Tired of peaks when recording?

skei · #1 10-17-2010, 05:18 PM

Yo, jazzers!
If you, like indeed me, have struggled like a maniac to get a good level on your guitar when recording on your pc or mac, then fret no more, for I just saw the light...
After doing all I could to get a good level on my latest tune, 4, the Davis one, I searched for 'free limiter for daw' and stumbled upon THIS.

Just go there and download your AU or VST component, install in the correct folder, and be amazed. I had to lower a number of individual peaks on the guitar track for 4, like all other recordings I care to remember. When I applied this plugin I could take away all adjustments on the level curve, and it is PRO!
Totally perfect! Like a trained engineer would make it, only FREE!!
Try it, you'll like it!
Peace
Skei

GodinFan · #2 10-17-2010, 05:39 PM

Dynamics = interest. If you keep peaking, turn down. Using a limiter lessens the musical shadings.

skei · #3 10-17-2010, 05:47 PM

True, dynamics is good, I mean those things I, and I guess lots of folks struggle with, like I have this problem with the low e-string, it booms all the time, even if I adjust pickup pole etc, so I must have a ceiling, a level it won't exceed, and this works. If you don't experience this when recording, I congratulate you, and please disregard the post, but I need this thing, and I'll use it to be able to get a good overall level in my home studio.
Peace
Skei (the appreciate dynamics but want to hear the stuff one)

billkath · #4 10-17-2010, 07:36 PM

Quote:

Totally perfect! Like a trained engineer would make it, only FREE!!

Absolutely wrong, my friend.

It's software.

What does that tell you?

Well-it should tell you that all you are doing is limiting an already clipped digital signal. Software compressors and limiters do their job AFTER Analog to Digital conversion. If you are going "over" without the limiter/comp it means that you have already clipped the A/D convertor. You are thus limiting/compressing a clipped signal.

That is unrecoverable (unless you want to start freehand drawing of the clips later, which is bloody stupid.)

What to do?

Record in 24 bit if you can-that gives you TONS of headroom over the noisefloor. Aim for peaks at around the -12dB mark. You can ALWAYS make up gain later.

THAT is what the pros do.

Not loud enough on recording?

Turn your monitors up.
Make up the gain later, once everything is recorded.

The other way to do it is to limit with hardware-BEFORE the A/D convertor. Not many people do that these days, as there is no need to hit tape like we used to with a hard signal, or to maintain a good SNR in a 16 bit system.

skei · #5 10-18-2010, 03:29 AM

Ok, I'm sorry for posting the link. Disregard it. My bad.

billkath · #6 10-18-2010, 04:30 AM

I does have it's place, though, Skei, but more as a mix tool, or even a mastering tool.

The way too deal with the odd peak that's louder than the rest (and thus stopping you getting any louder by using the volume slider) is to draw a volume envelope after recording, and dip a bit just before the peak and back up to zero afterwards. That'll just reduce the volume on just the actual stray peaks and sound MUCH more natural-no pumping, no artifacts. If you do decide to use compression then after that-whether for effect, or for reducing it's dynamic range, it will have a track to work on that is in control. That'll allow you to use a slower attack time and faster release on your compressor, which is how you get more punch into your music, introducing some dynamics in a good way. Good, because it lets the initial transients through.

KB_Blues · #7 10-18-2010, 05:27 AM

Quote:

Originally Posted by billkath

Absolutely wrong, my friend.

It's software.

What does that tell you?

Well-it should tell you that all you are doing is limiting an already clipped digital signal. Software compressors and limiters do their job AFTER Analog to Digital conversion. If you are going "over" without the limiter/comp it means that you have already clipped the A/D convertor. You are thus limiting/compressing a clipped signal.

That is unrecoverable (unless you want to start freehand drawing of the clips later, which is bloody stupid.)

What to do?

Record in 24 bit if you can-that gives you TONS of headroom over the noisefloor. Aim for peaks at around the -12dB mark. You can ALWAYS make up gain later.

THAT is what the pros do.

