Is recording at 96k/192k with UA 2192 Worth It?

[jrod9900]

Originally Posted by mixermarkb

I don't care to debate width and so on, but I've been working at 88.2 since doing an acoustic piano/tabla/native American flute project a few years ago.

I mix OTB, and in my world, the difference is far from subtle on harmonically rich sources. Rock guitars sound MUCH better to my ears, for example. Breakups are more distinct, the high end of stacked guitar tracks doesn't turn into a wall of fizz as easily, but it's way worth higher sample rates for me.

Drive space is cheap.

YMMV

Hey, Mark,

Really good input. I'm here iin STL and have heard a lot of good things about you from Inimical Drive. I just did some work with them on their new album coming out. I also just did some drum tracking wih Mark on Monday. They were really impressed with your work! I

I'm more of a songwriter/musician, but I have been doing some freelance producing lately and have been loving it. They told me I need to check out your room. We were actually talking about it Monday.

My curiosity is really peaked. My whole thing isn't that I'm against higher sampling rates, but I would want to know the cons are worth it. Like you mentioned, hard drive space is cheap.

I would really like to hear some of what you have done. I also would love to check out your room. The room I've been working at is great, but I know it would be nice to have a larger live room for drums occassionally.

Also, if you think that there is a track you've done at a higher sample rate that showcases the difference, I would definitely be up for popping down and check it out.

Thanks!

JROD

[kennybro]

Originally Posted by minipoodle

Try the higher rates and listen. Much more constructive than asking for subjective opinions from people with different ears. I own a 2192 and found that tracking at the 192khz rate made a massive difference, even over the 96khz rate. The results may or may not be the same for you.

yup. OP, Try it and make up your own mind from what YOU hear.

[jrod9900]

Originally Posted by toneguru

The better your ear the more 96k will matter.

The better your gear, the more the dif 96k makes will be revealed.

I've always felt everything makes a difference.

About a year ago, I participated in a double blind test. Not everyone in the room could pick out the dif from 48 to 96 but some could. To those that could, we appreciate the dif. Its a very small dif but still better sounding to my ears.

- Cheers

How was the test set up? Not trying to take anything away from it, but it seems like the tests I see are a 50/50 shot with like two samples. If there were say three tracks at 48k and one at 96k, and some people could tell a discernable difference, then I would definitely use that. Most tests I see are not very realistic.

I know with the Vintech preamps, they have three classic Neves and one Vintech. Like virtually nobody picks the Vintech every time. That's a test.

I would really like to participate in a sample rate test though. I would really like to see if I can tell the difference. I have a pretty decent ear and would like to think if there IS a difference, I would (should) be able to tell.

I can definitely pick out a 128 mp3 vs a 256 every time!
Haha

JROD

[audioman]

Im convinced that 96k might or will sound better.. My next project will be in 96k.. I like the comment made about the hard drive. They are not expensive and why not capture the best audio possible.. Are there a alot of people recording at 192k with the 2192. If 96k sounds slightly better.. 192k must sound amazing!!!

[travisbrown]

Originally Posted by audioman

If 96k sounds slightly better.. 192k must sound amazing!!!

Actually, the opposite. If people have a hard time picking out between 48 and 96, they will have doubly hard time picking out between 96 and 192. The granularity gets finer as you increase the rate.

But keep in mind there are smart converter designer people like Dan Lavry who disagree that rates above 96k are better, and can in fact deteriorate audio. Lavry believes this to the point that he doesn't make gear that operates above 96k. It's not just because he doesn't know how.

I don't know who is worried about hard disk space any more. That's not the issue with higher sample rates. It's about processing resources. Takes a whole lot more computing power to process 88.2, 96, 192, etc. So when you consider diminishing (or even negative) returns above 96k, you might opt to choose a sample rate that's appropriate for your setup and session.

[psycho_monkey]

Originally Posted by audioman

Im convinced that 96k might or will sound better.. My next project will be in 96k.. I like the comment made about the hard drive. They are not expensive and why not capture the best audio possible.. Are there a alot of people recording at 192k with the 2192. If 96k sounds slightly better.. 192k must sound amazing!!!

