compressing/mastering the ULTIMATE turd

[rgraves]

Hey all,

Got something I'm doing for a friend, he has an old set of recordings he did of 35 of his songs, just acoustic guitar and voice, and he recorded them all on a SONY tape deck recorder! I know, it's basically a lost cause, but they actually sound ridiculously good...it's amazing how all the levels are just perfect and the sound quality is actually quite good. It was recorded with no microphone at all, just the built in mic, haha.

So, a family member of his transfered the tape to 2 CDs and now I have it. What I'm wondering is, just for the sake of fun and to give him a little something I wanted to basically improve the volume. The poor guy has wanted all his life to record in a studio and has never had the money. I am going to finally record him in my home studio when he gets out here, that should be tons better than what he had before, but he recorded on that sony tape deck 22 years ago.

So, I went ahead and played around a little last night with the files and at the very least could do a normallization. But I wondered if anyone else might have some suggestions? I was thinking compression/or EQ, the thing is even though it sounds darn good, it of course has a fair amount of tape hiss or something like that.

In order to compensate for that, is there a frequency range I could try to EQ a bit? Should I avoid compression with fear that the noise floor would be raised? oR is there a way to set the compressor so that it won't bring up the hiss too much. I was going to try some basic effects to make it sound less redundant, like some reverb, but it's all one track, is there anything you guys would avoid??

Thanks for any help!

[SonicAlbert]

If they already sound ridiculously good the best thing might be to just leave them alone. There is audio restoration software that will reduce or eliminate tape hiss, and it would be far better to use that then try to eq out the hiss.

[rgraves]

Well don't get me wrong, I guess I should have clarified that it sounds ridiculously good all things considered. It definately doesn't have adiquate volume. I mean, the way I look at it is if I have to raise the volume on my stereo to the max just to enjoy the song, it's not loud enough.

I guess what I should have said was; what are the best things to try to do an audio restoration project. Obviously I'm not looking for radio results, just trying to do the best with the noisy situation I have

[travelin travis]

i doubt that you'll make much of an improvement if the mix volume is already weak and has alot of hiss. can you post a sample of what you're working with?

when i first started with the pc/recording thing, a friend of mine brought over a tape that his dad had recorded when he was a kid. actually it was a cassette copy of an old rtr recording. we played around with it for days in cool edit and ended up liking the untouched version better than any of our attempts to improve it. btw, it sounded alot like you described. low volume, background hiss, and a few clicks here and there.

chances are, compressing will bring up noise levels as will normalizing.

[timboZ]

How did the family member transfer the tape to 2 CDs?
If he just used a stock pc sound card I would see if you can get the tapes and do the transfer yourself on some better equipment.

[rgraves]

Yeah, I could post a sample...I've never done that but I'm sure it would be easy.

Well, I've already played with it a bit and normalizing it turned out fine. But that's just turning up the volume, I was asking more about trying some more advanced features such as compression and EQ and some way to reduce the tape noise perhaps

[timboZ]

If you do the transfer yourself and set the levels a little hotter you will have no need to normalize.

[Hi_Flyer]

Originally Posted by timboZ

How did the family member transfer the tape to 2 CDs?
If he just used a stock pc sound card I would see if you can get the tapes and do the transfer yourself on some better equipment.

that sounds like the best idea so far. do the transfer at 24 or 32-bit and make any attempts at noise reduction, compression or whatever processing at the higher resolution.

some of that audio restoration software is pretty good, if you know how to use it...

[chessrock]

Originally Posted by rgraves

So, I went ahead and played around a little last night with the files and at the very least could do a normallization. But I wondered if anyone else might have some suggestions? I was thinking compression/or EQ, the thing is even though it sounds darn good, it of course has a fair amount of tape hiss or something like that.

In order to compensate for that, is there a frequency range I could try to EQ a bit? Should I avoid compression with fear that the noise floor would be raised? oR is there a way to set the compressor so that it won't bring up the hiss too much. I was going to try some basic effects to make it sound less redundant, like some reverb, but it's all one track, is there anything you guys would avoid??

There are some de-noiser plugins out there. But they're pretty expensive. Something like Waves X-noise would be perfect, in fact.

