Forum Closed
The forum is now to new posts. All the historical content is still available to browse.
if you are looking for musicians to play with, please view the Bands Seeking Musicians list, or use the Musicians Directory
You can use our pages on social media to connect:
Upping the sound levels on a analog recording...
| Message Board > Music Chitchat - Heavy > Upping the sound levels on a analog recording... |
| _Griphin_ User Info... | It's cool finding the band I do, thing is I need to figure out how to up the levels on some of these recordings (there quieter then newer digital recordings), even though I normalize the recording, I'd like to find a utility which brings the levels up without shredding the original recording. Perhaps I'm overlooking something in WaveLab?!? - Tue, 21 Feb 2006 1:16pm | ||
| lonemonk User Info... | To do much more than normalize would be to change the original dynamics of the piece of music (Possibly quite significantly.) If you are transfering in from Cassette, try to get a better level on the way in. Without going over on the digital meter of course. Keeping a digital recording operating using as many bits as possible is key, not only for volume reasons, but also to retain an accurate representation of the original. Recording too low means fewer bits are used, which makes digital systems create errors, and any future volume boost will bring along more noise in the form of those same digital errors plus any noise already recorded from the tape. *Very* judicious use of a good compressor/expander can allow entire mixes to sound much louder than they probably do now, but this will change the dynamics of the original forever. Good compressor programming is a black art, and many recordings have been ruined by mis-use. From a Archival/Restoration point of view, it is a questionable practice. Philosophically speaking, an archival recording should be as accurate sounding as possible to the original. Heavy signal processing afterward will leave you with something different from the source. It might arguably sound 'better', but the original recording engineer and/or band may totally disagree with that assessment. A couple of things that might help before getting into other techniques are: 1. Invest in a decent soundcard, if possible one with actual (physical) level meters on the input in order to see real-time how close the audio is to going over 0db. or 2. Perform more than one test recording of the same piece and examine closely the levels in wavelab to ensure maximum record levels were used. (Keep the ones which are closer to 0db without peaking) I've also had success in picking the highest peaks of the sound file (Quite often louder than normal drum hits), and very carefully highlighting just the very peak of of the waveform and lowering that volume a cunt-hair. Sometimes this allows the Normalize process to bring everything up in a more even proportion. This is in fact a manual form of compression of sorts and is only adviseable when there is a very small number of very brief loudness peaks. (I would never do this over long parts of a song for instance, as that would destory the dynamics of the band playing the music) In short, its best to solve these issues long before the final stages of making CDs, etc. - Tue, 21 Feb 2006 4:28pm | ||
| Destroy_The_Flesh User Info... |  | ||
| _Griphin_ User Info... | FLOPPY DISC DEATH!!! Heh.. well yeah, I run filters over the tape stuff because it tends to make the recording sound better (eg. get rid of tape hiss for instance). I just wanted to try to up the levels so the tapes don't sound as quiet. - Tue, 21 Feb 2006 7:49pm | ||
| FLY ON THE WALL User Info... |  | ||
| _Griphin_ User Info... | Already did that, I upgraded to a P4-3 gHz. - Tue, 21 Feb 2006 8:49pm | ||
| Mi*coll* User Info... | The sound card is really important, as lonemonk said. And if you really want the recordings to sound right, though, you should have them mastered by a real mastering engineer-- no software can do it for you. - Tue, 21 Feb 2006 10:13pm | ||
| lonemonk User Info... | Too true Mi*Coll*. Many of the projects may not have had a mastering engineer to begin with! - Wed, 22 Feb 2006 10:04am | ||
