New Musical Journey – Part 02

The kind of joke I am capable of making is to rewrite the entire article to summarize it, just to say that respecting the instructions somewhere… It’s a bit like in the Legion series (Marvel by FX Studio), when the Shadow King (David Negahban) tells David Haller (Dan Stevens) that the Gods don’t follow the rules, they write them… Well, here too, I can decide not to respect a summary for the pleasure of not respecting it and that’s what I’m doing since I don’t respect the principle of the summary of the previous article… Or maybe I particularly like long introductions.

Anyway, in the previous article I told you about the difficulty for a musician who has inherited several musical repertoires to be able to reuse them, highlight them or even simply exploit them. It must be said that the means exist but that in all cases it takes time, a lot of time… And for a result that will essentially depend on one’s own ability to play or program each instrument. Suffice to say that apart from seasoned and experienced musicians, there are few people who are capable of it and I am not ashamed to say that I am not capable of it myself.

So, after explaining to you that in addition to the musical repertoires of 5 different groups I also have my own archive containing more than a hundred demos and that for 16 years I have been looking for a way to honor all this work without success, I ended the article by telling you that I came across SUNO and that AI-assisted creation clearly changes life. And indeed, it changes life but it also complicates it quite a bit! You’ll see, it’s going to hurt…

At the same time, I’m not going to bother with that anymore… I already know what it’s like to be at the forefront and to explore new paths under the astonished gaze of people who observe me and don’t understand why I allow myself to try things that, a priori, have no place being there… For that I refer you to the Shadow King’s response to David Haller, knowing that when you are a creator at heart you are the only master on board so you can allow yourself to make your own rules and experiments, provided however that you respect the more general and societal framework in which you evolve… The freedom of some stops where that of others begins!

Ok, so for those who are not yet familiar with my workings, know that when I make a sequel to something that I have cut into several pieces, I do not start again at 1. It makes no sense to do that since it is the continuation of the previous one and if you want to glue the whole story back together you will have to bother renumbering the parts while making sure that it is the right page whereas by doing what I do, it is weird but at least to rebuild it goes faster and it is clearer. Does that seem stupid to you? Go tell that to Korn, come on! On their album “Follow the Leader” the first track is number 13 and it is not even the continuation of the previous album… and that poses a problem for anyone?

It’s not like they came up with their band name after shitting on a corn cob…

So to resume, here I am immersed in SUNO, to make many songs with version 3.0 of the tool… I admit, I spend an infinite amount of time there but the result is more than correct… When I manage to agree with myself. Indeed, everything is not perfect and sometimes I hesitate on the style to give to a title, which means that I will have it redone several times until I find the sound that will convince me… And with 50 credits per day, it does not progress quickly. So I realize that I can maximize my production capacity… Yes, because to do things well, I have a full discord address, an old hotmail address, a more recent outlook address, a gmail address by obligation (thanks Android devices) and a second gmail address which is used by a bot on Star Wars: Galaxy of Heroes to allow the alliance to have visibility on its statistics. Add another professional email address that also has access to SUNO and in reality I have 6 accounts to make songs… And I didn’t even have to open them specifically for that: I had them well before discovering SUNO so… Did I cheat or did I use the resources at my disposal intelligently?

In any case, the result is that with 6 accounts I end up with 300 credits per day and that’s comfortable for composing! From April to June 2024 I manage to produce 57 valid songs from my personal repertoire with even a few new ones. The thing is, I refuse to use the “prompt” to make compositions. I find that as a musician there is no added value in generating music without having given at least your own lyrics and the structure of your composition in addition to choosing the musical styles… Too bad you can’t bring your own pieces into SUNO to have them completed by the AI… Indeed, some pieces can never be remade and completed, especially instrumental songs. Especially since what defines an instrumental song is the sound and chords that you used and if it is impossible to capture them to generate them then you lose part of what you could use. In any case, version 3.0 is already convincing even if to make pieces longer than 2 minutes you have to assemble several generations knowing that I am a fan of pieces longer than 5 minutes… Needless to say, it’s going to be annoying and believe me, it was!

Putting together small pieces to make bigger ones… the complete opposite of what you see here. 😉

So after the 20th title I am seriously considering taking out a subscription to SUNO, but at the same time I am doing my calculations. In version 3.0 I clearly need 5 generations on average per song. Each generation costs 10 credits. That gives me a minimum of 50 credits per song knowing that it is 2 generations at the same time. However, I am considering taking out a PRO subscription at 2500 credits per month. At 50 credits per song, that still gives me 50 songs that can be done per month! So, I am reviewing the possibility of a subscription several times and there is something that worries me a little… Indeed, I said that for each generation I need 10 credits since the generations are done by 2 x 5 credits each time on the “Basic” version of SUNO. So I will let you look at why I was reluctant to take out a subscription and you will tell me if you understand the same thing as me:

SUNO’s offerings, as offered.

