Importance of 96k vs 48k

  • Thread starter Thread starter notCardio
  • Start date Start date
N

notCardio

I walk the line
How big of a difference do you think recording at 96K makes over recording at 48K (or 44, for that matter)? I'm not talking about 16 bit vs 24 bit (or 20 bit), but specificaly 24/96 as opposed to 24/48? And yes, I understand the whole Nyquist frequency thing (even though I can't spell it), but in the realm of HOME recording, how much of a difference would it make to someone on budget HOME equipment (i.e., no Avalons, Neves, Drawmers, Genelecs, etc.). Things like sub-$250 mics, sub $200 pre, or board pre, run of the mill amps (Alesis, Hafler), and budget monitors (20/20's, Reveals, etc). And here's why I ask. I always assumed that there would be a huge difference, and this was reinforced by the guys at the music stores (who want to sell you something). I therfore thought that I should automatically eliminate anything that wouldn't do 96K from consideration. Then I started noticing that there were relatively expensive (to me) Apogee converters that did 24/48, not 24/96, and I began to wonder. Would some of the new USB interfaces that do 24/48 be good enough? So, I guess what I'm asking, is would 96k make a significantly discernable difference to a home hacker whose music will more than likely never be played for anyone other than himself and possibly a few friends?
 
I have gear (through no fault of my own) that will record at 24/96, as well as 24/48 and 24/44.1.

I've _tried it_. I really can't hear the difference betweem 48 and 96, and I don't like the difference between 44.1 and 48 with my gear. So, I'm staying at 24/44.1. I prefer the results of sampling at the native release-media rate, and *not* having to go through sample rate conversion. I *can* very definitely hear SRC artifacts at the current state of the art. SRC has a long way to go before its artifacts are less objectionable.

I can easily hear 24bit versus 16bit, as most everyone here probably can- that decision is a no-brainer. But the doubling of storage costs in going from 44 or 48 to 88 or 96 doesn't bring sufficient bang for the buck for me. Speaking strictly for myself: I just flat can't hear it. Sorry, guys: skippy can't hear monster cable or 96kHz sampling. Start digging my freakin' grave...

Now, undoubtedly there are folks here who will slag me off for being stone deaf. Can't argue that; there's no question whatsoever that I'm getting old. But you folks go sample at 192kHz all you want: the results I'm getting from 24/44.1 are _orders of magnitude_ better than I ever got from printing 185nWb/m on Ampex 456 on my lovingly-aligned 1/4" half-track mastering machine at 30ips. And that's *my* benchmark. I like what I hear; that makes it unnnecessary to engage in worry about a sampling-rate gap. To quote Dr. Strangelove: "You can't fight in here, this is the war room!"

In answer to your final question "will it make a difference to a home hacker?", I have to answer "Nope." But let's hear from other folks on this one as well- I'll probably be proven to be a stick in the mud...
 
i cant hear the difference from 24bit/44.1 and 24 bit/96 though im sure there is a pretty BIG difference and would probably be more evident with better gear (converters,monitors)than what im using....if i had the resourses to record at 96khz i probably would but until my gear can show a difference it isnt worth twice the space/half the track count......
 
I agree -- I have trouble hearing a difference at 24/96 compared to 24/44.1

I mostly mixdown to 24/44.1 as well....

Bruce
 
argh!!!
I'm even more confused!
Gidge, don't u use the audiophile 24/96? Isn't that taking advantage of the higher sampling rate?
 
Thanks for the replies, guys! I know I can always count on Blue Bear for a straight answer. And thanks for making me think of something I hadn't, Gidge. I assumed that 96k would take up twice as much space, but I hadn't thought about it limiting the track count because of cpu usage. But, since both of you use the Audiophile, which is what I was leaning towards before this new onslaught of standalone USB devices, I'm assuming this means that the Audiophile doesn't HAVE to record at 96k?
 
no, you dont have to record at 24 bit/96 khz.....you can record at any bitrate or sampling rate......i choose 24bit/44.1 because it takes up a little less the resources (storage space, hard drive usage,etc).....
 
