Request for this forum - Stop Locking Answered Questions

Ask a Question
1

Hi Imusician Team,

I’ve frequently encountered the same issues when looking at your responses to questions. You often consider a question answered, but many times the replies seem to be copy-pasted, sometimes not addressing the main question or redirecting to unrelated answers.

I suggest implementing a new system where multiple people can judge whether an answer is satisfactory. After reaching a certain threshold of agreement, the question can then be manually closed.

One major issue is that you haven’t addressed payout concerns related to YouTube CID. For instance, I haven’t received my missing 100 CHF from May 2024, and there haven’t been any updates on this significant issue.

Additionally, some users ask about AI-related topics in mixing and mastering, but the responses claim these are AI-generated, which is incorrect. As someone who works in Machine Learning and Data Science, I believe these questions should be answered by an expert, rather than using generic responses.

Another problem is with AMPS or credits. You’ve removed the option to use them for releases due to subscription changes, and you don’t offer reimbursements. To me, this feels like theft. It’s like buying a gift card, only to find out the company has removed all but one product, which may not be valuable to everyone.

I was initially very happy with this community and your services, which is why I accepted the yearly fee to continue receiving my royalties. However, your services are now facing numerous issues, including:

  • Product releases
  • Customer support
  • Payouts
  • Streaming analytics
  • Artist profiles

I encourage other users to respond to this thread as well, so the Imusician Team can improve overall and establish a more two-way communication channel, rather than just one-way communication.

4

Effectivement, cependant j’ai un abonnement amplify+ et en plus de payer plus cher globalement le service qu’autrefois, car je ne cours pas au model de sortie chaque mois que recommande Spotify… Je trouve ce forum très mal géré, d’autant plus que les “moderateur” qui sont des artistes comme nous, prennent un rôle de distributeur de priorité, ce qui devrait être au final le rôle d’Imusician ou de tous

Le forum au lieu d’être une foire aux questions, devrait rester ouvert à la communauté de répondre à tous et en tout temps… Reddit prouve que son système de voting améliore grandement l’efficacité pour voir les problèmes majeurs d’un groupe. Pourquoi pas tenter cette approche, et une fois les questions répondues, aussi proposer un boutton “This is Answered”, pour que chacun puisse voter et comparer sur la totalité des vues d’une question avant que l’admin décide de fermer le thread. La c’est vraiment mal organisé, avec des liens entre questions que les admins mettent et qui parfois doivent être répondue sur plusieurs threads.

7

@liamphanmusic,

One major issue is that you haven’t addressed payout concerns related to YouTube CID

→ This has been adressed multiple times. We are still waiting for YouTube to send us data & revenue but this is currently being delayed due to some changes in their revenue policies. We don’t know when this will be resolved unfortunately, so for now all we can do is wait.

Additionally, some users ask about AI-related topics in mixing and mastering, but the responses claim these are AI-generated, which is incorrect. As someone who works in Machine Learning and Data Science, I believe these questions should be answered by an expert, rather than using generic responses.

→ We are not here to discuss the nature of the technology behind the AI tools used to mix and master the tracks. The rules on our end are simple; if any AI-tool is detected in your release during the quality check, then we will reject it. We’re open to hear what you have to say about these tools (you seem to be qualified according to your work), but keep in mind that it won’t change the policies we apply.

Another problem is with AMPS or credits. You’ve removed the option to use them for releases due to subscription changes, and you don’t offer reimbursements. To me, this feels like theft. It’s like buying a gift card, only to find out the company has removed all but one product, which may not be valuable to everyone.

→ We understand this situation and we are currently working towards a solution for this. You will be notified in due time once we have more info to share on this topic.

I’ll leave this thread open as you requested, but to anyone commeting here: please use this only as a space to share your feedback and opinions, not to request help. If you require help on a release or anything else that requires the team’s attention, please open a separate thread so it can be managed properly by us and not lost amongst other comments.

Thank you.

4 Likes
8

This has been adressed multiple times. We are still waiting for YouTube to send us data & revenue but this is currently being delayed due to some changes in their revenue policies. We don’t know when this will be resolved unfortunately, so for now all we can do is wait.

I am sorry Carlos, but this wasn’t for me, multiple times I created a question about why you are talking about youtube CID while I am not making more than 300 streams from them. I am talking about 100CHF missing revenues out of 400, more than 25%, not 10CHF.

Please look up this threads: Payment Failure and your excuses - #3 by Debreu

Another problem is with AMPS or credits. You’ve removed the option to use them for releases due to subscription changes, and you don’t offer reimbursements. To me, this feels like theft. It’s like buying a gift card, only to find out the company has removed all but one product, which may not be valuable to everyone.

If you are looking towards a solution for this, then I can reconsider Imusician strategy of subscription yes, but please be more transparent about this before changing subscription of people… This is not how it should be in business, us consumers to have to alert you about those issues !

Additionally, some users ask about AI-related topics in mixing and mastering, but the responses claim these are AI-generated, which is incorrect. As someone who works in Machine Learning and Data Science, I believe these questions should be answered by an expert, rather than using generic responses.

I am not here to change your policy of course, but I am talking about multiple answers given about voice isolation, cleaning and others RX tasks done on stems, most of your answers (and colleagues) are clueless about what generative AI is, and why you should make it more clear in your policy what it is, and also propose a better pipeline for people to not fall into the wrong category of AI of course (so you have less work at the end also)

Thanks @Carlos for your answers, and also for letting this open. Please also consider the voting scheme for closing or not a forum thread / question, that might be worthwhile for all of us and you as well.

Best,

Liam

1 Like
9

Je me permet, si vous le voulez bien, de poser une petite question : mais comment diable fait-on pour détecter si une IA a été utilisée dans un morceau, vu que cela n’apparaît pas dans les metadonnées ? C’est fait au pif ? Veuillez excuser mon ignorance j’aimerais bien être éclairé à ce sujet. Peut-être pouvez-vous répondre Liam .

1 Like
10

il s’agit plutôt de tester les pattern, par exemple, dans un audio, ou de faire de la discrimination sur plusieurs modèles crée pour générer une AI par certaines entreprises leaders, et recréer un model de machine learning uniquement dédié à la classification binaire en “AI” ou “NOT AI” , donnant des données regroupées en entrainement et en validation. En soit, ce genre de modèle est à la base de ce qui permet au modèle génératif de mieux produire des outputs, mais utiliser un modèle uniquement dédié à la classification en label permet de mieux contrôler les faux positif par exemple ou le taux d’erreur de classification, en recherchant entre 1 et 2% en général. Cependant je ne peux pas confirmer que Imusician utilise un tel modèle, en soit dans leur intérêt car les artistes pourraient facilement commencer à contourner le système en place en faisant des recherches sur les performances du modèle en question.