So when I see that I understand that with my “Basic” plan which costs me nothing, I can receive 50 credits per day, which is the equivalent of a song for me. At SUNO, they consider that a generation is two songs! Yes but no, 2-minute songs are not everything either… It’s sure that if each generation counts for a song with 50 credits we make 10… now if we are realistic, we need on average 50 credits to generate a complete song with the generations that we are not going to like and those that are going to go crazy and be unusable or those of which we will only keep 10 to 40 seconds! So already when in the PRO subscription we are told “you can make 500 songs” understand that you can make 100 at best. Except that in your BASIC subscription, you make 2 generations at a time and in the PRO and PREMIER versions? It clearly says “10 running jobs at once”, right? So if we launch the generations by 10 instead of 2, then suddenly we need 50 credits to make a single generation, with the possibility of having 9 useless versions of the same song?

Presented like that, it’s sure that it doesn’t sell the dream at all! 50 credits per generation makes 50 pieces of songs and when you know that I need 5 on average to make a song that only makes 10 pieces per month! At that rate, we might as well stay on the “Basic” version… Except that in the “Basic” version, you don’t have the right to commercially exploit your generations and that’s a bit embarrassing… So I asked the question directly to the SUNO staff who took a few days to answer me completely off the mark! I asked them what was the point of taking a “Pro” subscription if my generations are done by 10 instead of 2, which will use up my credit 5 times faster… And their answer was “well in that case, if it’s not enough take a PREMIER subscription”. So yes, 10,000 credits is more comfortable but in the end it only makes 40 songs instead of the 10 that I can do per month with the PRO subscription knowing that with my BASIC subscription I can still receive 30 x 50 credits per day or 1500 credits. If I burn my generations by 10 credits, that makes me 150 pieces of songs or 30 complete songs per month knowing that I still need 5 pieces on average to make a song. Don’t you think that’s a bit of a screw-up somewhere?

And this is the effect it produces when I start talking about numbers…

So to summarize, to produce a hundred songs or more under the stated conditions, it will take me either 3 months of PREMIER subscription at $30 / month, or 3 PRO subscriptions at $10 per month. In PREMIER I “only” make 40 songs per month and in PRO I make 10 per month. That’s not much and I feel that there is something wrong with this calculation but it’s impossible to know what’s wrong… And the SUNO staff clearly did not answer my technical question when they potentially had another answer to give me that would have been more incentivizing… So I tell myself that it doesn’t matter, that I’ll find a way to finance all of this and that in the meantime I’ll work on my 6 accounts to release as many songs as possible that I will have regularized later. Indeed, I read somewhere in the general conditions that once you have a “PRO” or “PREMIER” account even if you go back to BASIC the songs remain commercially usable. So I tell myself that I would subscribe my 6 accounts once I have generated what I have to generate and like that it will allow me to regularize what I have already created to use it commercially. In the meantime, I will have to find a way to make the subscriptions at least partially profitable because otherwise I will not get by… If I want to be able to use what I have created I will still have to subscribe and so I start to think about the most appropriate way to regularize the situation.

So on June 2, 2024, I took the plunge and chose the Ulule crowdfunding platform! I submitted my project, explaining that I was doing crowdfunding to pay for my (or even my) subscriptions to SUNO in order to acquire the rights to commercially exploit my creations. Concretely, if the project is funded, I will be able to provide rewards (and everything has already been calculated, including packaging and shipping) and make donations to associations on the possible profits from sales, which assumes that the project is fully funded. Basically, I propose a project for less than 3,500 euros in total and I have planned digital and physical rewards. The next day, I receive the approval of the Ulule team… It’s done, my project is accepted, I will be able to prepare and develop everything so that it can come to fruition. I am happy but also a little worried… I am swimming in the unknown with a risk-taking that is certainly moderate but despite everything the risk of failure is still quite significant.

In 2021 I had already tried an experiment in the same genre although relatively different. In 2021 I was on the verge of publishing all of A&TDR on my personal website. A&TDR is a series of novels started in 2001 and written at the rate of one volume per year on average. In 2021, I have been writing for 20 years and I already have 20 novels that are still in a corner gathering digital dust. So one day, I decide to take the plunge into the interstellar void with the intention of getting a good cosmic slam in the face, or even an intergalactic slam. I take my worst manuscript, the one I wrote on a cell phone while I was on a toilet break in 2015 and which contains a very special story with a character who is not a hero at all. He is not even an anti-hero. He is downright bad and that is what is funny… Anyway, I send my manuscript and I expect to get a cosmic beating…

This is the general idea of ​​what a cosmic beating is!