Not loud enough on recording?

Turn your monitors up.
Make up the gain later, once everything is recorded.

The other way to do it is to limit with hardware-BEFORE the A/D convertor. Not many people do that these days, as there is no need to hit tape like we used to with a hard signal, or to maintain a good SNR in a 16 bit system.

+1 for this.

When setting up input levels I find it best to set my meters to show RMS (aka average) levels. I aim for -18dbs RMS. Like Billkath said, you can always make up gain later.

skei · #8 10-18-2010, 06:51 AM

Yeah, the way I used to do it, setting levels low enough to avoid clipping, then adjust individual peaks that would go red if I raised volume. The thing is I find this incredibly boring, and as I never intended to become a professional sound engineer I look for ways to simplify the process and still get an acceptable result. Not aiming for pro levels or so, just make it sound much much better than any of the recordings I have on my pages now, which it really does. I don't like compressor sound, but have had to apply it in the past. What I don't like about it is it brings up all those stray sounds that are not supposed to go up. The limiter, on the other hand, which I only use for mastering and post production, lets the dynamics remain, but sets a top level which it won't pass.
So, to sum it up; it helps me a lot, allows for dynamics below a set top level, doesn't bring up sounds like pick clicking etc that I don't want to have amplified, good for me, I tried a brief mix, could take away all compressor stuff, didn't have to reduce individual peak levels, got a much more natural sound, like it is straight from the amp, so I'll continue using it even if it's unpopular...

billkath · #9 10-18-2010, 07:22 AM

Yep-it's all a matter of taste. The Beatles used natural tape compression on their acoustic guitars all through their recording career, bouncing and bouncing whilst hitting the tape really hard to get that ultrasmooth rhythm sound. To me, nowadays, it sounds awful in isolation. However, as with all things in recording, it's within the context of a whole mix that one must judge. In that respect, the Beatles engineers got it just right.

What you need to know is that a limiter is simply a compressor with a high ratio-at least 10 to 1 but nearer infinity to one on some of the big boys like the waves L3- real brick-wall. Depending on where you set your threshold, and how the attack and release are set up, you may just be reducing those peaks. It CAN be done transparantly.

I know the post production work like doing your envelopes is boring, but believe me-with practice it can take less than a couple of minutes. I think it's a better option, especially for EQ'ing purposes, but as you're not doing this for general release, it's perfectly acceptable to do things your own way.

As with all advice about recording-listen to what people are saying, but ultimately make your own mind up about what YOU-as the artist-want. If this works for you, and you like that sound-good for you. It's not an exam, and there is no failure mark for not meeting others expectations of what constitutes a good sound.

skei · #10 10-18-2010, 07:54 AM

Good points, and just to point out what I mean, the two versions of Four are now up at the wix. At the Matrix there is only the new, with limiter. What I did more than adding the limiter is I narrowed down the piano and bass. Just so they don't use too wide a spectra of sound, leaving a more open room for the guitar to wander in. To my untrained ears it sounds much better now, which is why - at the wix - I call it 'Fourbest'.

billkath · #11 10-18-2010, 08:36 AM

As a critique-not of your musicianship, but purely on the recording- this is exactly the situation where I'd use envelopes. You are very slightly hesitating on some of the notes, and that's why you're getting volume differentials on different notes. It's sounds like you are not completely sure of the song-a really, really common thing . Do I think the limited version is better than the first one? Yes-I'd agree that it is. To my ears though, it could be improved enormously by eveloping it before some slight compression.

Listen-I'm on two weeks off from work, as the opera festival is happening in my home town. If you like, and if you could send me the individual wave files for each instrument, I'd be happy to have a go at it for you, just to see if it could be improved using my method. No charge, of course-it'd be a fun thing to do. If you're happy with the way it is now, of course, that'd be fine too. You could use YousendIt-- a free way to e-mail large files over the internet.
Let me know if you'd like me to mess with it. I really like the song and your playing, by the way.

fep · #12 10-18-2010, 08:52 AM

Quote:

Originally Posted by billkath

What to do?

