44.1 Khz vs 88.2 or 96

mnstudio

#1

15 years ago

I'm wondering if it is really worth it to work on 88.2 or 96 khz....
Considering that at final stadium I have to convert to 44.1 for mastering CD...
I particularly work with software samplers based arrangements (on Logic Pro), and samplers have substantially 44.1 or 48 recorded samples.....
But also when recording audio files, is it really worth it to record at 96 e then convert to 44.1?
I have a system which connects Logic through ADAT litghpipe to the digital Sony Mixer DMX-R100 which, if it worked at 96 khz, would work with halved channels and reduced funcionality.
So I wonder if it would be worhwhile for me to change drastically my system.
Any comment would be appreciated.
Tnx
mnstudio

July 23, 2008 @08:37am

kope

#2

15 years ago

Depends what you prefer. Clear precise sound or just to do a job :-). If song at self is good no one care about sampling rate. What is proven is that plugs are much better on 96000. This difference can hear anyone especially with plugs like reverb, delays etc. So, if you get material on 44.1 and convert to 96 you will get better result in mix stage then on 44.1 (personally, i need mach shorter time to accomplish job on 96 - not that much returns for details like on 44). I don't work with live recordings so someone else should answer for that part. You should consider that 96 KHz 32 bit wave is a hell of space, average project is around 5 Gb.

July 23, 2008 @10:58am

dcwave

#3

15 years ago

Oh not this again
Stick with a sample rate for your end target.
When I get a chance I will post the Dan Lavery white paper about sample rate and its diminishing returns past about 60Khz, as well the Bob Katz info AND an AES paper showing that in a double blind test people couldn't hear the difference between 44.1 ad 96. There are mega threads debating this on gearslutz, harmony central, REC and other places where top engineers hang out.

July 23, 2008 @01:12pm

JeffBarnett

#4

15 years ago

I've got a customer who works for a marine research company and records dolphins. They make meaningful sounds all the way up to 40 kHz. He needs to record at 96 kHz.
For the rest of us, who record humans, it's not worth it. There are (in my humble but correct opinion) no tangible up-sides, but there are some very real down sides. Even those who argue that they can hear the difference say that it is marginal at best, and I'd have to weigh it against these big disadvantages: Everything takes double the hard disk space and double the processing power. It will likely cut your maximum track count in half, and reduce the number of plugin instances you can run.
And when you're finished, you'll just convert it back down to 44.1k to burn to a CD, or even worse, encode an MP3.
A waste of time and resources. Not worth it.

July 23, 2008 @01:35pm

howie15

#5

15 years ago

I'll second that idea. I work with the guys who record the MN Orchestra and St. Paul Chamber Orchestra. They obviously work with some of the best converters, pres, and mics in the biz. The MN Orch is just moving to 48 next year because of a change in the broadcast system we use. I think acoustic recording like that could be one place it would matter, but after talking with these guys....it doesn't. They have tried 96 and 192 and it's not worth the hassle.
Howie J

July 23, 2008 @02:19pm

dcwave

#6

15 years ago

I've got a customer who works for a marine research company and records dolphins. They make meaningful sounds all the way up to 40 kHz. He needs to record at 96 kHz.

I wonder what they do with those sounds in order to actually hear them. I suppose they encode them into some type of binary and convert them to a lower frequency or something? Or maybe just use spectral-analysis and don't try to hear them at all.

July 23, 2008 @02:41pm

kope

#7

15 years ago

True is that bad sound can hide many mistakes. Also, 80% of professors on any academy newer take care about sound quality - they just listening what is played. Moving on 48 next year is only because they have market. 192 KHz equipment cost a lot - someone should give a money. It was shock for me when i discovered that i must change my el. guitar because is noisy on 96 KHz (Works fine on 48) - Simple, you can't change only sound card, all other equipment is under question mark (Your good old studio room too). Mathematics is of great value in acoustic but can't say to you what is the best for you.
There is only one truth - TRY IT and if you like it use it. If what you are doing brings to you pleasure and positive results do not change anything. Recently i got studio session (24 channels) of Bohemic Rapsody - Queen. Except Fred's voice there is nothing worth of attention (about playing and recording) but final result we all knows. Technically all job there is on the level of average city band but idea and song are great. I hope someone will replay this song but seriously.
mnstudio - i hope i helped to you with my comments. All today philosophy is to give minimum quality for target purpose. If you go to low you will crash on land and if you go to high you will spend too much money. But where is pleasure?
Forget for your units and try to do hole project on 96 KHz from scratch. Take some good software synts and make your own samples. After one or two weeks of work i can guaranty to you that you will newer wish to work on 44 KHz.

July 23, 2008 @04:08pm

JeffBarnett

#8

15 years ago

kope -
No offense, but do you understand the practical differences between 44.1 and 96k recording? Do you understand that in numerous double blind tests (a couple of which I've been a part of), there has been proven to be no audible difference between 48k and 96k? In fact, some types of processing can actually induce noise in the audible band if you run them at high sample rates, and that noise is masked at 48k. So I could argue that 96k recording actually sounds WORSE than standard resolution.
This isn't my theory, mind you. It's scientifically-tested, and backed by very solid physics and physiology research. I'd never advise anybody to work at 96k from scratch (with the exception of the aforementioned dolphin guy).
I won't deny, though, that if you spend the time and money to do everything at high sample rate, and you think it should sound better, then to you, it will sound better, and you won't regret your decision. Psychology is a powerful thing. But I would say that if you listen to the finished product, in a blind comparison, you won't be able to hear the difference. I can't. I've sat with some of the highest-paid golden ears in the business, and they can't either.

