite a few people asking Makie to make the aux sends switchable pre/post EQ and Dynamics. Currently they can be switched pre or post the channel fader but not pre or post EQ/Dynamics - I am not sure how they missed this one. (Please vote for it!)
Cheers
-
I just reloaded the saved show (save sometime back) for venue we just played at and realized that I don't have gating or compression enabled for the vocal channels. I must have inadvertently loaded another show or preset to turn compression (heavy at that) back on. The threshold was set for -20dB, 4:1 with soft knee with fast attack/release.
-
Yeah that'd do it :).
I imagine once you start to build up lots of shows and scenes, its going to get tricky managing them.
Cheers
-
My issue tends to be duplicate shows since I use one docked iPad (3rd gen) and one undocked iPad (4th gen).
-
"There you go again" by RR, that's Ronald Reagan not RK. Making uninformed and unsubstantiated statements like "I think you misunderstood what I said - compression DOES increase the likelihood of feedback." is how wife's tales get started and makes us FOH guys look bad. Compression does NOT increase the likelihood of feedback in fact it does the opposite as was intended. Let me give you a simple example in Audio 101. Let say that the feedback threshold for you gig is -7dB FS and your music rms is at -20 db that would leave you 13 dB of crest factor (the difference between the rms and the loudest peak in your music) before you trigger feedback. In my experience and most other Audio Engineers a crest factor of up to 12dB would be typical. There are exceptions mostly in bad music. So with this setup you have 1 dB headroom before triggering feedback. Now lets enable compression and use Harpman's settings of -20 dB threshold and 1:4 ratio. That would enable 13 x 4 = 52 dB more input before you reach feedback threshold. Other things would fall apart long before this would happen. Yes I know that it was only one channel that it would be diluted by the mix but for simplicities sake I'll consider it the main source. What Harpman forgot to mention in this post but did in another is that he had a gain setting of 7 dB in the compressor settings !!! Well an additional 7 dB would send this system screaming without the compressor, 6 dB is a 100% increase in loudness above the feedback threshold. The lesson here is that observation without understanding the basics or checking your settings can lead to the wrong conclusions. RR there is an old proverb "that if you want to be rich you don't hang out with the poor" so let me introduce you to Bob Katz. There is an excellent text by Bob Katz on this and many other aspects of audio in his book "Mastering Audio...". Heavy reading for most but if you pick up 0.001% you'll be light years ahead. Harpman your initial gut feeling was correct and yes the "Laws of Sound" are intact. Your compressor settings are another matter.
-
WK154, thanks for your great explanation and of course you are correct.
Given that this is a forum for the DL1608, not the mixer of choice for most audio professionals, I think RR's statement and my own are not all that dangerous.
Admittedly we should have said 'compressor' instead of 'compression'.
After all it is very common advice for users with feedback problems to suggest they try turning off any compressors.
Cheers
-
You hit the nail on the head with "After all it is very common advice for users with feedback problems to suggest they try turning off any compressors."
It's the common bad advise (other than a quick check for possible bad settings) that's the problem. The word usage "compression, compressor " is fine and "dangerous" I would normally say that came out of left field but in this case it's from down under (pun intended). You'd also be surprised what audio pro's work with.
G'day mate
-
I was with Harpman when he was experiencing his feedback problem recently and thank him for posting the question here. I had discovered that problem for myself with my own DL 1608 and band and suspected his problem was compression on the vocal channel. As a general rule of thumb then, is it advised that compression only be used very sparingly, if at all, on vocal channels? I've got ours completely disabled for the time being and we seem to be doing just fine without it.
-
As a general rule of thumb then, is it advised that compression only be used very sparingly, if at all, on vocal channels?
Yes.
-
The simplified way I've told people that using a compressor can increase the chance for feedback is only when applying makeup gain. Otherwise there is no change in feedback potential using a compressor since there is no change in channel or bus gain. It's pretty simple really. Gain is gain regardless of where it's applied in relation to mic level or speaker level changes. If you have a stable system and decide to turn on a vocal channel or bus compressor but don't add makeup gain, the stability will not change. Add some makeup gain, and things will change.
