Why stay under 0?

Discussion in 'Mastering' started by chavick, Mar 19, 2006.

  1. chavick

    chavick Guest

    So i used Sequoia samplitude to record, mix, and master. I am never happy with my mixes when i bring them down under 0. But i burned a cd that was well into +6dB and i didn't hear any problem on my consumer cd player. what is the reason for mixing under 0 and how can i get better sound out of my mixes without paying tons for mastering?
     
  2. Thomas W. Bethel

    Thomas W. Bethel Well-Known Member

    When you hit OdBFS on the digital meter you have filled up all the available spaces on the file and your file, if you could see it, would have nothing but "1s" in it. If you go over that then you are into clipping which makes wave forums look more like square waves than sine waves. This is NOT the same thing as guitar distortion and in analog you can go over 0 DB and still not be in too much trouble. There are lots of places were you can read about this. One place is http://www.digido.com.

    The second part of your answer is. This is what a compressor/limiter is for. They keep the PEAK level at or below 0dBFS but bring up the average level which if done properly will sound louder. This coupled with some eq and done by someone who knows what they are doing will make your tracks louder. The big trick in doing self mastering is a really good monitoring setup and top notch gear. If you don't have that then I suggest you find a good reasonably priced mastering engineer and go to him or her for your mastering. If possible sit in on the session. You will learn a lot and have a direct input into the mastering.

    There is no such animal as Seguoia/Samplitude. They are two different programs and are both under the Magix umbrella of companies.

    Best of luck! There is a ton of information here if you use the SEARCH button.
     
  3. David French

    David French Well-Known Member

    Maybe he was using something other than a digital peak meter? If he went 6dB above 0dBFS he woudl definitely hear problems!
     
  4. TVPostSound

    TVPostSound Guest

    There seems to be a confusion with your metering.
    Are you talking about 0dBVU....0dBFS???

    +6 does not exist in digital as 0 is max, and everything at 0 is clipped.

    You need to mix with a VU meter to evaluate relative loudness, and a peak meter to control....well peaks.

    Buy the Bob Katz book, Mastering Audio, the art and science.
    I no longer do music, but this book taught me a few lessons for TV, and DVD work.
    Its not just about mastering.
     
  5. dwoz

    dwoz Guest


    No, he wouldn't have problems. He'd be stinkin' filthy rich. The guy who figures out how to go +6db above 0DbFS is going to have the world by the oysters.

    :)

    dwoz
     
  6. David French

    David French Well-Known Member

    You know what I mean.
     
  7. RemyRAD

    RemyRAD Well-Known Member

    I would like to see a VU meter that goes to + 6??? You must have the only one?

    The one and only
    Ms. Remy Ann David
     
  8. JoeH

    JoeH Well-Known Member

    What a goofy thread this has turned into. :wink:

    I'll add my .02 as well:

    Perhaps he means to say that for most purposes, Samplitude/Sequoia is the general software app he uses. As we all know, they are for the most part the same software. I own both (V7 Samp, and V8 Sequoia), and I frequently say "Samplitude/Sequoia" as a shorthand when someone asks me which program I use for Recording & Mastering. (Generally, they've heard of one or the other; but MAGIX only gets a puzzled look in the meantime.)

    No matter what the corporate spin is, Samplitude is really just a partially crippled version of Sequoia; there's a handful of things they've disabled to make it more affordable. In my case, I upgraded to Sequoia to take advantage of some of these things. One of them is the Algorithmix toolset, "reNOVAtor' specifially.

    I do like that +6 idea, though......and when you've been soloing like 12, 13 minutes or so, it's great to be able to turn it up to +7! :cool:

    (with apologies to Christopher Guest....)
     
  9. Massive Mastering

    Massive Mastering Well-Known Member

    Going deeper - I would *think* that anyone who ponied up $3000 for Sequoia - Even if $1000 for Samplitude, would have a firm grasp on the difference between 0dBVU and -0dBFS.

    That being said -

    Any mix that actually *sounds better* with the top 6dB simply sheared off into digital oblivion has issues that we don't know about at this point...
     
  10. chavick

    chavick Guest

    the glass ceiling

    i guess i never really learned much about what the peakmeter means. i am totally aware that it is theoretically and experientially impossible to have information over 0 in the digital realm. my question is more about the way samplitude (i also use sequoia from time to time, this is why i combine them) is set up. i heard that sequoia does something to audio that peaks above 0. it must, because i have gotten readings as high as +9 (in the red on the master peakmeter) without hearing problems. i know what digital clipping sounds like, i know what overmodulation does. is there any explanation for this or should i call the patent office?
     
  11. saemskin

    saemskin Active Member

    don call the patent office just yet.
    Consider getting your hearing checked, or reading the manual of the software you might not be using/understanding properly.

    Anyhow, is it possible this program does some sort of auto-limiting, and he's just seeing 6 db of gain reduction, which on the UAD Plim read as "red" in the input stage.
     
  12. bblackwood

    bblackwood Active Member

    With the incredible dynamic range 32bit float affords (somewhere around 1500dB), it's easy to have 'illegal' signals (anything above 0dBfs) in software like Samplitude or Sequoia. It's very difficult to make Sequoia clip internally - you have to work hard to do it. So he could very easily be increasing the gain to +6dBfs on the meters, but that does not translate to the disc - it simply gets clipped.