Audiophile does not have to record at 96. It will happily serve you (or your recording software, I should say) at a lower rate, if you want.
Sorry Skippy, Gidge, Blue Bear, but I do hear the difference in sound when listening to music on CD, which was originally recorded using different rates. BUT the problem is, I have trouble deciding, which one sounds better overall. One song sounds better if recorded at 96, another - 44.1. For some reason it is very song specific. Guess, I will have to play with it more...
 
"i cant hear the difference from 24bit/44.1 and 24 bit/96 though im sure there is a pretty BIG difference and would probably be more evident with better gear (converters,monitors)than what im using" GIDGE

no need to apologize...i agree......

its not just the bitsize/sampling size....its what you do with it also.....
 
Put me down as another person who records at 24/44.1 by choice.
 
Then I started noticing that there were relatively expensive (to me) Apogee converters that did 24/48, not 24/96, and I began to wonder. Would some of the new USB interfaces that do 24/48 be good enough?

-------------------------

I never really thought about this, but I would suppose a good quality converter like an Apogee at 48 would sound much better than cheaper ones at 96. I'd think it might be a good idea if it's relatively inexpensive. Still, why even go that far if it's just for you and your friends?

As far as the usb interfaces go, I would make sure they were designed around the newer, faster usb ports. Those might even have more bandwidth than firewire (or pretty close), if I'm not mistaken.
 
I usually agree with Skippy (my hero) and Chessrock, so it's no surprise that once again I find their comments particularly apt and useful.

I just finished writing a response over on the "is hardware dead?" thread in which I argued that the future will likely belong to 24/192 or better. Nevertheless, my final comment was that the key point of "convergence" in digital recording is (or should be) not a bunch of computer sound-generating and processing capability, but the A/D converter.

In other words, yes, a high quality Apogee converter like the Rosetta running at 24/44.1 will give you better quality, more musical recordings, and will hold its value better and ultimately be more satisfying, than having the latest computer, fastest hard drive, most complex software and a cheap 24/96 sound card.

Perhaps I'm just remarking on the obvious here, and if that's the case, I apologize. This revelation is somewhat new to me.

Cardiodpotent did a great job of clearly defining the parameters of his questions, which have been answered beautifully by others above. What I'll paste in below is from Nashville recording engineer Lynn Fuston, and I want to make note that his comments do NOT apply to home recording. I offer them just for contrast (and perhaps a hint of the sound quality the next generation may enjoy).

Here's the URL for his article about bit depth versus sampling rate.

http://3daudioinc.com/3daudio_hi-res.html

Best wishes,

Mark H.
 
Hero? Good grief! I'm just another HR person with an opinion...

Interesting article. No great surprises there, to be sure. However, I would also like to throw in one cynical and cautionary note: it is very difficult to base judgements on the progress of a technology upon what is heard in a company's demo suite at AES, or at any other trade show. If I'm going to listen to a comparative demo for anything other than amusement, it has to *not* be run by the marketing department of any of the participants: it needs to be done by an independent body.

I've been to a number of AES conventions, both as an attendee and as a booth- (or suite-) monkey representing a manufacturer. The demos are usually pretty amusing, and they you can bet that they always show the manufacturers' product in the best possible light. Of course they do: the AES convention is about selling product and making money, so it's not likely to be any other way. The company I worked for did not rig its demo, because at the time they had no competition: they just set a production example of the instrument up in a quiet suite and had a professional playing it, all by its lonesome. That was sufficient.

I'm *not* claiming that the DCS folks had a rigged demo there- perhaps they didn't. Perhaps their product really is head-and-shoulders better than the current industry offerings. Or perhaps they just picked their examples of the other resolutions very carefully.... But the problem is that you can't _tell_. I've seen vendors who've used every trick in the psychoacoustic book to make their product sound marginally better than their competition: the normal technique is to make your whatever 1-2dB louder than the other guy's whatever, because _everybody_ knows that "louder is better". It's fun to carry a sound level meter with you, when and if you go to these, just to know who is stacking the deck.