Never heard of a pro project at 192k. None of the records in your collection were at 192k.

[travisbrown]

Originally Posted by psycho_monkey

Never heard of a pro project at 192k. None of the records in your collection were at 192k.

I think David Byrne did some archival conversion and remixing of old material at 192k some years ago. But Byrne also likes Powerpoint as an artistic medium.

[robertshaw]

Originally Posted by travisbrown

Hmm.

Width come from panned non-correlated channels in a mix and the ambient interaction of the two. How does a dual discrete channel converter increase or decrease width? I'd be concerned if it did.

Depth, presence or focus I can understand. Width, not so much.

width, length, height, circumference.... depth, whatever
semantics........
how about 3D?

whatever word you want to use
with 2192 96Khz gives you more space
it's sounds wider...

unless you use a 2192 you wouldn't know what I'm talking about anyway
not all converters sound the same at one or different sample rates

Originally Posted by travisbrown

Hmm.
How does a dual discrete channel converter increase or decrease width?

there is this technology called 'stereo' it "refers to any method of sound reproduction in which an attempt is made to create an illusion of directionality and audible perspective."

[travisbrown]

Originally Posted by robertshaw

width, length, height, circumference.... depth, whatever
semantics........
how about 3D?

whatever word you want to use
with 2192 96Khz gives you more space
it's sounds wider...

Of course it's semantics. Width means something different than depth. That's why they are different words that you use to convey different ideas. Before passing off as "semantics", you should actually know the semantics of the word "semantic."

So the signal that comes out of the back end of a 2192 is spatially different than what goes in? Ugh. What's wrong with it? Does it change the width of the audio if you are just using one channel? And if it doesn't on one channel, how does it do it on two?

Originally Posted by robertshaw

unless you use a 2192 you wouldn't know what I'm talking about anyway
not all converters sound the same at one or different sample rates

What about if I have used one and still don't know what you are talking about?

Why would converters have different sample rates if they sounded the same?

Originally Posted by robertshaw

there is this technology called 'stereo' it "refers to any method of sound reproduction in which an attempt is made to create an illusion of directionality and audible perspective."

Right. There's also this thing called reading the post. See that bit about non-correlated channels? That's fancyspeake for stereo. There's nothing about discrete channels that should add or take away from stereo - they just channel two mono signals. If a converter unintentionally changes the stereo image, it is ineffective or broken.

Quick test: Put a correlation meter inline on your input signal and also on the output of the 2192 at 96k. If they read the same, then there ain't no widening going on.

[LeMauce]

Hitting the "sweet spot" of your converters is one of the best things I learned next to summing stems out to a summing box to improve my "sound". YES mixing properly is number 1. But the other 2 brought me 15 a 30% improvement when proper working (gain staging levels inside AND outside the daw). the punch and seperation between instruments is just ridiculous compared ITB summing (protools, nuendo, logic) when working with a lot of tracks. So when summing go for 16 or more tracks. Others wise stay in the boxie.

So yes. Try to record to 96Khz and threat it with respect inside your DAW when working with plugins.

[psycho_monkey]

Originally Posted by travisbrown

I think David Byrne did some archival conversion and remixing of old material at 192k some years ago. But Byrne also likes Powerpoint as an artistic medium.

Fair enough. There's always an exception.

[toneguru]

Originally Posted by filipv

Could you please describe the blind-testing technique?

How many samples did you listen to?

It was a "double blind" test. That is, the scientist/musician did not know himself which sample he played until after each segment was finished and the audience (40 engineers) raise hand or not as we all guess the sample rate. Then the answer was revealed. As I recall, he played two versions of the same recording but at the two rates. Then we (and he) would guess which was which. It was a while ago, maybe two years ago now that I think about it.

We had maybe a dozen different dual recordings to discern. Maybe more.

The environment was not optimum nor the monitoring but still the diff was present about 70% of the time to about 70% of the audience. With better monitoring and acoustics, the number would probably rise. Although one guy with golden ears got it right every time. I think I got it about 85% of the time. To me the reverb tail was most telling along with high end silkiness.

[audioman]

I think I will stay put on 24/44.1 and 24/96 I dont feel 192k is necessary. gonna start tracking a project right now in 96k lets see how it comes out.