-
+1 ... and the same holds true for EQ's if you decide to boost a range of frequencies all hell can break loose if you don't compensate for the overall level increase, especially if you're close to your feedback limit.
-
The simplified way I've told people that using a compressor can increase the chance for feedback is only when applying makeup gain. Otherwise there is no change in feedback potential using a compressor since there is no change in channel or bus gain. It's pretty simple really. Gain is gain regardless of where it's applied in relation to mic level or speaker level changes. If you have a stable system and decide to turn on a vocal channel or bus compressor but don't add makeup gain, the stability will not change. Add some makeup gain, and things will change.
Seems like that's what happened in my case Greg. Too much make-up gain and your right, gain is gain no matter how you look at it :)
-
I was with Harpman when he was experiencing his feedback problem recently and thank him for posting the question here. I had discovered that problem for myself with my own DL 1608 and band and suspected his problem was compression on the vocal channel. As a general rule of thumb then, is it advised that compression only be used very sparingly, if at all, on vocal channels? I've got ours completely disabled for the time being and we seem to be doing just fine without it.
Jeff, thanks for chiming in here. When your at a live, your first goal is to isolate where the feedback is occurring (I start muting channels). My next step is cutting frequencies (typically mids around the 5K range), but in my case it wasn't doing anything. This is when Jeff suggested killing the compression on my GF's vocal channel based on what he had experienced. When I went back through my saved shows the next day, that's when I realized I had the wrong show loaded. I name my shows based on band and venue. For example "BE@AR" is Barefoot Express @ Anthony's Ristorante. The issue with using two iPads at gigs it that you still end up with duplicate shows and I wasn't paying attention to the one I loaded. I had 2 BE@AR and the older one had compression on on the vocals and the other didn't. I must have had this issue before and disabled compression on the vocal channels, but who remembers over 6 months ago :).
Thanks all for you feedback here (no pun intended LOL). I tend to gravitate to EQ as being the feedback issue because 90% of the time, it is. When faced with feedback issues, my goal is to isolate and remediate as quick as possible and analyze later.
-
"There you go again" by RR, that's Ronald Reagan not RK. Making uninformed and unsubstantiated statements like "I think you misunderstood what I said - compression DOES increase the likelihood of feedback." is how wife's tales get started and makes us FOH guys look bad. Compression does NOT increase the likelihood of feedback in fact it does the opposite as was intended. Let me give you a simple example in Audio 101. Let say that the feedback threshold for you gig is -7dB FS and your music rms is at -20 db that would leave you 13 dB of crest factor (the difference between the rms and the loudest peak in your music) before you trigger feedback. In my experience and most other Audio Engineers a crest factor of up to 12dB would be typical. There are exceptions mostly in bad music. So with this setup you have 1 dB headroom before triggering feedback. Now lets enable compression and use Harpman's settings of -20 dB threshold and 1:4 ratio. That would enable 13 x 4 = 52 dB more input before you reach feedback threshold. Other things would fall apart long before this would happen. Yes I know that it was only one channel that it would be diluted by the mix but for simplicities sake I'll consider it the main source. What Harpman forgot to mention in this post but did in another is that he had a gain setting of 7 dB in the compressor settings !!! Well an additional 7 dB would send this system screaming without the compressor, 6 dB is a 100% increase in loudness above the feedback threshold. The lesson here is that observation without understanding the basics or checking your settings can lead to the wrong conclusions. RR there is an old proverb "that if you want to be rich you don't hang out with the poor" so let me introduce you to Bob Katz. There is an excellent text by Bob Katz on this and many other aspects of audio in his book "Mastering Audio...". Heavy reading for most but if you pick up 0.001% you'll be light years ahead. Harpman your initial gut feeling was correct and yes the "Laws of Sound" are intact. Your compressor settings are another matter.
WK154, thanks for the support here. What your saying and Greg is spot on with MG. I can totally see that. I'm pretty sure I have that book by Bob Katz. Very good reading!
-
+1 ... and the same holds true for EQ's if you decide to boost a range of frequencies all hell can break loose if you don't compensate for the overall level increase, especially if you're close to your feedback limit.
