Music Player Network
Previous Thread
Next Thread
Print Thread
Page 1 of 7 1 2 3 4 5 6 7
Joined: Jul 2001
Posts: 1,740
I
Imagine Offline OP
Platinum Member
OP Offline
Platinum Member
I
Joined: Jul 2001
Posts: 1,740
I'd be interested in knowing your results...

Dean

Joined: Dec 2002
Posts: 10
M
Member
Offline
Member
M
Joined: Dec 2002
Posts: 10
We are almost exclusively electronic instruments and samplers and audio manipulation and found no audible benefits of using 96kHz.

We simply used the same MIDI-sequence and re-recorded at different rates. Everything else setup identical to the last decimal (ProTools HD w. dithering mixer, Sony Oxford EQ/compressor, nothing else).

I know other people are getting opposite results, that o.k. But for us the discussion was over after this.

As for consumer media I think we should try to realize the potential in 44/16 CD before anything else. Very few releases seem to do this, but the few that does shows that the problem is not 44/16.

Joined: Sep 2000
Posts: 944
R
Gold Member
Offline
Gold Member
R
Joined: Sep 2000
Posts: 944
Given the choice I prefer 96 over 48 for multitrack recording.

One comparison I haven't done for myself is to compare a project done at 48 that is up-sampled to 96 right before the mix versus one that is 96 from the giddyap to the whoa.

I recently mixed a track that was done at 48 and just before printing the mix saved a copy of the session at 96 and played it back through the desk - there was a noticeable improvement in impact and clarity. Probably nothing more than better analog performance by ProTools at 96 than at 48. Still it was better.

Before that I worked on the new Bruce Cockburn CD and that was originally going to be done analog, but we had a cheap Mackie recorder hanging around that did 96, and when we checked our tracks back from both the Studer and the Mackie the Mackie (to our amazement) had a much better sound than the 2". This was not the case with the Mackie at 48.

So we did the whole project at 96, including mixing via SSL.

Comparing one track in 48 and 96 I think the difference, given good quality converters in and out, is minimal. But I definitely find an improvement for whatever reason when I put up numerous tracks together - they seem to combine better through the analog mixer.

The difference I've experienced is not subtle, current popular wisdom notwithstanding.

JW

Joined: Sep 2000
Posts: 15,398
10k Club
Offline
10k Club
Joined: Sep 2000
Posts: 15,398
I've had similar results to John's at 96, with this addendum (in my experience): when you simply sum tracks without doing any DSP, the differences are minimal. But once you start adding plugs, doing gain changes, etc. the differences between 44.1 or 48 and 96 become obvious. It seems that DSP works better at the higher sampling rates. Therefore, like JW I'd also be really interested to hear a project that is upsampled prior to mixing vs. one that was tracked at 96 from the beginning. I'd wager that THOSE differences would be very subtle.

Interesting comments about the Mackie vs. 2" - I haven't compared those two, but I have compared 2" to Pro Tools HD at 96 and still prefer analog for tracking, although I like the HD converters much better than the 888's!

Joined: Jan 2001
Posts: 2,915
Likes: 79
But who is number 1 ? ...
MP Hall of Fame Member
Offline
But who is number 1 ? ...
MP Hall of Fame Member
Joined: Jan 2001
Posts: 2,915
Likes: 79
Just did a live acoustic jazz combo using a MOTU 1296 and DP at 88.2 KHz. I concur about plug-ins seeming to sound better at the higher rates. Because I had changed hard drives and hadn't had a chance to re-authorize my Waves stuff, I was leaning heavily on DP's native plug-ins, and was surprised at the good results I got at 88.2.


"I'm just a confused musician who got sidetracked into this damned word business..." -Hunter S. Thompson

Stephen Fortner
Principal, Fortner Media
Senior Editor, Music Player Network
Former Editor in Chief, Keyboard Magazine

Joined: Jun 2003
Posts: 97
T
Senior Member
Offline
Senior Member
T
Joined: Jun 2003
Posts: 97
am very impressed with 96 on protools hd, but also at 44.1/48 impressed... compared with 888's at least. don't know about 192... don't think i can deal with the storage!!