[robertshaw]

Originally Posted by travisbrown

Of course it's semantics. Width means something different than depth. That's why they are different words that you use to convey different ideas. Before passing off as "semantics", you should actually know the semantics of the word "semantic."

So the signal that comes out of the back end of a 2192 is spatially different than what goes in? Ugh. What's wrong with it? Does it change the width of the audio if you are just using one channel? And if it doesn't on one channel, how does it do it on two?



What about if I have used one and still don't know what you are talking about?

Why would converters have different sample rates if they sounded the same?



Right. There's also this thing called reading the post. See that bit about non-correlated channels? That's fancyspeake for stereo. There's nothing about discrete channels that should add or take away from stereo - they just channel two mono signals. If a converter unintentionally changes the stereo image, it is ineffective or broken.

Quick test: Put a correlation meter inline on your input signal and also on the output of the 2192 at 96k. If they read the same, then there ain't no widening going on.

we are talking about sample rates here. You are (trying) to argue something different

[Apollo Soul]

Originally Posted by LeMauce

Hitting the "sweet spot" of your converters is one of the best things I learned ...

Anyone got an informed opinion of just where that "sweet spot" is on an apogee duet 2 specifically?

[travisbrown]

Originally Posted by robertshaw

we are talking about sample rates here. You are (trying) to argue something different

Oh, sorry. For a minute I thought you wrote that the 2192 sounds wider at 96k. I'm so embarrassed.

Originally Posted by robertshaw

96k on the 2192 is much wider than 44/48
it is worth it

[filipv]

Originally Posted by robertshaw

with 2192 96Khz gives you more space
it's sounds wider...

what does "sounds wider" mean? more pronounced stereo effect? less crosstalk? greater dynamic range? Please clarify.

[travisbrown]

Originally Posted by filipv

what does "sounds wider" mean? more pronounced stereo effect? less crosstalk? greater dynamic range? Please clarify.

I already asked, but got accused of semantic obfuscation.

Originally Posted by robertshaw

width, length, height, circumference.... depth, whatever
semantics........
how about 3D?

whatever word you want to use
with 2192 96Khz gives you more space
it's sounds wider...

[robertshaw]

Originally Posted by filipv

what does "sounds wider" mean? more pronounced stereo effect? less crosstalk? greater dynamic range? Please clarify.

I think the best way is to describe it as adding like a waves S1 plugin or another stereo widening plugin to a mix, although the plugins sounds cheap and fake. Higher sample rates on 2192 are more natural, unlike a plugins. Some how its more wide like you just took you mix and spread it out wider. It also gives it a more 3D feel, more depth like as if you were listening to a live band and you are in the audience or something. Going from 48 to 96 on the 2192 would be like going from mono to stereo, although a much lesser drastic degree of course. That's the only way I can put it into words

I can only speak for the difference between 48 and 96 on a 2192 not all converters are equal in this sense. Although I do recall when using a Bigben on a rosetta and it had the same sort of phenomenon just not as intense as 2192.

I did try going from 48 to 96 on an RME ADI-2 one time and it made no difference. The 2192 is a special converter. unless you have heard one and used one day in and day out words cannot really describe how awesome it is. And one cannot speculate the difference between 44 and 96 on a 2192 unless having experienced a 2192 on both 44 and 96 first hand.

Just because someone's $200 maudio interface sounds the same at all sample rates, doesn't mean all converters sound the same at all sample rates

[audioman]

Originally Posted by robertshaw

I think the best way is to describe it as adding like a waves S1 plugin or another stereo widening plugin to a mix, although the plugins sounds cheap and fake. Higher sample rates on 2192 are more natural, unlike a plugins. Some how its more wide like you just took you mix and spread it out wider. It also gives it a more 3D feel, more depth like as if you were listening to a live band and you are in the audience or something. Going from 48 to 96 on the 2192 would be like going from mono to stereo, although a much lesser drastic degree of course. That's the only way I can put it into words

I can only speak for the difference between 48 and 96 on a 2192 not all converters are equal in this sense. Although I do recall when using a Bigben on a rosetta and it had the same sort of phenomenon just not as intense as 2192.