So to be honest, the cosmic beating did arrive, but not at all as I had imagined. My manuscript generated interest precisely because it was a bit out of the ordinary and because it told a somewhat unusual story. Without going into details, I was offered to publish “Ac Dominum Malum” on my own account and there too I did my calculations… Given the amount I had to put up to launch my book, I would have to sell at least 300 copies to cover my costs. Well, do you know how many I sold in total? 3. The cosmic beating is there: the people I sent the manuscript to found it worthy of interest, at least enough to include it in their catalogue and then in the end it didn’t sell and that’s mostly my fault because I didn’t advertise the book at all. I didn’t prostitute myself on social media or even on YouTube to try to sell my book, even though knowing how to promote your work is part of the job. I’ve said it often enough: I don’t like being in the spotlight. If I have to do it, I do it by default, but otherwise I prefer to stay hidden in the shadows.

So obviously, it was a real commercial failure and I can only blame myself for not having defended my work enough… But on the other hand, it was just a test with the worst of my novels to give me a good reason to stop A&TDR once and for all and throw everything on the web out of spite after 20 years of working on it. In reality, this failure did me a favor because even if the book didn’t sell, at least it had potential and it was in the Éditions Vérone catalog for two years. For that alone, I can’t consider the experiment a total failure because it means that what I’m writing is valid and worthy of being in a publisher’s catalog. Knowing this made me want to continue my work on A&TDR and finalize the last remaining parts… I took a break from the “PrEvIoUsLy UnReLeAsEd” project but as soon as it’s finished, I could then finalize and fold A&TDR for good. Whether I take on a new cosmic beating or not, it doesn’t matter because I’ve already proven that what I was doing was valid so that’s enough for me.

And here is the book that you will no longer find anywhere else… “Ac Dominum Malum”.

That said, the failure of this book was predictable and it was for me a first test to also understand the mechanics of publishing and see how it works. It served as a validation test for me and it was passed successfully since the book was not purely and simply rejected. Could the fact that it is self-published have a lot to do with it? Probably, although I doubt that Éditions Vérone would let just any crap go through under the pretext that people pay to have it published. There is still a minimum of quality to have to maintain credibility as a publisher so I don’t think they accepted my crap just because I paid for it… Apart from the fact that I think it’s the worst of my novels, it’s not without interest… It would even be rather interesting in relation to the point of view of the character who shares his thoughts without restraint as if we were in his head, which is a rather special immersive experience, I must admit. That said, if I managed to accept and digest the failure of this book project, I think I would have more difficulty dealing with the failure of the “PrEvIoUsLy UnReLeAsEd” project.

The main reason is that behind this project I would still like to try to do a good deed for causes that are really close to my heart and if it doesn’t work out then I could always say that I would have tried but it would leave me with a bitter taste of failure because I would really feel like I did all that for nothing. This is perhaps what is bothering me the most at this stage: we are not necessarily obliged to like what I do and at the limit, if people tell me “I hate what you do” I could not blame them because tastes and colors are not discussed… But in this case instead of not putting a few euros into my creations, if you could directly make a donation to an association that would suit me very well too. The only difference here is that I offer content in exchange for a payment of which a part of the profits will be donated in the form of donations whereas when we make a donation well… We don’t expect anything in return. At least, that’s how I always imagined things, even if the reality is a little different.

When your daughter’s favorite song is “Coucou Hibou” the choice of platform is no longer really a choice…

I don’t usually compliment the people around me, but I have to give credit where credit is due: the Ulule team is really there to answer my questions and give me advice on what I’m proposing. I have a lot of discussions with a person from the team in charge of my project and his advice is not without common sense. It must be said that when it comes to communication and marketing, I’m really bad. I know that you can’t be good at everything, but here I’m completely crap and I don’t feel comfortable in this area at all. Communication and marketing is clearly what I call “full light”. That’s what’s supposed to put the spotlight and focus on what we’re doing and here… How can I say it? I’m far from being an outstanding lighting designer. Fortunately, I am guided by the Ulule team who advise me and challenge me on certain points and very sincerely, this is the first time that I feel supported in one of my projects. It is so unusual that I even have trouble believing that it is real… Ok, they are only doing their job but they do it well and for once I have someone experienced at my side who knows how to tell me when I am deviating and that, believe me, is “priceless”.

