What Sample Rate Do You Work In?

[John_C]

ahjteam just showed pretty well that higher sample rates are beneficial when working with a lot of plugins.

When my profire 2626 arrives i'll try and do a blind test with one of these as my signal source

 

[James Murphy]

Using pictures and theoretical text is nice, but I made an audiovisual experience for you:

there's something wrong with that... can't reproduce anything like that result on my system. i have a/b'ed guitars recorded at 44.1, 48, 88.2, and 96... can't hear a difference.

 

[Metaltastic]

Anyways, we all know this could go on forever, back and forth. I'll leave it at that and say we should let these arguments be hashed out by the fags at gearslutz.

Quoted for truth my good man!

 

[ahjteam]

there's something wrong with that... can't reproduce anything like that result on my system. i have a/b'ed guitars recorded at 44.1, 48, 88.2, and 96... can't hear a difference.

I was using the standalone version so I thought it might be the reason, but I tested this out with Cubase too and came out with same results, only thing I changed was the samplerate: (do pardon my sloppy playing)

http://www.ahjteam.com/upload/sampleratetest_amplitube2duo_44khz_24bit.wav
http://www.ahjteam.com/upload/sampleratetest_amplitube2duo_96khz_24bit.wav

edit: and James, I didn't record guitar amps, I recorded the DI-signal. Recording the amps might be a totally different thing

 

[John_C]

I was using the standalone version so I thought it might be the reason, but I tested this out with Cubase too and came out with same results, only thing I changed was the samplerate: (do pardon my sloppy playing)

http://www.ahjteam.com/upload/sampleratetest_amplitube2duo_44khz_24bit.wav
http://www.ahjteam.com/upload/sampleratetest_amplitube2duo_96khz_24bit.wav

edit: and James, I didn't record guitar amps, I recorded the DI-signal. Recording the amps might be a totally different thing

you seriously didn't change any settings at all other than the sample rate? That's an insanely large difference you've got there, and it's still there if you downsample the 96k one down to 44.1

 

[Mutant]

Hey guys !

The thread title says "Work In" not only "Record At".

Many VST plugins dont oversample internally so going to 88200 really gives them a huge boost in quality.

Just check it yourself.

 

[James Murphy]

perhaps it's a vst issue.... i only use TDM and RTAS... and i can't reproduce those hugely disparate results.

 

[ahjteam]

The thread title says "Work In" not only "Record At".

What, you don't work while you are recording? :Smug:

 

[ahjteam]

perhaps it's a vst issue.... i only use TDM and RTAS... and i can't reproduce those hugely disparate results.

Actually, my guesses are that it is the amplitube2duo is to blame. I couldn't reproduce it with any other ampsims. There was a very minimal difference with the guitarsuite jcm900:

http://www.ahjteam.com/upload/sampleratetest_guitarsuitejcm900_44khz_24bit.wav
http://www.ahjteam.com/upload/sampleratetest_guitarsuitejcm900_96khz_24bit.wav

But atleast it seems that SOME ampsims like higher samplerates better.

 

[Mutant]

There was a very minimal difference with the guitarsuite jcm900:

It happens to be one of my favorite ampsims

 

[Mutant]

What, you don't work while you are recording? :Smug:

Mutant said: "not only"

 

[aramism]

i always do 48/24

i don't worry about downsampling or "math" because i leave that to mastering where they are using very high quality converters anyway intended for that. the several ME i have spoken to basically tell me the same as several people here. i don't master stuff i record/mix so i'm not really worried about it.

 

[James Murphy]

converters don't have a thing to do with SRC unless it's being done in real time through them.

 

[John_C]

a question........is oversampling by a factor of 2 and then downsampling back to the original going to have zero effect?

 

[Mutant]

a question........is oversampling by a factor of 2 and then downsampling back to the original going to have zero effect?

[edit]Experimental data seems to say yes.

0 effect.

(from 88200/32 to 176400/32 and back to 88200/32 then invert and mix with original = digital silence)

 

[Styvo]

all i can say is TLDR.



I use digi 192's regularly and do this, it definately makes a difference - is that difference the difference between a good album and a great one? no, but it does sound slightly better. My home interface is 44.1/48 max, so i mix in 44.1 at home.

Thanks man,
I have a fair understanding of digital audio theory,i understand the principles discussed,but out of personal experience i wanted to get a view of users opinions.
I appreciate the input!
My conclussion......44.1/24 and when clients request it,i'll satisfy there request and keep it at 48/24.

 

[gabriel g.]

No problem with downsampling when going from 88.2->44.1

1.
A higher samplerate is in theory always better!!! no questions about that.
But the question is: What is the end result. What is needed in the end. How can we get the best result: the audio-cd.
For example: A real SUPER-35mm film (used for most cinema blockbuster) has a definition up to 12k (12000 picture points horizontal - pixels) but in most ways it is downsampled to 4k (4000 pixels horizontal) or more often 2000pixels horizontal because the most cinemas show movies in this definition, and the blu-ray has this defnition (1920x1080 pixels)

This means, there is no reason to do things, just only because you can , because there is no need for.
You save disk-space and have no quality loose in the end, because of the end result.

2.
In cubase sx3 I had the same strange problems with lower samplerate on windows, but never on OSX and logic.the audio sounds better in the sequencer because of the bit depth .

I think, a audible difference between sample rates is just the result of bad digital filters, because we dont hear higher frequencies then 16.000-20.000hz.
Nyquist cuts at ± 22.050hz, so there is nothing audible!!!no harmonic content hearable for us humans.
There are no perfect filters, they always have a steepness in dB. No analog, and no digital filter is perfect.
Maybe the bad audible filter starts at 18.000hz and ends at 22.050 so some people can hear a difference. But with good filters this shouldnt be a problem.

Please correct me if I´m wrong. I learned and learning this stuff in collage but I´m still learning!!!

 

[kalony]

i was working with 44,1/24 too

but i recorded a session few months ago for a client who gave me the order to record in 192 khz. the plan was, that i record that band, and he will mixe it than.
i talked to him and got him down to 176,4 khz. its still brutal high and better for downsampling.
since it was not me who should mix it, we recorded a lot of posibillities and versions. the sessions were hughe and the songs 9 minuits long

in the end the band asked me to mix it. but i am working native, and i cant realy handle 40-50 tracks in 176 khz with busses and plugs and stuff, so i copyed the sessions to 44,1 khz down.
i made a bounce of both session formats, but couldnt hear a difference. but i think the difference might come when you use a lot of plugs.

i am still working with 44.1/24 because the bands folder on my harddisk is over 200 GB large now and the band loves the mixes i did in 44,1

i would probably record small sessions/projects in higher resolutuion, if i would own a HD1/2/3

kalony

 

[kenboy]

I use to give a crap till I realized that most music these days are either listened to through ear buds, computer speakers (myspace, Pure volume), or bought via itunes which is MP3, so at the end of the day it doesn't really matter. You pristine recording is going to get sonically raped either way you slice it.

Just my 2 cents though.

 

[kalony]

yeah and that just shows, that technique is not the big deal. the most important thing is still to get the right feeling on tape!

kalony