44.1 Samples, 96k project?--Audio Quality

Discussion of music production, audio, equipment and any related topics, either with or without Ableton Live
Post Reply
cabletone
Posts: 337
Joined: Thu Jul 19, 2007 3:55 pm

44.1 Samples, 96k project?--Audio Quality

Post by cabletone » Wed Jan 12, 2011 5:11 pm

There are 2 audio quality issues when running high sample rates with 44.1 samples in a project, one positive, one negative.

Positive: Higher quality of plugin performance at higher sample rate.

Negative: Degradation of samples from ableton's conversion to higher sample rate (using ableton's 'high quality' mode).


So there is a trade-off when working at a higher sample rate.

My question is: for those of you who are concerned with audio quality, do you find working at a higher SR worth the cost in audio quality of ableton's upsampling of 44.1 samples?

My choice is to work at 96k as I think it sounds better and I do not notice a difference in the audio quality of the samples. However my ears are not very good and it could be a psychological thing as well.. So I am wondering what others are doing.

Angstrom
Posts: 14755
Joined: Mon Oct 04, 2004 2:22 pm
Contact:

Re: 44.1 Samples, 96k project?--Audio Quality

Post by Angstrom » Wed Jan 12, 2011 6:27 pm

most plugins worth their salt are oversampling anyway, surely?
Most will 4x oversample their filters, etc. Oscillators are usually wavetables, or in a few cases calculated on the fly without aliasing. There are a few processes such as cross-modulation that are usually handled less well, of course. But in general decent plugins can oversample independently of the host.

I wouldn't upsample on the fly, not at all. If you really want to do this you should go for a doubling (44.1 x2 = to 88.2khz) rather than 44.1 x 2.17687075 = 96Khz .

That calculation right there tells you all you need to know. There is a lot of audio being 'made up'. Not worth it at all IMO

Khazul
Posts: 3185
Joined: Wed Feb 23, 2005 5:19 pm
Location: Reading, UK

Re: 44.1 Samples, 96k project?--Audio Quality

Post by Khazul » Wed Jan 12, 2011 11:45 pm

Mastering engineers like to up sample to 96k regardless of what you give them, but I havnt come across anyone (worthy of note) for ages who works at anything other than 44.1, if we ignore very very expensive high end commercial studio systems.

And yes - 96 can sound a tiny bit smoother in the high end especialy using processed samples - that is until you throw and soft high cut eq at 14.5K over the high end then it make no difference.

And given a wall full of shelves of music processed in such a way that seems to have a nice smooth clear open high end on them - guess it cant be a bad idea for smoothing things over and getting rid of any digital harshness that creeps in due to some unpleasent processing that was perhaps beyond your control (or caused by resampling).
Nothing to see here - move along!

3phase
Posts: 4648
Joined: Fri Sep 26, 2003 3:29 am
Contact:

Re: 44.1 Samples, 96k project?--Audio Quality

Post by 3phase » Thu Jan 13, 2011 3:09 am

cabletone wrote:There are 2 audio quality issues when running high sample rates with 44.1 samples in a project, one positive, one negative.

Positive: Higher quality of plugin performance at higher sample rate.

Negative: Degradation of samples from ableton's conversion to higher sample rate (using ableton's 'high quality' mode).


So there is a trade-off when working at a higher sample rate.

My question is: for those of you who are concerned with audio quality, do you find working at a higher SR worth the cost in audio quality of ableton's upsampling of 44.1 samples?

My choice is to work at 96k as I think it sounds better and I do not notice a difference in the audio quality of the samples. However my ears are not very good and it could be a psychological thing as well.. So I am wondering what others are doing.
i would say yes.. it is just better for your mood when not even trying to keep tracks bit transparent..
just accept samplerate conversion and use it at the highest possible quality..
accepting it also allows to do the tracking on any samplerate you like and swithch higher at a later production state..

bit transparency in live is not really easy to keep or reliable enough.. the brown henke dither or Samplrateconversion can ocur on speed changes.. even when going back to original tempo it can stay on...
Using loop makers can produce it on the smallest wrong move..
It is allways on in slave sync..
On sliced tracks and sample instruments..
when using groove quantisation

so its actually a samplrate conversion based sound engine concept
so the quality of the SR algo pretty much defines the the sound signature of the program.
and this sound signatur gets way better on higher samplerates.. the stretching sounds better..and the samplerate conversion itself does..

best possible soundquality in the box i was able to achive yet was with rewiring live into a 96 k daw.. and running all masterstems inside the host daw bit transparent.. and all live internal modulated tracks benefitting from the higher samplerate and additional mix features of bigger daw´s

and..there is a rumor that avoiding abletons mix bus helps..

actually running live with samplerate conversion on 96k into an analog summing device sounds pretty cool..
somhow the grain of the SR pretty much can be seen as a tape distortion like fx ..at least on electronic sounds. forces more mojo out of the analog inputs..
but you need at leats 16 channels to really benefit from that.. analog summing of stems is not really doing it as good than analog summing of single tracke and fx returns... and as more complex a submixed file is, as more samplrate conversion hurts On the single sound it gets part of the sound.. on the stereo mix it eats room information and transient precision

So especially for samplrate converted tracks analog summing helps.. removes a bit the clogging togther fx a mix of samplerate converted files gets in the box.. makes it more 3 dimensional and seperated again..
mac book 2,16 ghz 4(3)gb ram, Os 10.62, fireface 400,

Post Reply