1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

OK So recording in 96k is awesome...

Discussion in 'Recording' started by Guitarfreak, Mar 29, 2009.

  1. Guitarfreak

    Guitarfreak Well-Known Member

    Except it makes my computer cry and cut its wrists. :D It's so nice though, mixes sound brighter and livelier, and I can actually keep the bass frequencies in and not have to make super heavy cuts all the time just to attain clarity. Sorry this isn't much of a forum topic, but all my old recordings were in 44.1 so I felt I had to share it with you all. LOL

    On a side note, I've heard of people/studios getting rid of all unnecessary background activity on their macs, effectively "prepping their macs for recording" and how would one do this. Aside from the obvious, close apps, close internet connection.
  2. Link555

    Link555 Well-Known Member

  3. Guitarfreak

    Guitarfreak Well-Known Member

    Ya Rly.

    At least in this one mix I'm working on. Out of curiosity why do you ask? Or are you just breaking them?
  4. MightyFaulk

    MightyFaulk Active Member

    I think Link asks "really?" because in theory there shouldn't be a difference in bass clarity between 44.1kHz and 96kHz. Perhaps your recording techniques have improved?

    As for the "mac recording preparation" I would suggest doing just what you said: closing all other apps and stay offline. I wouldn't recommend any special "hacks" to stop other services on the computer as this could cause some stability issues. It is also important to tweak your buffer settings in your recording application as well as managing how many plug-ins you are running in the app at once. Still having issues after making these adjustments? It might be time for an upgrade. Recording with a 96kHz is more than double the processing your computer must do from 44.1kHz, keep that in mind.
  5. Link555

    Link555 Well-Known Member

    What am I breaking?

    No I was just confused because your statement is very different from my experience.

    Would you like to learn more about sampling rate and bit depth?

    Amplitube Uno
  6. Guitarfreak

    Guitarfreak Well-Known Member

    My techniques may have improved recently, but the amazing thing is that I was taking DI of electric guitar, so there was no amp settings or miking techniques for variables. Here's what happened...

    I recorded one track at 44.1k. Then had to roll off the bass at 130Hz at 12 dB/Oct and it still sounds bassy.

    I re-recorded the same piece at 96k. The resulting track was bright enough that I only had to roll off at 80Hz at 24 dB/Oct. Even then the final product was pretty bright and not bassy at all. Maybe too bright.

    I know the bass roll-offs are basically pretty similar. And I wasn't having a problem with bass clarity, just with there wasn't ever much treble frequencies. Does it make a difference that it was recorded DI?
  7. MightyFaulk

    MightyFaulk Active Member

    Perhaps recording at 96kHz is allowing you to capture more of the sonic qualities of the guitar than 44.1kHz. That's a highly debated topic. Just curious, have you tried a 48kHz sample rate on any recordings?
  8. Guitarfreak

    Guitarfreak Well-Known Member

    No, But I've been meaning to, This was my first venture to other sample rates. I plan to try out 48 and 88.2 in the near future.
  9. IIRs

    IIRs Well-Known Member

    Were you using an amp simulator? That might explain the difference. Sonic differences between raw recordings at 44.1 and 96KHz should actually be quite subtle. The biggest advantage to higher rates when mixing is that its like upgrading your plug-ins: the ones that don't use oversampling will suddenly have loads less aliasing, while the ones that do oversample will suddenly have better anti-aliasing filters with a higher cutoff.

    But sometimes you come across a plug that wasn't designed properly for high sample rates, and which sounds dramatically different at 96KHz. An obvious example would be an EQ or filter plug that assumed a 44.1 KHz sample rate when calculating coefficients: the plug would still work at 96KHz, but the filters would be set to slightly over twice the frequency they were supposed to be at...
  10. jammster

    jammster Active Member

    Hello Guitarfreak,
    I use a PreSonus FireStudio and Logic pro 8 as well. I utilize 96k recording most of the time and I do like the sound.

    When I first started using this setup I noticed a greater latency with my system at the 44.1k setting. I got the buffer down to 128 and it really started to have less latency when I kicked it into 96k.

    I now always use 96k when I am tracking by myself. If I am tracking someone else I often stick with 44.1 unless there is a special reason to utilize 96k.