I started by preparing my page and my rewards, and I made good progress but I had doubts about a certain number of things. Because while I was doing the work to prepare the crowdfunding project I followed the evolution of SUNO and the least we can say is that there was a change in the days before and after the start of my project. They have already moved to version 3.5 for the generation and apparently it seems that there are features that change on the PRO and PREMIER versions and that would not be present on the BASIC version… And there, I no longer have a choice. I base all the achievements of my crowdfunding project on SUNO… The tool is changing and I am going to take out a subscription without really knowing what is in store… And somewhere, the share of risk has just increased because if ever the tool were to reduce its features or change its terms, I would be likely to find myself in the water having to tell the people who would have supported my project that I misjudged the risks and that I will have to reimburse them all…

Who? Taxes of course! (this all smells like a tax audit)

But that’s not really how I see things, so I’m going to take a little break on Ulule and schedule the campaign until September 2, just to have time to subscribe to at least a PRO account and test the latest changes… That way, if my project ever crashes due to a lack of tools, I’ll be able to prevent carnage from happening… You’re either a project manager or you’re not: my daily job is to anticipate, to have a short, medium and long-term vision, to find solutions or workarounds when there are no solutions and to always prepare for the worst-case scenario. There’s also a part of communication, negotiation, arrangement but on the other hand I don’t do marketing. That’s the role of a salesperson while the role of the project manager is to place themselves between the end client and the teams and to make sure that everything goes smoothly and according to the rules. He must referee, be educational and adapt to his audience. In short, I will take a PRO subscription and I will know exactly whether or not I can continue my project…

Alongside my generations of madness, I am looking for a way to make my subscriptions profitable and I have a completely idiotic and crazy idea. I take a Microsoft Excel file and I start to make tables… 10 then 12. Here is the concept: in 10 years of music, I have produced almost one album of demos per year or even a little more. So I can redo everything in the form of a unique collection of 10 to 12 albums; I checked, I have enough content to not run out of demos before the 7th or even the 8th album knowing that SUNO opens wide the doors of creation assisted by artificial intelligence and that concretely for a guy all alone who does not know how to play all the styles, who is limited in free time, who does not master all the instruments and who will cut back on his hours of sleep to carry out his projects well it is ideal. And to top it all off, I already have a name for the project since my demo compilations from that time were already called “PrEvIoUsLy UnReLeAsEd” and they were numbered, sometimes even with their own title (from number 4 to number 10).

I have a logo, a start to a project with enough raw material to not run out right away and what’s more, it’s part of what I’ve wanted to do since I was forced to stop making music! What more could you ask for? But before launching a project, I’m going to do what I do in my everyday professional life: I’m going to find out. Because yes, it’s not all about trying to get funding and sell your music… Maybe you have to think in terms of support. So OK, the dematerialized could do the trick but for that we will have to go through online sales platforms and I have absolutely no idea how to go about doing that… On the other hand, I know exactly how to go about making physical format (CDs, Vinyls, USB sticks) and so I start my search for suppliers and my simulations on Microsoft Excel to know how much it costs me to press 300 CDs, to repackage them in order to be able to send them back by post. Same for vinyls and USB sticks. In the end I have 3 suppliers that I have kept for whom I have my prices. I still add the cost of the subscription and a small safety margin in case the prices change between the beginning and the end of the project and I’m ready!

I can assure you that this is not the most pleasant part of the job but it is absolutely necessary before considering a crowdfunding campaign with rewards. You have to take everything into account; supplier costs, packaging costs and shipping costs and calculate based on mass and quantities to make sure I have a consistent price. My main goal is clear: I would like to rework my old demos and make them evolve using artificial intelligence. But intellectually it works on me because I am a musician first and foremost. I know what the cost of effort is when you compose songs, play them with a band and it takes 20 to 30 rehearsals for the song to be well put together… I know what it’s like to defend your compositions in front of incredulous people who are discovering you and who are not necessarily going to adhere to what you do… I know what it’s like to go on the tour of concert bars to try to get a date and then to do your own advertising campaign (and at the time, the support without social networks was the basic homemade flyer that ate up your printer cartridges) everywhere in the area where you play, to contact the media in the hope of getting a bit of coverage… In short, I’ve experienced all that and somewhere I haven’t forgotten what an effort it takes…

The image I have of the value of effort through the myth of Sisyphus.

And me, what is the effort that I make compared to real musicians when I use a tool that certainly requires time and involvement, but much less than what it takes to release a real album of totally organic music (with real instruments)? Here too, I know what it is and what it feels like to spend hours recording, mixing, remixing, re-recording, arranging… I did that for quite a while and I admit that it is extremely time-consuming. So I understand very well all these revolted musicians who say “but guys, do you realize that with AI anyone can do anything and that we who are struggling like crazy for real find ourselves in direct competition with people who may have never touched the slightest instrument, who know how to make prompts but who have no real experience of music?” I can’t say that they’re not right and that composition by artificial intelligence won’t create problems at this level. A priori there should be enough for everyone but in an increasingly saturated offer it becomes difficult to be able to stand out and even live from your passion. And users of artificial intelligence should not compete with real musicians. And that’s where I realize that I find myself right on the border between the two!