| Microphonic User Info... | I've never used WaveLab before, but you should have some luck with a compressor/limiter, whose job it is to manage the peak levels that fool normalizing. When you normalize audio, it just finds the single loudest sample in the file (say it's -6 dB), and then cranks the whole file up by a uniform amount so that the highest peak doesn't exceed 0dBFS. With a compressor, set the threshold *below* the highest peak(s), so that the compressor is active on big peaks like kick and snare hits in a finished mix. Then set the compression ratio, expressed as x:1 to get the desired amount of compression. If we go with the earlier example of having your highest peaks at -6dB, and let's just say the rest of the mix is riding around -12dB, and your threshold is set at -10. So at a 2:1 ratio, every 2dB of level over -10dB only comes out as 1dB of level. So say you set your ratio to 5:1, and if your attack time is set as close to instantaneous as you can get it (or it will automatically be instantaneous if you're using a limiter), theoretically no signal hotter than about -9dB is getting by. You can see you just gained a few dB of headroom. Adjust the release time setting to taste, bring the make-up gain control up about 9dB, and you're off. And remember, since the decibel scale is logarithmic, 6dB equates to a doubling or halving of perceived volume, so if you've gained 6dB, you're laughin. Oh, and if WaveLab doesn't have an appropriate plugin built in, check out KVR or some such website for some of the great free plugins floating around out there. Cheers Dirk - - Sun, 26 Feb 2006 10:46pm | ||
| mactac User Info... | ok, i've done a lot of this. a lot of mixing, a lot of premastering & a lot of volume maximizing. the key is to follow a few methodical steps if you want to maximize the volume of an analogue recording, this is what you need to do: Keep in mind that it depends on a whether on not you are working from a stereo mix or seperate tracks... you'll be able to do a much better job if you have seperate tracks. if you do, start here, keeping in mind what everyone has told you about how to use a compressor, sound card, etc: Keep in mind that by bringing up the volume, you will reduce dynamics. to a point, this is often a good thing. after this point, you'll be squashing it too much. use your ears. it often requires years of practice to be able to to this, and it is just as much an art as a science. 1. first transfer everything over to your computer. preferably, transfer it all at once with a multi-channel sound card if possible so you avoid phase problems. 2. listen to each track seperately & get a feel for what is going on 3. on tracks that don't have instuments that have slow attacks & decays (eg keyboard pad sounds), run them through a noise gate. with precussive sounds (ie kick drim) you can be more agressive about the attack & release times... be careful that you're not cutting off reverb tails, etc. you don't want to mess up the sound, just get rid of some of the noise floor if you can. 4. now, you want to apply a bit of compression to each track seperately. you're going to have to experiment - it's always a tradeoff between dynamics and volume. 5. now, when we compress each track, we're going to do it TWICE.. yes. twice. the first time, compress each track (& use the makeup gain in the compressor to bring it up as loud as possible without clipping) by around 3db or so. this will get rid of many of the peaks 6. second pass, this is where some art comes in. you'll want to compress percussive sounds more (ie kick drum) and non-percussive sounds a bit less (slow, ambient sounds). Pay particular attention to any reverb here. compression makes softer sounds louder, so make sure the reverb or any ambience is not getting too loud in relation to the original sound. you'll want to go anywhere from 0-6db here. some things (ie vocals, bass) sometimes sound good with lots of compression, but sometimes they don't. it REALLY depends on the style of music - rock vocals are often very compressed, jazz vocals are not for example. 7. NOW, listen to the mix - you may need to re-adjust the relative volumes now because the mix might sounds different. make sure it all sounds *btter*. if it doesn't , go back & see what you've done & adjust 8. ok, so you've got you individual tracks compressed. You may want to subtly eq a few tracks to slot them in a bit... ie make the bass blend with the kick so they don't exist in exactly the same frequencies. if you have many instruments piled up in the same frequencies, you'll have a hard time making a loud mix without that one frequency clipping. MIxing s also an art form & makes a HUGE difference on how loud your final mix can become 9. Now, we work on the full mix (this is where you start if you're starting from a stereo mix, after it's in your computer). You want to tame the big peaks again. 3db compression is a good place to start. 