I'm sure Mr. Fuston reported fully and accurately on what he heard. My concerns apply only to what was made available for him to hear.

That's why folks like Harvey Gerst and other truly independent reviewers are so important to the industry for us viewers-from-the-sidelines. If they can gather products, test them, and render an opinion, I find that very useful. But I don't put much stock in manufacturer demos, or manufacturer claims...
 
We are advised often regarding equipment reviews to 'read between the lines.' I find it very interesting to compare the drive to 96k, and now 192k, with the fact that there is a fairly significant portion of the pro audio community (including some that build the stuff) that hear no improvement, and/or a difference but not better, and/or don't think there is even sound theory why it even could be any better. A lot of the reasons for it are near the realm of 'it could, or should be better', 'maybe it's because'... There is also the concern that it is opening up whole new sets of problems in the ultra-sonic regions that could make things worse if handled incorrectly. The one manufactuer that mentioned this subject indicated it was the popular market that was expecting to have this feature(s). In the mean time it seems it is we, the users who see it being offered befor us as the next greatest thing. (And surely we can trust ad agencies not to sell us anything we don't really need, right?)
One guy said it well; (paraphrasing) ...sometimes we expect the output to be BETTER than the input. This can be done, but it is not, by nature, high fiedelity.
So, what do yall' think? Is this is a pretty big priority to the implemention of your craft (and equipment budget alocations)?

...the preacher in 'Blazing Saddles' after they shot his bible up..."Your on your own boys...":)
 
24/44.1 here for all the above stated reasons. I just don't hear enough of a difference to justify allocating the additional resources.
 
I appreciate Skippy's and Mixit's points, and I should have spotted both the *potential* for a bit of trade-show tomfoolery AND the possibility that higher resolution (from a digital perspective) could bring new, unforeseen problems and not just auditory bliss. Hmmm. Much more to think about. Thanks, guys.

In contrast, though, I have to say that my very limited experience has been that product demonstrations by vendors and manufacturers are more often than not truly awful-sounding, in no small part because of their lack of control over the venue and ambient roar. It seems to be the exception if anything sounds good at all.

However, I do recall that a piece of "holographic ambience-creating" outboard equipment I purchased in the mid-80s to try to bring some life to early CDs did in fact have about a 2dB boost (in addition to some phase information that appeared to expand the soundfield), and it was the rare visitor indeed who didn't think it sounded dramatically better (even though living with it became tiresome after a relatively short time).

Unfortunately, it is the gimmicky "dsp equalization," the sizzly highs and heavy one-note boomy bass that define "good sound" for too many consumers, who will make their stereo purchase decisions based on listening to a system for less than 10 minutes. I own two pairs of Rogers speakers, the diminutive LS/35a and the more substantial Studio 1a with upgraded caps, and I think either would be hard to sell at Sears or Circuit City.

Perhaps Skippy would be more comfortable with the title of "mentor" rather than "hero." Often, you are able to help me cut through the confusion and consider the facts more clearly. To you and all like you, I am grateful and glad to be associated with you through this forum.

Best wishes,

Mark H.
 
A little info to pass on...

I'm no pro by any means but I also had this same question and asked a very reputable mastering engineer I met through a friend of mine. This guy has done some fairly large stuff for companies like Virgin, etc. Basicly I asked if I could come by and hear the difference on some final product on his gear and he, of course said I could. I never went....I should though. Anyway.........read it carefully!

His answer via email:
------------------------------------
"... as for the hype over 96k... well, yea it does make a difference
when all (or at least most) other things on a project are as close as they can be to being "perfect"... 96k will make a
difference. Will 96k make a big difference on a project recorded by most engineers at any one of our local studios?? To
be honest, probably not. Will it make a difference if someone like a Tom Lord Alge, Bob Clearmountain or Steve Albini
(not that Steve would ever use digital) is doing a mix in a really kick ass studio? Absolutely.

