such excellent contribution to the reputation and credibility of your camp.

I'd love to hear your in-depth breakdown of what "bit-depth" amounts to in terms of real-world scenarios where you would potentially be at risk for "losing some".

Go ahead, school us.

No, if you add a limiter, you destory the dynamics/transients regardless. Use no limiter and exceed 0dBFS digital and your signal clips.

Every mastering engineer I know sends back the file or a note to ask for a file with at least -3dBFS digital peak headroom.


Do the test yourself:
Have a file not peaking higher than -3dB in 24bit, render in 16bit - no limiting, no dithering/noise shaping. Then analyse it with a tool like InspectorXL to see the true peak. The red LED should go off since you have hidden overs.


There is a reason why 0dBFS is "the end" - it's the upper limit of your equipment. Some even starts to go haywire at -1dB, or even -2dB. Hence the -3dB headroom.

Same as with audio CD render. A lot of firms started to go as high as -0,1dB. This however doesn't keep you save of hidden overs and depending on your age/model of your CD playe, your DAC starts to barf. 0,3dB is more save, but chances are you still get overs with the wrong used limiter and much too high loudness. I tend to use *gasp* 0,5dB and not go higher than K-12 as of late. This resembles around -11dB to -10dB with a 300ms RMS meter, or about DR-10 to DR-8 with a Dynamic Range Meter.


The rest was pretty much covered by both Funkybot and cron.



Not wrong, but also not right. See "hidden overs" and "intersample peaks".




Also not quite right. What makes a "dynamic sounding mix" are the transients. If you leveled in properly and mixed accordingly as well, you have a wonderful mix with no noise.

Additionally: the Noise Floor of a 24bit recording is way below -85dBFS - barely noticable to normal ears. Dare I even say lower than -95dB, depending on your ADC/DAC of course. So if you take a headroom of 18dB and a footroom of 25dB with a 24bit maximum dynamic range of 120dB into consideration, you still have a dynamic range of 77dB(!). This is a lot considering that most music nowadays has a dynamic range of 10dB the most. Even less if run through the radio.



Yes of course, but you still miss the transients.

Example:
http://youtu.be/3Gmex_4hreQ

This one has a sublink to Bob Katz's "Loudness War" video as well (Loudness War - War and Peace)





With lossy encoding, you do. With lossless not. At least this is what I encountered.




Not of your system is properly calibrated - no volume adjustments needed.



Yes, really.
Vinyl can't utilise hot masterings like, say, Death Magnetic at -5dB RMS. The rails would be funked and you can forget that you can play it whatsoever. More suitable is K-14.

Tapes as well. Of course you can press the signal into the reds and get the "tape saturation" (due to softclipping and slight compression) effect, but it is not desired. This is why the Dolby Noise Reduction system was made in the first place: healty level, less noise.



I said it before, I say it again. There are two sides of the medal. I opted for the limitation, you might not. It's your thing.

Dude, this makes no sense at all. You do not lose anything. If you provide a 24bit export at a sensible level (be it -0.1dB or -10dB), this is more than enough headroom to make a 16 bit master without losing any resolution. Unless you're recording external sources with a noise floor, you are not gaining resolution by working close to 0 in the box. That's the only "old myth" I'm seeing here.

I'm not talking about compression (neither is anyone else AFAIK) i'm talking about peak levels.

funny to hear such a query from someone professing the importance of mastering.

Compyfox: didn't know about "hidden overs" tbh. results of stochastic variables in file rendering or inexact dB monitoring in DAW? gonna check some of my tracks with the trial version of inspectorXL.

funny it should take so long to get a rational response. show's I'm at least half right.

I love this site simply because it's filled with video game PC builders that think they are musicians and audio experts because they know how to illegally download software. Do any of you AES haters have ANYTHING professional you have contributed to the world of music or are y'all just "you mad, bro?" bedroom dwelling haters??? It's so silly it's seriously funny

professional? no but my music (both digital and acoustic) kicks your butt out of the window.

I've also contributed way more advice to this forum than you have. stop being irrelevant.

also remember guys: profession is cock sucking.

Yes. It is seriously funny. I simply can not stop laughing at this banquet of comedy. Haha. Ha.

you lose bit depth. even if inaudible in the end product, it's a greater loss than the pride in sticking to analogue-age directives.

as said, bit depth should be considered the relevant headroom.
Last edited by qa2pir on Sun May 13, 2012 11:38 am; edited 2 times in total

And your contribution to this subject so far has been..... what exactly ?

You really don't! You have 144dB of dynamic range in a 24bit signal. This is SO much room to boost before you start compromising the 96dB 16 bit gives you. I'm not sticking to any kind of analogue-age derivatives. I've never owned a piece of analog gear. I'm simply saying that tickling 0dB during 100% ITB mixes has no advantages unless you enjoy occasionally checking whether you need to turn your master down.

True that but i'm doing that for my own sake not the mastering engineer. I'm not saying you should hit close 0 but when people say you must leave some headroom for mastering the burden of proof is upon them to say why that is.

And so far i haven't heard a single technical explanation to why that is. A few good ones for my own workflow but not a single one yet has anything to do with the mastering engineer. But i'm willing to learn so bring 'em on everyone.

Same here. I'm also doing it for my own convenience so I never have to bother checking if I'm clipping. I don't have a clue why mastering engineers request this either. All I can really think of is that it's to discourage people from applying their own limiting or whatever before they send their track off. Still, if it stops them adding an extra 15 minutes of studio time to the bill for 'preprocessing' (i.e. turning it down), then I'm listening.