July 23, 2008 @04:25pm

JeffBarnett

#9

15 years ago

I wonder what they do with those sounds in order to actually hear them. I suppose they encode them into some type of binary and convert them to a lower frequency or something? Or maybe just use spectral-analysis and don't try to hear them at all.

They play it back at 24 kHz, which drops everything by 2 octaves, comfortably within the audible band.

July 23, 2008 @04:27pm

kope

#10

15 years ago

.

Is not so simple :-)
What i was talking here is not conversion or simple comparison. What i noticed is that complete project sounds better. Truth is that plugs operate on high sample rates (they do not care what you can hear) . As i mentioned is weary simple - just try. Don't play with single wav. We all know that is weary hard to compare two speakers. Make complete project on 44 and then repeat the same on 96 (I mean start second time from weary beginning). When i bayed my sound card i converted complete actual project to 96 just to see possibilities of my new card. After one hour of playing i closed project disappointed. I opened project again (regularly on 48) to continue my work and after short time i comprehend that there is difference when i started manipulation with plugs. Most of good plugs read wav in advance and calculate average whatever and they are limited by time (1 or 2 s). Don't say to me that is same if they take 44 or 96 K examples.
It's all about mix stage. As i say before, i do not work with live recordings and i don't have this experience so i can't discus about this (no meter how good in math i am). There is one thing which i can say. Voices recorded in 44 is much harder to improve by plugs then one recorded in 96 ( with real 96 equipment). Again voice of experience - don't ask me why is it.

July 23, 2008 @05:30pm

dcwave

#11

15 years ago

There are two problems I have with what you are describing, Kope.
1 - the equipment listed in your chain is not going to reproduce 96kHz accurately at an electronic level and your monitors, even if your DACs were of a high enough quality to have a 96Khz sample conversion back to analog are overpriced consumer speakers (most cheap near fields are) that are not going to let you hear the minuscule changes that a good DAC would reproduce. I can almost guarantee that you would hear a much more drastic difference by using a high quality external clock on your system
2- As far as I know from reading VST specs (and I could be wrong) they are operating at the sample rate of the host software and do not up-sample for processing if they did they would be far, far more expensive than they are do to the additional filters needed to keep strange aliasing and noise from being re-introduced back into the sound stream; some plug ins will increase their internal bit depth for the processing which will make a difference in quality.
Ethan Winer has an article explaining why people perceive a difference when testing their theories on their own. Most of it is the placebo effect.
But if you think you hear a difference then good - keep doing your way. At the end of the day the only question you need to ask is - did the client like it enough to come back? If yes - who cares about how or what you used / did to get the end result.
If I get a wild hair, and find I have the time - I will post some science behind why 96Khz and above is hype for most cases. There is a law of diminishing returns and a trade off between price and benefit.

July 23, 2008 @06:06pm

kope

#12

15 years ago

I do not care what going on in small black bugs on sound card and who convert what. I'm producer and sound engineer. My way is weary simple: plug and listen. Yes, may chain is cheap and probably produces many mistakes. Like all sound engineers - when i accomplish some stage in may work i always do check on different systems (car, TV, Daily and Night listening on Hi-Fi systems, producing in different cafes). So, please give a brake with theory. Also this price thing is outdated (prices are down if you compare market before 5 years - so wake up), For me, producing music is some kind of art and if i can reach something to realize my ideas - there is no question about money. Also we are not in hobby we are in business and if you chew about some 1000 Euros for basic tool better change branch.
The reason way i react so nasty (i apologize in advance - but discussion mast to have sparks otherwise is bore) is because "other side" newer listen - I try this and is good. Main misunderstand here is that computer is not any more only recording machine - is fu...ng good instrument and processor for sound and you can ignore it anymore. Technology going forward (192 KHz knocking on door) and there is no room for sleepers. Have nice mixing - i'm out of this discussion - pointless - nothing new to say .

July 23, 2008 @11:05pm

dcwave

#13

15 years ago

Like I said - if you hear a difference that's great - you have golden ears and should be schooling the top engineers and equipment designers that do not hear a difference.
And you are correct - discussing something subjective, like what you or I hear, is pointless - discussing something repeatable, and observable in order to insure that correct information about how or why something works has value - but even then, I am becoming such grouch lately I might just say - whatever

July 23, 2008 @11:34pm

Bispo

#14

11 years ago

Just new here, but i agree with Kope. I`m making several tests daily and the plugins behavior at 96khz are amazing, I hear the details on tape emulation and reverb are beautiful in 96k. definitely heavy load and a lot of hardware is required to do job right, specially trying to emulate good analog rig. For myself there`s no going back to 44/48Khz

February 22, 2013 @05:17am