.

[rgraves]

Could someone suggest any audio restoration software as I am unaware of it.

I'm quite certain it was transfered using some stock PC card blah blah blah.

I don't own a tape player and don't really plan on getting one just to help him out, haha. Basically the goal would just to be to give the thing some life, I mean here's what I'm thinking: for example if he throws one of his songs in a mix of other songs for his friends on a CD, they have to turn it up to max volume to hear it and then get shocked to crap when the next song comes on at 130db blasting them to hell. OK, so I normalized it, that problem is solved.
Next, every song more or less sounds the same, it was all recorded in the same session, and the slow ballad sounds as dry as the fast swing song. Would any of you guys try throwin some 'verb on a ballad and something else on a different song? Considering it is not 2 seperate tracks that is...

The tape hiss isn't bad, but it's noticable of course. I will post a file when I get back home tonight.

[mshilarious]

Compression is going to be potentially dangerous because of the noise issue. Definitely start with a denoiser. Then fix any EQ problems. I shudder to say it, but a bit of exciting, or maybe just some high boost might be called for some enhancement. See if you can do some stereo widening, or perhaps simply use a nice but subtle stereo reverb. Finally, bring volume up as appropriate, but try to avoid reducing dynamic range.

[peopleperson]

Post a sample.

[rgraves]

Originally Posted by peopleperson

Post a sample.

Hey, you guys got it. I won't be off work for another 10 hours unforetunately, but I will post a sample or 2.

[LemonTree]

Originally Posted by rgraves

Hey, you guys got it. I won't be off work for another 10 hours unforetunately, but I will post a sample or 2.

If you want the hiss removed make sure the sample you post up starts a second or two BEFORE the song starts....it's easier to grab the hiss on it's own from the silence then remove that from the actual song.

[Reggie]

Unfortunately, removing tape hiss is never transparent. It covers such a big frequency range that you are almost positively going to lose some of the important stuff if you try to reduce the hiss by very many db's. I don't know of anything free you can use, but Soundforge (Sony Noise Reduction), CoolEdit/Audition, Waves, and a few others have noise reduction utilities.
Since it was recorded on the built-in mic, the audio probably doesn't have much useful information above 10-12K or so; so I guess you could try rolling off the highs up there and maybe lose a little hiss.
And don't do any compression, as this will cause the hiss to go up and down and be very annoying (or relaxing, like the ocean I guess...). You can use some fast peak limiting to get a couple extra db of volume if you like. And EQing the frequency balance, that will be up to your ears and abilities.

[rgraves]

Originally Posted by LemonTree

If you want the hiss removed make sure the sample you post up starts a second or two BEFORE the song starts....it's easier to grab the hiss on it's own from the silence then remove that from the actual song.

OK guys, here it is...umm, whoever did the transfer from tape started each song like on the first note, so unfortunetly there is no pre-song time at ALL.

http://www.users.qwest.net/~fisherdennis/

[mshilarious]

OK I played around with it for a while. Used a Denoiser with about 9dB of noise reduction, didn't do too much damage there, but then I added highs back at 10kHz. I "essed" the vocals just a hair. Then I covered it all up with reverb

I wouldn't try to compress this at all.

[kylen]

Somebody mentioned getting a better transfer, that's a good idea. I use a Mytek96 for that little old thing and bring it in at 24bits and either 44.1K or 96KHz depending. Then downsample it if necessary using a good minimum-phase SRC converter R8brain:
http://www.voxengo.com/product/r8brainpro/

As far as NR software I use Redunoise to eliminate most of the noise especially fading in/out of the song:
http://www.voxengo.com/product/redunoise/

In Sonar5/Adobe Audition you can assign an automation envelope to the "Mix" knob in Redunoise and turn it off during the song or turn it way down when the music is louder than the noise (why have the NR plug remove unnecessary stuff?). In that mode it's non-destructive meaning you don't have to make decisions about NR settings then apply EQ/Comp to a rendered wav file. I just put Redunoise last in my plugin chain and use it whenever necessary. It does eat some cpu cycles though so you might have to lock the track down and apply EQ/Comp off in a buss somewhere...

Anyway - it's VST software for in the box kind of stuff.