Which samplerate for your DAW? [a sort of poll]

erminardi · Post by **erminardi** » Wed Oct 18, 2006 2:20 pm

In my Saturn synth test I've found a big difference between 44.1Khz, 48 and 96 (at 32 bit).
At 96Khz the sytnh sounds almost like i.e. Pro-12 (!!!) but the polyphony, obviously, becomes low.
At 48 the filters sounds a lot better than 44.1
Wich is the best samplerate for for a DAW in your opinion? And why?

-44.1 is CD compatible but a little "poor"

-48 is better but requires more DSP/CPU resources for the same FX/Synth and requires, at the end of the work, a bit quantizing.

-96 is awesome, the best, but dramatically heawy and a little instable.

Your opinion?

Mr Arkadin · Post by **Mr Arkadin** » Wed Oct 18, 2006 2:36 pm

44.1 at 24 bit does me fine for the hassle factor (CD compatibility). Maybe on the next generation cards 96 won't be such an issue, in which case i'll give it a whirl.

arela · Post by **arela** » Wed Oct 18, 2006 2:36 pm

I've been DVDized, so i have stepped up from 44,1 to 48 kHz.
Better to resample down for CD than resample up for DVD.
My guitarplaying wasn't any better with 96kHz so i leave it for now

ARCADIOS · Post by **ARCADIOS** » Wed Oct 18, 2006 2:39 pm

96 khz quality is better for sure if you intend to apply many plugins on your basic sounds, synths or live instruments.
the fact that all will be downsampled to 44.1 doesn't mean that the project must be mixed and mastered at 44.1. the higher the better for working on the project and applying effects. some say that 88.2 is a good choice of working if the final result will be a cd.
in scope the instability of 96khz makes it imposible to work with.
it is a matter of pci as well i suppose.
but if you could work in this sample rate i wouldn't say that it would be good for i do not know how good is scopes clock at this high. scopes clock is probably ok on 44.1 or 48, but thats it.
i think that choosing 96 shoulcome out of wordclock especially for that work like apoggee, prism etc.
a few months ago i made a question here in z about using external clock via syncplate with scope in 192khz.
the question still remains. will scope work as slave in 192khz?
ps. syncplate specs say up to 400khz!!!

astroman · Post by **astroman** » Wed Oct 18, 2006 5:10 pm

On 2006-10-18 15:39, ARCADIOS wrote:
...some say that 88.2 is a good choice of working if the final result will be a cd.
...

for the 99th time
the 'even division' story is bullshit and only tells that the person in question has a severe lack of knowledge about digital signal processing - and can be safely ignored (in that context...)

cheers, Tom

- oops, that may read ambiguous - I refer to 'some', not Arcadios

[ This Message was edited by: astroman on 2006-10-19 01:46 ]

Post by **garyb** » Wed Oct 18, 2006 10:21 pm

24bit 44.1k.
what more could you want?

bill3107 · Post by **bill3107** » Wed Oct 18, 2006 11:00 pm

the sharper, the better for me. For human ears (for UFOs i don"t know), i would say that i have noticed differences between 44,1 and 48. And the dsp usage difference is not so huge so i would go for 48. Remember that the key point is RECORDING quality rather than samplerate. The choice of mic, distances, preamp, etc... all this will give you THE sound. Increasing the samplerate can be just useful when apllying a lot of FX as it works on thinner slices...

One can say that eventually you will have to get a 44,1 samplerate. Yes... but 10% of 1000 gives more data/info than 10% of 800...

Jo

[ This Message was edited by: bill3107 on 2006-10-19 00:01 ]

Post by **garyb** » Thu Oct 19, 2006 12:01 am

...until the aliasing from the downsampling wipes out the improvement...

that's right, good recording, mixing and mastering technique is more important.

erminardi · Post by **erminardi** » Thu Oct 19, 2006 12:36 am

On 2006-10-19 01:01, garyb wrote:
...until the aliasing from the downsampling wipes out the improvement...

that's right, good recording, mixing and mastering technique is more important.

Not if U go for analog out at 48 and return into digital at 44.1 (a pass trough for a neutral setting tube preamp could be a good thing).
ps. I have 2 PCs with Scope system.