A lot of EQs have a makeup gain control, but I can't recall any situation where I've used it. And I almost never use an EQ boost on house or monitor graphics, cuts only. The one exception is a local historic theater I do sound design in for plays occasionally. The installed speaker system is a early 70's era Altec home brew box with HF horns that don't have anything above 4k. I've got a linear boost on the house graph starting at 3k gradually increasing on each band ending at +10dB at 20kHz in an attempt to get some sibilance and "air" out of the headset mics. Even that wasn't enough. I had to boost the highs on all the individual mic too. If I did that with any decent speaker system, the feedback with omni headset mics would take off in a jiffy. But not with this system.
-
My own approach even on the old graphic EQ's is to set them in cut only selection if they had it. Never used boost myself other than to experiment. I find that fixing (or trying) mic position or technique is time better spend. The right mic for the job does wonders and changing acoustics in a venue you have no control over is of course impossible. Yes Divas are a PIA at times but some actually come around and learn something new. I use compressors for limiters only for such Murphy events as mics being knocked over (NT3 two month ago) or them being dropped.
-
My usual compression setup when resources are available is to use a combination of channel and bus compression. For vocals, I'll set a very low compression ratio on the channels using something between 1.25:1 - 2:1 & medium attack/release with the threshold set to kick in so there's slight compression during "normal" singing. Then the bus compressor for the vocal groups is set more as limiter with 6:1 ratio with high threshold and fast attack/release. I do a similar thing with guitars and horns as well but the channel compression is usually more around 3:1 and the bus compression too. Keeps things under control without killing dynamics. For kick/snare, I'll run those to a group bus and compress mildly, no more than a few dB of reduction unless the drumming bangs the crap out of the kit occasionally. Toms get their own group compression too in similar fashion to the kick snare group.
-
My usual compression setup when resources are available is to use a combination of channel and bus compression. For vocals, I'll set a very low compression ratio on the channels using something between 1.25:1 - 2:1 & medium attack/release with the threshold set to kick in so there's slight compression during "normal" singing. Then the bus compressor for the vocal groups is set more as limiter with 6:1 ratio with high threshold and fast attack/release. I do a similar thing with guitars and horns as well but the channel compression is usually more around 3:1 and the bus compression too. Keeps things under control without killing dynamics. For kick/snare, I'll run those to a group bus and compress mildly, no more than a few dB of reduction unless the drumming bangs the crap out of the kit occasionally. Toms get their own group compression too in similar fashion to the kick snare group.
Unfortunately, no bus compression on the DL1608. Was that one of the feature requests? Would be nice to have overall compression on the main mix.
-
There are compressors on the main and all the auxes, swipe the geq down and u find it
-
There are compressors on the main and all the auxes, swipe the geq down and u find it
Wow, I've had this unit since November and never knew. Thanks!!
-
There are compressors on the main and all the auxes, swipe the geq down and u find it
Wow, I've had this unit since November and never knew. Thanks!!
Yep sometimes reading a manuel helps lol
-
Yeah, RTFM :lol: So, Greg chimed in with his recommendation for bus compression. Anyone else care to share in regards to bus compression. Noticed the defaults are Hard Knee, 0dB Threshold, 2:1 Ratio, Fast Attack, Release @ 250ms and 0dB Makeup Gain.
-
Basically anytime there is gain reduction you need more gain to get it as loud as it was without the gain reduction. That's fine until the gain reduction goes away when the signal is reduced or stops, then it's feedback city :( .
-
Exactly RR :)
I had a gig tonight that got cancelled, so I decided to spend some positive time on this forum that I enjoy from time to time.
Here's what goes on when more than one vocal or instrument is routed (assigned) to a group with a compressor on that group....
We have Mary singing soprano, Jill singing alto, Tom singing tenor and Jack singing bass.
Even if all four vocals have their own compressor, (ideal) since all four vocals are routed to a group with additional compression we probably never end up with good sound and here's why...
If all the singers were singing sensibly and with reasonable dynamics that's great, but what happens when the soprano Mary hits a note that takes everyone's head off....