Joined: Nov 2002
Posts: 24
J
Member
Offline
Member
J
Joined: Nov 2002
Posts: 24
i am working on a song at 192 right now...

drums 4-tracks(D112, 57, Royer 121, Coles 4038), bass (Great River DI), acoustic gtr (C12VR), elec gtr Royer 121), lead vox (C12VR), 4 backing vox (Neumann U47 feti), perc 2-tracks (Royer 121 & Coles), ...etc...

much less need for EQ period....open and summs nicely with out too much effort...

front end...Great river MP2-NV, Avalon 2022, Avalon 2055, Audio upgrades modded 160A

about half way through 4:30 song and file size is 2.5 gigs...

PTHD3, 192 I/O, AVammo 100gig 7200rpm firewire drive...on G4 533 single, 512 ram, 1 18" flat panel monitor

i think so far that drives are cheap and it is worth the extra storage space...

i also only run out 24/192 analog to Avalon 2055 - L2 Hardware - HHB CDR-850 with L2 being the A/D for the HHB for home mastering

i am doing all this at home by myself (i am not a good drummer but i am trying to get good other tracks so a drummer can play for me) .

i was going to go Radar and a sony desk but right now it is cool to try this and see if it can hold up to 2" and Radar....

it is also not bad having the limitation of 24 tracks as i think as time goes on PT will up the track count and so will Radar at the higher rates...

happy high resing....

peace john

Joined: Jul 2001
Posts: 1,740
I
Imagine Offline OP
Platinum Member
OP Offline
Platinum Member
I
Joined: Jul 2001
Posts: 1,740
Here's another question. Are there any benefits to recording at 88.2, instead of 96?

Joined: Jun 2001
Posts: 175
G
GD Offline
Senior Member
Offline
Senior Member
G
Joined: Jun 2001
Posts: 175

Joined: Jul 2001
Posts: 1,740
I
Imagine Offline OP
Platinum Member
OP Offline
Platinum Member
I
Joined: Jul 2001
Posts: 1,740
David, then why is everyone on Gods green earth recording at 48. Shouldn't they be recording at 44.1?

I've been recording 24/48 for many years without any inaccuracies.

Joined: Mar 2000
Posts: 7,950
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Mar 2000
Posts: 7,950
ive been at 48khz for a long time. nothing suprizing dropping down here either. and i rip my mp3's at 48khz source.

but lately i have had the urge to fully move up.... just have to get a HD192. i have been thinking of going to 96khz in the meantime.


alphajerk
FATcompilation
"if god is truly just, i tremble for the fate of my country" -thomas jefferson
Joined: Sep 2000
Posts: 1,277
A
Platinum Member
Offline
Platinum Member
A
Joined: Sep 2000
Posts: 1,277
David, you wrote,

"If you are targetting CD or MP3 at 44.1k, you should record at 44.1 or 88.2. Resampling from 48k or 96k to 44.1k will result in an innacurate final (putting it nicely).
Same thing for targetting 48k, use 48 or 96."

Are you just making this up, or do you have some info I don't have? I have been researching the src issue for a year now, and the vast majority of experts dissagree with your statement. Only ONE agrees, that being Dan Lavry, and only when using a synchroneous src designed specifically for src'ing such multiples. The overwhelming opinion is that, with most src's, there will be no difference.

If you have some concrete info on this, please lay it on me. I am most interested in the subject.

You also state, "And regarding 44/48 vs 88/96, as mentioned, you will get no actual "audible" gains (in that humans can't hear over ~16-20k) other than perhaps filtering differences on the convertors,"

Even if the first half of that statement is true (some would still agrue against it, despite one very long thread last year) the second half is mighty important. With most current converters, including the Digi 192,filter artifacts are quite noticeable at 44.1K. with the Digi box in particular, there is a very noticeable improvement at 88.2 or 96K.
---------------------

Of course, if you are going to mix in the box and then use an src on the resulting stereo file, you may gain back SOME artifacts. However, a non-realtime, software src has a much better chance of giving you an unobtrusive filter than any real-time AD converter.

