1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Logic 8 44.1 samples and 96 KHz Audio

Discussion in 'Logic 8' started by HKC, Apr 16, 2009.

  1. HKC

    HKC Senior member

    I wonder what people do when working with 44.1 samples while recording with a higher bandwidth. Is it even possible to playback files recorded in 96 KHz along with standard sample files from the EXS. I suppose that some would say that 88.2 will do the trick but since ýou can´t go directly from 88.2 to 44.1 I would be surprised is you could go the other way without any loss of quality.
    .....or do people simply stay with 44.1 because of this. I was planning to upgrade my converters until this crossed my mind.
     
  2.  
  3. Pete Thomas

    Pete Thomas Administrator Staff Member

    My EXS samples are all kinds of sample rates as many are imported from very old AKAI S1000 library. They play fine at any sample rate on Logic.

    At least the 44.1kHz samples won't sound worse with a 96kHz project! (I hope)
     
  4. alienimplant

    alienimplant Senior member

    I don't think it's a big deal. There is no 'loss' of quality, just nothing gained. And if you then process with effect, then you are benefiting from the higher sample rate because of the reverbs and other effects.

    On a side note, a lot of pro studios use 88.2 instead of 96 if the target output is a CD at 44.1 because there is no interpolation, just a 2 to 1 bit drop. I would only use 96 if your project is targeting DVD instead (96 is 2x 48, so your final DVD would be either 96 or 48, which are standard rates for DVD).
     
  5. Pete Thomas

    Pete Thomas Administrator Staff Member

    I didn't know this was still considered an issue. I would have thought the result of downsampling from 88.2 would be the same as from 96. Maybe the sums are more difficult, but how does it affect quality?
     
  6. alienimplant

    alienimplant Senior member

    I believe it is still very relevant. Think about the logic that must be applied in the conversion. There is little adjustment to the bits in the 88.2 to 44.1 conversion. There is no interpolation—at the most, some simple averaging between the first of each pair of samples and its counterpart. It's very easy, neat and tidy math.

    But 96/x = 44.1, x = 2.1768708...

    You will never in a million years convince me that this conversion is sonically equal to the 88.2 to 44.1 conversion, as the 96k conversion requires that all the bits be recalibrated to the new sample rate timeline. It's a far more complex and less accurate conversion.
     
  7. Pete Thomas

    Pete Thomas Administrator Staff Member

    Sure, it's more complex and I won't try to convince you that one is better than the other. But I don't think complexity of the conversion necessarily relates to the accuracy. But maybe it does.

    It all comes down to what your ears tell you, I know I couldn't tell the difference but I totally respect that you may be able to (I'm getting on a bit now and I'll be the first to admit my ears aren't what they were!)
     
  8. alienimplant

    alienimplant Senior member

    Agreed. I just thought it was interesting that industry veteran engineers often state in magazine articles that they are opting for 88.2 for CD projects. It also makes sense to me logically. I've never done an A to B comparison myself, so it may be moot as you say. As I get my digital studio more polished sounding though, I can definitely hear the difference in dithering algorithms, for example. My ears are becoming more sensitive to digital artifacts, so I think I'll opt for the 88.2 mentality unless I'm doing a film score. But I'm still running at 44.1 currently. And to be honest, my 44.1 to 48 conversions for DVD sound fine to me, but again, I haven't been too worried about it. One day soon, I'm sure I'll care a lot more. I anticipate that my future optical media releases will be in 5.1 surround and my stereo releases as MP3s and the like. So I might be recording at 48 or 96 very soon. Regardless, my final mixes from 44.1 are on par with major label releases recorded at either 88.2 or 96k. As you said, it's the ear and what the engineer does with his ears that matters most.


    But between you and me, my CDs are usually heard on better systems than my DVDs. I haven't made my video surround system a priority yet. I paid $99 for it, lol.
     
  9. HKC

    HKC Senior member

    So there is no problem with playing back 44.1 samples through converters that are set to 96 KHz (or 88.2). Can you also mix files that are recorded in diffenrent frequensies then. Like a few 44.1 mixed with a few 48 etc.
     
