Let’s say my project and sound card are both set to 48kHz, but I’m using samples recorded at 44.1kHz. When I place these 44.1kHz samples into my arrangement without converting their sample rate to 48kHz (as Cubase suggests), the audio plays faster. I understand why this happens, basically, for each second of audio, Cubase expects 48k samples, but the sample only provides 44.1k. To make up for the missing samples, it pulls data from the next second, essentially speeding up the playback.
So, my first question is: when Cubase asks to convert the sample to 48kHz, what kind of interpolation does it use to fill in the extra samples? I assume this conversion is an interpolation process to add the remaining 4k samples, but I just want to know what type of interpolation is applied “behind the scenes”.
My second question is related to the scenario where I don’t want to convert the sample rate and let the sample play faster, as it naturally does when placed in a 48kHz timeline. Is there a way to have the sample play faster but still maintain its original pitch?
Well not as part of the Import/Conversion process.
But once you have Audio in Cubase (which you should always convert to the Project’s rate) then there are a variety of ways to manipulate both the Pitch and Audio Event’s Length independently.
The simplest would be to set the Select (Arrow) Tool option to Sizing Applies Time Stretch and drag the Audio Event’s lower right corner to shorten the Event.
Thanks @Martin.Jirsak for the information.
Is there any documentation available online about Cubase’s internal resampling algorithm you are referring to?
Thank you @raino , for the extra tip… it might come in handy I guess. But, I agree, it’s best to convert everything to the project’s sample rate right from the start.
The good thing is that Cubase gives the user the flexibility to either keep samples at their original sample rate or convert them to the project’s sample rate. You can also choose whether to match the project sample rate to the sound card’s rate or not, which adds more flexibility. However, one need to understand what is happening when switching and changing… this flexibility can lead to issues if not handled correctly i guess. In contrast, other DAWs like Ableton automatically convert samples to the project’s sample rate and lock the project sample rate to the sound card’s rate, limiting that flexibility.
I remember having seen some graphics of different apps performing resampling. With the exception of one or two outliers (not Cubase) they were all pretty similiar.
It also seems to me that some 15 to 20 years ago somebody found a resampling algorithm that just worked very well with little CPU usage. It seems since then the entire industry is using that.
Resampling quality was a big issue in the 1990’s but is pretty much a non-issue nowadays. It’s just good.