I've recorded some acoustic stuff at 96k and it's been really nice.... a lot better than at 44.1... that being said I just
mastered a record from dat (16bit, 44.1) that sounded really awesome... so go figure."
--------------------------------------

For what it's worth...hope it helped!



heylow
Rock Jedi/Indie Snob
www.heylowsoundsystem.net
 
It's probably at that level that only with those pesky _real_ double-blind tests at matched levels will you ever really know, human nature being what it is. :)



"...I'm with Ashley. Let's nuke'm from orbit.
....It's the only way to be sure."
 
skippy said:
I have gear (through no fault of my own) that will record at 24/96, as well as 24/48 and 24/44.1.

I've _tried it_. I really can't hear the difference betweem 48 and 96, and I don't like the difference between 44.1 and 48 with my gear. So, I'm staying at 24/44.1. I prefer the results of sampling at the native release-media rate, and *not* having to go through sample rate conversion. I *can* very definitely hear SRC artifacts at the current state of the art. SRC has a long way to go before its artifacts are less objectionable.

I can easily hear 24bit versus 16bit, as most everyone here probably can- that decision is a no-brainer. But the doubling of storage costs in going from 44 or 48 to 88 or 96 doesn't bring sufficient bang for the buck for me. Speaking strictly for myself: I just flat can't hear it. Sorry, guys: skippy can't hear monster cable or 96kHz sampling. Start digging my freakin' grave...

Now, undoubtedly there are folks here who will slag me off for being stone deaf. Can't argue that; there's no question whatsoever that I'm getting old. But you folks go sample at 192kHz all you want: the results I'm getting from 24/44.1 are _orders of magnitude_ better than I ever got from printing 185nWb/m on Ampex 456 on my lovingly-aligned 1/4" half-track mastering machine at 30ips. And that's *my* benchmark. I like what I hear; that makes it unnnecessary to engage in worry about a sampling-rate gap. To quote Dr. Strangelove: "You can't fight in here, this is the war room!"

In answer to your final question "will it make a difference to a home hacker?", I have to answer "Nope." But let's hear from other folks on this one as well- I'll probably be proven to be a stick in the mud...
Click to expand...


Amen, amen, amen. I was doing 24/48 for a long time and downsampling to 16/44.1 with noise shaping and dithering etc... all the while thinking that although there are artifacts and occasionally very slightly clipped (or boxy sounding) audio that was cleaner in 24/48, I was getting the best resolution at 48 and retaining most of it with downsampling. While I dont honestly hear a dif bet 44.1 and 48 my logic was that when you stack up the tracks the dif becomes more pronounced. In any case I never tested this theory but recently got I converted a song to 16/44.1 and it sounded like shit bec of the conversion, the cymbals lost their depth and definition, the reverbs sounded dirty, etc So Ive opted to stay with 44.1. Glad to see Im not alone.
 
I had the opportunity to mix to and monitor a recent mix I did at 88.2KHz sampling rate. What a treat!!!

More importantly, the mastering possibilities in the digital realm really opened up! The DSP was MUCH smoother to my ear, as it very well should be! I heard NO artifacts halfing the sampling rate to 44.1 using the standard Wavelab SRC, and certainly I could have probably improved that with a better SRC. Alas! No money for that right now, but will be sure to have a box for it in the future.

You DON'T have to take my word for it, but I believe 88.2 is better for music. The resulting file is better, the DSP I do in "mastering" is better.

My .02 worth on this topic. I have posted many times in the past about this issue and linked to other sites where people who are regarded as "experts" in digital like the higher sampling rates, and indeed hear the difference. I suppose for a home recording, 44.1 is just fine. But I try to get the edge where I can, and 88.2KHz capability IS an edge that I can hear.

Ed
 
You must log in or register to reply here.
Back
Top