How to prevent squashed snare ???

[kingliar]

Hi all,

Maybe some of you know this problem : snare sounds great in the mix!, but at the final stage during mastering (finalizer), the snare is getting squashed and it sounds like there's a "hole" in the mid-range of the song (like there's no energy/power in the mid-range)

My question is : how to get a loud master (-9dBfs) without squashing the snare?

I'm probably not the first to ask this question and I know it has everthing to do with compressing/limiting. But how to do it ?

Does the mix has to be "fixed" or can it be done during mastering?

[Gravity8058]

The way to keep the snare popping involves running the mix through a large number of high quality analog outboard devices. Think about it. What do all great mastering houses use? They carefully hit each unit just right to soak the transformers and tubes just right, this enables them to use very little compression and limiting.

Unfortunatley, there is no cheap way to acheive this result other than sending your stuff to a good mastering place. Good luck

[Andy Krehm]

Originally Posted by Gravity8058

The way to keep the snare popping involves running the mix through a large number of high quality analog outboard devices. Think about it. What do all great mastering houses use? They carefully hit each unit just right to soak the transformers and tubes just right, this enables them to use very little compression and limiting.

Unfortunatley, there is no cheap way to acheive this result other than sending your stuff to a good mastering place. Good luck

Amen, brother! That's the best and most simple way to describe the "elusive" technique that everyone covets.

I might also add high quality digital outboard gear as most good mastering engineers use digital units in conjuction with their analog gear, such as Weiss, TC 6000, etc.

Let's keep your post handy and copy and paste it into the many queries about this and similar subjects (why do I lose my kick?) as well as the many "how to make my album loud without losing all the dynamics" posts!

As always, having the gear doesn't guarantee results but when you have an experienced engineer with great gear and an excellent studio, the odds are that it will be a great job.

[brethes]

Originally Posted by kingliar

Hi all,

Maybe some of you know this problem : snare sounds great in the mix!, but at the final stage during mastering (finalizer), the snare is getting squashed and it sounds like there's a "hole" in the mid-range of the song (like there's no energy/power in the mid-range)

My question is : how to get a loud master (-9dBfs) without squashing the snare?

I'm probably not the first to ask this question and I know it has everthing to do with compressing/limiting. But how to do it ?

Does the mix has to be "fixed" or can it be done during mastering?

If you do your own mastering (using the Finalizer), adjust the mix to taste...
If you use a good mastering engineer, ask him to do a test with your mix and ask his advice if the results are still not what you expect. Unfortunately, in order to achieve very high levels, the trade off usually happens with the snare and you may have to increase the levels of the snare in your mix even with the best mastering engineer unless your mix is top notch with good separation and pristine dynamics in the first place. The "hole in the mids" is a different problem and probably a sign that you should get someone experienced (and hopefully using something better than a Finalizer!) to do the mastering.

[Musiclab]

First off listen to records, even old records have a certain amount of the transients shaved off the snare, so a certain amount of that is a "record" sound. They way to alleviate losing your snare is to use 2 buss compression in the mix, and start the mix with the compressor on and mix into the compressor

[lucey]

Originally Posted by kingliar

My question is : how to get a loud master (-9dBfs) without squashing the snare?

That should be doable. Is your mix too compressed? Are your EQ balances off? Is your center channel clearly defined, from low to high? Mix it well and send it to a pro. Most people will listen to your mix and make specific comments as part of their service.

Sell that Finalizer while there is still a market for them.

[MattGray]

Originally Posted by Musiclab

First off listen to records, even old records have a certain amount of the transients shaved off the snare, so a certain amount of that is a "record" sound. They way to alleviate losing your snare is to use 2 buss compression in the mix, and start the mix with the compressor on and mix into the compressor

I'd have to disagree with this advice, in my opinion it's a lazy approach to mixing & what you end up doing is over compressing the final mix because you keep pushing things up to attempt getting it through the compressor. This now gives you a louder more squashed snare (& mix) with no life or transients left in it. By the time it gets to the mastering engineer it's too late, the life has already been sucked right out of it. I've seen this all too many times. It's best to get your mix sounding just right without buss compression or limiting & then let a reputable mastering engineer get the right balance & levels using the best processors designed for this job - transparency. Now if you want to use a certain buss compressor for artistic reasons such as it's colour or sound then this is different. But if you want to keep your snares don't use traditional buss compression!

