Help - Search - Members - Calendar
Full Version: Setting An Input Recording Level
GMC Forum > Discussion Boards > VINTAGE GMC > Community Activities and Tutorials > Recording > Recording & Mixing Classes
tonymiro
Hi all,
I'm just gearing up to write the next two tutorials in the on-going series on how to record. As part of this I was playing back a few recordings that people have posted in GMC - members and instructors. In so doing one thing I've noticed is that a lot of people, members and instructors, seem to like to record 'hot'. That is they like to have their levels at or very close to 0 dBfsd. In quite a few instances recordings are over 0 dBfsd.

Now for personal use/practice and so on it's really up to you what you set your levels at. It's also been fashionable (but not necessarily right) to record 'hot' for sometime now. A lot of people think that recording like this is great because they believe that it makes their mix louder and so it is more likely to stand out from a quiet one.

Notwithstanding the above my advice would be not to record hot - either individual tracks or the mix - prior to mastering. If you record hot you are getting closer and closer to saturation and you may well clip. Digital clipping is not nice. If you want to hear clipping then load up Reaper and go through some of the downloaded takes - quite a few clip: when they do you get some very unmusical distortion. Also if you record hot, whether itb or not, your faders will be closer to the extent of their travel. It's known that faders actually offer better performance in the first part of their travel then at the extreme. Further whether you record itb or not at some stage a mix is summed and recording a lot of tracks hot can result in summing issues.

Also as you approach 0dBfsd you reduce the dynamic range available for adding eq, reverb and so on. Keep in mind that something like eq can be used to increase the gain of a frequency band in a track - if you are close to clipping and you eq gain you will clip. Apart from effects, if you have several individual tracks that are all close to 0 it is possible to saturate and you will end up with a dynamically compressed mix. The closer you get to 0 (or worse exceed it) the less natural your mix will sound.

If you are recording with the intent of producing a CD/MP3 or similar then you will probably send your mixdown to be Mastered. Let the Mastering guys normalise to unity - they almost certainly have better speakers and outboard then you have and probably a lot more experience.

As a suggestion, focused on guitars as its GMC, you should try to set your gain so that your average level is at -10 to -12dB and your occasional peak is at -6. This will provide you with more headroom to add/adjust reverb, eq, gates and so on and you will be less likely to clip/distort and generally saturate your track and mix. (Note- most home studio setups have sufficeint nois floor on a good audio card to do this.) In brief the art of recording is not to produce the loudest recording but one that sounds as good as possible.

The above is just my 2 pennies worth. If you like recording 'hot' fine but I personally believe that a lot of 'hot' mixes would sound better by lowering the gain.

Cheers,
Tony
Andrew Cockburn
Great advice Tony - you can always normalize, but you can never undo digital clipping.
tonymiro
Absolutely Andrew though personally I'd leave normalising until as late in the day as you can. (I've seen people applying brickwall limiters and normalisation to the recording part of recording/mixing long before they even got near mixing never mind mastering ohmy.gif )

One other bit - most people here are likely to be recording a distorted guitar - clipping that (even analogue clipping/tape saturation) is likely to result in a mess. You will lose dynamics like pick attack and lots of subtle nuances if you ride the gain on a distorted signal too high and your wave form will end looking like a series of rectangular boxes - not good IMHO.

Cheers,
Tony
Andrew Cockburn
Thats what I meant actually, meaning you can always normalize (late in the process) but you can't undo it (if you do it earlier), I was agreeing but rereading it was a little ambiguous! With 24 bit systems today, headroom and noise is much less of an issue than it used to be with 16 bits, so keep your headroom high as long as possible just as Tony says!
tonymiro
Might be worth stickying this Andrew - or I could cut and paste it into the sub - to avoid it getting 'lost' in the rest of the on going discussions?

Cheers,
Tony

ps I'd be interested in hearing from anyone who wants to put the alternative view for recording 'hot' as none of this is set in stone - it's largely opinion and what you get used to doing and what you think sounds good...
Andrew Cockburn
I Moved it into your forum as it is a kind of lesson smile.gif
tonymiro
laugh.gif
Sure np.