Instead of just Mary's vocal being lowered by "Her" compressor, she has hit the threshold of the group's compressor, and all four voices are lowered. Compressors are dumb, they have no clue about high notes or low notes, they are completely unbiased to frequency and therefore can only reduce volume by shear increase of level (voltage). Using some compression on the master fader of the 1608 can help to minimize potential dangerous levels to your speakers. It is always a tradeoff.
When Mary sang that obnoxiously loud note, we did not want her bad singing to effect the volume of the other three voices in the group, but that's what you get when you think it's good to put a compressor on all the vocals, or all the drums, or whatever. It's not usually a good idea. When things like this happen, the unexperienced sound operator has no clue what is really going on.
Compressors are typically used to ensure that a reasonable dynamic range is heard by the listener without having to worry about wigs flying off now and then or straining to hear soft passages.
A compressor (when set properly) can catch a nasty blast of volume far quicker than any sound person could ever anticipate.
A properly set compressor on a vocal minimizes, to a degree, the dynamic range of the vocalist. Mary is all over the place, one minute she's fine then she's whispering, then screaming, and all the while never uses any amount of microphone control.
When we compress Mary's vocal channel, the first thing that most engineers grab is the threshold knob (Mackie's slider), then the ratio, then attack and release times if available, and finally the makeup gain sometimes labeled output. Think of threshold as "threshold of pain". Usually when we lower the threshold knob (counter-clockwise) (Mackie's slider to the left) we are lowering the threshold of pain meaning.... we are not going to tolerate even the slightest abuse of loud screaming from Mary, if we use a very low threshold, virtually everything she sings will be above the threshold, so if we have a reasonable setting for the ratio knob, somewhere between 3:1, and 6:1, the compressor will minimize her volume when she goes nuts on us. Of course if we have a significant amount of compression on Mary, guess what???? now we can't hear her, sooooooo, we use the makeup gain (output) on the compressor to put her volume back up to where we can hear her. So when Mary is singing, we look at the compressor levels. The 1608's compressor has a red level between the input level and output level. The red level works backwards, showing decibels of attenuation. We can clearly see the input level as it came into the compressor, we also see the amount of gain reduction (red) and then to the right of that we see the resulting output level. When Mary is singing a loud passage the compressor will do it's job and take the volume down, but when she sings softer, the compressor releases (lets go) and allows the mic level to be uncompressed, kinda like riding her fader if you knew when she was going to blast off. Since we increased the output (makeup gain) when we were setting the knobs during soundcheck, this is when feedback could rear it's ugly head. YES, too much compression (with makeup gain added) (DOH!) ::) will more than likely cause feedback. In these situations, when she is not singing, and her mic is on her mic stand pointing at the drums, we will get a very ugly sound from her mic picking up the drums that she is standing in front of or was.
Finally, attack and release.
The attack knob is used to adjust exactly what it says "attack" how quick do we want the compressor to begin compressing when Mary hits the threshold with a nasty loud note. Let's say she sings a line that goes like this....Happy birthday to you,... happy birthday to you,... happy BIRTHday dear Johnny,
if we had the attack set to a half second, (500 ms).... BIRTH would blast out uncontrollably, if she held the word BIRRRRRRTH for at least 500 ms. then the compressor would lower her volume.
If we had the compressor's attack time set as fast as it could be, the compressor would have taken care of the loud BIRTH regardless of soft lyrics before the nasty loud note.
Now let's talk about release time.
The release knob adjusts how quickly or slowly the compressor will let go once Mary has blasted a note. If we have the release time set too long like one second, we won't hear the rest of her words for that legato line. If we have the release time set too short, the compressor will let go super fast between words and will cause a nasty sound that engineers refer to as "pumping".