I would say a large portion of getting the snare right is finding the right tone to begin with. Set your mics up so that you are getting the body of the snare drum sound not just hot levels with a whimpy thin sound. You want to be able to hear the snare sustain with enough attack on the front to give you cut on the front end & enough body on the back end so it sits in the mix when everything else is in there. Short plate reverbs with the right pre delay can help the body sustain a little more if it's used in moderation. The best way is to use a combination of close mics & room mics. Don't EQ it in the same range as the vocals or guitars & don't be afraid to leave some low mid meat in there.

As for the mastering side of things, limiters can do more harm then good when it comes to drum transients. To get it just right you need to use high quality processors & don't rely on only one to get the final level. Using a few stages incrementally to lift the overall level is key in getting a more transparent result.

One of the tools I use to get me part the way there in mastering is tape compression (the most transient friendly compressor around!) & one of the reasons why drums sound so good through tape. Tape compresses the transients in a musical way & brings up the body of the drum sounds without squashing the life out of them, something that digital processors have trouble emulating. There is very few compressors or tape emulation boxes that can do this anywhere near as well as the real thing.

Disclaimer... Now there are tape machines & then there are mastering quality tape machines, they all sound different as does the various tape formulas & widths. The machines need to be serviced, calibrated & aligned properly in order to sound their best. Sure it's a lot of money & effort but the sound is very rewarding...

This is one reason why you pay a good amount of money for quality mastering, it's not cheap to buy & maintain the best gear & how do you put a price on experience?

Matt

[lucey]

agreed, many records dont have any 2 mix compression, unless you count the analog tape

[kingliar]

Thanks a lot for your info/help......but...... if I understand it correctly :

What you're saying is; that it cannot be done in the digital domain !???

[Bump Music]

My trick: Trigger a very tight and thick snare underneath your snare. Very snappy. Then monitor with a Peak limiter from time to time. The unlimited mix will sound off balance, i.e the snare will pop out too much... but once it hits the clipping or limiter, it will pop nicely.

Careful though, this can backfire big time if used too much or incorrectly.

P.S. I've heard MANY records mixed by very "in demand/popular" mixers and mastered by "in demand/popular" mastering engineers and they STILL have this problem. So it's not as simple as getting the right people and right gear.

[MattGray]

Originally Posted by kingliar

Thanks a lot for your info/help......but...... if I understand it correctly :

What you're saying is; that it cannot be done in the digital domain !???

Getting it to pop in the mix is the easy part, getting it to pop at -9RMS using just a finalizer is damn near impossible. If any one could do it easily with a 'finalizer' then why would anyone pay for a mastering engineer?

To answer your question, yes it's possible to achieve the result you want all in the digital domain, but you still need great digital tools & know how to increase the levels a bit at a time through a number of processes & this is usually unique for every mix. So there is no single remedy or technique that will achieve your result unfortunately.

Matt

[Brad Blackwood]

Originally Posted by kingliar

Thanks a lot for your info/help......but...... if I understand it correctly :

What you're saying is; that it cannot be done in the digital domain !???

I think a better answer is that it shouldn't happen at all. Different mixes require different methods but there's no reason the snare should be radically changed.

This question is about the driver, not the car.

[Musiclab]

Originally Posted by MattGray

I'd have to disagree with this advice, in my opinion it's a lazy approach to mixing & what you end up doing is over compressing the final mix because you keep pushing things up to attempt getting it through the compressor. This now gives you a louder more squashed snare (& mix) with no life or transients left in it. By the time it gets to the mastering engineer it's too late, the life has already been sucked right out of it. I've seen this all too many times. It's best to get your mix sounding just right without buss compression or limiting & then let a reputable mastering engineer get the right balance & levels using the best processors designed for this job - transparency. Now if you want to use a certain buss compressor for artistic reasons such as it's colour or sound then this is different. But if you want to keep your snares don't use traditional buss compression!

Matt

Respectfully I disagree completely. When a mix is not compressed and it goes to mastering, your mix proportions can change , reverb levels etc. When compression is applied later the transients will change depending on how much compression the ME
will have to apply. Using a quality 2 buss compressor properly can help alleviate that. The reason why mixing into the 2 buss compressor is important is so that you're hearing from the start what the mix buss compressor is doing and you're reacting to it as a component of the mix. You'll see what's happening to the drums and verbs and the dynamics of the mix. Lazy approach? hardly since you should now be looking
for whats really needed to make the mix dynamic. I do agree that using analog tape does really nice things to just about everything,which is why I use my 2" deck every day. And using a finalizer
is not in any way shape or form, mastering, what that box is doing to your audio when you're hitting it real hard is not nice.

