Increasing Samples Quality

Discussion in 'Mixing and Mastering' started by MaXe, Jun 30, 2018.

  1. MaXe

    MaXe Member

    Joined:
    Aug 30, 2016
    Messages:
    62
    Likes Received:
    18
    Assume I have found a bunch of samples that are not crap -they have a decent quality- I want to make a track with them. How can I increase the quality of those samples except using fair amount of noise reduction ( if the samples have some noise ). My question is, how do these producers out there get those super clean sounds? I check out sample packs and they have some high quality sounds, it is like even if they have captured those sounds with high quality equipment ( That I doubt all of them have high quality - expensive gears ) how do they increase their sound quality to sound clean?
    When I load commercial patches in synths commonly used for different genre I here lots of noise and unwanted frequencies. Part of it may be fixed with subtractive EQ but again, it does not compete with those quality samples Professional producers use.
    I want to increase my sound quality. If you know sound processing tools that I might have missed or you have tips for me that you may not want to say it over here tell me in PM.
    Please enlighten me, may god bless you!
     
  2.  
  3. MaXe

    MaXe Member

    Joined:
    Aug 30, 2016
    Messages:
    62
    Likes Received:
    18
    Any ideas?
     
  4. mercurysoto

    mercurysoto Audiosexual

    Joined:
    Nov 23, 2011
    Messages:
    1,108
    Likes Received:
    855
    Location:
    The bottom of the rabbit hole, next to Alice's
    Some noise is inherent in sound, it’s just the nature of the beast. I don’t think sounds can be upsampled (there can’t be added anything non-existing in the first place). What you might appreciate in commercial mixes is sounds that are well blended in a mix so that the noise and unwanted frequencies get eq’ed out and masked by the other elements in the mix. The use of saturation and exciters create some harmonic content that will make a track stand out. Transient shapers do that, too. I think the sample sound is only one part of the equation.
     
    • Like Like x 2
    • Agree Agree x 1
    • List
  5. dbmuzik

    dbmuzik Platinum Record

    Joined:
    Apr 15, 2013
    Messages:
    393
    Likes Received:
    194
    Location:
    Detroit, MI
    It's perception of quality. A 24bit kick drum could be perceived better by the ear at 14bit. Reverbs and delays have a tendency to increase the perception of quality as well. Isolating lead instruments by cutting a lot of the lows "and highs" so they have warm, tight, and unobstructed clarity is another example. The list goes on. Post editing is key.
     
  6. relexted

    relexted Member

    Joined:
    Jun 18, 2011
    Messages:
    165
    Likes Received:
    19
    Location:
    Amsterdam
    That super clean sound starts with proper room acoustics.
     
  7. Baxter

    Baxter Audiosexual

    Joined:
    Jul 20, 2011
    Messages:
    2,145
    Likes Received:
    1,046
    Location:
    Sweden
    Great acoustics, great mics, great pres, great outboard/HW, great converters, layering, good usage of width and depth, M/S processing, etc.
    Basically most "pro samples" have gone through a sound design-, DSP- and mastering process.
     
  8. dv8r171

    dv8r171 Member

    Joined:
    Sep 24, 2016
    Messages:
    35
    Likes Received:
    11
    It depends on the sample sound. Are they short transients like drum sounds? Gate them. Are they longer acoustic sounds like pianos guitars? in those cases some noise is good. Some samples sound more natural with a certain level of noise. Spend more time worrying about the actual music than the samples. If the song is great, no one will give a shit about low level noise.
     
  9. Tyler Fingerle

    Tyler Fingerle Member

    Joined:
    Jun 9, 2018
    Messages:
    23
    Likes Received:
    18
    If you are resampling the samples in a sampler, then you are going to lose quality and possibly add noise the further you travel away from the original sample in pitch. This also goes for changing a sample's timing (especially if you are changing it's timing and wanting it's pitch to remain the same)

    If you want a good sampler then check out Izotopes Iris 2.

    As for increasing the quality of a sample, the only way that I know of to save or increase the quality of a sample is to mic up one of your monitors and re-record the sample at a higher sample and bit rate, but I only do that when I am absolutely desperate because it takes a lot of EQing (I really wouldn't recommend this unless you have a good mic that you are familiar with, and a room that is acoustically dead, or else you will just waste a lot of time. I really only do this for when I need to transpose a sample further than the original sample will allow. Sometimes I can get it to go a few more semitones, or increase it just a few more beats per minute).