I'm having good results with Barbabatch. I will continue recording at 88.2K, not 96K, even though both barbabatch and Digi both advise that with their respective src's there is no difference with 96K-to-44.1K. This is because I hope to eventually find an src optimized for converting 88.2 to 44.1 without bumping the numbers up first.

Dan Lavry says he used to make one, but it has been discontinued. More research needed.....

Joined: May 2002
Posts: 96
D
Senior Member
Offline
Senior Member
D
Joined: May 2002
Posts: 96
David, quite bold statements.

I record, edit and mix on my Digi HD rig with 192i/o's at 44.1, 48, and 96 nearly every day. Most sessions are 48 or 96.
There is a definate difference between 96 and 44.1/48. The top end is much more open and I don't have golden ears, unbelievable speakers, or a well tuned room. I also find I need to EQ less at 96k.
I first noticed with my Avalon737 and it's 32k airband. At 96k, I could tell audible frequency interaction a lot earlier than at 48k. i.e. 3db, rather than 6db boost.
there are many studies and test out there talking about audible interaction of octaves beyond human hearing. I really don't think it's a black and white issue.

Some of the work I do is mixed and SRCed entirely in the box. and a lot of it is mixed in the box and mastered out. So even if my final delivery format is 44.1k I want the highest practical deliverable resolution. All the professional mastering engineers I know of run the mix through true analog gear. So I want the mastering engineer to have the highest fidelity master he can. If I'm doing it myself, I run the final mix out of PT, into a analog comp (if applicable) and then analog EQ for final sweetening, then into PSX100 to hardware L2, to DAT
It's almost like saying, we're going to watch the movie on VHS so there is no need in recording to film, digi beta, or beta sp, lets just start out on VHS.

But probably as much as the audibility, I work at 96k for future compatibility, for DVD-A, DVD, etc. I would rather pay a couple extra bucks for hard drive space, and have a little less DSP, then have the material individually upsampled from 44.1 or 48k, even 88.2 up to 96k.

However, if you're working with pure electronic music, I would think the results would be hardly noticeable unless ran through outboard comps and EQs that may somehow introduce nearly non-existant transients...?!

Joined: Oct 2000
Posts: 80
D
Senior Member
Offline
Senior Member
D
Joined: Oct 2000
Posts: 80
I went 96k for a while.. With heavy distorted guitar bands the difference is not that much. So I went back to 48k.. The convertors on HD sound a lot better then the mix ones..

For some CD work I will still probably do 96k as the mastering place I use takes the 96k data files so it only gets SRC'd during the mastering.

The less hard disk space and the more plugins makes a difference..

Joined: Jan 2000
Posts: 1,032
Platinum Member
Offline
Platinum Member
Joined: Jan 2000
Posts: 1,032
Quote:
Originally posted by Allan Speers:
You also state, "And regarding 44/48 vs 88/96, as mentioned, you will get no actual "audible" gains (in that humans can't hear over ~16-20k) other than perhaps filtering differences on the convertors,"

Even if the first half of that statement is true (some would still agrue against it, despite one very long thread last year) the second half is mighty important. With most current converters, including the Digi 192,filter artifacts are quite noticeable at 44.1K. with the Digi box in particular, there is a very noticeable improvement at 88.2 or 96K.
I don't know how many people here have had the chance to record identical source material at both sample rates (48 and 96K). I have. I was tracking a project and had two 96K RADARs in the room, so I recorded the rhythm tracks to both machines, one at 48K and one at 96K. At the end of the day, I could play back both at identical levels, and with A/B input switching on the monitor side of the console, I was able to hear the same identical mix sourced from either machine.

The 96K made a difference in ways that I did not expect. On bass guitar, the difference was extremely audible, with the bottom end more solid at 96K. On distorted electric guitars the difference was extremely audible, the 96K actually sounding warmer. On acoustic guitar, the 96K sounded easier, more effortless, more open. On drums, the 96K sounded more sterile. Maybe you could call it accurate, but it was less punchy, maybe just less distorted? On acoustic piano, though I could hear a difference, it was so slight that there was no clear choice. That came as a big surprise to me.

