Jump to content

Limiter while recording vocals.


Misha

Recommended Posts

This subject is something I am fighting with for years. Yes, I tried to use all those techniques to reduce the level differences. The only one that helped a little bit without destroying the performances was more distance to the microphone.

I know there are those restrained vocalists that can be recorded without any pain. Those takes are very balanced and even-tempered, but sometimes boring! On the other hand there are those singers that use a wide range of dynamics and there it is very difficult to get an even recording. Sometimes there are just single syllables that are too loud. Yes, you can adjust the input level to their highest peaks, but then there are still those uneven results that take a lot of time to be adjusted manually afterwards. You cannot just use a compressor or limiter to even them out.

To cut a long story short, in such cases I use my old dbx166 dynamics processor (from about 1990) right after the microphone and that helps a lot!

Link to comment
Share on other sites

3 minutes ago, marled said:

singers that use a wide range of dynamics

Marled, bingo! That is what I am battling with :)

I looked at some DBX stuff. Prices are very reasonable. The thing is, I am not worried about "beautification"  of sound by getting something that cost as much as used car,  but something that would do the job clean, without audible artifacts. I do not want to kill the tone of the nice mic that I am using, just so the compressor kicks in when needed to suppress clipping. 

P.S. Grant, thank you for reply.  V.Rider is a nice little plugin, but it is used on post recording, not in live recording.

Link to comment
Share on other sites

You ha

11 minutes ago, Misha said:
22 minutes ago, marled said:

singers that use a wide range of dynamics

Marled, bingo! That is what I am battling with :)

So you need to ask yourself why is a  singer using a wide range of dynamics. If he is just an  out of control ignoramus who expects an interventionist engineer to cobble together a "normalized" gain envelope on the digital track, then compression may save you some work. But what if he intends to go from a whisper to a scream in the same song as a matter of style or musical expression. Compression anywhere is going to thwart that intent by narrowing the dynamic range of the recording more than the performance that is its source. To put it in perspective, the digital representation of audio on a 16 bit CD is capable of representing a dynamic range of 96 dB--that is roughly the range in dBspl from a whisper at six feet in a quiet room to the threshold of pain. The usual loudness range for singers is about 30-80 dB with a smash the glass opera singer maybe getting to 100. Presumably your singers are not singing so loudly without amplification that they are hurting themselves or those near them as they perform, so there should be more than enough dynamic range in the digital realm. You might have a problem if your microphones etc. are not sensitive enough to deal with the quietest parts when the loudest parts are quiet enough to avoid clipping. If the singer (or his engineer) expects the volume in his headphone feed to be normalized regardless of the loudness at his mouth hole, then it might be better to put a compressor on an analog circuit from mic to headphone and record the full dynamic range, at a lower volume for later tweaking. 

Link to comment
Share on other sites

slartabartfast,

Lol! I am that ignoramus, that likes to experiment  with different vocal ranges in one song for the sake of musical expression and just have some stupid fun! I believe that is still legal :)  Yes, sure I will lose some dynamics at very peaks using compressor, but it is better than having some  clipping ..in my case.  Yes, I tried software that magically "heals" clipping, but I do not like the result of that.  My clipping issue is not that serious., around 5-10 seconds total in a single tune (If I would of left it as is, without re-recording  sometimes with maaaaany takes) . 

When I record vocals, it is semi-improvised process,  and it seems to me (and I am sure some people here know exactly what I mean) that occasionally   the best sections are the ones screwed up with clipping. Then, when I start at fresh takes,  I usually lose the  drive needed to recreate same thing, and the end result is not as hot as intended.  So instead of creative process I get  choreographed stage dance thing. 

P.S. I use a decent, sensitive, large diaphragm mic, and that is a mic of choice.  

 

 

 

 

 

 

Link to comment
Share on other sites

It's really up to the singer to employ good mic technique. I often recommend k.d. lang videos, she plays the mic like a slide trombone.

Try making the meters as large as possible in CbB, then have the singer sing while looking at them. Hopefully the end result will be learning to correlate distance from mic with volume.

Link to comment
Share on other sites

Craig, true.

Sometimes I  jump a few feet away from mic and then right back to it, just to get the phrasing the way I want :) 

I am not a pro (or semi-pro) musician by all means, just a hobbyist doing music thing between work & family choirs. Once I compose/record a tune, I rarely go back to re-sing it or re-mix it.   I like the process of creating new material more than production. Since it is open discussion, may I share tune I cooked in spring, that gave me very hard time (and oh boy, did it made me dance around the mic to avoid clipping! :) )?  Maybe somebody kind enough could suggest if hardware compressor or/and limiter would help me in similar scenario with clipping issues, before I buy one. There used to be a shop to rent gear close by, but unfortunately it closed, otherwise I would just rent the thing and try it.  Here is the link:

https://www.bandlab.com/bluebeat13/tongue-my-foe-57637b27

Thank you in advance,

Misha.

 

 

Link to comment
Share on other sites

9 hours ago, Misha said:

Lol! I am that ignoramus, that likes to experiment  with different vocal ranges in one song for the sake of musical expression and just have some stupid fun! I believe that is still legal :)  Yes, sure I will lose some dynamics at very peaks using compressor, but it is better than having some  clipping ..in my case. 