    Whether or not it's audible has as much to do with the density of the content as it does with his hearing...
     
  13. TrilliumSound

    TrilliumSound Active Member

    Absolutely agree!
     
  14. Thomas W. Bethel

    Thomas W. Bethel Well-Known Member

    One of my mastering clients came up with a GREAT solution to the level problem. As he put it "just raise the "0" point and we can have more room to work" I tried to explain to him it was not QUITE that easy but he was off to write his congressman or the RIAA about his rights to have more level. Anyway what is 0 dBFS is a fact not a fictional level like a VU meter is. It would be nice if everyone would use the same meters to check their levels but that has about as much chance as a snowball in he!!.
     
  15. saemskin

    saemskin Active Member

    I dont follow this, 0 is 0 right? Is this program using a different metering system? Can you/someone explain this more?

    What's the point if a cd is still going to clip those peaks and give digital distortion. Arent you better off working around that when you have the opportunity to fix it, ie at the mixing stage?
     
  16. bblackwood

    bblackwood Active Member

    Well, it's probably agood idea for you to do some research as explaining floating point math (were I really even capable) is beyond the scope of a mastering forum...

    The basic concept is this - 32 bit float is a 24 bit word with an 8 bit exponant which allows the 24 bit word to slide around over about 1500dB of range to avoid clipping. So even though 0dBfs is still full scale, there can be data above it, so to speak, with-in the confines of the 32 bit world. Since 0dBfs is always 0dBfs, anything above that will get chopped off in any fixed playback device (such as your 16 bit CD).

    It's not just metering, it's actual 'headroom' above 0dBfs.

    Clipping is one way to get level. It's everywhere. Sometimes it's more 'transparent' than limiting, sometimes not so much.

    I doubt any mixer has better tools for getting level than I do, and I suspect most serious mastering rooms can make the same claim. Making something loud isn't a part of mixing, it shouldn't be addressed there.
     
  17. Thomas W. Bethel

    Thomas W. Bethel Well-Known Member

    From the web

    IEEE 754-1985 Standard for binary floating-point arithmetic often referred to as IEEE 32-bit floating-point. A standard that specifies data format for floating-point arithmetic on binary computers. It divides a 32-bit data word into a 24-bit mantissa and an 8-bit exponent.

    http://www.rane.com/note153.html

    http://service.steinberg.net/knowledge_pro.nsf/ce35c8761a02a318c1256c8b004b87b6/7dd303e48ba0d149c12569ff0053968c?OpenDocument

    What it basically boils down to is this;

    You have more head room working in 32 bits IN YOUR COMPUTER but once you move outside of your computer or piece of equipment you run into the real world and you lose all the headroom. Headroom chopped = distortion.

    I am sure others will chime in.
     
  18. FifthCircle

    FifthCircle Well-Known Member

    People for the most part have hit it squarely... The 32 bit float in Sequoia allows for aparant levels to go over 0dbFS. In many cases, you won't necessarily hear it (I don't know how Sequoia deals with that conversion back when going to 16 bits), but sometimes you will. When you look at a bounced out 16 bit waveform where you have consistently been above 0, you will see clipping. You also get a big warning in your bounced file.

    This is a fundamental argument that I've had with the manufacturers on many occasions when asking for better metering on the buses. They say "32 bits won't clip" and I respond by saying "what happens when you go back to 16 bits?"

    For that matter, I've clipped the summing bus in Sequoia many times. Give me a multi-tracked a capella choir. When some of their chords sum together (due to harmonics, etc...), the levels can go out of control.

    --Ben
     
  19. Reggie

    Reggie Well-Known Member

    OK, so does Sequoia somehow slide the overshoots down to zero when exporting or something? Most other programs will just do the clipping thing if you export when your main bus levels are bouncing above 0, even if it sounds OK when you are listening within the program.
    Or could it be that Sequoia just does a really decent internal clipping, as opposed to other softwares I have used?
    But I guess the question remains, why in the world would you do this instead of using a nice brickwall peak limiter? As someone mentioned, the UAD PrecisionLimiter is pretty good as long as you don't get crazy with the release time. Or if you want to get even more experimental, send your mix out of your D/A converter outputs (whatever they may be) through some kind of outboard device to increase gain, and feed your hot signal back into your A/D converters - letting the level shoot into the red occassionally. This is what they mean by clipping the converters, and every once in a while it can sound not terrible, depending on the converters. It at least beats clipping within the software, IMO.
     
  20. FifthCircle

    FifthCircle Well-Known Member

    I see it more often as something that happens with playback rather than in the bounce. When I see it go over 0 when I'm mixing/mastering, I almost never hear it (save the occasional choir-YIKES!), however when that session is bounced, there is more of a chance that you will hear it.

    I think if it is just a few samples over 0, you don't hear it, but obviously if you are hovering over 0 for extended times, you'll hear it much more clearly in your 16 bit file.

    That was my point... (I'm not always clear before I've finished my pot of morning coffee).

    --Ben
     

Share This Page