Cheers,
Tony

guess I'd better do a lesson on normalising pdq then wink.gif
Andrew Cockburn
Just stick a fancy title on this post and you're there smile.gif
tonymiro
TNX Andrew smile.gif .

Title is now about as fancy as I can get at 3:30 in the morning smile.gif .

Cheers,
Tony
ballistic31
hi well I hva e a few questions on this topic..........cause i ave been recording alot recently...I must say i have bee recording a bit on the hotside of things.....I have noticed...that the master just gets bombarded with the mix..and is just in clip mode the whole time.............I have tamed down the input but then i notice my wave form is very thin seems like not much activity there .............so is this a problem as far as dynamics go .......I have also found that getting a good mix without clipping a mic on a drum kit mainly cymbals was a huge chalange in the drum booth but i got that worked out .... I have allways read that anything over 0db...was just defeating the purpose in a digital world because anything after that is just lost space signal with no gain except noise..... is this true ?

I think my biggest question is ........what should a proper wave form look like ........I do know you dont want one that looks like it is the whole track width wide ........ that is a bit unpleasent to hear... I must say i am addicted to the chalanges i face in the home recording world but help and tips is allways nice...with the learning curb........

take care ballistic
tonymiro
Hi Ballistic:

If your master is peaking in the red then the mix is too hot. IMO if its peaking in orange (and personally I'd say yellow) it's too hot. On most digital meters with a 'standard' calibration if you get the master peaking over 0 that's hot. (Meters often have different calibrations though - if you listen to the a 'hot' mix it will sound odd and distorted.)

Track width can be altered and rescaled on pretty much every DAW I've seen so the extent to which a wave fills a track lane, or appears not to have much activity, doesn't mean a great deal to me - it depends on how it's been scaled. What I'd be more interested in is the dynamic range. Basically you need as wide a dynamic range as you can get and that means having as low a noise floor as possible with a ceiling where your peak signal doesn't clip both at recording and mixing. -10/-12 rms and -6 peak should be fine for virtually any DAW. If you need more range you should try to lower the noise floor - removing hiss etc from your system and/or getting a better sound card - rather than increasing the gain and recording 'hot'.

I know of some pro recording engineers who actually record with an rms of more like -18dB and peaks at -12 in a DAW. Their argument is that they find that to be more 'open' and natural as the closer the DAW comes to 0 the harder and more congested it can sound. That sort of thinking though really only works on a really quiet system - for most of us -12/-6 should be fine.

Drums are a challenge to record as they are very dynamic. A single drum hit goes from near silence to full scale very quickly. You need to set the gain so it doesn't peak in to the red at all or you will get a distorted drum. 0 and over in digital will produce digital distortion and a clipped and heavily compressed signal. So you're reasoning for staying below 0 us spot on.

What a wave actually looks like depends on the instrument/source you are recording and to some extent the type of music you are trying to record/mix/produce. A clean wave form though should produce a series of peaks and troughs - troughs where there is silence and sudden peaks as the instrument is hit/strummed, hammered... The peaks may have different intensities/heights depending on how loud the signal gets and the width of the rise from silence to maximum will depend on the actual instrument (all sound sources have different characteristics for their attack, decay sustain and release.) A distorted guitar will give a wave that gives little in the way of peaks and troughs because a distorted signal has a compressed dynamic range and usually the signal doesn't get the chance to fall to silence before the next note/string/chord is played. (A very distorted guitar wave can look pretty much like a rectangular box from the start of the passage to the end.) A clean guitar will give lots more peaks and troughs because the range will go from near silence to maximum as you pluck the strings, slide up and down and so on.

One of the difficulties with a heavily distorted signal is if you have to edit the wave form. If you can't see where a note starts and ends (no picking attack, no peaks and troughs etc) then its really difficult to see where you can splice the wave, where the beat is and so on. Other difficulties, as already mentioned, are that you have so much less room to work in with regard to adding effects, eq, panning laws and so on and whilst you can normalise a signal easily its not as easy to remove digital distortion.

One thing that is nice about DAWs is that you can record clean and add a distortion post recording. If the distortion effect doesn't work, fine you still have the clean guitar track, just change the distortion model. Record a distorted guitar and if you don't like the distortion well re-record the whole take wink.gif . I think I'm right in saying that much of Pink FLoyd's/Dave Gilmour's more recent stuff has been recorded clean into the desk and distortion added later with maybe some reamping.