[Methlab]

Most guys i know, before they send off to Sterling, or a place like that, follow a rule. It's snare up, guitar down. This is to compensate for the mastering crush. So make sure your snare is a little louder in the mix then you want before you squash your mix.

[lucey]

The usual thing with a mix that's lacking in snare after mastering is that it's lacking in snare before mastering ... and has often been compressed to death already. So from that standpoint (a mix that's too compressed already) then sure, turn up the snare to compensate for mastering. But that's really the backasswards way to go.

Better to make a dense, powerful mix that still has some life to it. The best loud masters come from mixes with some transients left in.

[Pixel25]

Originally Posted by MattGray

One of the tools I use to get me part the way there in mastering is tape compression (the most transient friendly compressor around!)

I dont' agree totally with that statement. Tape has it's benefits in mastering...sometimes. It's defeinitely not something i would default to on every single session, as some engineers do (mostly to try and impress clients) I used to use an ME that bounced everything i gave him to tape (on a really well maintained ATR machine.)
On so many cases it smeared the drum transients much more than a traditional compressor. In other cases it worked just fine, however I finally found an ME that was flexible and made an informed choice based on the mix.
The fact that you say it's the most transient compressor ever, does not sit well with me.
If it was the best thing, then many more top MEs would use one. Don't get me wrong, I'm not bagging the technique, it's just that i find it a little over rated.

[Pixel25]

Originally Posted by Methlab

Most guys i know, before they send off to Sterling, or a place like that, follow a rule. It's snare up, guitar down. This is to compensate for the mastering crush. So make sure your snare is a little louder in the mix then you want before you squash your mix.

I've never, ever altered my mixes in order to compensate for mastering. I think that's a joke. If my ME used that much compression that it altered verbs and other such things, I'd kick him in the ass. a 'good' ME should be able to work with a 'good mix', end of story.

[MattGray]

Originally Posted by Pixel25

I dont' agree totally with that statement. Tape has it's benefits in mastering...sometimes. It's defeinitely not something i would default to on every single session, as some engineers do (mostly to try and impress clients) I used to use an ME that bounced everything i gave him to tape (on a really well maintained ATR machine.)
On so many cases it smeared the drum transients much more than a traditional compressor. In other cases it worked just fine, however I finally found an ME that was flexible and made an informed choice based on the mix.
The fact that you say it's the most transient compressor ever, does not sit well with me.
If it was the best thing, then many more top MEs would use one. Don't get me wrong, I'm not bagging the technique, it's just that i find it a little over rated.

I don't use tape on every job nor do I recommend others use it on every job, I merely stated that it was one tool that I often use & find more successful then typical 2 buss compression. I don't use tape where the mix has already been to tape somewhere along the line, or if the mix already sounds thick or too dense. But for the jobs that are thin, harsh or not well controlled & need some body, warmth or even on some that are really nicely mixed ITB (which is about 70% of my jobs these days), tape can make a great difference to the end result.

As I stated a lot depends on the machine, the electronics on the machine, what operating levels the machine is set up for & how hard the operator hits the tape. It's kind of like dialing in a compressor, it takes time to get that magic setting. I've got 2 decks here a Studer A807 1/4" which tends to compress transients more & sounds a touch honky. This gets used (not that often) at lower speeds with some music (usually stuff that's all mixed in the box) that needs a little character or for bands chasing after an old school sound. The other deck I have is the 1/2" ATR-102 with ARIA electronics, I compared the stock & ARIA electronics & chose the ARIA for it's detail & transient response. This gets used on jobs that aren't over compressed or too dense to begin with & it's sound is cleaner & punchier then the Studer.

As for compression I'm not against it, I use it all the time but usually prefer parallel compression over traditional brute force 2 buss compression.

Matt

[MattGray]

Originally Posted by Musiclab

Respectfully I disagree completely. When a mix is not compressed and it goes to mastering, your mix proportions can change , reverb levels etc. When compression is applied later the transients will change depending on how much compression the ME will have to apply.

