What is a good output path?

[mij sang]

I own a Mac 5 Leopard , PT 7.4.2. & Wave Lab 5. I desire a very smooth and clear sound. I,m not sure what path I should take b4 I bounce. For now I "convert after bounce", burn a wave then import into Wave Lab where I process conservatively. I render as a PCM (uncompressed wave)
Should I be using AIFF instead of wave files? Is it better to use Power Dither instead of "Convert after bounce"? How Much Head room should I leave for Wave Lab? -5? I also own one Avalon 737. I'm not sure how to use this on the Master channels or during mastering. I know It's a big question. I do appreciate any advise on a good process or an effective tool.

[daeron80]

It doesn't matter whether you use AIFF or WAV files. They sound exactly the same, all other things being equal.

The Avalon 737 is not really a mastering tool. For one thing, it's mono. And it's designed for tracking and mixing, not for mastering. If you had two, it would be possible to master through it, but it might be difficult to be sure you had exactly the same settings on both channels. And you wouldn't want to compress one side of a stereo mix separate from the other.

Smooth and clear sound result from keeping a clean, solid analog signal path during tracking, and maintaining the greatest digital resolution possible during mixing and mastering. IOW, if your original tracks aren't smooth and clear, you're not likely to get there by mixing or mastering techniques. It's something that's preserved more than added to the sound. But even if your original tracks sound great, it's easy to loose the smoothness and clarity by not understanding bit depth and sample rate, and what happens to the bits during digital processes.

You should track and mix at the highest resolution your system will support. With PT LE, that means always 24 bit. If your session only has a few tracks and not many plug-ins, you can probably track and mix at 24-bit/96 kHz. If you need many tracks, you might need to use 24-bit/48 kHz to handle it. It can still sound excellent if you're careful.

When tracking, try to keep peaks below -12 dBFS. If you shoot for -18, you'll usually come out about right. Higher peaks on percussive material will suffer a loss of punchiness. On all tracks, leaving that headroom will help you achieve a clearer mix.


Preserve full resolution at mix down - no dither, no bit reduction, no sample-rate conversion. If you tracked and mixed at 24 48, mix down to 24 48. It's usually good to leave about 3 dB of headroom for mastering. A little more is fine. Import the 24-bit/96 or 48 kHz file into WaveLab and do your processing. Output the result to 24-bit 44.1 kHz without dither, using the highest quality sample rate conversion algorithm the program provides. Open the resulting 44.1 kHz file in Wave Lab, dither it down to 16-bit with a good quality dither like POWr, and save as a different file with a different name at 16-bit/44.1 kHz (PCM uncompressed, of course).

Dither should be used any time you reduce bit depth (like going from 24 bit to 16), and only then. Ideally, it should only be applied once, as the very last thing that happens to the audio before it's put on CD or converted to mp3 or whatever. 24-bit resolution should be maintained up to that point. It's important that the sample rate conversion be done at 24 bits if you want good clarity.

[daeron80]

Also, the term "burn a wave" makes me wonder whether you're needlessly involving a CD. If you mean you're burning a WAV to CD then importing from the CD to Wave Lab, you can skip that step - just import the WAV straight from hard disc.

[mij sang]

Thanks David, It seems your the only one who can help me. When I track , -12db is the point where it peaks at 24/96 bit? Is db different than dbfs? At -12db Would'nt I sample less of my signal? I'm a little confused. Your input is very helpful.

[mij sang]

I have Wave Lab on a different computer (PC). A like a Different set of speakers for reference.

[DrFord]

I have been told by another member of this forum (who I believe got kicked for foul language) that proper voltage on your studio gear (meaning using it at the ideal settings) will get you a singal of roughly -18db to -16db.

One of the nice things about the new meters in PT8 is that they show you the 0db line, and then I believe 12 db above. Meaning that you shouldn't really be trying to max out your meters, you should be trying to get anywhere between the -18dbs and 0db (unity gain, as in the in equals the out). The other reason to shoot for below 0 is so that you have a chance to use makeup gain from plug-ins and other outboard devices. This will allow you to really push some devices without distorting or clipping your track. If you are recording at -3db and try to give 3db of makeup gain on a compressor like say the Fairchild, (tube amps) you wouldn't get to impart the color and tone of the plug in as much as if you turned it up until the outbound amp is really pushing.

Typically Daeron knows what he is talking about, and I have agreed with his on just about every post he has made.

I am in favor of having your tracks nice and loud individually and using up the 24 bit depth to it's fullest, but if you try to track to loud you really will find that you loose alot of punchiness especially on drums and things that are typically very transient based. Get louder tracks through good mixing and compression, so that you have the full throw of the faders to make more discerning movements.

On mastering, I used to Convert after bounce to save processing power. Now I don't need to. Daeron is again right that you should only dither at the last possible moment, and only when you are converting bit depths. I however have gotten into the bad habit of mastering in the same session as my mix. That being said my "major clients" get invoiced out to a Professional Mastering Engineer with all the great gear, which I am not and have not. You'd be amazed at the difference it makes. So another piece of advice might be to really focus on making quality, even, good sounding mixes, maybe with some brickwall limiting to help you get a better overall track volume and leave the real mastering to the pros, at least until you get to watch a pro work on a mix of yours so that you can learn what they do and why.