10. now, you want to appy compression to the mix again, 0-6db. play around, listen for pumping, paying attention to the dynamics making sure you are not squeezing the life out of it. 11. EQ... balance the mix - sometimes compresssion does strange thingsw to the eq levels. sometimes adjusting eq before compression works well, sometimes afterwards. play around. i can explain what's going on here if you'd like... but i need to move on :) 12. now, a good multiband limiting or compression will maximize the volumes further. Keep in mind throughout all of this that you don't want to screw up the sound. listen with your ears, not with your eyes on your computer ! Also, the quality of the compression will play a HUGE role in the final result. experiement with different compression plugins, some will sound better for different steps. Also keep in mind that people spend YEARS perfecting and it will take you a long time to do your first mix. some people have "tin ears" and never can do it right. If you need some help, le tme know - or perhaps I can do a sample mix for you so you have something to go by. BY all means, learn to do it yourself - it's a big job, but you can make a good recording sound amazing if you do it right. - Tue, 28 Feb 2006 8:08am | ||
| mactac User Info... | oh, and don't *ever* normalize. all it does is destroy data. you lose resolution when you normalize. the olnly time you should ever normalize is when you're done, but if you did everything right, you shouldn't have to. use the makeup gain on your compressor plugins to get the level up to close to 0db peaks. don't normalize !! - Tue, 28 Feb 2006 9:28am | ||
| _Griphin_ User Info... | You know that would be a good idea, if the levels of the recording weren't so low to begin with, I normalize to bring out the sound actually. And well I would agree with you, but normalizing brings out stuff in the recording. Usually the recording gets normalized after I have recorded before splitting up the tracks. I dunno about losing resolution, course everyone I've done work for has said good thinks about the recording. Here's how a typical session works... Run WaveLab v4.0>Record live analog>Normalize>Save recording as file>Split up each track as to the songs on the recording and save individual tracks>Run Steinberg Clean v4.0>Run filtration to get rid of tape hiss and other things on each track>Run RazorLame and compress the tracks to MP3 (I can tell people like Scott Henderson are gonna cringe after reading that). I can usually tell if levels are shredded, my EPI speakers are really good to hear screwups, and unlike my failed attempt at restoring Dreadnaut, stuff sounds good. Course, I used to do analog recording without processing with Clean, it sounded horrible (eg. Mission of Christ-Silence in Grave (Yellow Tape) had this annoying hissing (bad tape copy) but after cleaning it you can actually hear most of the instruments unlike before (just ask Ross Bay about my first Mission of Christ attempt). I should look into your point about destroying data. Without running Clean on the recordings I find that the recordings don't sound as good as they do cleaning the recording up. - Tue, 28 Feb 2006 10:39am | ||
| zakco User Info... | Interesting discussion.... Just for the record, ALL digital processes result in rounding errors. This goes for normalizing as well as EQ or dynamics processing, even volume and pan automation. Theoretically, the less math you perform, the less errors and subsequent quantization distortion you subject your audio to. Now back to the real world....Whether or not we can actually HEAR this is another debate altogether. Suffice to say that any artifacts introduced by normalization pale in comparison to the mess that mp3 encoding creates. Translation: chances are your audio has bigger problems than normalization.... Having said all that, in my experience, normalizing is usually an unnecesary step considering that a good peak limiter (like Waves L2) will give you all the make up gain you'll need while setting the output level where you want it. Personally I'd keep things at -0.3 dbfs peak level. That will allow some headroom for cheap DA converters and keep mp3 encoding from pushing things above 0dbfs. (mp3 encoders often add a couple 10ths of a db gain during the process) In regards to your analog recordings, I don't see how anyone can tell you how to approach EQ/Dynamics/Noise Reduction without actually HEARING the source material. If there's one thing I've learned about audio it's that no two situations are the same... Mactac, Microphonic and the others have offered good, well intentioned advice but it is really a shot in the dark without hearing your recording. (BTW, this is NOT a stab at anyone, just my opinion) My advice to you would be to pick a couple of songs, book an hour with an experienced mastering engineer. Ask questions, watch and learn.... Best of luck, -Zak - Tue, 28 Feb 2006 12:25pm | ||