Listen to one of your favorite pop recordings with a vocal, all too often, the compressor on the vocal has a release time that has been set way too fast by the "seasoned mix engineer" and every time the singer takes a breath, the breath is actually louder than the singing. There was so much compression on the vocal that when she stopped singing, the compressor let go, then the singer took a breath, but the breath was not loud enough to meet the threshold, so it wasn't lowered in volume like the singing was,.... I hear this all the time. :facepalm:
For most vocals that are not out of control, I recommend not using any compression, I did't say don't use compression on vocals, it completely depends on the vocalist. Mixing music in a studio is completely different, especially when the vocal track has no other instruments leaking into her mic, and feedback is not an issue since most are wearing headphones. In a studio environment, you can use a compressor to make her sound any way you want... punchy, smooth etc. For vocalists that are not so good with mic control or vocals who weave in and out of the mic some compression will help, this way we don't loose too much as they back away from the mic, and smooth the sound out when they suddenly get too close or loud.
4:1 is a pretty common "Ratio" setting. Watch the display at the top of the window showing exact numbers as you move the sliders for the various settings. You can also click in the display window and enter exact amounts via the iPad's keyboard. The threshold setting is going to completely rely on your input gain setting,.... (gain knobs on the 1608).
I suggest a very fast attack time and a release time of at least 200 milliseconds, this will virtually eliminate any chance of pumping, and we won't have to worry about missing sibilance and diction.
Now on a completely different note 8) compression on bass is commonly used to help increase the sustain especially upright acoustic basses. Since the bass note immediately starts to decay after plucking the string, the compressor would gradually release the compression which would make the bass notes appear to sustain longer, and notes on different strings would sound more consistent. This is where a compressor is used to change the way a particular sound sounds. Good engineers also use compression on snare and kick, as well as toms sometimes to increase punch. This can be accomplished by using a slow attack time then adding the sound of the compressed drum to the uncompressed track or channel. This is sometimes referred to as parallel compression, an old studio trick but one that its not known by many. This can be achieved with the 1608 if you run the kick into an XLR Y cord and feed the kick mic into two different channels, one channel would be named kickcomp and the compression settings would be set very radical, but the volume of this kickcomp channel would be added judiciously when needed for extra punch. Same thing goes for snare, brass, anything that you want to be more aggressive sounding.
Compression, (when set properly) is our best friend
when set poorly, it is our worst enemy.
Final conclusion to this advice which you can take or leave,..... if you don't completely understand all the settings of a compressor, you're better off not using one.
Dan
-
RR: That's fine until the gain reduction goes away when the signal is reduced or stops, then it's feedback city :( .
Am I reading this wrong? Gain reduction goes away by what means, the signal (vocal or instrument)? I interpret this as the loss of sound will cause feedback! The gain reduction in the compressor is a setting you need to change manually and certainly not during performance. Please explain.
-
Like wow, I don't know how to respond to that. Your compressors don't have a gain reduction meter so's you can see how that varies dynamically with the "key" (usually the input) signal's strength?
-
DpDan : Let me say that it's always good to know your talent and their idiosyncrasies but its not always possible. I agree with you on never grouping compression and its best not to even try to get coloration or other effects by using a compressor. In my opinion its a waste of time. In Mary's case use it as a safety net (limiter) and tell her not to quit her day job. Four part harmony needs balance and no FOH engineer I know of could be fast enough to deal with that. That's were mic technique and there own sense of balance separates the men from the boys. I know having dealt with that for too long. It's a pleasure to actually mix for a group that has that under control like the Alley Cats or the Dukes of DooWop to name a few. I'm still chewing on the rest of the tricks you mentioned.
-
^ Yup :)
-
DpDan : Let me say that it's always good to know your talent and their idiosyncrasies but its not always possible. I agree with you on never grouping compression and its best not to even try to get coloration or other effects by using a compressor. In my opinion its a waste of time
In theory it may seem like using group compression for vocals may cause a situation where one vocal "pushes down" the rest. But in practice this doesn't appear to be the case. I leaned the method of compression on vocal channels and groups from someone considered to be one of the best FOH engineers out there and I haven't heard a bad mix at one of his shows yet. There are also cool tricks you can do using VCAs when using bus compression, but that's another discussion.
It's a pleasure to actually mix for a group that has that under control like the Alley Cats or the Dukes of DooWop to name a few. I'm still chewing on the rest of the tricks you mentioned.