I did try going from 48 to 96 on an RME ADI-2 one time and it made no difference. The 2192 is a special converter. unless you have heard one and used one day in and day out words cannot really describe how awesome it is. And one cannot speculate the difference between 44 and 96 on a 2192 unless having experienced a 2192 on both 44 and 96 first hand.

Just because someone's $200 maudio interface sounds the same at all sample rates, doesn't mean all converters sound the same at all sample rates

great explanation!!

[MrVelvet]

To me, it's worth it recording at 96/192 kHz with the 2192. I go for 192 for most projects meanwhile with my 2192's, but it also got me in trouble from time to time, regarding track count with my DAW or some plugins that can't do 192 kHz.

I found that with 192 kHz, I can stay ITB and get great results with plugins, as the high end tends to be less harsh when treated with an EQ (which is probably due to the higher resolution the plugin is fed with?). I'm happy with my ITB mixes that I don't always reach for some OTB stuff.

But then again, it's all about the recording. I feel that working at 192 kHz requires special care to make it worth it, otherwise it's just a pain in the ass. If the recording of a band isn't really great thanks to me or the band, then I am a bit embarrassed to have worked at 192 kHz and didn't make use of the extra potential in there...

Bottom line probably is: If the session turns out great sonically, then think of 192k

[filipv]

Originally Posted by travisbrown

I don't know who is worried about hard disk space any more.

Nobody worries about hard disk space. What IS worrying is the sheer amount of misconceptions concerning digital audio, misconceptions that go all the way into pro circles.

[ARIEL]

Here is a link for you , scroll / look down and see the part on sampling rate . This fellow has recorded some great albums and he even is involved in designing gear as well . Lots of good info UnderToneAudio

[timlloyd]

Originally Posted by ARIEL

Here is a link for you , scroll / look down and see the part on sampling rate . This fellow has recorded some great albums and he even is involved in designing gear as well . Lots of good info UnderToneAudio

From that paragraph - "The multi track recordings were mixed in pro tools". If the point of the test is to evaluate the differences between 44.1k and 192k ADC, then ... concern #1 ... were any plugins used in the mixes? If so, then the test has failed because it's also evaluating plugin performance at different rates. But it doesn't specify either way.

Concern #2 - it's also evaluating the performance of the converter units' DAC on playback. What's it intending to evaluate, one or the other or both? This isn't specified - what's the hypothesis that's being tested?

Concern #3 - it's only actually evaluating that particular unit's adc and dac. Presumably this was a Digi 192 IO, but it's not actually specified unless I missed something while skim-reading.

Try not to draw any conclusions from it that aren't actually supported by the provided info. Regardless of how rigorous and illuminating that part of the test was for those taking part at the time, what can we who have access to only the info on that site learn from it? Nothing really.

Lack of specificity in evaluative tests should throw up a big warning sign for everyone imo. Unfortunately this isn't the case for many, eventually leading to the situation filipv mentions after people extrapolate on insufficient info. Perhaps I'm unduly grouchy today, perhaps not but it strikes me as rather irresponsible to make such watered down and unspecific information available from such a location.

[Syncamorea]

Originally Posted by timlloyd

concern #1 ... were any plugins used in the mixes? If so, then the test has failed because it's also evaluating plugin performance at different rates. But it doesn't specify either way.

That and the test involved transferring an analog tape recording to PT. If your goal is to understand the effect of sampling frequency, I don't know that analog tape is the ideal source - it is valid as part of a suite of tests but a live mic feed is the way to go for the core of such a comparison. Why? Because the tape deck and tape itself are adding their signature to the signal path and I don't think that's the intent of a sampling frequency comparison.

[popmann]

A) it is the high water mark for recording

B) being not converted to PCM, it is pristine in that it's being changed to PCM for the first time in this test.

C) related to the last--it would be better to record something live, right? But you can duplicate that. You can pull up the tape, 44, and 192 transfer simultaneously on an analog desk and compare the "original" with the digital conversions. The only way to do that with a live mic is "live"--live feed to one channel, looped 44 on another, and looped 192 on the third. That's a cool test, but it has no replays.