GL

D

[daeron80]

Thank you, DrFord, for the compliment. Some of what I know, I learned from reading your posts here, you know! (But I have been at this PT thing professionally since v 3.1, and a few years of 2" tape prior.)

Originally Posted by mij sang

When I track , -12db is the point where it peaks at 24/96 bit? ... At -12db Would'nt I sample less of my signal?

mij sang, yes, you want to be conservative with your tracking levels. I was told by a late, great engineer - whose name escapes me at the moment, oddly - anyway, he was one of these guys who built expensive custom gear for top musicians, and people would send him U-47s and C-24s to fix/mod and so on. He told me that all A/D converters have built-in limiters, and when they detect a peak rushing up toward 0 dB FS, they engage in order to protect the analog circuitry from damage. It's those limiters that kill the punch when you track hot to digital. I've never seen that info anywhere in print, but I trust the guy, and experience seems to bear it out.

When you're recording at 24-bit, you have lots of headroom to work with. Each bit doubles your resolution, so 24-bit is 256 times as precise as 16-bit! To look at it another way, each bit theoretically adds 6.02 dB of dynamic range, so tracking with peaks at -48.16 dB is theoretically the same resolution as tracking to 16-bit with peaks at 0! In reality, converters have more of a noise floor than that. But still, at 24-bit you can have peaks below -24 and still have greater resolution and signal-to-noise ratio than if you had recorded at 16-bit and maxed out. So, don't worry about maxing your tracking levels. Leave plenty of headroom, and you'll come out sounding way better than if you had pushed it.

Originally Posted by mij sang

Is db different than dbfs?

I guess you could say that dB FS is a kind of dB. dB are always measured relative to something. In an analog system, a common reference is 1 volt. In a digital recording system, the most common reference is the maximum level the system can represent, Full Scale. All levels are relative to that. Or it can be set a certain number of dB below the mathematical max. Not sure, but it sounds like they may have done that in PT 8? That would be cool.

[mij sang]

Thanks daeron / DrFord, Your information makes good sense. I do know of a good mastering studio in town. I have a private studio. Some of my songs are not worth sending out to be mastered. Is Wave Lab the only game in town for mastering? I would like to stay in the Mac world. I can buy better speakers for my PC or buy a mastering tool compatable with Leopard. I read that "Audacity" was good (free) but I just can't believe it is as good as Wave Lab. Thanks again gents.

[daeron80]

PEAK is a great Mac tool. If you get the full Pro XT version, I think they call it, you get some really clean EQ and multi-band compression plug-ins, and some other goodies. Combine that with native VST support, and you've got all you need, on the software side.

Then again, a lot of guys master in PT. You have the disadvantage of having to work with split mono instead of interleaved files, but that's not horribly inconvenient. Get the fxpansion VST-RTAS adapter, and there's a world of free plug-ins to explore right in PT, a few of which are killer.

[DrFord]

Originally Posted by daeron80

He told me that all A/D converters have built-in limiters, and when they detect a peak rushing up toward 0 dB FS, they engage in order to protect the analog circuitry from damage. It's those limiters that kill the punch when you track hot to digital. I've never seen that info anywhere in print, but I trust the guy, and experience seems to bear it out.

Here's what my input converter is
http://www.apogeedigital.com/products/ad16x.php

SoftLimit: Maximize Levels, Minimize Overs
One of the greatest difficulties in digital recording is getting an adequate level without clipping and unwanted distortion. Soft Limit is an analog peak limiter that enables the capture of an additional 4-6 dB of level without going into an ‘over’ condition. Short of purchasing a very expensive stand-alone compressor/limiter, there is currently no product on the market that handles this process better than SoftLimit.

Pro's... well it's Pro and sounds pro.
Con's... when I first got it I was like... YEA PUSH THOSE LEVELS! And then I listened back and my drums sounded like crap. Ok, so I went back and pulled the faders down and now they sounded good again. Especially when I was coming out of an MPC where the sound of the outbound converter is soooo good... why ruin a good thing. I digress...

I recently got a copy of the new T-Racks 3 Deluxe, and I would recommend checking it out if only getting a demo, or watching the videos on mastering audio. It's funny how many simple tricks out there really make a big difference. Just from watching the video I put my hand to my head and said "Why have I never tried THAT before..." Example was when they put a Tube Compressor on the master buss at 1.2:1 ratio with a 100ms or so attack and just maybe -3dbs of gain redution, if you can hear the compressor's threshold working its too much.

I have Wave Arts Master bundle, and it's cool enough, but I found the Final plug (brickwall limiter) did not to my ears sound as good at the same settings as McDSP's 4000. I would hugely recommend getting the Project Studio bundle from McDSP for any LE user, as the price is only $250 for studio staples I have been using since OS9 Pro Tools Mix Plus.

And Daeron, I learned how to tweak a tape machine once...
Although I would LOVE to get my hands on a nice 1/2" to track my mixes to.