| mactac User Info... | my advice was some general advice to use as a starting point.... which it is. if you follow it, paying special attention to listening, it will do a good job of helping him with his questions. obviously there is no substitute for experience, but you have to start somewhere. paying a mastering engineer (and by the way, we're talking about pre-mastering here) $200/hr to help you might be out of reach for him , I don't know. "My advice to you would be to pick a couple of songs, book an hour with an experienced mastering engineer" ... i've done a bunch of remixes for a number of labels. I've done the premastering on a bunch of commercially released tracks. This is why i offered to do a sample mix so he has something to go by if he wants to learn.. .. my offer still stands - Tue, 28 Feb 2006 12:42pm | ||
| zakco User Info... | Mactac, I just offered some advice, same as you. You seem to have taken my post personally. By "experienced mastering engineer" I meant just that, nobody in particular. I wasn't suggesting you were unqualified. I'm sure you'll do a fine job for him. No need to get defensive.... Have I intruded on your "turf" or something? -Zak - Tue, 28 Feb 2006 1:05pm | ||
| _Griphin_ User Info... | Can they fix a 2 track recording? Also, why bother paying someone $200 per hour when you can learn how to master yourself and save the money. Perhaps if I was planning to release a comp and make meager funds, then I'll look into remastering, but well, since I lose money with my projekt... :) - Tue, 28 Feb 2006 1:18pm | ||
| mactac User Info... | " No need to get defensive...." I wasn't. I was clarifying. - Tue, 28 Feb 2006 2:05pm | ||
| zakco User Info... | Griphin, I understand your desire to learn to do this yourself and save a few dollars. That is how most people find themselves getting started in this biz, myself included. In the spirit of DIY, you may find this forum to be helpful: http://recforums.prosoundweb.com/index.php/f/31/0 Brad Blackwood from Euphonic Masters is the moderator and there are several serious heavyweight MEs that hang there. Bob Ohlsson Dave Collins, Bob Katz to name a few....these guys are VERY experienced career MEs and are really generous with their time and knowledge. I've learned a ton from reading their posts and the've answered many questions I've had along the way. It's a great resource. Check it out.... -Z- - Tue, 28 Feb 2006 4:35pm | ||
| _Griphin_ User Info... | Oh hey Zakco, it's not the fact that I wanna save a few dollars, it's the fact that if you gonna try to copy something, you should always do it correctly, therefore you save a ton of work. (?!?) Besides, if I get good at it, then I could make a lot more then what I presently make money wise, and if your passionate about something, then by all means work at doing just that. - Tue, 28 Feb 2006 5:14pm | ||
| zakco User Info... | Agreed on all points.... And do check out that mastering forum. Cheers! -Z- - Tue, 28 Feb 2006 5:22pm | ||
| _Griphin_ User Info... |  http://www.meritline.com/meritline-digi-deck-convert-tapes-mp3-copy-digital-music-cassette.html They've developed an actually drive which lets you play tapes and record data from them, but you can't record information to the tapes in the drive. That's so cool! - Tue, 28 Feb 2006 6:11pm | ||
| Mi*coll* User Info... | "don't ever normalize" I don't want to be a dick, but that is ridiculous advice, I must say. Especially when you go on to advise using make up gain. Do you think that they are different processes in the digital domain? Because they aren't-- both are amplification. And if you are working with 32 bit files (which you should be, and then dither down to 44 khz, 16 bit for release), there is absolutely no problem. Zakco is absolutely right about all processing resulting in rounding errors, but seriously, I wouldn't worry to much about any of that stuff if you are planning on doing a home mastering job. Your lack of a pristine monitoring environment is going to be the biggest bottleneck by far, no matter what. Real mastering engineers have sometimes upwards of $100,000 monitoring setups, not including room treatments and, of course, years of experience and great ears. In general, my advice is to pay a pro or to do as little as possible besides normalization. Maybe limit the songs a bit to get a consistant level on the recording. And, lets be honest, "loudness" results from the listener turning up their amplifier, not from something inherent to the CD. --Mike - Wed, 1 Mar 2006 5:12pm | ||