Record in 24 bit if you can-that gives you TONS of headroom over the noisefloor. Aim for peaks at around the -12dB mark. You can ALWAYS make up gain later.

THAT is what the pros do.

+1, a pro told me that also... never had any clip worries again.

skei · #13 10-18-2010, 02:40 PM

Ok, you hear the difference. Of course lots could be done, but for me it's actually ok as it is. But thanks for the offer. The thing is I just wanna have acceptable & audible quality, I do actually have a pal from the old Rocker days who has a studio, where we used to do our tapes back when, if I really need some help I might as well approach him. But thanks again for the offer, and yeah, it is a great tune but then again, Miles was a genious, no less...

Quote:

Originally Posted by billkath

Listen-I'm on two weeks off from work, as the opera festival is happening in my home town. If you like, and if you could send me the individual wave files for each instrument, I'd be happy to have a go at it for you, just to see if it could be improved using my method. No charge, of course-it'd be a fun thing to do. If you're happy with the way it is now, of course, that'd be fine too. You could use YousendIt-- a free way to e-mail large files over the internet.
Let me know if you'd like me to mess with it. I really like the song and your playing, by the way.

billkath · #14 10-18-2010, 04:19 PM

No problem at all.
Getting your friend involved would be a good idea. You'll learn more in a couple of hours watching someone than can ever be learnt out of books. I suppose learning the guitar is just the same-a good tutor can teach you more face to face than you can learn from books in the same time.

reb · #15 10-18-2010, 10:58 PM

i am hesitant to even ask....how does one add gain later?

i watch the lights (no meter on my behringer 802), and cut clip whenever i see/hear it. adding gain later would be a great topic for those of us completely new to mixing. i have yet to figure out 'how do i mix any instruments into this and have them synched in time'.

frankly, skei, i would be proud to have done either version. the first sounds 'raw', like a club band. i would be fine with doing something that good personally.

skei · #16 10-19-2010, 05:05 AM

And I did a new remix this morning, enveloping et al. It's more even, the worst peaks are gone. Did it in Reaper, followed by amplification in Audacity, just up to
0 Db. I also edited out a few sloppy mistakes, so it sounds better all over, a slight multiband compressor, set to 'analog' preset, as I don't want to mess it up. It's ok now, as good as I care to and am able to get it.

Regarding the whole discussion about limiter/compressor and such, anybody who's read my posts knows the sound I aim for is a Wakenius type sound, not too much dynamics volumewise, smooth, dark. That's the sound I've always wanted but have only managed to get on one recording, the 'Nightingale' version. That's recorded in Ableton Live. Even if I get that sound out from the microcube, I just can't seem to get it on recording other than in Ableton. Which seems odd. I'd prefer to get that sound in Garageband or something, cos Ableton is not my favorite recorder.
Well, the new improved take is at the wix, where it's simply called 'Four', and at my matrix, labelled 4.
Peace.

billkath · #17 10-19-2010, 05:55 AM

Quote:

Originally Posted by reb

i am hesitant to even ask....how does one add gain later?

i watch the lights (no meter on my behringer 802), and cut clip whenever i see/hear it. adding gain later would be a great topic for those of us completely new to mixing. i have yet to figure out 'how do i mix any instruments into this and have them synched in time'.

frankly, skei, i would be proud to have done either version. the first sounds 'raw', like a club band. i would be fine with doing something that good personally.

There's lots of ways. The simplest is to just turn up the track fader!!
But-there's a few things you need to know.

First is-Normalize.

It is not a good idea, when mixing.
Why? Well, what it does is-it looks through the whole track to find the loudest bit, it raises that sample to the highest it can go, and every other sample by the same amount.
Normalizing doesn't affect a tracks dynamics in any way whatsoever.
Now-if, at one point during the song, one hit is much louder than the rest of the song, normalize won't really help get more PERCEIVED volume. You'd get the same effect by simply raising the fader on that track.