Agreed that mixing a band that's truly pro and has good control of their own dynamics is a pleasure. I knew another group called the Alley Cats, but they weren't doowoop. They were punk rock :)
-
OK RR one more time sloooowly: Lets take the whole post.
Basically anytime there is gain reduction you need more gain to get it as loud as it was without the gain reduction.
That's cool and pretty obvious and you have meters in the DL compressor for that labeled IN and OUT and the GR in between is just another Mackie way to let you know that compression is happening (Gain Reduction GR clever!). That's a good thing and gain hasn't changed going thru the compressor (IN and OUT are equal). The assumption of course is that your gain structure is correctly set and your away from the feedback limit by at least 3 dB FS preferably 6 dB. Also that your setup gets no feedback without the compressor engaged under normal levels or at least up to the compressors set threshold limit. Here's the problem statement:
That's fine until the gain reduction goes away when the signal is reduced or stops, then it's feedback city :( .
When the GR indicator goes away you should be at or below the threshold limit and unless your gain structure is incorrectly set there should be no feedback. I hope this clarifies things.
-
GregC; " There are also cool tricks you can do using VCAs when using bus compression, but that's another discussion." That's the CATEGORY I'm trying to convince RR to open up for just such reasons. Real experience and NO Monday morning quarterbacking. Contribute or be gone.
-
Real experience and NO Monday morning quarterbacking. Contribute or be gone.
Me? I've been doing it real world for years.
-
DpDan more observations: The open mic problem with pickup of drums can be addressed with the gate, of course mute on the channel will do the job. Done successfully with automatic mixers all the time and referred to as NOM (number of open mics). This when properly set should eliminate the drum pickup problem and also reduce potential feedback. It can also in Mary's case be a problem since she's all over the place with loudness cutting her in and out when too soft (below the set threshold). Voice lessons from an opera trained instructor and mic technique training would do wonders. A headset mic may help as well. I get the impression reading thru your explanations that the compressor lowers the sound level when activated but it really doesn't it only reduces the increasing level by the ratio set. When you drop below the threshold the compressor is out of the way. Even when nearing the threshold compressor effect is hardly noticeable. A 3 dB difference is considered noticeable by the ear. It also does not provide any kind of delay other than a few hundred usec for processing and certainly not noticeable by the ear. Again 3000 usec + before good hearing can pick up any difference. Hope this clarifies some of the explanations.
-
I think the problem people run into is running the compressor on the microphone where it's in a constant state of gain reduction. Then as stated by other posters the operator starts raising the channel output to get the vocals back in front of the mix or raising the make up gain on the compressor. This can be made even worse if the threshold is set so low that the on stage volume is coupled into the vocal mic contributing to some of the compression. When the band stops and the threshold is no longer reached in effect the gain of the mic has been raised by so many db. It's not a problem with the compressor, it's a problem with the usage.
I typically run my mic compressors where they're kicking about 25-50% of the time, basically on peaks. I don't like the 'squashed' lack of dynamics sound
-
wk,
you're preaching to the choir
-
Depends on the choir.
-
The open mic problem with pickup of drums can be addressed with the gate, of course mute on the channel will do the job. Done successfully with automatic mixers all the time and referred to as NOM (number of open mics). This when properly set should eliminate the drum pickup problem and also reduce potential feedback.
I've found gating of vocal mics to keep out other instruments & stage wash to be pretty ineffective, especially stages with loud drummers. The snare will always trip the gate if it's the loudest thing on stage. Also, even if you can get the gate to behave well enough to only open when the vocalist is on the mic, I find it disconcerting when the gate does open because it also causes a jump in the level of stage wash which then pops off when the gate closes again even if the gate is set for a relatively low level of attenuation. IMHO, it's better to use mics with high off-axis rejection like Audiix OM6s and 7s. If you have vocalists with poor mic technique on a loud stage where they can't use such a mic and have to go with a more forgiving SM58, no amount of processing is going to fix it anyway. It is what it is.
-
Agreed I don't use gates except in very uniform background situations and then rarely. I certainly agree with finding the right mic for the job and stopping the problem at the source instead of band aids down the road. In the above case mute is the most effective.