    Just simply increasing the sample rate of a sample won't ever work because when the original sample rate was set; at exactly half that value is where the lowpass filter is set in order to prevent aliasing (for instance if a sample rate of 44.1khz was chosen then the sample only retains the frequencies below 22khz, and 48khz retains everything below 24khz, etc...), and increasing the bit depth of a sample might not develop unwanted noise at first, but it almost certainly will when you begin re-sampling it (and best case scenario, it sounds exactly the same...)

    Most artists develop their clean sounds through vigorous mixing and mastering. Try your best to avoid compressing the samples (this includes limiters, which is the same thing as compression), as compressing and limiting turns up the volume of your sound floor, bringing any subtle noise to an audible level, this also causes you to lose dynamics which is what most people mean by when they say "it sounds clean."

    Your dynamics are the difference between the levels of your quietest frequency and your loudest frequency. Compression simply turns down the volume of your loudest frequencies, which causes your main sounds to get lost in the mix and can lose their place as being the center of attention. You begin to develop all types of distortion as well as a lot of white noise.

    Also, it is important to note that artists don't always use samples. The only samples I use are things live vocal chops, pre-drop vocals, kicks, snares, etc... but when it comes to things like leads, synths, basses, pads, live instruments, and lead vocals: nothing beats creating or recording them yourself.

    Also, one last thing to note, when you are eqing your samples... Try turning them up as loud as you can without it red lining at any point during the sample, and then try EQing the sample by turning down some of the frequencies that are preventing you from turning it up any louder (sometimes you'll have to automate your EQ as levels can change) this is a good way to help increase your dynamics. The key here is that the sample still sounds good, but also allows you to turn it up just a hair more, which gives you some headroom to do with as you please.

    Oh, and make sure that you correct any DC offset, and filter off the lowest frequencies that your are not using (which should always be anything at least below 30hz, sometimes higher) these frequencies can cause you trouble when you begin distorting, and compressing your sounds

    I probably left something out but whatever, this book is long enough haha
     
  10. m9cao

    m9cao Producer

    Joined:
    Oct 3, 2012
    Messages:
    293
    Likes Received:
    78
    re-record those samples. use high quality analog signal gears, then record those sample at final stage using high quality outboard recorder which has best level ADC. the high quality analog signal lines has good fidelity, but it could filtered some noises or something others unwanted stuff or some usable stuff, high quality signal cable can also filtering unwanted but you should waste some money to find which one is not the fake consumer hifi waste.
    all digital procedures, plugins, and low quality outboard gears should not works.
     
  11. No Avenger

    No Avenger Audiosexual

    Joined:
    Jul 19, 2017
    Messages:
    1,124
    Likes Received:
    791
    Location:
    Europe
    Hi guys, I admit I don't use samples, so I don't have any experience with it, but I wonder how you will really enhance the quality. If you have, for instance, a 8bit sample with 22kHz sampling frequency, how will you make it sound like a 24bit sample at 44,1kHz? Or how to make a 128kbps mp3 sound like a wav with 16bit at 44,1kHz? I don't know of any procedure with which this would be possible.
     
  12. Tyler Fingerle

    Tyler Fingerle Member

    Joined:
    Jun 9, 2018
    Messages:
    23
    Likes Received:
    18
    I somewhat tried to answer this in my above post, but I can elaborate. Unfortunately, no body can really give you a good answer on this because when things are mixed down in a lower quality, you are losing data. As I explained above, sometimes you can save a file (especially if your concern is the bit depth) by simply playing it through your monitor, and recording the sound from your monitors. By doing this, you are essentially deleting it's bit depth since bit depths don't exist in sound pressure waves emitted from your speaker cones. You can then re-record it at a higher bit depth, and sampling rate... However, there is a downside to this... It takes a lot of prep, and you risk introducing background noises, new phases, and reverb that weren't originally there, but if you are desperate and want to try this technique, here's how:

    First, you have to have a good studio microphone, but you also have to be familiar with the pick-up on your mic. If you want to check the spl pick-up on your microphone, play a pure sine wave through your monitors, and automate it so that it plays from 22khz down to 29 hz at a consistent rate (the slower it goes, the more detailed it will be), you will then record this and what you will most likely find is that your microphone doesn't pick up all of the frequencies at the same level. You will want to create a pre-set equalizer to adjust this so that the levels are more even all the way across. (This is correcting both the microphones pick-up as well as your speakers spl [or sound pressure levels] so that the sample played retains it's levels.)

    next you will need a room that is acoustically dead. This means that your sibilance feedback from things like reverb and background noise is minimal. You can test this by playing a quick transient sound with a quick attack and release, recording it, and zooming in to see if the recording has a reverb like tail that isn't in the original sample.

