I understand the concept of sample rate and bit rate etc. But i get confused as to how this applies to the action of creating a track from scratch. Let me give you this scenario:
Firstly i record a vocal using my liquid channel preamp and i set the sample rate at 96khz. Then my daw soundcard will have to be set to 96khz, also my workstation environment in this example Apple Logic will also be set to 96khz. I guess then my question is in order to achieve 96khz quality my chain of soundcards plug in fx, outboard fx, daw software, will all need to be set to the same sample rate? What different would it make if say i recorded the vocals in 96khz and then import to a 44.1khz project and use 44.1khz plugin eq's and compressors? And anyway if i record with 96khz it still needs to be downsized to be put on CD as CD can only handle 44.1khz. Can somebody explain to me the concept clearer.

Yes - you need to keep the rates the same throughout the recording process. If you want to use a 96k file in a 44.1 project, you can convert it to the correct rate and use it, and you won't necessarily retain any improvement in sound quality. General rule of thumb - pick a a resolution / rate and stick to it.

As to working at a higher rate rather than the target rate, the simplest analogy is with tape. In the old days, we recorded on 2" tape to get the best possible sound quality, then mass produced casettes. Same principle applies - record to the best that think is needed and your whole recording and mix will sound better than if you cropped it. There's a difference between making all of the sources sound less good than just degrading the quality of the finished mix.

If you get my drift...

--------------------
Veni, Vidi, Aesculi
(I came, I saw, I conkered)

I'd add that there's probably little point working at 96khz unless your studio is absolutely the bollox, with a fantastic room, real high end outboard, and mics that cost thousands... with that quality throughout the signal chain. Otherwise you'll be forcing your computer to work far harder for a sound benefit that will probably be inaudible.

24bit at 44.1 khz is good enough for 90% of projects, and a lot of us, including me, are still working at 16bit, 44.1 khz, which for a lot of electronica is still adequate. Film music has to be recorded at 48khz. 44.1 khz is the standard for CDs.

As the previous poster said, all your digi gear has to be running at the same rates, and choose the unit (probably the soundcard) with the best clock and jitter suppression to act as the master clock for everything else. Working out of sync will introduce all kinds of phase problems and timing inconsistencies which will really crap up the sound, whatever rates you choose.

--------------------
http://anotherfineday.bandcamp.com/ http://anotherfineday.co.uk http://apollomusic.co.uk

There are some guiding principles to apply.

1) work at 96kHz sample rate if your system will accommodate that without losing too many channels/processes.

There is little point going higher than 96kHz. The sonic benefits are negligible, but the increased demands on clocking accuracy escalate enormously as do the processing power and storage requirements.

However, there are significant benefits to working at 96 compared to 44.1 or 48. Reconstruction and anti-aliasing filters sound considerably better; EQ precision and sound is better, dynamics are more accurate, and metering is considereably more precise. It is for these (and a few other reasons) that increasingly plug-ins run at double sample rates (oversample the data onthe way in and downsample on the way out).

2) There must only ever be one clock master in your system. Everything else has to slave to that. In many cases, you can often configure the system so that you can change the entire clock rate simply by flipping the physical or virtual sample rate switch on the master device.

3) Downsample to 44.1 for CD as the last but one process, once you have mixed, edited, complied and mastered your project at 24/96, then sample rate convert to 24/44.1 and finally reduce the wordlength with dither to 16/44.1.

4) If you system can't adequately support 96kHz operation, then run at 24 bits on the interfaces and 44.1 for a CD final output or 48kHz for a film or video-based final output. The extra wordlength enables you to work with a far more practical degree of headroom, and maximises the sound quality. Within the DAW, use the 32 bit float option whereever possible.

Hope that helps.

hugh

--------------------
Technical Editor, Sound On Sound