Hi everyone,
I decided to make a dedicated post to introduce the Sonarpad Audio Description feature, which is available on Windows, Mac, iPhone, iPad, and Android.
Just like I previously did on Windows and on Mac, I have brought to the mobile versions of Sonarpad the ability to create audio descriptions for any kind of content: movies, TV series, YouTube videos, personal videos, and much more.
What makes Sonarpad different is the way the audio description is generated.
Sonarpad uses an AI model to listen to the dialogue, understand when people are speaking, detect the silent moments in the video, and insert the audio description only during those gaps. The goal is to make the result feel much closer to a professionally produced audio description, instead of simply placing descriptions over the original dialogue.
There are two ways to use the feature.
1. Free mode with your own Gemini API key
You can enter your own Gemini API key and use the audio-description system for free.
However, with the free Gemini tier you may not always have access to the latest available model. Gemini may also temporarily limit free usage when there is high traffic or high demand.
2. Sonarpad AI
Alternatively, you can use the Sonarpad AI service. In this case, I provide you with a paid API key.
The costs are intentionally kept very low. Creating an audio description for an entire movie costs approximately β¬3.
Most of the processing is performed directly on the user's device, which helps keep the service affordable. Other AI audio-description services can cost the user many times more. For example, similar services such as the one offered by the developer of PiccyBot can have significantly higher costs.
The Sonarpad audio-description system also includes several advanced features.
You can create a character catalog, which is particularly useful when audio-describing an entire TV series because the system can recognize and consistently describe recurring characters.
You can also resume an interrupted audio-description project instead of starting again from the beginning.
Once the audio description has been completed, you can save both the video and the generated audio description, so you can watch or listen to it again later, including together with friends or family.
This feature has already been used extensively in Italy, and I currently have around thirty paying users in Italy using the Sonarpad AI service.
Sonarpad itself remains free.
I decided to introduce the optional paid AI service mainly because of the limitations of the free Gemini system. For example, several users had difficulties creating their own Gemini API key, while others encountered temporary free-tier restrictions.
In Italy, this feature has helped make something possible that many blind people had wanted for a very long time.
We have traditionally had far fewer professionally audio-described movies and TV series than English- or Spanish-speaking audiences.
With Sonarpad, blind users can now create audio descriptions for popular TV series and movies that have never received an official Italian audio-description track. They can also audio-describe YouTube videos, personal recordings, family videos, and many other kinds of content.
For me, this is one of the most important goals of Sonarpad: helping to fill a gap that has existed for years and making audio description much more accessible, flexible, and affordable for everyone.
For those who don't know the Sonarpad project yet, you can find all the information on the sonarpad.com website,
https://sonarpad.com/
, where you can download the version for Windows, Mac, and Android. For iPhone and iPad, you can find Sonarpad directly on the Apple Store.
Comments
this is awesome!
Thank you for continuing with this project. Great feature. I already use Sonarpad, and what I like about it is that you can describe without that typical censorship
Instagram
Does this work with Instagram reels?
Lord1995
Thank you for the compliments; in fact, the audio descriptions created with Sonarpad are unrestricted, because it is up to the user to decide which content to audio describe. Sonarpad has been trained so that if content cannot be described by Gemini, it will still provide a description, because it is necessary for a blind person.
Blackadder
Hi, yes of course, it is also possible to audio describe an Instagram reel. Naturally, you have to download it somehow. At the moment, you cannot download them from Sonarpad, but you can directly audio describe YouTube videos. If you download an Instagram reel and import it into Sonarpad, you will be able to audio describe it.
Great!
Iβm soo pleased thereβs a number of apps in this field that compete and try to make the process as cheep as possible but leave the quality high!
I have only one question: if Iβm discribing a video of a person talking nonstop, is there a way to make it automatically pause the video for the discription to be read? It would probably be an extra good option for more convenient discription of dynamic TikTok videos and things like that.
Also, what if the video is too short to be able to fit the audio track with the discription, will it keep playing the remaining discription after the video ends?
