Podcast - Is AI mastering ay good?

In recent years we have started to see a huge increase in artificial intelligence in just about every industry. Within the music industry we have started to see it in the form of Ai mastering services such as Landr, emastered and Cloudbounce all of which boast instantaneous mastering services that provide a professional level master without the need for a mastering engineer. These companies are trying to provide a cheaper alternative for one of the most expensive stage of any release. the only problem is that they used to be, you can easily pick up a decent mix on fiverr for £50. So if this is the case then the question is, which option is better for producers and has the recent situation given these kinds of services the platform they needed to solidify there place in the industry. In this podcast we want to talk about exactly this as we try and work out there worth from both a professional and an amateur perspective. 

Throughout the podcast we felt that we came to the conclusion that they do have a certain place within the industry. For content creators like youtubers or other online presences, these kinds of services can be perfect as they will produce a decent sounding master without needing any kind of pro audio knowledge. But when you look at it from more of a pro audio perspective they seem to fall short on both quality and value for money, But as we have seen the reviews are massively varied. Some sources say they are awful while some cant stop giving their praises. It clearly shows that these AI tools have a lot more learning to do before they start completing with a real experienced pro. 

Creating the podcast was harder than expected with the recent situation, Corona virus meant that the podcast couldn't be recorded in the same room like they normally would. This meant that the recording had to be done remotely which wasnt ideal. The way we got round this was using the following format for recording at home. We decided to all set up our own personal mic's, mine was an SM-58, while Angus and George had NT1A's. We then ensured out mic outputs were being recorded through out own DAW's. once this set up was ready we then did our podcast through teams but used the recordings on our DAW's to create the final product. In this way it ensured that we could do a good conversation without being worried about any lag from teams or any low quality audio. Doing it this way meant there was a little more editing to do as i needed to line up all 3 channels of audio to ensure they were all in time but with a bit of fancy editing this worked well and to a high quality.  

Check out our podcast and let us know what your opinions are on these services.

 

references:

Deahl D, (2019) How AI is solving one of musics most expensive problems, the verge, blog, article, available at: https://www.theverge.com/2019/1/30/18201163/ai-mastering-engineers-algorithm-replace-human-music-production (accessed on: 10/12/2020)

Inglis S, (2016), Landr, cloudbounce & the future of mastering, Sound on Sound, Blog, article, available at: https://www.soundonsound.com/techniques/landr-cloudbounce-future-mastering (accessed on: 10/12/2020)


Older Post Newer Post


Leave a comment