Second-compression/limiting
A compresser/limiter acts like an automatic volume control. When a track reaches a certain volume it starts to turn the volume down. It can stop the volume of a track exceeding a certain level. Depending on how you set the controls it can do this very quickly. you can set it so it will only turn down the volume of just the very loudest bits, and then release for the quieter bits. By doing this it allows you turn the whole track up, as what used to be the loudest part is no longer as loud.
It is affecting the tracks dynamics-the difference between the loudest and quietest bits. Disadvantages are that it MAY sound, if the controls are not set just right, pumpy, squashed, or lifeless and it affects the tracks frequency content, meaning you may need to EQ more.

Third- Envelopes.
Envelopes are another word for automation. It draws a graph line across the waveform, which is connected to the fader. You make little adjustments to the volume of the track by putting in little dips and peaks along the graphline. When you play the track back, it's like you were mixing in realtime, adjusting the fader up and down at exactly the right time. Now-you are affecting the tracks dynamic range, but not in the same way as using a comp/limitier. Frequency content is not affected. It enables you to turn up the overall volume of a track and even it out.

Fourth-and the most important---Do nothing-get volume during mastering.
Think about this-it's quite obvious, but it's something most newcomers miss completely. It really doesn't matter how loud a track is on it's own, as long as it's not stupidly quiet.

All that matters is the balance BETWEEN the different tracks is right. If that balance is right one can simply turn the volume of the whole song up once it is mixed. Once you have lots of tracks you are going to have to drop the individual track faders back anyway, because the cumulative volume of their summing would clip the master outputs if all tracks had individual volumes above a certain level.

Sounds quiet? Yeah-so turn up your monitors or headphones. You'd be stunned at how quiet the volume of a song is when coming out of a professional studio. They do that because they know that it's going to be mastered, where overall EQ and volume will be applied by the mastering engineer. Mix engineers job is to make the song sound the best it can, irrespective of overall volume.

So-Best advice.
Don't worry about tracks in isolation. All that is important is to get the mix right-the relative EQ and volume of each track in context with the other tracks. Once you get that right you'll bring up the volume, using a combination of the above methods of the overall song.
Remember-recording (called "tracking"), editing, mixing and mastering are all SEPERATE processes. Don't try to do them all at once. Take a break between doing any of them.

skei · #18 10-19-2010, 02:08 PM

Good run-through!
Fine you mentioned the Normalising thing. Had no idea. Somebody told me long ago one should always normalise when mastering, so I did. I won't anymore if it's no good...

Peace
Skei (the abbeynormal one)

billkath · #19 10-19-2010, 03:21 PM

I didn't say not to normalize in mastering. I said it's not a great idea in mixing. Too very different processes.
Many engineers will normalize the master as the second- last process,after sample rate conversion but before dithering down to 16 bit to eek out the remaining last bit of volume. They don't normalize to 0, though-it's to -0.2dB.

skei · #20 10-19-2010, 03:42 PM

Ah, you see how easy it is to misinterpret things if not really understanding the process...

Peace
Skei (the will now have coffee at long last one)

skei · #21 10-23-2010, 03:58 PM

What I did now is I fixed a good sound, just like the one out of the amp, or as close as I am able to get. I used no limiter, just compression, not too much - to avoid freaky sound sensations, and some reverb. Did this in reaper, as I noted the sound lagging behind in ableton, can't remember the word for it, oh yeah, 'latency', I believe that's the term. None of that in reaper. So now I have a set sound, perfect, or as close to perfect as my very limited capability allows. Will use that for all recordings hereafter, then master in audacity, simple and good app. easy to apply the 'amplify' function to up volume to a set level, 0db or just below, without using any artificial stuff, oh, and i do the 'normalize' thing here too. I just fiddled around to once and for all get a good sound setting saved for future use, and will now concentrate fully on learning the play-bit of it all. That feels like the really important bit, but a sound to accept feels like something that might help in the process. I think I have managed to get that now, after sucking in all this knowledge and trial and error et al...