So anyone that tells you that 96K is irrelevant because we can't hear above 20K, either A) is quoting conventional wisdom and has done no listening of their own or B) can't hear.


Lynn Fuston
3D Audio Inc
Home of 3dB
Joined: Mar 2003
Posts: 219
N
Senior Member
Offline
Senior Member
N
Joined: Mar 2003
Posts: 219
The threshold for what we hear between 48 & 96k are not inherent to those sampling frequencies. I suspect that the "difference" we hear (aside from converter performance & characteristics) is a result of what happens at sampling above 50kHz.

Ask the Earthworks guy. Or the Sweetwater guy. Or the RN dude, or the Oxford guy.

NYC Drew

Joined: Sep 2000
Posts: 15,398
10k Club
Offline
10k Club
Joined: Sep 2000
Posts: 15,398
I don't see why anyone has difficulty believing that higher sampling rates would sound better. I think as Lynn says, where people get hung up is in dwelling on the "we can't hear above 20K" part. Certainly mathematically, anything below 20K can be represented with complete accuracy by a 44.1K sample. Nor do we have to get stuck in esoteric arguments about how we can somehow "perceive" stuff above 20K even if we can't hear it. Even if we could, there's pretty much a 100% chance that something or other in the signal chain, either in the recording or the playback, is going to be incapable of capturing or reproducing above 20K in the analog stage.

The benefits of higher sampling rates have everything to do with the fact that things in the real world don't work the way they do in theory or in a laboratory. Apart from the fact that there are going to be many differences between converters in the analog stage, there are a lot of places things can go wrong in the process of digital encoding and decoding, jitter being the most obvious, poor software design in plugins and other DSP being another. Add in inaccurate or incomplete metering in some cases, multiple conversions when using analog outboard gear, pilot error, etc. etc. and it seems obvious that living right on the edge of what can mathematically represent a signal is not a great idea.

It seems kind of a no brainer to me that redundancy in sampling is therefore a good thing, and "we can't hear above 20K" and its corrollary "but we can FEEL above 20K" are red herrings.

Joined: Oct 2002
Posts: 1,646
W
Platinum Member
Offline
Platinum Member
W
Joined: Oct 2002
Posts: 1,646
Quote:
The 96K made a difference in ways that I did not expect. On bass guitar, the difference was extremely audible, with the bottom end more solid at 96K. On distorted electric guitars the difference was extremely audible, the 96K actually sounding warmer. On acoustic guitar, the 96K sounded easier, more effortless, more open. On drums, the 96K sounded more sterile. Maybe you could call it accurate, but it was less punchy, maybe just less distorted? On acoustic piano, though I could hear a difference, it was so slight that there was no clear choice. That came as a big surprise to me.

Now, to my 3 functioning brain cells, this is remarkable...Does this perhaps foretell the dawn of multiple sampling rates withing the same project? That would REALLY be confusing, but very cool...48K drums, 96K acoustic instruments, 192 vocals...

If the technology ever develops, does this mean that we will start using sampling rates selected for a cpscific voicing, just like with preamps, mics, etc in today's world? (and I just THOUGHT this stuff was expensive now...)


So What ARE We Gonna Do With 8 Tracks...Fire the Arranger?
Joined: Feb 2002
Posts: 479
Senior Member
Offline
Senior Member
Joined: Feb 2002
Posts: 479
I have tracked at 96k and 192k, it sounds fantastic, I was surprized at the difference, more headroom, cleaner, more detail. I can't say enough about using high sample rates- Just do it.

Joined: Feb 2000
Posts: 2,184
gm Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Feb 2000
Posts: 2,184
You know, I hate to be the sand in everyone's KY, and I'm well-advised to pick my battles, but...

O.K., I can often hear the incremental improvements in some sources when going from 48 to 96k and then to 192k. But...uh...honestly?? Not always. First, I don't regard the improvement as audible - or, if audible, necessary - for all sources. Secondly, I don't regard the high octave as mandatory for all uses against given delivery costs.