| mactac User Info... | "I don't want to be a dick, but that is ridiculous advice, I must say. Especially when you go on to advise using make up gain. Do you think that they are different processes in the digital domain? Because they aren't-- both are amplification" you missed my point. of course they are both amplication. (i've *written* audio apps btw). The point is, don't amplify *twice*. it literally multiplies any roiunding errors. of course if you are working at 32 bits, it doesn't make much difference- in any case, it is arguable, but not "ridiculous". the fact is, many people are still working at 16 bits, and if the level is low, it can make a big difference. the point being, again, is that you don't send audio through more gain stages than necessary. you use the makeup gain instead of using the normalization, THEN the makeup gain. it's unnecessary. calling me ridiculous is not helping the person who posed the question. and no, it is not ridiculous - it's common knowlege. in any case, i stand by what i said... you almost never need to use it. >"loudness" results from the listener turning up their amplifier" well of course. but "perceived loudness" is different. doing a good premastering job can make musich sounds *MUCH MUCH* louder. There is a reason people go through this process, knowing full well that you can just turn it up.... i'm not going to bother explaining why, because i'm sure you understand this concept. In any case, trying to help. the person asked how to get their cd louder. there have been many people who replied, with some great advice & tips... Anyone who wants any more info from me, email me privately, i'm done posting in this thread :) - Thu, 2 Mar 2006 1:08am | ||
| Destroy_The_Flesh User Info... | One the subject of mix/mastering, does anyone have any good book recommendations? Or any good books for sale - Sat, 4 Mar 2006 2:27pm | ||
| zakco User Info... | Two books that I've found helpful: "Mastering Audio" - Bob Katz http://musicbooksplus.com/bmastering-audio-scienceb-p-4355.html?osCsid=aec579c83783a1f2d8d3aebdb8e94594 "The Mix Engineers Handbook" http://musicbooksplus.com/bthe-mixing-engineers-handbookb-p-2116.html?osCsid=aec579c83783a1f2d8d3aebdb8e94594 I've also heard good things about this one: http://www.mixingwithyourmind.com/index.htm -Z- - Sat, 4 Mar 2006 3:00pm | ||
| _Griphin_ User Info... | I totally have to find that book written by Bob Katz. - Sat, 4 Mar 2006 3:58pm Edited: Sat, 4 Mar 2006 3:59pm | ||
| Mi*coll* User Info... | Yeah, Bob Katz is a big name. Him and Bob Ludwig. I hear that that book is good. - Sat, 4 Mar 2006 4:11pm | ||
| Destroy_The_Flesh User Info... | Thanks - Sat, 4 Mar 2006 6:35pm | ||
| _Griphin_ User Info... | Yes, Katz I've heard about before (perhaps I heard about him in Guitar World). Thanx for the suggestions Zakco. - Sat, 4 Mar 2006 8:37pm Edited: Sat, 4 Mar 2006 8:39pm | ||
| zakco User Info... | the Bob Katz book is great. Very technically oriented though. More focused on the science rather than the art I think... I learned alot from reading it. -Z- - Sat, 4 Mar 2006 10:20pm | ||
| _Griphin_ User Info... | Always remember, when it comes to audio on computer, it all has to do with math. - Sat, 4 Mar 2006 10:45pm Edited: Sat, 4 Mar 2006 10:45pm | ||
| odb User Info... | toomuch compression and gating and normalization is the diffrence between bland mainstream albums and exciting indie albums. don't take the character and feeling out of your music like that. - Sun, 5 Mar 2006 7:10pm | ||
| Mi*coll* User Info... | ...that and the music... - Sun, 5 Mar 2006 10:10pm | ||
| _Griphin_ User Info... | http://www.cdmasteringservices.com/dynamicrange.htm -=- This might explain why analog is more quiet then digital. Awesome conversation so far (I still remember it). - Wed, 12 Apr 2006 2:43pm Edited: Wed, 12 Apr 2006 2:44pm | ||
| lonemonk User Info... | I remember running across that RMS/Peak level RUSH comparison before. A very good example of the trend being discussed in the rest of the article. Its not so much the difference between analog vs digital per se, but more to do with what people feel they need to do to modern recordings to make them somehow standout. Radio (and CDs) killed the recording star as it were. 'Volume/loudness' are relative terms of course, without much meaning by themselves. (Mactac already covered that) When most of us suggested that compression might not be the best way to go, it was precisely for the same reason that your article is talking about. Its all too easy with modern programs to make a recording sound as LOUD as you want, but the dynamic range is seriously comprimised by doing so. Whatever nuances there were are all gone. Used sparingly and knowledgeably, compression is no bad thing. - Thu, 13 Apr 2006 8:31am Edited: Thu, 13 Apr 2006 8:43am | ||
| _Griphin_ User Info... | Yes, the Rush CD had to be remixed cause the levels were too high, I dunno, it sounded OK to me but it's not a CD I'll remember that much of. - Sat, 15 Apr 2006 11:59am Edited: Sat, 15 Apr 2006 12:01pm | ||
|
|||
Not logged in
Log In / Register (optional)