    Not everyone will agree that running 96k is an improvement over 44.1. It does work the processor harder and eats up twice the memory and if you don't notice a reasonable difference then it just plain does not make any sense to utilize it.

    Just remember that when you finish your project at CD quality 44.1/16bit your going to lose that extra little bit of 'air' that 96k / 24bit brings. I think you can still achieve that 'air' in different ways. There are no "rules" other than do what is best for you.
  11. Guitarfreak

    Guitarfreak Well-Known Member

    Hmm, very wise. That's cool how the higher fidelity makes the plug-ins work better, it makes sense. I did use an amp sim on the 96k one, but after bypassing all plugs on both projects and A/B them, the 96k is still brighter and more balanced sounding. Thanks for the info. What exactly is anti-aliasing as you refer to it? And how would I know if my plugs can handle higher resolution rates? I have to go to work now, but I will def check back later tonight.

  12. IIRs

    IIRs Well-Known Member

    Thats quite a big question!

    "Aliasing" is what happens when you feed a signal with a frequency higher than Nyquist into a digital sampling system.

    Nyquist frequency = half the samplerate. So according to the Nyquist theory, the highest frequency you can succesfully represent with a samplerate of 44,100 Hz is 22,050Hz. Any signal components with a frequency higher than this will be read incorrectly and emerge as lower frequencies below Nyquist. This is Aliasing, and it sounds horrible.

    Anti-aliasing refers to any strategy used to remove* or reduce aliasing. Eg: your analog to digital converters use anti-aliasng filters to remove any components higher than Nyquist before digitizing the signal.

    Aliasing is also caused by any digital process that adds harmonics: if those added harmonics are higher than Nyquist they will emerge as horrible high frequency modem noise. The worst cuplrits are distortion or overdrive plug-ins, but aliasing can also be added by compressors and limiters (especially with fast time constants) or EQs (especially with HF boosts.) The most common technique used to overcome these problems is oversampling: the plug-in recieves its input at 44.1 KHz, then up-samples it by 2x or 4x or 16x or whatever multiple the developer deems neccesary so that the processing is occuring at a much higher samplerate. This means that some, or most, or all of the aliasing occurs at frequencies higher than 22,050 Hz, and can therefore be removed by anti-aliasing filters before the signal is downsampled back to 44.1 Khz.

    Hope that makes sense...?

    * of course I meant to say prevent, not remove. Aliasing cannot be removed once it has occurred.
  13. Codemonkey

    Codemonkey Well-Known Member

    roll off the bass at 130Hz at 12 dB/Oct.
    roll off at 80Hz at 24 dB/Oct.

    You go on to say that "the rolloffs are pretty similar".
    If doubling the rolloff and lowering the frequency it occurs at by almost a full octave is "similar" then I guess so.
  14. Guitarfreak

    Guitarfreak Well-Known Member

    Yeah, I never took frequency theory. I meant that they both have the same endpoint and just the slopes and starting points change. But other than that...eh?

    Wow, that's a lot of info IIRs. Is there any way I can hear aliasing in action? Like an A/B with and without aliasing. Or how can I create this myself.
  15. MadMax

    MadMax Well-Known Member

    Listen to a low bandwidth/high compression MP3 (hell... ANY eMPty3 for that matter) and/or create an MP3 of the same characteristics... compare to a full res WAV.

    Granted, it's not exactly the same aliasing, but in principle, the algorithms are close enough to illustrate the effect.

    It's a mathematical representation of data that doesn't exist, but becomes a known data plot due to transients beyond Nyquist.
  16. IIRs

    IIRs Well-Known Member

    The easiest way to hear aliasing is probably with a very simple virtual analog synth: a 'perfect' sawtooth wave would in theory have harmonics extending to infinity... an analog synth would have its bandwidth limited by the quality of its components, but digital works differently; there is no problem creating a 'perfect' sawtooth shape, but the infinite harmonic series then gets folded down into the audible frequency range as aliasing.

    You may be able to find a synth that offers switchable oversampling (try ASynth: http://antti.smartelectronix.com/ ) or you could use any simple free synth and switch sample rates. Play a raw sawtooth wave up high with no filtering, and listen for all the nasty chirpy artifacts. Then increase the sample rate, or the oversampling setting, and hear how much purer the sound becomes.