Don't get me wrong, I like working in wide-band - I immediately feel more at home EQing in the high octave with it's concomitant reduction in artifact (from the filters, mainly) on voices, acoustic guitars, cymbals & etc. I just think it's often an over-indulgence. For instance, I defy anyone to hear the difference in a Fender bass recorded both at 48k & 192k. Sorry Lynn, my dear friend, I really don't believe you'd be able to hear the difference between sample rates on electric bass guitar in a supervised double-blindfold test. I just don't. And, I'd be willing to put money on it. A lot of money. Any takers?

Frankly, the main perceived benefit these days of 96k and 192k recording is when you're talking audio, not listening to it.

George


George Massenburg

http://www.massenburg.com
Joined: Jun 2003
Posts: 12
M
Member
Offline
Member
M
Joined: Jun 2003
Posts: 12
George, this may be hard for you to believe, but I have a LynxTWO sound card, and truly, the lower midrange sounds very much fuller when recording at 96 vs. 44.1. The difference is closer to running the sound through a Fatso than it is to being inaudible, to my ears anyway. Don't know why, but it is.

However, I have also been listening to the recently released 3D Audio ADCD, and must admit that some converters sound great at 44.1.

Certainly a Prism AD-2 at 44.1 sounds better than a LynxTWO at 96 (and I mean across the whole frequency spectrum), but seeing as how I am stuck with the Lynx for now, 96khz it is.

Joined: Sep 2000
Posts: 944
R
Gold Member
Offline
Gold Member
R
Joined: Sep 2000
Posts: 944
This is starting to sound familiar, isn't it?

Most of what I do is at 48k as well, but when I can manage 96 I do it.

The disk/bandwidth overhead is a drag to be sure.

Although I know the technical implications, when the session is running I'm not really thinking technically. I'm thinking about making music and enjoying the process - and that's where I've found that 96 has 48 beat. I just find it's more fun to deal with the sounds at higher res.

I haven't been able to compare 192 in any meaningful way, but the one time I used it (on the tree and deep mikes for an orchestra) it certainly sounded good.

Also I'm not completely sure I wouldn't get enough of a benefit by working the tracks at 48 and up-sampling them right before mixing. The difference I hear when doing that is not subtle.

That's when using a console, rather than mixing in Pro Tools. I don't own an HD system yet, so I can't vouch for the difference in virtual mixing.

JW

Joined: Dec 2002
Posts: 109
P
Senior Member
Offline
Senior Member
P
Joined: Dec 2002
Posts: 109
Quote:
Originally posted by gm:

Frankly, the main perceived benefit these days of 96k and 192k recording is when you're talking audio not listening to it.

George[/QB]
Isn't that the truth. I find it increasingly difficult to get people to just listen objectively anymore. They would much rather argue about technical minutiae. Not just the engineers and producers, but EVERYBODY. It's no longer just whether you like the music, but if you approve of it's method of creation. Very frustrating.

Thanks George and Lynn for voicing those rather unpopular opinions. I came to similar conclusions when a/b testing identical sources as well. Marketing and popular opinion have spoken however.

Joined: Jan 2000
Posts: 1,032
Platinum Member
Offline
Platinum Member
Joined: Jan 2000
Posts: 1,032
Quote:
Originally posted by gm:
You know, I hate to be the sand in everyone's KY, and I'm well-advised to pick my battles, but...

O.K., I can often hear the incremental improvements in some sources when going from 48 to 96k and then to 192k. But...uh...honestly?? Not always. First, I don't regard the improvement as audible - or, if audible, necessary - for all sources. Secondly, I don't regard the high octave as mandatory for all uses against given delivery costs.

Don't get me wrong, I like working in wide-band - I immediately feel more at home EQing in the high octave with it's concomitant reduction in artifact (from the filters, mainly) on voices, acoustic guitars, cymbals & etc. I just think it's often an over-indulgence. For instance, I defy anyone to hear the difference in a Fender bass recorded both at 48k & 192k. Sorry Lynn, my dear friend, I really don't believe you'd be able to hear the difference between sample rates on electric bass guitar in a supervised double-blindfold test. I just don't. And, I'd be willing to put money on it. A lot of money. Any takers?
George (aka Mr. KY Sandy),

I think I still have the files on my Masterlink. I'll check. Even if not, I'm tracking on Tuesday with two identical RADARs in the same room. I could do the same test again. It was indeed electric bass.