Thank you in advance :)
eyesfreesight
Hello, thank you for your questions. Yes, I have also thought about this scenario and there is only one answer. There is an option to insert exceptional pauses when it is not possible to find technical gaps during the audio description. This option is also useful for movies where there is too much dialogue and there is no room to insert audio descriptions. So, this way, Sonarpad will look for at least one free second and insert a pause there to add the audio description. This way, if a video has non-stop dialogue, Sonarpad can still insert the audio description.
question
hello, i downloaded the file Sonarpad-macOS-AppleSilicon.dmg, is the audio description featcher in this post in this app or is it a separate app
ziixies
Hello, the feature to create audio descriptions with SonarPad is available on Mac, Windows, Android, and iOS. You downloaded the Mac version, so if you go to Tools with VO M and navigate with the arrows, you will find the Create audio description feature. The first time, you will need to enter your Gemini API key, or contact me to get one. In addition, from Tools, if you open the Play streaming media feature, you can search for any video from YouTube. Next to each result you will also find Create audio description, which will download the video and take you to the window to create an audio description, with the video already prepared and all the settings ready to audio describe that video.
Ambro
Bug report and feature request
Hi, I've submitted 3 to 4 bug reports already using the report button from inside the SONARPAD app for iOS. I'm on the latest greatest iOS and SONARPAD versions.
First, can we please add the ability to connect to our Google Drive also? Right now I see dropbox but no Google Drive to fetch our files from.
Second, I want to use the SONARPAD personal AI service I think it's called, according to the post will cost money which is fine with me. But while I see and edit field to paste, my SONARPAD personal code for the AI service, I see absolutely no way to locate with VOICEOVER on how do they even subscribe or pay for the service, which presumably will give me a code which then again I'll have to presumably enter only one time in the foreman field. With the app developers please ensure that once set up user doesn't have to re-enter this code every single time and the app will remember this code for the personal Sonarpad AI service? Circling back, how the heck do we pay for this? SONARPADAI service, do we need the personal AI code then and if so, how etc.?
Third, under system/utilities, and then routes and also a couple of other places which I can't quite recall right now, the app must remember the previously selected as in when the app was used the last time for that particular feature, the country and the navigational mode. Right now every time it's set back to Italian and driving so it's a major pain to change that every single time. As I mentioned, this is also true among some other sub menus in the app currently when it all revert back to some default every single time I reopen the app which is not acceptable so the solution of remembering the last selection should be universally implemented across this app.
Fourth, NSONARTUBE to watch YouTube content, do we have the ability or rather does the app allow me as a user directly paste the URL of a particular YouTube video so that sonar tube can directly go to it and I can start playing it in SONARTUB along with all your description? It's a huge pain to have through go through search/research using the search field all over again in SONARTUBE because many of us including me have direct YouTube video page hyperlinks already saved somewhere in bookmarks on one of our devices.
Fifth, due to lack of any clear accessible instructions to help me with some of my points above and others, once I successfully have purchased and enabled the SONARPAD personal AI service on my iOS app, could you please explain if I should suddenly start seeing a start audio description or some such button appear inside SONARTUBE whether I am playing YouTube videos or playing video files that I already have? Would really appreciate some clear instructions and clarifications and expectations with regards to usage here of the personal AI service description.
Please note above I'm talking about the AI audio description in the couple of points above and not the text transcription of the video which is also one of the selections somewhere on the app.
Sixth, if I choose the edge DTS voice and currently I have chosen HTTS and then Libby UK English as my voice, should we expect there to be a lack time between what's playing on screen on the video in the app and the actual description as in the audio description read out? It says "HTTS (online)" so I'm wondering if this is like a cloud based TTS where the TTS processing is happening in the cloud which sounds frankly little bit scary to me because it directly means lag and slower at performance.
This is what I have for now and will continue looking for your responses here, please. Thanks.
Cannot fetch videos from iOS photo library
Hi, under media entertainment, using the ad videos button, the only options are what's available on the "files" app which is a native Apple app. Which means iCloud drive and file stored natively on the iOS device. Device do I see a way to import videos from the iOS photos app. Conversely, SONARPAD app does not appear on the iOS share sheet from within the Photos application and I'm trying to share a video sitting there. This needs fixing please. Most people when they take videos from the iOS device, it automatically get stored in the photo library inside the iOS photos app. Thanks.
Prateek Dujari
There are several misunderstandings here, so let me clarify a few things.
1. Google Drive
Sonarpad currently supports Dropbox directly. Google Drive can already be added to the iOS Files app, so files stored there can be accessed through the standard Files picker.
2. Sonarpad AI
There is no in-app subscription or payment page. To activate Sonarpad AI, you need to contact me directly. I will provide your personal code after activation.
You enter that code once. Sonarpad already saves it, so you do not have to enter it every time.
3. Routes and saved selections
I agree that options such as country and navigation mode should remember the last selection instead of resetting to Italy and driving. I will review this.
4. Direct YouTube links
SonarTube currently focuses on search. Adding the ability to paste a YouTube URL directly is a reasonable feature request and I will consider it.