    Aliasing also occurs visually btw. Think of old western films with waggon wheels that appear to turn slowly in the wrong direction: this is because the frequency of the wheel's rotation is greater than half the camera's frame rate. You can see the same effect here with helecopter blades.
  17. bobbo

    bobbo Active Member

    i've tried 96k recording when i had my 896hd setup as my main interface and output, and was still using the crappy ass bigknob for monitor controlling, and i use adam a7 monitors, recording into dp5.13. and i didn't notice a noticeable change from my normal 48khz 24bit setup. and i actually almost preferred the 48khz setup. the one thing i can say though about sample and bit rates is that i def need/rather record at 24bit, not for the sound quality but for ease of mixing, and i had only realized that when i had accidentally recorded a session in 16bit 44.1. so thats the only issue bitrate/sample rate hands on info i can give.

    now the big change in what i heard back from recording/mixing, was getting my new monitor controller with built in d/a converter, which opened up a whole new way of hearing and mixing. and soon i hope to be getting the analog summing going as well for better mix clarity.

    and i'd also like to bring up, about the whole bass frequency thing, why are you cutting so much bass out on the final mix, if you have bass problems, you need to either lay off adding bass to individual tracks by using a parametric eq and sweeping around to find the bad bass build up, or setting your monitor speaker's "room" shaping settings to be bassier so that you don't turn up the bass too much when tracking/mixing.

    i also feel too, that only very professional recording facilities (or very high quality smaller setups with very high end ad/da converters, mic preamps, mics, monitors, room treatment, and rooms) should use anything above 48khz. recording in a higher quality sample rate isn't going to give you a better "sound" its really more likely if anything, point out your setup's flaws.

    what preamps are you using, just the interface preamps? what kind of a/d d/a do you have, just the stock interface's? how about the monitor speakers you're using, are they very "hyped" speakers, made to sound "good" instead of sound flat and honest? how about your room, is there a build up of bass where you were tracking? did the bass player just crank his bass knob and turn down the highs and mids to get a more "ballsier heavy sound" or did the guitar players turn their bass on their amps all the way up to sound more "brutal". poor settings on the instruments can result in bad finished product.

    you're obviously doing this for just a hobby for yourself and maybe some preproduction for a band you're in, so as long as you can hear whats going on, then recording at higher sample rates is extremely pointless. save it for when you go to the "real studio" where they have real gear, and not an 8 channel presonus interface...
  18. Guitarfreak

    Guitarfreak Well-Known Member

    Wow, that's quite the post. Yes, since I posted this a while back I realized that it was my amp/room that was causing the bass problems. Since I got the amp off the ground I have been able to get much cleaner signal going in. Currently I'm only using my PreSonus FireBox and the pre's that came with it.

    Yes you are correct, my speakers are hyped. They're Altec Lansing. Not bad for listening but for tracking/mixing...NO. I am debating putting some money down for a nice set of KRK 8's, but I'm afraid that after I make such a big purchase like that I might get buyer's apathy and get bored with recording. That or I'll feel the need to get a whole bunch of other stuff to go with it.

    lol. 8 channel please, mine only has two lol, although I somewhat regret it and think maybe I should have gone with the 8 input FP10 for a hundred or so more. :?
  19. intchr

    intchr Guest

    When you're recording, how are you controlling your volumes? When I first started out I thought the general idea was to get my levels as close to 0 dB as possible, and did this for a long time. I had however failed to realize that multiple tracks and the phenomenon of amplitude doubling among frequencies were always making it an up-hill battle in the mixdown, and I could eventually get the mix to sit right but not without some pretty drastic EQ work and cannibalizing my headroom via the use of compression.

    After a good amount of research however I started to record and pre-mix all my levels to around minus 6-8 dB and that's been immense in improving the overall quality of my mixes. That gives every frequency on the board room to expand, which has improved the dynamics of my tracks and allowed each track to retain more of its original characteristics since my EQ nudging doesn't have to be as fierce. Moreover, if you're looking to force a certain track up front, let's say a vocal track for example, you've got that wiggle room to make a solid 3.5khz bump and push it ahead of your other tracks, without putting yourself at risk of having a multi-band compressor duck anything else in that range significantly later on down the pike.

    If you know this stuff already then awesome, maybe that will help someone else though. Ciao!
  20. audiokid

    audiokid Staff

    intchr , great advice.

    Good topic.

Share This Page