So how much money are we talking about? I've been looking at that new GML 2020, but I'm a little short on cash. Would that be "a lot" enough money?

I'll go check for those Masterlink files now.


Lynn Fuston
3D Audio Inc
Home of 3dB
Joined: Jul 2000
Posts: 477
R
Senior Member
Offline
Senior Member
R
Joined: Jul 2000
Posts: 477
Funny how things change. Remember those 900 post horror shows a while back. I was flamed endlessly for saying screw the math, it felt better and was easier to get great stuff at 96. God bless people actually using their ears.

As for differences in low mids, etc, the thing I've found is that all of the convertors are very different beasts at the different sample rates. I'm assuming this is in the chips themselves, because the audio componentry is all the same. Anyone with any tech knowledge of such things?

The most real tangible difference is in the upper octaves. Adding some 16k presence? At 96, twist away, but at 48, you'd better be careful. I also find that dynamics are different- much more open at 96. Maybe the less hardcore filtering lets the audio through more smoothly.

my 2 cents.

Joined: Jan 2000
Posts: 1,032
Platinum Member
Offline
Platinum Member
Joined: Jan 2000
Posts: 1,032
I'll have to re-record them. They aren't on the drive anymore. That was about a year ago (maybe more) when I did that test. I'll have acoustic piano, drums, electric bass, electric guitar, acoustic guitar and keys.

I'll try to record samples of as many of them as possible. The format is RADAR with Nyquist converters. I'll do one at 24.48 and the other at 24/96. I can then transfer the files to Masterlink when we're done.


Lynn Fuston
3D Audio Inc
Home of 3dB
Joined: Feb 2001
Posts: 1,041
V
Platinum Member
Offline
Platinum Member
V
Joined: Feb 2001
Posts: 1,041
Your choice of microphone and it's placement influences the sound more so then your decision of whether to use 44.1, 48, 96 or higher sample rates. At least, that's what my ears tell me. Choice of sample rate as a way to influence the sound is way down on my list of sonic priorities. That being said, I record at 24/44.1 and mixdown to analog & digital.

Joined: Jul 2001
Posts: 709
A
Gold Member
Offline
Gold Member
A
Joined: Jul 2001
Posts: 709
from robdarling

Quote:
As for differences in low mids, etc, the thing I've found is that all of the convertors are very different beasts at the different sample rates.
Very different beasts, exactly. I even know a converter that sounded worse (less transparent) at higher sampling rates. This was 2 yeaers ago when every manufacturer offered 96 kHz converters.

From my experience, the better the converter the more difficult is it to hear differences between higher and lower sample rates.

Processing (EQ, dynamics) at higher sample rates is a different story. Especially EQs have advantages at higher sample rates - mostly. There is an EQ from Daniel Weiss which is also good at 48k.

Joined: Sep 2000
Posts: 15,398
10k Club
Offline
10k Club
Joined: Sep 2000
Posts: 15,398
Quote:
Originally posted by adebar:

From my experience, the better the converter the more difficult is it to hear differences between higher and lower sample rates.
Yep, exactly. That's why it's no surprise to me that guys like George don't hear much difference - they don't ever have to deal with the relatively shitty converters that some of the rest of us do. \:D

Quote:

Processing (EQ, dynamics) at higher sample rates is a different story. Especially EQs have advantages at higher sample rates - mostly.
Yep, totally agree.

Joined: Feb 2000
Posts: 2,184
gm Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Feb 2000
Posts: 2,184
Quote:
Originally posted by 3D Audio:
I'll have to re-record them. They aren't on the drive anymore. [...]

I'll try to record samples of as many of them as possible. The format is RADAR with Nyquist converters. I'll do one at 24.48 and the other at 24/96. I can then transfer the files to Masterlink when we're done.
ah-HA!!! You're not testing only differences due to sampling rate changes! You're testing different filters! You testing different clocks!

Q.E.D.

George


George Massenburg

http://www.massenburg.com
Page 1 of 7 1 2 3 4 5 6 7

Moderated by  gm 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.5