Q: I have a problem with longer recordings (for example, 45 minutes).
The transcriptionist correctly transcribed the entire recordingš . But, for example, "Simple Clear" can break off and not be continued. Can this be improved somehowā
Do you think of a way to summarize longer audio? So that it doesn't cut the summaries in half. And it continued where it stopped. . When can you take care of this?
To improve the transcription process for longer recordings where phrases like "Simple Clear" might break off, you can try the following approaches:
1. Break Down the Audio into Segments If possible, split longer peinado recordings into smaller chunks (e.g., 10-15 minute segments). Many transcription systems handle shorter clips more accurately, reducing the chance of cutting off words or phrases.
š¤ I miss easier editing and sharing. Something that has, for example, Letterly ā¾https://appsumo.com/products/letterly/ ā¾https://youtu.be/4Y7-shLFpu0?si=ZpfOGreQf5o_SZjv
Simple Clear format doesn't work well. š¤ Do you have any idea how to fix itā
Or maybe you could share Prompt that is responsible for it because generally this format doesn't work for me.
Q: Have you seen this:
https://appsumo.
com/products/letterly/
Interesting, especially these features: ā¾Native iOS and Android apps ā¾Screen-off recording ā¾Offline recording š¤ Are you thinking about something like thisā
https://appsumo.com/products/salad-transcription/ Could you please implement so that we can additionally use this API too: https://salad.com/transcription Relatively cheap than competing APIs.
A: It's eventually using same Whisper Large 3 behind the scenes, we are on Whisper v2 and will move to large v3 when OpenAI provides it directly via their key.
Q: Rising maintenance costs with the arrival of new users.
Did you calculate it correctly? š¤ Do you have a contingency plan in case costs exceed your living expenses? Maybe we need to raise the price on Appsumo? Or the possibility of using some alternative, cheaper technologies. Or the ability to connect your own OpenAI API Key code. š¤ It would be a pity to lose your solution: Speech to Note.
I also love Speech to note and how you constantly add more great features, thanks a lot! Like Jack said, bringing our own API key would be brilliant and a Win-Win: You guys save money since users using their own key pay for AI use, but your interface around it is still so much help and way better than anything else, I have tested a lot. And lots of users will love the own API key because it's better privacy for us ā really anyone who took the time to dig a little deeper should only use their own API key (that's why I use your transcription but sometimes not the custom prompts/formats, but copy it over to another tool and paste the result back in). Should be very fast and easy to implement ā give us a text field to enter our API key settings, and for every query you send when QUERY_Key is your standard API key, just add IF (USER_Key != NULL) then QUERY_Key = USER_Key Done.
Q: I have a problem with longer recordings (for example, 45 minutes).
The transcriptionist correctly transcribed the entire recordingš
.
But, for example, "Simple Clear" can break off and not be continued.
Can this be improved somehowā
abhishek_ux
Jun 28, 2024A: This we have tried and will highly depend on context window getting improved in future iterations.
Will keep you posted.
Share Speech To Note
Verified purchaser
This still doesn't work for longer recordings.
Do you have an idea or a plan on how to solve itā
Still haven't taken this up. We will see what to do for this one. For now it continues to be as is.
Verified purchaser
š¤
Maybe it should be divided into smaller parts and processed?
Verified purchaser
Do you think of a way to summarize longer audio?
So that it doesn't cut the summaries in half.
And it continued where it stopped.
.
When can you take care of this?
I don't have a timeline for this but we will try this in the coming days for sure.
Verified purchaser
š¤
Apparently, Google recently announced that it will be possible to process very long texts in their model.
Maybe use thisā
Quality of the output isn't that great with Gemini Pro atleast when it comes to creative outputs.
We are considering the direction of multimodality though in the coming updates.
Verified purchaser
This may only apply to longer transcriptions.
š¤
And only use it if it exceeds the word limit that could not be processed by Open AI.
Verified purchaser
And what does it look like with Claude 3?
To improve the transcription process for longer recordings where phrases like "Simple Clear" might break off, you can try the following approaches:
1. Break Down the Audio into Segments
If possible, split longer peinado recordings into smaller chunks (e.g., 10-15 minute segments). Many transcription systems handle shorter clips more accurately, reducing the chance of cutting off words or phrases.