If a mastering engineer doesn't know what they're doing, then perhaps I would agree with this statement. The good ones make every endeavor not to mess with the balance of the mix & this includes snare level, musical dynamics, stereo width vs centre. The challenging part is doing this while getting the rms level up to be competitive. This is where mastering engineers pull out all the tricks in their bag to make this process as transparent as possible & this is often what separates the boys from the men. A lot depends on the mix we receive, if the mix is already quite compressed then this can make achieving the above goals for a mastering engineer even more difficult because you've already taken away some of the life/dynamics. There is some techniques that can reverse this a little but it's often a band aid.

I tell you what, next time you're mixing do 2 mixes of the same track, one with your 2 buss compressor approach & then another without it (let that be the only difference). Start with mixing the track without compression first & then from there do your compressed mix approach. Don't think to much about how it will sound in mastering as this will work against you (preconceptions can be dangerous), just balance the levels exactly how you would want them as if it was going to be released that way without further treatment & keep your peak levels 2db under 0dbfs. Send both of them over to me & I'll give you a free demo master & let you decide which one you prefer. Only condition is that you share your findings here, sound fair?

The best masters (& mixes) I've done haven't had 2 bus compression, limiting or clipping before the mastering stage. Not to say that you can't get a great master from a mix that has been mildly compressed but this is usually 2nd place to one with no compression.

Matt

[Methlab]

Originally Posted by Pixel25

I've never, ever altered my mixes in order to compensate for mastering. I think that's a joke. If my ME used that much compression that it altered verbs and other such things, I'd kick him in the ass. a 'good' ME should be able to work with a 'good mix', end of story.

Well, maybe you don't mix metal then. The guy I'm referencing has a lot of records coming out this year under WB, so I don't know if it's "a joke" to him.

[Ecktronic]

A cowboys way of fixing this would be to slap a limiter on your mix at mixing stage, and then play about with the snare volume and EQ so it sounds cllose to how you want it to sound.

Before playing about with the snare I would bounce your mix with the limiter on it and repeat with different limiter settings until you get the volume you want your mix to be at. Then start playing about with the snare. Probably you might want to boost the mids and cut some low mids maybe. And boost the overall level of the snare.

Eck

[Musiclab]

Originally Posted by MattGray

I tell you what, next time you're mixing do 2 mixes of the same track, one with your 2 buss compressor approach & then another without it (let that be the only difference). Start with mixing the track without compression first & then from there do your compressed mix approach. Don't think to much about how it will sound in mastering as this will work against you (preconceptions can be dangerous), just balance the levels exactly how you would want them as if it was going to be released that way without further treatment & keep your peak levels 2db under 0dbfs. Send both of them over to me & I'll give you a free demo master & let you decide which one you prefer. Only condition is that you share your findings here, sound fair?

The best masters (& mixes) I've done haven't had 2 bus compression, limiting or clipping before the mastering stage. Not to say that you can't get a great master from a mix that has been mildly compressed but this is usually 2nd place to one with no compression.
Matt

The mastering engineer I normally use is Scott Hull, I never worry or compensate for mastering since I know that if I give him something good, it will always come back even better. I used to mix for years without 2 mix compression, now I don't. The hole in your challenge is I can't mix the way your describing, for the reasons I've previously stated, slapping my SSL FG384 on at the end will change the mix, so if I mix with the SSL out it will be a different mix than with it in, since engaging the SSL WILL change things. This is why it's important to use it from the begining of the mix.
So if you have another Idea how we can do this let me know, I have some mixes coming up on tuesday and wednesday

[RockManDan]

i think there may be a littl ebit of confusion here between the two techniques. In my experience, the idea is not necessarily to mix to a limiter on the 2buss and leave it there and print it. The idea is to slap a limiter on the buss and mix to it (to a degree) then take it off and send that mix to be mastered, yielding a mix that will be able to translate better through a few more db of gain reduction. I myself find that it yeilds better results. Its kind of a ballpark thing. Giving the mastering engineer a sqashed mix is stupid, as theres not much for him to do, but giving him a vibrant mix that is a little hyped in some areas that will be more resilient to compression might help him come to the overall mix dynamics/loudness ratio that you're looking for. at least in some cases.