At the end of the day its really important to get your input levels right, right from the start before you start recording. You're better of spending a few extra minutes setting up and reducing the input gain to avoid clipping then recording too 'hot'. Record too 'hot' and there is little you can do except scrap the take and start it all again with the levels set up more appropriately.

Hope this helps,

Cheers,
Tony
Andrew Cockburn
QUOTE (tonymiro @ Oct 14 2007, 02:26 AM) *
I think I'm right in saying that much of Pink FLoyd's/Dave Gilmour's more recent stuff has been recorded clean into the desk and distortion added later with maybe some reamping.


Dave has been reamping for years, back to at least Wish You Were here days maybe earlier - I read about this somewhere but I don't recall exactly where, could have been Sound on Sound, or maybe Nick Mason;s Autobiography.
tonymiro
Interesting - hadn't realised he'd been doing it that long blink.gif .

Cheers,
Tony
ballistic31
thank you very much for the great piece of advice,,,after reading that i went back to my mix.......and reduced my tracks by 3 db....and got everything in a fine working order...I took off all the effects and also creared all of the volume and panning envelopes........started just with a clean mix the main ryhthm guitar was distorted .........but i could see what you ment about the way different wave forms look like after a closer look..........I got all my levels equaly matched clearly about 3 db or so under zero.........and my faders where down around in the - 12 db i think...........but what a difference in controll over the mix....the trim controlls where nice and low as well...............and the master had plenty of headroom.........the mix im working on is the one i have been posting called creeps........slowly getting the bugs worked out of the mix and you definatly helped me out with it........I must say the age of digital is definatly sensitive to work with lol not like the old days of the 4 track tape eaters lol

thanks once again..................for your help......

ballistic
tonymiro
Pleasure Ballistic smile.gif .

Cheers,
Tony
Majickal
there are a few schools of thought on the 'hot or not' scenario - this is my take on it.

Record Hot and Turn Down - Record the track at the highest level possible without clipping. Why? When you turn the track down you have reduced your noise floor of the track and given yourself the maximum dynamic range to play with. The wider the dynamic range the more real the sound of the instrument. The lower the level you record at the less the dynamics of the instrument play a part and the closer to the noise floor of the system you are. This may not seem like a problem with one track, but add another 23 tracks and you suddenly end up with a load of unwanted noise. The further away your noise floor is from your target sound in terms of level, the more effective gating will be as well.

Fader position: Analogue vs Digital

In the days of analogue/analog wink.gif gear turning the fader down was a big no no, as unity gain (the 0db, point, not to be confused with the infinity point/lowest point of the fader) was the ideal location for the fader as input gain = output gain. As the fader was moved either up (adding gain) or down (reducing gain) the frequency response of the electronics changed due to the logarithmic nature of the fader and therefore changed the properties of the track. Most digital systems, especially those that use 'in the box mixing' do not exhibit the same problems that analogue faders did.

So to sum up, turning a digital fader down does not result in the same sort of corruption of sound as happens in the analogue world.

The real question that should be being asked here relates to the quality of the converters that are being used. For example really good 16bit converters can sound much much better than poor 24 bit converters. The dynamic range of these devices can also be brought into question as well. This directly affects the initial quality level of the recording.

Most consumer, prosumer grade A/D (analogue to digital converters) do not have peak limiting on them, so it is prudent when recording sources with wide dynamic (drums, vocals, slap bass, etc..) ranges to pop a compressor in-line (or insert before the A/D) of the pre amp. This will protect the A/D from unwanted clipping and perhaps save that magic take from the one louder pop that would of otherwise ruined a perfect recording.

To summarise, i would suggest the following:

Recording at lower levels reduces the recorded dynamic range of the source material and bring it closer to the noise floor.

Recording at higher levels requires more attention to detail but yields higher quality results (imo), though additional investment in outboard equipment is required.

So, choosing the right balance of louder and softer recording levels is part of what being a recording/sound engineer is all about.