5. AI audio description
AI audio description and video transcription are two different features. Once Sonarpad AI is activated, the protected audio-description functions become available in the supported areas of the app for videos and YouTube content.
6. Edge TTS
Edge TTS being online does not mean the finished audio description will gradually go out of sync. The narration is generated while the audio-described file is being created. Once the MP3 or MP4 is finished, the narration is already part of the resulting media. An online voice may make the creation process slower, but it does not introduce Internet latency during playback of the completed file.
Regarding the Photos report: videos from the iOS Photos library can already be used with Sonarpad. From Photos, choose Share and Save to Files, then open the video from Sonarpad.
Direct Photos-library selection or adding Sonarpad to the Photos share sheet would simply make this process more convenient, and I can consider adding that in a future update.
Requesting ways to get AI for the audio description from the app
Hi,
I am writing to request more information on how to get access to the Sonarpad AI for video description. I sent a message using the contact form on your website but have not yet received a response. I am located in the US and Iβm interested to know how much the code would cost, is it a monthly or yearly charge, and how much subscription can you get with the amount that you purchase. Please feel free to either respond through contacting me on the website here or replying to my feedback that I sent on your website. Thank you so much for the awesome work that you do and Iβm excited to try out this feature of the Sonarpad application.
Mike
Hi Mike,
thank you very much for your message and for your kind words about Sonarpad.
I did see the feedback you previously sent through the website, but unfortunately your email address was not included, so I had no way to contact you directly.
Sonarpad AI is a prepaid service designed for people who want to use the AI audio-description features without having to create and manage their own Gemini API account.
It is not a monthly or yearly subscription. You simply purchase the amount of credit you want.
As a general reference, creating the complete audio description of a full-length movie usually costs approximately β¬2.00 to β¬3.00, depending on the duration of the video and the amount of AI processing required.
Because the cost can vary from one video to another, there is no fixed number of hours associated with a particular amount of credit. However, even a relatively small balance can normally be used for several movies.
The credit does not expire. It remains available until you use it, and when your balance runs out you can simply recharge the same Sonarpad AI key. You do not need to purchase a new key every time.
The Sonarpad AI key is personal and works directly inside Sonarpad. It is not a general Gemini API key and cannot be used in other applications.
The service is already being used regularly by more than 30 users, both in Italy and internationally. For example, Prateek has already purchased and activated his own Sonarpad AI key.
Since I do not have your email address from the feedback you originally submitted, please either send me your email address or contact me directly at:
[email protected]
I can then send you the payment information and, after payment, activate your personal Sonarpad AI key and provide the instructions for entering it in the app.
Thank you again for your interest in Sonarpad.
Best regards,
Ambrogio
fyi
3β¬ is like 200 units of money in my country so no thank you. and if you download a video to your computer (and probably phone) then upload it to chat gpt, then ask it to describe it, it does scene by scene. at least mine does. Without losing 200 something
Sysskuu
Hi Sysskuu,
I completely understand that β¬3 can represent a very different amount of money depending on the country, but I would like to clarify one important point.
Sonarpad itself is free and will remain free.
You can already create audio descriptions completely free of charge by using your own Gemini API key. The only difference is that the free Gemini tier may have stricter limits, may not always provide access to the latest model, and Google may temporarily restrict usage depending on traffic or quotas.
Sonarpad AI is simply an optional alternative for people who do not want to manage their own Gemini account, billing, quotas, or API key, and want to use the best model.
The price is also not an arbitrary fee that I decided to charge for Sonarpad. It mainly reflects the actual cost of using the paid Gemini models required to process the video.
You are of course also right that tools such as ChatGPT can analyse an uploaded video and describe its scenes, and if that works well for your needs, that is absolutely a valid option.
The purpose of Sonarpad is slightly different. It does not simply return a list of scene descriptions. It analyses the original audio, detects dialogue, silences and available gaps, generates descriptions that fit those spaces, creates the narration, and produces an audio-described result designed to behave more like a traditional professional audio-description track.
This is also one of the main differences compared with simply uploading a video to ChatGPT. ChatGPT can describe the scenes, but it is not specifically using Sonarpad's silence-analysis system to determine the exact points where the narration should be inserted. Because of this, it cannot place the descriptions into the available silent gaps with the same level of timing precision that Sonarpad is designed to achieve.
So the paid Sonarpad AI service is entirely optional. Nobody needs to purchase it in order to use the audio-description feature.
Best,
Ambrogio