Or, in alternative, this free stuff... any opinion about it? (I'm not so exepert to recognize all the nuances & aliasing issues or audio downsampling): http://www.voxengo.com/product/r8brain/

[ This Message was edited by: erminardi on 2006-10-19 01:41 ]

djmicron · Post by **djmicron** » Thu Oct 19, 2006 12:38 am

44.1/24 bits

especially on scope, there is not substantial quality difference between 44.1/48/96 khz.

erminardi · Post by **erminardi** » Thu Oct 19, 2006 12:46 am

On 2006-10-19 01:38, djmicron wrote:
44.1/24 bits

especially on scope, there is not substantial quality difference between 44.1/48/96 khz.

Mhh, the synths IMO sounds better (mainly the filters) at higer res.
Maybe not the flagship synths like Minimax/Prodyssey/Pro-12/B2003/ProTone/ecc. because they are already internal oversampled.
But almost all other free, basic or old devices seems to sound better.
Maybe a blind test poll here could solve this matter...

wayne · Post by **wayne** » Thu Oct 19, 2006 12:47 am

44.1/24

astroman · Post by **astroman** » Thu Oct 19, 2006 12:56 am

On 2006-10-19 01:36, erminardi wrote:
...Not if U go for analog out at 48 and return into digital at 44.1 (a pass trough for a neutral setting tube preamp could be a good thing)...

a complete DA/AD cycle plus passing lots of mediocre parts and an obscure amp will most likely have an even heavier influence.
It's not just the core of the conversion chip - there are op-amps, capacitance, cable inductivity, influences from power supply.

If you can afford gear with only the least possible impact on the signal, then you (usually) can also afford a complete 96k recording chain...

cheers, Tom

Post by **garyb** » Thu Oct 19, 2006 1:00 am

On 2006-10-19 01:36, erminardi wrote:
Not if U go for analog out at 48 and return into digital at 44.1 (a pass trough for a neutral setting tube preamp could be a good thing).

then you have an extra digital to analog and analog to digital conversion. i can't believe that's better than being at 44.1k in the first place.

hey, whatever makes you happy with your work. you may like an external processor so much that the degradation from the extra conversions may be worth it. the effect from an external box may still yeild an improvement in the overall character of a recording, even as it's degraded(ever so slightly).

btw- the voxengo plugin is a good one. still, the problem is not really 100% solvable by some plugin. it's about math, probabilities and the error that uncertainty must cause. once again, the error from downsampling is small, but it's there, just like the difference between 44.1k and 48k.

as discussed endlessly, filters probably will work better at higher sample rates because it's easier to make a good sounding filter at higher sample rates(it has to do with resonant frequencies and the fact that certain errors in a poorly designed filter are frenquencies beyond the audible range at very high sample rates). for this reason many plugins' filters oversample.

high sample rates are nice, except that for very minor gains in sound quality(not enough improvement to make a recording go platinum or not), all the computer's resources and storage are used prematurly.

*edit* yes, Tom, exactly.

[ This Message was edited by: garyb on 2006-10-19 02:01 ]

Shroomz~> · Post by **Shroomz~>** » Thu Oct 19, 2006 1:18 am

On 2006-10-19 01:56, astroman wrote:
If you can afford gear with only the least possible impact on the signal, then you (usually) can also afford a complete 96k recording chain...

That would be so sweet to be completely 96k. But then AFAIK not all Scope devices can run at 96k, so...

For the moment ... 44.1/32bit when using Vdat. 24bit when not.

katano · Post by **katano** » Thu Oct 19, 2006 3:13 am

44.1/24

tgstgs · Post by **tgstgs** » Thu Oct 19, 2006 3:37 am

audio CD is 44.1 / 16 isnt it??
if your song needs 'more' to sound good forget the song!

sending good vibes at 44.1CD / 48DVD from vienna

tgstgs · Post by **tgstgs** » Thu Oct 19, 2006 3:45 am

have to add:
when you get more producers in your studio on 96/32 do it but think of that:
most of the people listen to at mp3 with shitty speakers;
so that is a good mix if it even sounds good there in mono!!!

greetings

erminardi · Post by **erminardi** » Thu Oct 19, 2006 3:50 am

On 2006-10-19 04:45, tgstgs wrote:
have to add:
when you get more producers in your studio on 96/32 do it but think of that:
most of the people listen to at mp3 with shitty speakers;
so that is a good mix if it even sounds good there in mono!!!

greetings

Yes, yes, sadly is true

fidox · Post by **fidox** » Thu Oct 19, 2006 4:06 am

44.1/24