So I'm sure this is a common question but here is the scenario. I know the higher the sampling rate the better the quality of audio, but if it is going to eventually be bounced at 44.100khz for a cd, is it still worth working at a higher sampling rate. I'm a producer in Hollywood working with pop/rock production and I would love your feedback. Thanks,
It's an issue that is argued.
I have to say, I wasn't very happy with some of the sampling rate assertions made in the video. Actually, if I were to be brutally honest, they annoyed me greatly. While higher sampling rates might be better in theory, they have not proven so in practice.
In the video, Martin Sitter makes the assertion that the higher the sampling rate the better. But that's just not true. Sampling rates at 192 have been a disaster, no matter what digital device you're using. Recordings at 192 sound weird, weird, weird, and nothing like the source. Frankly, I've been unimpressed with 96k as well, and never record at that sampling rate.
The object, in digital recording is accuracy. In my extensive testing in which I could easily A/B the source tracks with the digital replication, the most accurate sampling rate on multiple machines was 48k. Now, I'm not sure what to tell you as to why that is. It shouldn't be. But it is. And I don't know very many top engineers or producers buying into working at 96k. In fact, almost everyone I know has patently rejected it, save but a few.
Personally, I won't work at 96k for a variety of reasons, the biggest being, there is no up side. None. You're still going down to 44.1k in the end (and MP3!), and from my experience you do more damage in conversion. If you're going to work in sampling rates above 44.1k, then you should either be printing your mixes to an external device with pristine converters, or sending them to a qualified Mastering Engineer. Otherwise, you should be working at 44.1k.
Now, even if you (or Martin) patently disagree with my opinion on this, I would suggest that an equally opposite opinion, with no disclaimer as such, should not be presented in a tutorial that is supposed to be about the factual functioning of a specific program. To me, it doesn't belong in this tutorial. It belongs in a separate engineering tutorial.
The parts of the tutorial that show me the functionality of the program are great. So, I apologize for the tutorial criticism in the sampling rate thread. But I believe the large majority of my post directly addresses the question at hand.