Verified purchaser
What is the idea for this.
š¤
There is no answer or has this been resolved somehowā
Verified purchaser
Maybe you could use Simple Clear Deepseek for summariesā
š¤
Verified purchaser
Simple Clear
"I'm sorry, I canāt assist with that request."
š¤
Verified purchaser
Is this issue resolved yet or not?
Q: What about further developmentā
š¤
No new functionality or anything new has appeared since September 19, 2024.
Share Speech To Note
Verified purchaser
š¤ā
We are working on a new update. Coming soon. Apologies for the gap. We are not running away ;)
Verified purchaser
šš
Verified purchaser
And when is the iOS-App ready?
Q: What are the development plansā
Share Speech To Note
Verified purchaser
š¤
I miss easier editing and sharing.
Something that has, for example, Letterly
ā¾https://appsumo.com/products/letterly/
ā¾https://youtu.be/4Y7-shLFpu0?si=ZpfOGreQf5o_SZjv
Hi this has been taken care on mobile app for Android. iOS coming soon.
Q: šIt would be nice to have a notebook that collects all the knowledge
e.g. like https://notebooklm.google/
Share Speech To Note
Q: šØļøChat with Transcription
Would it be possible to implement a chat function with Transcription.
So that I can ask questions similarly to ChatGPT
Share Speech To Note
Q: Simple Clear - this format still does not work for longer recordings.
Cuts in half and stop. Are you thinking about doing something about itā
abhishek_ux
Jun 28, 2024A: We will have to rework on this for now it will have it's limitations.
Share Speech To Note
Verified purchaser
Maybe the problem is in Promptā
What do you thinkā
Maybe if we knew what it was like, we could experimentā
š¤
Verified purchaser
Any idea for thisā
Because this format annoys me that it doesn't work properly. š¤
Verified purchaser
Perhaps processing in smaller fragments could be implemented in such a case.
And finally combine everything into 1.
š¤
Verified purchaser
š¤
Hi,
Is there any way to solve this problemā
It's very frustrating when the summary ends halfway through.
Verified purchaser
Use maybe GPT-4o mini
https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/
Verified purchaser
Simple Clear format doesn't work well.
š¤
Do you have any idea how to fix itā
Or maybe you could share Prompt that is responsible for it because generally this format doesn't work for me.
Q: Have you seen this: https://appsumo.
com/products/letterly/
Interesting, especially these features:
ā¾Native iOS and Android apps
ā¾Screen-off recording
ā¾Offline recording
š¤
Are you thinking about something like thisā
Share Speech To Note
Verified purchaser
actually i get it and is good and easy to use. With speech to note i can not access i don't know what happened is buugy
Verified purchaser
Letterly is low comparing to Speech to note
Q: Moreover, i wanted that i play some video on laptop and it records it.
But i am not able to as it does not work without microphone thing. Is there any way i can use that?
Share Speech To Note
Q: āPossibility to add own Prompt extension.
This is an additional text, formula that could be added to your prompt to achieve what I want. It would be about clarifying the statement.
Share Speech To Note
Q: An interesting API has appeared recently.
https://appsumo.com/products/salad-transcription/
Could you please implement so that we can additionally use this API too:
https://salad.com/transcription
Relatively cheap than competing APIs.
abhishek_ux
Jul 6, 2024A: It's eventually using same Whisper Large 3 behind the scenes, we are on Whisper v2 and will move to large v3 when OpenAI provides it directly via their key.
Share Speech To Note
Verified purchaser
š
Verified purchaser
But with Salad Transcription API I could now buy 100h/month at once in Lifetime :)
https://appsumo.com/products/salad-transcription/
Q: Maybe we could also feature the option of only š¤audio recording without transcriptionā š¤ To quickly record ...
one by one on the dictaphone while away from home in the car, etc.ā¼ļøā¾āŖļø
Share Speech To Note
Q: What's nextā What are your development plans now that you are not on Appsumoā š¤
abhishek_ux
Jul 4, 2024A: We continue to develop next line of features. We will announce our latest roadmap soon via email.