I am afraid you missed my point. I am not saying that singers who employ a wide range of volumes in their performance are ignoramuses (ignorami?) or somehow doing something illegal--quite the contrary.  My own personal taste in music is to have something when you are done that you cannot listen to in a moving car because the softest stuff falls below the wind and traffic noise threshold without the loudest stuff damaging your hearing--the kind of product turned out by every orchestra or serious non-pop singer in the history of music since the stone age. What I am saying is ignorant is to expect a real dynamic range to be reflected in a final recording that smashes the variance into a band of a few dB.  If you want that kind of flattened volume range, then as a singer you should learn to produce it rather than depend on some feat of technology to achieve it from the original performance, and if you want to savor the beauty and expressiveness available in a wide dynamic range, then you need to use technology to capture it without exceeding the capacity of the technology. 
 

Look music-on-the-go and dodging-beer-bottle-venues  have been major drivers of the loudness wars, and we are used to hearing stuff that is so compressed that it no longer qualifies as high fidelity. In fact radio stations for years have routinely compressed already squashed recordings and many streaming services and MP3 players routinely do the same. The days when people valued going to a sit down venue where alcohol was available only during intermissions and listening to a decent stereo in the dark filling the silence of their homes are probably long gone. Singers of pop music have in fact learned technique that makes that narrow-range product easy for engineers to produce. If the singer wants to make that flat-volume product, and still manages to clip or overload the equipment without falling below the noise floor he qualifies as out of control. Given the capability of decent equipment and the dynamic range of digital representation he also has a voice that is capable of more volume than a normal human. If he values the exploration of the range of loudness available to the human voice I applaud him. In any case, if you want to have the option to squash your performance in the box while maintaining the ability to let it stretch, then capturing the performance by recording less hot is the better way. 

You should realize that clipping in the digital realm is dependent on the total power at all frequencies (including those too low and high to hear), and that mixing all your tracks together is basically the process of addition. Perhaps the most effective way to avoid clipping in the final mix while getting something that is reasonably "loud" for today's listeners is to use  is to use equalization in the box to filter out power at frequencies that are less important or are already occupied by other instruments so that they are not distinctly audible in the mixed track. That process subtracts dB from the final mix without requiring that loud become soft at least subjectively to the listener. You can then raise the volume of the entire mix without exceeding the clipping point. But early compression removes all the frequencies more or less indiscriminately and limits your options for that important technique in mixing.

Link to comment
Share on other sites

martsave,

I will look, but the project is on my backup drive somewhere, I do not have it on PC.

slartabartfast,

I did not miss your point, I was just teasing. Thank you for taking the time on this. It was a nice read! I get the points you are trying to make loud and clear. My in-laws are professional, (the non pop) musicians, so I know exactly how much guts it takes to have a perfect control of your performance, so the music is balanced. 

To be truthful, I know that my mixes are total disaster and my tunes are half baked. I am not aiming at public, this is just for me, friends and a very random listeners. I enjoy just writing stuff in a sketch form.  As soon as I grasp the general idea of the tune and record a sketch, that is totally fine for my purposes. 

2 hours ago, slartabartfast said:

You should realize that clipping in the digital realm is dependent on the total power at all frequencies (including those too low and high to hear), and that mixing all your tracks together is basically the process of addition.

That is true, but not specifically relating to specifics of vocals.  At times I would want close up, sort of gentle presence and harsher vocals in same track. I usually succeed in recording how I want it, but it requires a lot of experimenting and re-takes, a lot of them to be discarded because of clipping.  Again, this is not a huge problem at all.  I know that compressors are not a magic bullet by all means, but at least I would like to know from somebody who knows if it could help me reduce minor clipping. And if so, which one I should aim for? The compressor or limiter/compressor combo, tube or diod... I am on a budget, but do not mind used gear and rather get something better used than shrink wrapped  but of less of a quality.

 

Thank you.

Link to comment
Share on other sites

Martsave, No I did not. I mentioned that it is a large diaphragm and it is a mic of choice. 

I do not want this thread to wind way off from topic :)

It is not the mic or setup or gain level. It is me who is singing stuff too soft and too loud in same tune in the small room, which results in occasional clipping. All I want to know is if a compressor/limiter will help to address some of clipping.

Thanks.

Link to comment
Share on other sites

I used an outboard compressor/limiter for years but eventually sold it.

The reason is that such devices really only work well when you have somebody watching it, which isn't possible when you are both performer and engineer. I wouldn't know how badly the track had been compressed until after the fact, and if it did turn out overly squashed I'd have to turn down the gain and do it again - the same process I'd have followed had the compressor not been employed.

The downside, of course, is that you run the risk of going too far the other way: having the gain too low and ending up with excessive noise/room in the recording.  I battle that using Greg's "caveman" solution, recording quiet and loud parts separately. Most of the time, that works fine. In the rare situations where it doesn't, I try to address it with mic technique, singing like I'm live on stage, usually with a handheld microphone.