  10. No, that won't work. Softsamplers do the conversion internally, so you don't need to worry, but playing back one and the same audiofile on an audio track at 44.1 or 48 will result in pitch and length madness (a 44.1 file will be higher and faster under 48, obviously).

    - Sascha
     
  11. leytonnz

    leytonnz Senior member

    i wish i could find the posts and articles that relate to the whole 88>44 vs 96>44 conversion thing.. but the ones ive read say that the maths is not more complex, its just maths ( we just think its less complex to divide by 2).. and doesnt make a difference.. but as pete said-let the ears decide.
     
  12. HKC

    HKC Senior member

    Aha that´s what I thought. So the EXS will work but samplegrooves in audio won't, I simply could see how they could. It's something that I have never seen discussed before and not a bad reason to stay in 44.1.
    I mean I offen have clients who bring in a few files that they have made at home. At the moment a singer is doing preproduction in my studio and he likes to takes the work home to figure out the vocals and bring back the files to me. That wouldn´t be possible without conversion then....
    Suddenly I'm losing interest in the extended bandwidth.
    There´s so much talk about dithering, has anybody ever compared how 44.1samples sound when played back in 96 compared to their native format.
     
  13. Well, as long as you're not doing the final recordings with all these SRCs involved, there's the excellent and free Audiomove which one could use when collaborating with someone working at a different sample rate.
    Still, sticking with one common sample rate throughout a project obviously is a way less stressful thing.

    Fwiw, I'm not sure how much CPU juice you usually have left, but if you use a lot of software synths and FX, working at high sample rates can still bring the most modern machine down to its knees rather easily. For instance, I have a few (gorgeous) Zebra 2 patches that almost eat up one of my 2.4 GHz cores when played with a little polyphony in live mode. They would become next to unuseable at any higher sample rate (and Zebra is pretty fair on the CPU).
    Yeah, especially with some soft synths you will clearly benefit from working at higher SRs as you'd basically avoid a lot of the chances for aliasing to occur, but most recent synths offer proper internal interpolation algorhythms, so they're usually fine at lower SRs, too (unfortunately this is not exactly true for the EXS, which suffers from pretty bad interpolation, especially on patches not using much keyzones).

    - Sascha
     
  14. HKC

    HKC Senior member

    I'm glad that I thought of this before upgrading my converters. Recording in 96 KHz would make life more difficult for me and since it costs quite a bit to upgrade my converters and very little is gained by doing so (except for a little reduced latency) I have decided to stay with what I already have.

    BTW I just tested the RME QS8 and I find it really difficult to tell the difference between 44.1, 96 or 192. I tell myself can hear a little something but it´s simply not possible to say if it´s better or not.
    Anyway the end listener will still be listening back to it in 44.1/16 or more likely MP3 so we're at a point where technology has become better than what is needed and I for one won´t pay the price (yet).

    Thanks for the input, it has saved me 4500 euros.

    PS I have to say that this list works far better than I had expected, nice work.
     
  15. alienimplant

    alienimplant Senior member

    Ah, interesting.
     
  16. alienimplant

    alienimplant Senior member

    I agree. I'm still working at 44.1. I just track any live audio at 24 bits. I think there's a lot of hype over this issue. I don't mean to contribute to it. I just can't help but wince at the theory that a 96 > 44.1 is equal to an 88.2 > 44.1 conversion. I'm trying to let that seep in my thick head now.

    The only time I've ever heard that 96k made a lick of difference in critical listening tests among top engineers was when classical music or music with similar timbres were being recorded and mixed. Regarding synth music, feedback was completely diversified and lower sample rates often won listening tests.
     
  17. jlproductions

    jlproductions Member

    Yes I have also read (not sure quite where at this point) that the whole concept that downsampling to 44.1 from 88.2 involves less math than 96k is a falacy....however I am no mathematician and it is interesting hearing your thoughts on the topic. jamie
     
  18. alienimplant

    alienimplant Senior member

    I'll bet 96k sounds a hell of a lot better to cats and dogs though. 44.1 probably sounds to them like 22k rate sounds to us. Just ask Subwoofer Cat!
     

Share This Page