cerberus wrote on Sat, 13 January 2007 03:12 |
i believe that we are better off making decisons by listening to the musical results of all our actions, not watching for whether a red light is on or off! that is bogus toil; and as devo has said "toil is stupid"!
|
I couldn't disagree more. When I started recording, the most "digital" thing in the studio was an Atari ST, an SP12 sampler, and a 16 bit DAT machine. The idea of listening exclusively to the musical results to determine our actions is a really bad idea. In the analog world, just like in the digital world, if I listed to a guitar track by it's self. I may think to myself... "It sounds good, so it must be good". But then, I add 5 more guitars, a bass, two lead vocals, 8 BGV's, a keyboard, Hammond, and acoustic.
But now... I've added all those tracks... all with their own small... maybe un-hearable noise floor. They all multiply on each other to make a horribly loud noise floor. As an amateur engineer.. I'm baffled to why this happened. Or worse... I still don't notice it, but a week later, the client calls me up to say that they are at the mastering studio and that after the ME gets it "loud" the noise floor is so obnoxious that they are going to have to start over from scratch.
However... as a well trained and experienced engineer, I know a noise floor exists whether I hear it or not. I also no that when two signals combine, they increase in volume. So, I know to maximize my signal to noise ration as much as possible when recording.
Same goes for digital. I need to know how the math is working. I need to know what it's limitations are. I need to know what the buss (sorry William... I come from the "buss" camp) does when pushed vs. not pushed. 32bit float his limitations. 48 bit fixed (PT) has limitations as well. I need to know what they are. I've also got to get this signal back out somewhere. I have a very definite limitation on my output D/A. I need to know what that is, too.
Again... one clipped signal may not sound bad... you may not notice it. But 32 clipped signals is going to start to sound pretty harsh.
I'm also a professional.. that means I have paying clients. I need to know how this is going to work, because I have to work fast. They are paying by the hour. I don't need to realize halfway through the day that my mix buss is getting smeared because I'm stressing the math. Or that I'm way to hot for my output D/A and I need to make my clients wait.
If you are working on your own stuff at home... have at it. Experiment all you want.
Quote: |
i think that lack of imagination is another reason for bad engineering... "by the book" "paint by numbers" .. sure i'll tell you what frequency is "presence".. but who cares if nobody is buying it? yeah, we have problems.
|
Don't confuse knowing what you are doing with having a lack of imagination. Knowing what you are doing and what the gear is doing sets us free to be full of fantastic imagination. Nothing ruins creativity like like stumbling upon and unknown and not knowing how to fix it because you don't know how the system, the math, the electronics, etc. works.
Quote: |
fact: analog gpes to eleven. if we want to emulate analog-like responses, then our system has to go to eleven, all the way... no bottlenecks, none of that bullsh*t is necessary.
|
Digital can go to "11" as well. You just have to know where 11 is. Don't for a second think that Analog doesn't have a ceiling. It does. And the thought that it doesn't also appears to be a trend amungst amateur amongst. Hit an API too hard.... you'll see what I mean.
Quote: |
why keep your signal below -6? you protools users are acting like chicken littles! what is so scary about the uppermost bit of a 24 bit recording? why should the ones and zeroes contained in the "most significant" bit be less valuable to us than the data from any of the other bits? bits=information... music, that is.
|
There is no "more significant bit" than another.
Quote: |
a.f.a.i.a.c...it's all part of the music, once the signal passed the a/d converter, it would be a crime to throw any of it away.
|
That's just crazy talk. Unless you have a track with 144 dB of dynamic range or better (which doesn't exist in the musical world), then you are dropping bits on one side or the other. Heck... from the threshold of hearing (for an infant that hasn't had hearing loss) to the threshold of pain, is only 120 or so dB. So, even dropping the top 6 dB is going to give you 18 dB more dynamic range that human ears can even handle.
And that's just from the A/D and D/A. The PT internal mix buss has 288 dB (48 bit) of dynamic range. So internallly... dropping 6 dB still gives you 162 dB or more dynamic range than you'll ever need. On a 32 bit float, you have 192 dB total (or 66 dB more than you'll ever need).
So... drop 6 db and be "safer" on my clipping (and if you are tracking... there aren't any floating point A/D converters) and not give up any dynamic range. Or push the envelope to get one more dB and probably end up with cumulative distortion that causes "harshness" and confusion in your mix. Your choice.
Floating point has an expense associated with it, too. Each time it has to shift it's bit range up or down, there's potential for loss. It's not 100% efficient. Just like multi-processor machines. Each time a processor has to shuffle off the work load, it loses efficiency due to the math involved. This is fact. I'm not saying that the costs outweigh the advantage. I'm just saying it's not a flawless system.... but that's not what this conversation is about, so I'll get back on track...
BTW... I, just like you, was taught from the beginning of my career to saturate all the bits. But that was also when things were 8, 12, or 16 bits. The max dynamic range of digital back then was 96 dB and the converters sucked and were noisy... so you pushed it. I thought this way up until we started discussing this a couple of months ago. I know the math, so I understood Terry and WW's arguments. It made since, from a physics stand point, so I tried it. Sure enough... it made a HUGE difference in my recordings.
BTW... I often go between PT and Nuendo. 75% of my work is PT, because I prefer it, but I do a lot of work out of a room that only has Nuendo. I tested this on both platforms. It sounds better to keep plenty of digital overhead on both systems. There's two reasons for this... one, the A/D converters are not floating point. Two, if I'm not making the floating point math kick in, then the system has to work less and less chances for error.
OK... that wa a long post, but I had to weigh in.
-Tom
PS. Please understand that the comments about amateur engineers has NOTHING to do with you (cerberus) and I am in no way implying that you are amateur or don't know what you are doing. I have no idea what your skill level is, so it wasn't remotely aimed at you. I'm strictly arguing the post.