Categories
Artificial intelligence AI-generated Audio Loudspeaker MidJourney Speakers

Loudspeakers Generated by Artificial Intelligence – How AI Can Boost Your Creativity

Artificial intelligence in loudspeaker design

Artificial intelligence is a tool that loudspeaker designers can use when they want to think outside the box and come up with modern and unique designs. We will focus on image-based AI in this text because that is easily available to the public and does not require engineering considerations. Although AI-generated images seem to take cues from existing designs, AI can merge features in a novel way. For example, merging loudspeakers and furniture or blending the loudspeakers with the interior design. Instead of specific technical or topological solutions, think of it more like a mood board or general family of concepts. The greatest benefit of artificial intelligence in the design process is generating a multitude of new concepts in a short period of time.

artificial intelligence loudspeaker
Clever combination of furniture and loudspeakers. Posted by Lucas D. on the Facebook group AI Loudspeaker and Stereo Design.

How to get started

One of the most popular tools currently is MidJourney, which creates “unique imagery from short text descriptions”. You can start using AI algorithms like MidJourney to generate new designs by simply coming up with your own unique prompts. The easiest way to start using MidJourney is to join one of their newbie channels, create your own server and invite the bot there (see video at the end of this article).

Another option is using DALL-E, which is oriented more towards generating art, but also image editing. DALL-E is developed by OpenAI, who are known for GPT-4.

A great upcoming alternative is Imagine: AI Art Generator by Vyro AI. It is available for iOS and Android operating systems. The free trial has video ads, limited settings and watermarks.

artificial intelligence loudspeaker
AI-generated image when prompting for a sphere-shaped loudspeaker.

Contrary to what some guides tell you, direct messaging prompts directly to the MidJourney bot does not work in the free trial version. You have approximately 25 free prompts (0.4 hrs GPU time) without a subscription. Upscaling, generating variations and repeating a prompt (“rerolling”) uses up your GPU time quota. The most inexpensive subscription plan is $8/month at the time of writing this and it gives you 3.3 hrs of GPU time and ownership of results.

If You are not a Paid Member, You don’t own the Assets You create. Instead, Midjourney grants You a license to the Assets under the Creative Commons Noncommercial 4.0 Attribution International License./…/By default, Your images are publically viewable and remixable.

MidJourney Terms of Service

Examples of AI-generated loudspeakers

Coming up with useful prompts can be challenging and the output is seldom what you had in mind. That’s the beauty of it though. To give you an idea of what kind of results can be obtained, we have listed some of the more successful prompts and their corresponding results. They are by no means the best that can be achieved, but rather a result of simply experimenting with AI. We have included some comparisons MidJourney vs. DALL-E to give you an idea of what kind of results to expect. First, let’s give both image-generating AIs something ordinary like:

/imagine small loudspeaker with scandinavian design in a livingroom
artificial intelligence loudspeaker
Results from MidJourney
Results from DALL-E
Result from Imagine

Next, let’s ask something more specific:

/imagine studio monitor loudspeaker with waveguide in recording studio environment
artificial intelligence loudspeaker
Result from MidJourney
Result from Imagine

Since RD Physics develops 3D-printed loudspeakers, let’s see how well the AIs understand that and how creative they get:

/imagine 3D printed loudspeaker in livingroom
artificial intelligence loudspeaker
MidJourney
DALL-E
Imagine

You can also upload images to your Discord server and use the URL of those as part of your MidJourney prompt or upload an image to DALL-E and create variations. Below you see what happened when we uploaded a photograph of our Circular Sound Eikosa speaker and placed it in another environment using MidJourney and asked DALL-E to create variations:

/imagine <image URL> on a table in a scandinavian livingroom white colour
artificial intelligence loudspeaker
Placing into another environment with MidJourney
Asking DALL-E to create variations

As you can see above, MidJourney will often mess up the shape of the original input photograph. For some reason, the pentagonal face of the speaker always becomes hexagonal and the icosahedral shape is also lost. You can also directly ask for a shape in the prompt (with varying success):

/imagine small icosahedron shaped loudspeaker in livingroom
artificial intelligence loudspeaker
MidJourney
DALL-E
Imagine

Future Outlook

We asked ChatGPT to explain the potential benefits of AI in loudspeaker design and here’s what it had to say (edited for brevity):

Firstly, AI can help to optimize the design of loudspeakers for specific applications. By analyzing data from simulations and experiments, AI algorithms can identify the optimal shapes, sizes, and materials for loudspeakers that are intended for different uses.

Secondly, AI can help to reduce the time and cost of designing and testing loudspeakers. With AI, simulations can be run to test different design parameters, allowing designers to quickly identify the best designs without the need for expensive and time-consuming physical testing.

Thirdly, by analyzing large amounts of data, AI algorithms can identify patterns and correlations that might be missed by human designers. This can result in more precise designs that perform better and are more consistent across different manufacturing runs.

Overall, using AI for designing loudspeakers can result in more optimized designs, faster design iterations, and more accurate and precise designs, all of which can lead to improved sound quality and performance.

ChatGPT

Conclusion

The easiest way to implement AI in loudspeaker design is to use a visual tool like MidJourney. It does not consider acoustics, materials, or 3D dimensional structures. Knowledge in those fields is needed to evaluate the feasibility of each concept created. Nevertheless, AI-generated images can be of assistance in the ideation phase when one would typically use sketching or other quick visualization methods to explore new designs.

artificial intelligence loudspeaker
Posted by David M. on the Facebook group AI Loudspeaker and Stereo Design

At first glance, these ideas may seem way too complex to fabricate. That is true if confined to traditional woodworking methods. However, 3D printing is a technology that can help in making these intricate designs become reality. The design freedom that 3D printing allows is not limitless, but unparalleled. Many of the AI-generated loudspeaker designs could not be fabricated economically in any other way. For advice regarding 3D printing in audio, we recommend our DIY blog and the Facebook group 3D Audio and Hi-Fi Projects. Have fun experimenting and remember that the AI only knows what it has been taught and only does what it is told.

Below is a video explaining how to set up MidJourney with Discord.