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

Sample Rate Question

Discussion in 'Recording' started by Evil_Dave, Mar 21, 2009.

  1. Evil_Dave

    Evil_Dave Guest

    Having some problems with my mix downs… talked to a few people and I am starting to think it’s the sample rate conversion. I run a MOTU 828, pretty nice 3.36GZ duel core rack mount Audio PC, have Mackie board as preamps, I record with stereo matched pair of NT1As as overheads and vocals, SM7S, Beta 52, Audix D6, so its not the microphones, board, computer or preamp where this sound is coming from. It seems to be coming from the conversion.

    Just wondering because we recorded our demo in 96K on Neunedo 2.2 and I wondering if we need to retrack it either at 44.1 or 88.2 so that the dither does not effect the sound.

    I am thinking when a recording at 96K is converted that portions of the signal are lost in an uneven manner and recording in 88.2 would be better as there would not be a rounding issue when the dithering the files sample rate. Me and my band mates I trying to figure this out.

    I think this is what is causing a digital background distortion in my mixes off of my DAW system. Let me clarify not overload noise but almost like circuited noises really faint in the background, this noise is not present in the 96K recording when mixed and isn’t present when we record onto tape. So does conversion from 96K to 44.1 cause this type distortion in the mixdown.
     
  2. Greener

    Greener Guest

    I had a lot of issues with down-sampling, definitely go with 88.2 if you're going back to 44.1.
    I thought the world of 48k until I had to make 44.1 CDs, it did my head in.

    A trick that can be used to get from weird sample rates to something usable is to utilise bonus quality converters to analog then re-sample at the desired frequency. mastering Eng's do this a fair bit when they convert to analog and use high voltage boards, use analog outboard gear then re-sample the final product at the target rate.

    Just remember, 48Khz and 96Khz are for video not audio.
     
  3. Greener

    Greener Guest

    I should point out that ME's do the re-sampling with deluxe equipment and you may not be too happy adding another conversion with cheap gear, this will add noise.
    /Edit: Also, dithering is noise shaping, not down-sampling. They are two different things.
     
  4. Boswell

    Boswell Moderator Distinguished Member

    If you've got your demo recorded at 96KHz 24-bit, choose a mastering house that can deal with that rate and wordlength (i.e. almost any that isn't some guy with a PC in a basement).

    The ME will down-sample to 44.1 KHz and dither to 16 bits for CD writing using better equipment and algorithms than you will have access to.
     
  5. BobRogers

    BobRogers Well-Known Member

    Before you send this to an ME, I'd do some experiments on your own. That's the whole point of owning all of this equipment. See if you like the results better recording at 44.1 or 88.2. And remember, you do expect it to sound different (and worse) when downsampled to 44.1 from any higher sample rater or you wouldn't have wasted all of that disk space and cpu power in the first place. Of course, if your goal is to sell CD's the bottom line is what they sound like in the final 44.1/16 copy.
     
  6. Evil_Dave

    Evil_Dave Guest

    So what I am getting is, uless you are going to have mastering that can convert the sample rate properly that its actually almost pointless to record at a higher sample rate than 44.1 because that will be your final result in the mix.

    With this is mind why would you even make 48, 96 or 196K an option on your basic sound card and not tell people stuff like. For me this mean retracking the entire CD at 44.1.
     
  7. BobRogers

    BobRogers Well-Known Member

    I'm not saying that at all. First, I really find it hard to believe that recording at 96 and then converting to 44.1 (if done properly) would sound worse than recording directly to 44.1. It might, but that's for you to decide. A great many people are convinced that it sounds better to keep everything at as fine a resolution as possible for all processing until the final downsampling (and not just people who use top of the line equipment to do the job).

    What I am suggesting is that you record some reasonably complex sound that you can reproduce as exactly as possible at different sample rates. Downsample them all to 44.1/16; put them on a CD; push random; listen. This is the most important way to compare the sample rates - at the level of the final product. Unless you are going to invite everyone into your studio this is the way they will hear it.

    I will be surprised if you find that the source recorded at 44.1 sounds better than the recording at 96 and downsampled. But you may decide (as I have) that it's not worth recording pop sessions with large track counts at anything higher than 44.1. I do record acoustic/classical music (where I'm using 4 mics + maybe a few more spots) at 88.2. To me the benefit of the higher rate is small. In the classical recording the cost is also small so I do it. In a large multitrack recording the cost in disk space, cpu issues is higher. The difference in sound isn't worth worrying about plugin counts and buffer sizes.

    The whole point of investing in all of this equipment rather than leaving these questions to the pros is to be able to use your own ears and make your own judgments.
     

Share This Page