    Next you have to consider if the sample is stereo or mono, if it is mono then you only have to record it once. If it is stereo, then you need to record each channel separately in mono (while making sure that absolutely no variables change between recordings such as moving the mic or adjusting levels, and ect...), and then combine them together in your DAW (and FYI any misalignment will cause some serious phasing issues, so precision is super important.)

    It's important to remember that if this is done correctly, the sample won't sound necessarily better, in fact your goal is to make it sound just like the original, however, the file will be saved in a true 24 or 32-bit format, simply converting a sample from 16-bit or 8-bit to 32-bit free float doesn't truly make the file 32-bit (while the file will sound EXACTLY the same as is) if you try to resample the sound by either increasing or decreasing it's pitch, timing, or both, it WILL generate more artifacts and distort it's original formants and harmonics much quicker than the original file (this is why I never ever adjust the bit depth on samples before I make any timing or pitch adjustments.)

    This technique will not make MP3s sound better (and most likely, long samples such as entire songs are not recommended for this.) and this won't salvage samples that are such a mess that they sound like sh*t... This is simply a technique I use if I need to seriously edit a sample's pitch or timing more than +/- an octave, or more than +/- 10 BPM... It can sometimes allow you to push a sample just that much further while retaining it's fidelity in sound.

    Most of the time, this isn't worth the effort, but I have come across a few instances where I will have a stem file of a vocal track saved in 44.1khz 16-bit, and I will re-record it at 88.2khz 32-bit so that I can pitch it up a couple octaves and keep it's original sound so that it doesn't sound like a chipmunk. I'm sure it won't always work, but at least for me it has.

    Oh, and I know that this is off topic, but i know that some people wonder why I always use sample rates divisible by 44.1khz... The reason for it is because the sample rate of 48khz was purposefully designed to be as incompatible with the original 44.1khz sample rate due to patent wars back in the day. If you convert a sample from 44.1khz to 48khz, even with today's computers, the impossible sample rate needed in order to perfectly convert a sample from one to the other is 7,056khz which is RIDICULOUS. I have chosen to always stick with 44.1khz, because it is divisible by significantly more numbers (1,2,3,4,5,6,7,9,10,12 and even though it's not truly divisible by 8: 44,100/8=5512.5, which is far better than 48,000/7 which is directly responsible for their incompatibility...) however, if i had it my way, i'd say that we should all use the sample rate of 55,440 since it is an anti-prime number... meaning that it has more divisors than any number below it. it is divisible by everything from 1 to 12 (it stops at 13...) so yeah... you learned something.
     
  13. No Avenger

    No Avenger Audiosexual

    Joined:
    Jul 19, 2017
    Messages:
    1,124
    Likes Received:
    791
    Location:
    Europe
    @Tyler Fingerle, thank you very much for your effort and long and detailed answer. :bow:
     
  14. Tyler Fingerle

    Tyler Fingerle Member

    Joined:
    Jun 9, 2018
    Messages:
    23
    Likes Received:
    18
    lol I wish I could sum things up, but there's such a fine line between leaving out important details, and putting in so many details that no body is willing to read the posts... That's why I prefer to make videos, because audio is just too complex.
     
  15. Matt777

    Matt777 Platinum Record

    Joined:
    Oct 17, 2015
    Messages:
    446
    Likes Received:
    252
    Interesting reading from all of you guys :bow: ..and @Tyler Fingerle, I managed to get through ;), so thanx!

    I have to agree with @dbmuzik though, that perception is really important. Like if you take a filter sweep in e.g. classic housey tunes, that basically degrades the (original) sound.. but when bypassed, it comes out super clear. Same with rev. dry to wet, mono to stereo, the "phone" freq. fx, aso.. contrasting that is def used a lot in productions (BT comes to mind, as one of the "magicians"..;)
     
Loading...
Similar Threads - Increasing Samples Quality Forum Date
Increasing DAW Speed tips and tricks. DAW Feb 23, 2016
Retaining levels when increasing FX dry/wet balance Mixing and Mastering Sep 5, 2011
Question about multiple articulations as separate samples Kontakt Jul 2, 2018
Orange Tree Samples Sale (including New Libraries) Software News Jun 21, 2018
Virtual Acoustic Guitar Showdown: Orange Tree Samples Evolution Kontakt Instruments Compared Software Reviews and Tutorials Jun 8, 2018
Loading...