also, it should be noted that not all ITB limiters are created equal. Put a mix through 2 different plugins will yeild 3 very different (often in crtical ways) results. Throwing an l2 on a mix and dragging the threshold down to get your RMS louder will often sound crappy, and you'll completely lose your snare, hence the popular term 'l2 abuse'. However there are other ways to do it ITB and still have your transients sound good. The Sonic Timeworks Mastering Compressor, while a few years old, to my ears sounds miles better than the L2. One of its perks among fans of this plugin is 'YOU DONT LOSE YOUR SNARE', and I agree. Ive been able to get results of -9 RMS and still have very prominent kick and snare using this plug. There are others out there of course, but they are not all created equal. If you're really into working ITB, doing some researching and smart buying habits will help.

also another ITB trick is using a transient designer, such as Waves TransX (perhaps designed by them because they felt guilty about unleashing the l2 on the world). It allows you to set threshold and sustain parameters and basically hypes your transients to a crazy level and really makes your mix snap. Doesnt sound that great by itself (its a fairly annoying and sharp sound), but used before a 2buss limiter can give you miles more punch and attack at the same RMS level than you would get simply squashing a normal mix. And since you're not 'raising the snare' or 'equing' the kick to compensate, using the limiter doesnt alter your mix decisions. The limiter is technically shaving off more db but not for nearly as long as it would be if the whole drum track were turned up, thus reducing audible pumping in the mix if you really like to squash. Its not perfect for every situation, but on mixes that are otherwise mixed properly, it can stave off the death of your drums when it gets limited.

just some thoughts from my experience. ymmv
-dan

[Bang]

a question was asked in this thread, and there is a specific answer.

If you want your snare to stay punchy and hot in a mix, you should rely on little to no limiting. Probably none would be best. A peak limiter will make that snare go bye bye. If you mix it louder in the mix to compensate, you'll have to push the limiter even harder and you'll get the same effect, if not worse.

The answer? Do what Grundman, Sterling, Weinberg, Tom Baker and every other top mastering engineer does. Use clipping. If you need to, then slap on a db of limiting, probably won't hurt as long as you use something decent like a Voxengo limiter.

[Musiclab]

Originally Posted by Bang

a question was asked in this thread, and there is a specific answer.

If you want your snare to stay punchy and hot in a mix, you should rely on little to no limiting. Probably none would be best. A peak limiter will make that snare go bye bye. If you mix it louder in the mix to compensate, you'll have to push the limiter even harder and you'll get the same effect, if not worse.

The answer? Do what Grundman, Sterling, Weinberg, Tom Baker and every other top mastering engineer does. Use clipping. If you need to, then slap on a db of limiting, probably won't hurt as long as you use something decent like a Voxengo limiter.

I would avoid limiting completely, compression on the other hand can HELP make your snare punchy. BTW they do alot more than just clip the converters, they're using some very serious compressors

[Bang]

Originally Posted by Musiclab

I would avoid limiting completely, compression on the other hand can HELP make your snare punchy. BTW they do alot more than just clip the converters, they're using some very serious compressors

Some do, some use hardly any. I've seen Tom Baker and Ted Jensen use hardly any, in fact, Tom just basically uses his Manley to let the tubes absorb a bit, the needle moved about .5 db.

If the question is how do I keep the snare loud, the answer is simply don't limit and use compression (if needed) and clipping to get 95% of your volume. I'm not disagreeing with you Lou, I'm just saying compression completely depends on the mix.

[jayfrigo]

Clipping is no magic bullet either. It may sound good on the CD, but the codec for iTunes or other web delivery often sounds worse, and it often sounds worse on radio too (see the Orban/Foti article). Generous clipping has a tendency to make processes down the line more problematic. You should try to find a happy medium.

While most mastering engineers use a little clipping here and there, most also typically use a little bit of limiting too. The closest thing to a "trick" is to get a little bit from a lot of places. Don't rely on clipping, don't rely on limiting, and don't rely on compression. Use a little bit of each when appropriate. You may even do a little bit on two limiters instead of everything on one, and/or a bit of analog clipping plus a touch of digital clipping as well. Use your ears and see what combination works best for the song at hand. When using the digital domain for this, non-linear processes like these are usualy better at 2X sampling rates so that some of the junk is swept out with the trash when you filter with a good SRC back to 1X rates.