As Tony pointed out earlier in this topic, making sure the input meter is not clipping is critical to a good recording. The input signal should be peaking no higher than -3db.
Hisham Al-Sanea
yes its right hot records finally you are getting distorted track so you have to start with normall level and to leave
ever thingsd to the final mixing because the track pass in many periods in mixdown so finally this track should have a warm sound
Ivan Milenkovic
These tuts are great TOnyMiro, thanks for sharing your mighty knowledge! smile.gif
Nemanja Filipovic
great topic...I have a question....I am waiting on my new comp...and I am going to do my album here at home(exept guitars and vocals)...but....need your opinion.... I alway put wawes L2 ultramaximizer in master before any thing...he keeps my 0...(no clip...and bus my volume up)...but now I will do mastering in big studio...so do I need L2 anymore becose of the mastering to be.....or do I just mix up with master fader down(-10 or -12)....?
Nemanja Filipovic
Hi tony I was wondering if you can help me fith this question up steirs..thanks:)
tonymiro
I've never used the L2 Nemanja but from what I've been told about it it is good - well all Waves stuff is generally good.
A large pro studio may have AD/DA that is better then the L2 however but the L2 is as good as Apogee and RME from what I've heard and it's going to be a sound that you're used to. It's brick wall limiting is v good as well and has look ahead. Personally I don't think it would hurt to have it as a back up to whatever the studio has but I would leave the limiting until the Mastering stage when everything is summed and set the gain at the console for the recording and mixing as you're suggesting.
Cheers,
Tony
Nemanja Filipovic
thanks....it studio they have L2 hardware in rack...so....I will wait to the masteroig process..thanks again
Ivan Milenkovic
I think that you send your recordings on mastering without anything on the master bus, but Tony can confirm that..
Nemanja Filipovic
QUOTE (Milenkovic Ivan @ Feb 18 2008, 12:52 AM) *
I think that you send your recordings on mastering without anything on the master bus, but Tony can confirm that..

yeah...was interested in the subject a little more so I went to my producer...he sad if you do not plan to take your songs on profesional mastering in big studio you shoud put some limiter...but the again it is important wich one...L2(software,in rack is diferent story)gives to much color to the song...and the better solution woud be Timeworks or LA2A..and if you deside to master in big studio limiter is not a good solution...
OrganisedConfusion
QUOTE (Andrew Cockburn @ Oct 13 2007, 01:25 AM) *
Great advice Tony - you can always normalize, but you can never undo digital clipping.

Sometimes clipping can be a good effect and several bands have used clipping to actually add to the song. An example is Territorial Pissings by Nirvana. The vocals clip but that is the idea of the song and what they were going for.
Nemanja Filipovic
QUOTE (OrganisedConfusion @ Feb 18 2008, 06:06 PM) *
Sometimes clipping can be a good effect and several bands have used clipping to actually add to the song. An example is Territorial Pissings by Nirvana. The vocals clip but that is the idea of the song and what they were going for.

off course...whole Stadium Arcadium(RHCP) is cliping,Califonications to...SOAD..and manny more...but they have the comfor...and limiters and maximizers do not just make your volume up and not cliping...they have there own color...that is why hardvare limiters cost 3000$ and even more....Tony?
tonymiro
QUOTE (Milenkovic Ivan @ Feb 18 2008, 12:52 AM) *
I think that you send your recordings on mastering without anything on the master bus, but Tony can confirm that..


Generally yes spot on Ivan though with stems being used more and more and with the rise of the home studio it seems that things are changing on this. The mastering engineers I know aren't happy to be sent stems that have been 'premastered' though. Guess time will tell as to how much we all move to stemming.

Also all limiters etc will have their own 'sound', or 'voice' or 'colour' etc, it's one reason why a big commercial studio will have potentially a lot of outboard so that they can pick and choose which sounds best for a particular track. Most producers though have their own favourites for particular genres.

Cheers,
Tony
Nemanja Filipovic
thanks Tony....buy the way...what is the best hardware limiter in your opinion
tonymiro
I don't master v often Nemanja - much prefer to send stuff to a mastering studio. However when I've done it I've had good experiences with Chandler. That probably means I'm a fan both of analogue and maybe a more European as opposed to US type of 'voice' smile.gif .

Cheers,
Tony
Nemanja Filipovic
thanks Tony...always helpfull...
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2014 Invision Power Services, Inc.