And that's another solution that took me a long time to figure out: putting away that pristine condenser mic and using a handheld dynamic instead. Dynamics are inherently more forgiving, and unless you're recording a delicate ballad, the advantages of condensers is minimal-to-none. Lots of successful pop and rock singers record exclusively with SM-58s or SM-7s. These days I've abandoned the 58 in favor the better but still-affordable Sennheiser e945

Oh, and one other thought. If you don't have any acoustical absorption around your singing space, get some. That greatly extends the practical dynamic range of your microphone by reducing the room sound, allowing cleaner recordings at lower gain settings. My "vocal booth" consists of freestanding gobos stuffed with 3 inches of rigid fiberglass. The result is a very, very dry recording that I can subsequently turn up as needed, limited only by the microphone's own (quite low) internal noise floor.

Link to comment
Share on other sites

On 9/9/2019 at 9:57 PM, Misha said:

Gswitz.,

thanks, however this does not answer my question on using a limiter in live scenario, while singing.

 

Just put the effect of your choice on the channel. Preferably a low latency one. I don't understand why this is hard. Reduce the input gain and raise it with vsts before sending to the mains. Now you don't clip and it sounds great. You might use the cakewalk compressor on the track and the same with a high ratio for a limiter on the master out.

Edited by Gswitz
Link to comment
Share on other sites

 bitflipper,

Thank you for responding.  I guess most of us use some kind of "caveman" approaches. This was one universal approach that Greg / you suggested: "recording quiet and loud parts separately". I have one more that I can share. I have two tuned mics and what I tried is placing one mic  2-3 feet from another, so when clipping occurred on the closest mic, the one further away was more forgiving...."Caveman ate some sugar" approach.  

8 hours ago, bitflipper said:

putting away that pristine condenser mic and using a handheld dynamic instead.

Ohhh, no no. This is my baby, not giving up my children.  I do have 58 from years back but the tone does not compare for vocals to the one I have on the stick. Room is not the greatest for recording, but more than  ok for my needs. 

Bottom line, there is always room for improving, techniques, treatment but what I am after is trying to reduce minor clipping before it is recorded to Cakewalk when I sing quiet and loud parts in one take. So the question still remains: If outboard compressor/limiter will help to eliminate some clipping before it hits PC. 

--------------------

Gswitz,

You are talking about post recording. I am interested in live scenario.  Limiting/compressing "While" you sing...

 

 

 

 

Edited by Misha
Link to comment
Share on other sites

No, I'm talking live. You sing... it goes into channel 1... into Cakewalk... compressed... sent to the master out... limited there... out to the interface... out to the mains.

Why does this not work live? The only problem would be not being low enough latency. Latency would be driven by your computer and interface and choice of compressor/limiter.

I have at least 10 options personally that work fine for this with low enough latency to perform live. So do all the other people trying to advise you.

Most folks here may have better gear now. So, for example, my RME interface has compressors on input and output so I could do it without going through the daw. Same with delay and verb. 

But I could go through the daw. No problem. So can you.

Back me up here, Bit? Anyone?

Edited by Gswitz
Link to comment
Share on other sites

Just to add to Bit, if you sing loud loud, the dynamic will sound better. Being soft and breathy into a dynamic hurts. It brings out the words of the mic. Being good and loud it will sound much better. Also, if you are good and loud, your level will be more consistent. It will help you compress yourself as it were. When you sing softly then get really rolling the variance is much more significant than when you go from louder to loudest.

Long ago I made a soft recording with a dynamic and it sounded drastically different from my ribbons. Then, with the amp cranked, noticed the difference is much less obvious.

Edited by Gswitz
Link to comment
Share on other sites

Just in case I was crazy I just did it with the Sonitus compressor. I have my sample buffer at 64 which gives me 4.1 MSec round trip. So from the time I say a word to the time it comes out of the speakers is 4.1 MSecs. It's equivalent to having the speakers 4 feet further away. This is totally workable live. I just tried with a 48 sample buffer and this gets me down to 3 MSecs.

I'll grant that I don't record through Sonar because when Sonar drops out it stops. I never want the recording to stop during a live performance so I use RME Digicheck. Digicheck will clip all tracks by a sample buffer if the computer can't keep up, but it never stops recording. To me, this is preferred. 

If you are using a laptop with DPC issues, you will have to set the sample buffer much higher and this could be too high for live performance. But, we can probably help you resolve DPC issues so you can run at a lower sample buffer. Even on a cheap laptop, recording a heap of tracks with no (or only a couple) of FX is usually no issue at all.

It's another thing if you're going to drive the processors hard with soft synths etc. Then I raise the buffer as much as I can stand.

Edited by Gswitz
Link to comment
Share on other sites

Gswitz...

Maybe I am going crazy... 

Please see attached. This is while recording... Input meter is active Sonitus is dead (while recording!) 

If I press stop->rewind->play Sonitus will compress on playback, but not at the time when recording is actually happening....

Are you saying it is possible... to use software limiter while you recording live audio?

 

 

 

 

51 minutes ago, Gswitz said:

 

 

screen.jpg

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...