As for snare specifically, if you want a loud snare, mix a loud snare. I wouldn't chase mastering that hasn't even happened yet, and try to make a bunch of changes based on what I think might happen in mastering. Make it sound like you want it to, and a good mastering engineer will preserve as much of that as possible within the confines of the level you demand. The loudness wars definitely impact snare, and kick too, but all is not lost. The best way not to lose all your dynamics is to, ahem, not lose all your dynamics! Those pesky laws of physics...

[MattGray]

Originally Posted by Musiclab

The mastering engineer I normally use is Scott Hull, I never worry or compensate for mastering since I know that if I give him something good, it will always come back even better.

Sure, I agree & this should always be the case with a quality mastering engineer whether a mix is sent in compressed or not.

I used to mix for years without 2 mix compression, now I don't.

Did Scott use to master those uncompressed mixes for you back then?

The hole in your challenge is I can't mix the way your describing, for the reasons I've previously stated, slapping my SSL FG384 on at the end will change the mix, so if I mix with the SSL out it will be a different mix than with it in, since engaging the SSL WILL change things. This is why it's important to use it from the beginning of the mix.

Ok put it this way, what are the main things that you change when mixing with the SSL inline that you wouldn't of changed when mixing without it? You mentioned higher snare levels & less reverb, is there anything else?

So if you have another Idea how we can do this let me know, I have some mixes coming up on tuesday and wednesday

Ok lets do 3 mixes, the 1st mix your normal way, with the SSL inline like you've always done. The 2nd mix is easy, turn off the SSL & print this mix without it. Then with the 3rd mix leave the SSL off & listen to the mix again, listen for anything that sounds unbalanced i.e. snare is now too loud, vocals are a touch loud, guitars & effects are too quiet. Adjust these to the level you want to hear them but remember the idea is that the uncompressed mix sounds the way you want the final balances on the master to be (but don't compensate for the extra level that will ensue). The theory is there should be no need for you to compensate for the snare or the reverb levels as they should be pretty much the same after mastering. The master should be very similar to the mix once level matched in terms of snare & fx levels.

Send the 3 mixes to me & I will treat each one individually & we'll post the results blind & let people choose which one they prefer. Once the votes are cast I'll reveal which mix was which. Sound like a better plan?

Matt

[MattGray]

Another perspective on my thoughts... This was a reply to a PM I received on this subject & thought I'd share my reply here as it seems relevant to this discussion.

Definitely steer away from limiting, clipping or 'plug-in' 2 bus compression before sending it to mastering. If you have a nice (preferably expensive) analog compressor that imparts a particular sonic character that you want on your mix, then by all means use it in moderation. My suggestion would be to use longer attack times & if possible use auto release or set the manual release so it exhibits less audible pumping. Make sure it's set to a lower ratio (1.2:1 up to 2.5:1) & set the threshold for half a db to 1db of average gain reduction & you should be ok. Another alternative to using your prized compressor on the 2bus is 'parallel' compression. This involves summing an uncompressed version of the mix with a heavier compressed version & blending the 2 to taste. This can be done on an analog board or in a DAW that has plugin delay compensation.

As for final levels & loudness, this is best left to the mastering engineer to achieve. There is no need for the mix engineer to use compression or limiting to assist in this purpose. A good mastering engineer has the techniques & gear to be able to do this with less audible penalties. In my experience compression & limiting always sounds best if it's applied towards the end of the chain (the mastering chain). One of the reasons is because compression & limiting introduces distortion. This distortion can be acceptable if there is no further processing to be done, but if it gets sent to mastering for further processing the distortion can quickly become audible & reach unacceptable levels. Adding further to this problem, as Jay mentioned is that once this problem mix/master reaches the radio processors or Mp3 codecs, it becomes even more distorted.

Matt

[RainbowStorm]

Originally Posted by kingliar

Hi all,

Maybe some of you know this problem : snare sounds great in the mix!, but at the final stage during mastering (finalizer), the snare is getting squashed and it sounds like there's a "hole" in the mid-range of the song (like there's no energy/power in the mid-range)

My question is : how to get a loud master (-9dBfs) without squashing the snare?

I'm probably not the first to ask this question and I know it has everthing to do with compressing/limiting. But how to do it ?

Does the mix has to be "fixed" or can it be done during mastering?

You prevent that by giving the drummer the appropriate drum sticks and telling him to not beat the hell out of that snare. Then you let the compressor do what's necessary to create a controlled transient output. You get problems with a squashed snare when it is played and compressed too heavily.