Share Speech To Note
Verified purchaser
š
Q: Can I use this to dictate into my Outlook EMAIL or O365 EMAIL or into Word directly?
abhishek_ux
Jul 3, 2024A: As of now, no but since this can be used in browser, taking a copy of content and pasting it there should be simple enough.
We have android/ios app and chrome extension in plan next. But we are not thinking about addons for Desktop products of Microsoft for now.
Share Speech To Note
Q: Are you planning to change e-mail address to another oneā Maybe then I would buy additional codes.
š
abhishek_ux
Jul 3, 2024A: It won't be a straight forward piece but we will assist you with change Jack. So go ahead if you want.
Share Speech To Note
Q: šWhen you plan your foldersā I already have some notes and would like to organize them.
abhishek_ux
May 14, 2024A: We are working on it as we speak Jack. Looking into possibilities of how we can bring this on desktop mode first. And then on mobile
Share Speech To Note
Verified purchaser
š
It would be necessary to collect Speech to Note results in thematic foldersš
Correct will be on it for sure
Verified purchaser
It would be nice when starting the recording to decide in which folder and what topic it would be about.
To sort it somehow.
Verified purchaser
šFolders would be useful
Then you could implement š¤Automation for folders.
ā”ļøIf a new recording appears, then do this...ā”ļø
Q: Have you thought about a Chrome extensionā For example, so that I can reply to emails in Gmail šØ by speaking directly.
Inspiration:
https://www.getmerlin.in/feature/chatgpt-for-gmail
abhishek_ux
Jun 28, 2024A: Yes already working.
Share Speech To Note
Verified purchaser
Are you planning or already have a Chrome Extensionsā
Q: Mobile app already built up?
abhishek_ux
Jun 28, 2024A: Not Yet.
Share Speech To Note
Q: Rising maintenance costs with the arrival of new users.
Did you calculate it correctly?
š¤
Do you have a contingency plan in case costs exceed your living expenses?
Maybe we need to raise the price on Appsumo?
Or the possibility of using some alternative, cheaper technologies.
Or the ability to connect your own OpenAI API Key code.
š¤
It would be a pity to lose your solution: Speech to Note.
abhishek_ux
May 14, 2024A: Still early days. We will rather love the support of you buying additional credits š. That will be live soon.
We will stop LTD soooooooon.
Will try to invite people from community itself who have believed in us so far to get us to more people.
Share Speech To Note
Verified purchaser
And when can we expect additional creditsš³ to be available for saleā
The purchase of additional credits will be coming soon. By next update for sure.
Verified purchaser
I see you can buy it now.
ā¹599.00
But it is not converted into other currencies, e.g. USD.
š¤
Verified purchaser
I also love Speech to note and how you constantly add more great features, thanks a lot!
Like Jack said, bringing our own API key would be brilliant and a Win-Win: You guys save money since users using their own key pay for AI use, but your interface around it is still so much help and way better than anything else, I have tested a lot.
And lots of users will love the own API key because it's better privacy for us ā really anyone who took the time to dig a little deeper should only use their own API key (that's why I use your transcription but sometimes not the custom prompts/formats, but copy it over to another tool and paste the result back in).
Should be very fast and easy to implement ā give us a text field to enter our API key settings, and for every query you send when QUERY_Key is your standard API key, just add
IF (USER_Key != NULL) then QUERY_Key = USER_Key
Done.
Sure will look into it for sure.
Verified purchaser
In addition to your own OpenAI API key, will it be possible to connect other ones, e.g. Claude 3.5 from Anthropic is a good oneā
Yes we have discussed this in community and the option will be there soon. We will be working on it later.
Verified purchaser
šš
Q: Can you use this via your iPhone?
Like recording speech to text while you are on the go?
abhishek_ux
Jun 28, 2024A: Yes. It works on browser but you can use
Share Speech To Note
Verified purchaser
Yes š
Q: I bought one code when there was just 1 code at USD29, is it possible to buy a new one and add it to my current account?
abhishek_ux
Jun 28, 2024A: Yes
Share Speech To Note
Verified purchaser
š
You can purchase another code and add it in the panel.