This is what happens when you are torn between opposing choices… You end up burning yourself to death!

It’s complicated for me because on the one hand I totally understand that real musicians can complain and find that getting help from Artificial Intelligence is distorting the concept of music itself since we no longer need to bother learning music theory, chords and an instrument… We don’t even need to rehearse or make an effort: the machine does it for us practically effortlessly. On the other hand, I, who am a musician and who could hardly do otherwise, tell myself that concretely I don’t owe them anything either. My part of the job has been done and there as a creator if I decide to use artificial intelligence as a composition tool given the possibilities offered by the tool it would be really stupid to deprive ourselves of it because we have scruples about people who are not wrong but who are not necessarily in the same situation and who do not necessarily want to admit that there is real potential to use artificial intelligence for those who have limitations and who have struggled for years. So I decided to split the difference (literally and even laterally): Half of the gains that I could make on the profits will be donated to associations.

I would have a better conscience if I knew that I was stealing the bread from the mouths of my fellow musicians for a good cause. Of course, I am becoming a competitor but in a totally different category. I do not have the ambition to become famous or to make a career in music. No, I thought about that when I was 20 and I wanted to do something constructive with my life. Creating songs and playing them is something quite constructive for me… I never had this need to be famous or recognized for what I do; I am a man of the shadows and I hate being put in the spotlight. Clearly it is not in my nature and it has been like that since I was a child: I think I have already said it but I am a lunar person, not a solar person. The night is my kingdom and it is in the darkness that I am most comfortable. What is also paradoxical and quite disconcerting is that when I was making music in a group I was more often on the front line than I should have been. Some colleagues didn’t like the first line at all and since someone had to do it, I was often designated by default to do that… Announce the songs, communicate with the audience… I wouldn’t say that I didn’t enjoy the experience, but honestly I wasn’t any more comfortable than my fellow players, except that with me it was less obvious.

That’s a very good summary…

So to summarize: I am going to launch a crowdfunding project that will allow me to market my demos brought up to date thanks to artificial intelligence by offering physical rewards. I have already prepared everything in advance and I have decided that from the profits made, I will donate 50% of the profits to associations so that it is used for something. This will be my way of contributing to society and making amends for my project that could cause harm to real musicians to whom I owe absolutely nothing and who will in any case consider that I make soulless shit while making them unfair competition. Ok guys, I know, it is assumed… Shadow King, do you remember? Well, now I don’t want to hear anything more and the case is closed. I will do what I have to do if I earn a few bucks on my projects and I will make a point at the end of the project but for the moment I have to put together the crowdfunding file and that is the first step that begins: looking for a platform that could accept this kind of project… No because as for musicians who get angry quite quickly when we talk to them about music generated by artificial intelligence, a musical collection project comprising several albums relatively full of songs could be a bit big to get across… but I will try: nothing ventured, nothing gained!

And there, it’s the same: I spent a certain amount of time and a certain amount of time before finding the participatory platform that accepts the type of project that I intend to set up. That is to say that even if artificial intelligence is becoming more democratic with Chat GPT, Copilot, SUNO and others that are increasingly easily accessible and that are gradually transforming our uses, there is still a huge gap in the systems currently in place, whether it is crowdfunding platforms or even music protection. Since we are talking about it, did you know that SACEM literally says “fuck” to everything that has been generated by artificial intelligence? For them, there is no pure creation by musicians with instruments: it is the machine that does it and that is enough to say that it does not deserve to be protected and that we absolutely do not give a damn. I would like to tell SACEM with all my heart that it can go to hell with its stupid decisions, since as a musician my instrument is artificial intelligence knowing that behind it I have already done the preliminary work and that from the moment we can upload real pieces of our own compositions it will clearly change the situation since at that time there will be real initial composition work that will be incontestable. Except that I am not going to say it at all and keep quiet because the day SACEM understands that it can also make money on songs generated by AI from the moment it agrees to protect artists who legitimately use it to complete an initial work, I will have to severely cough up with my little 10-album joke… Especially if they are also doubles but shhhh. SACEM, FRIENDS! ME NOT AN ENEMY! YOU DON’T SHOOT ON SIGHT EVEN IF I SAY BIGGER BULLSHIT THAN ME! Thanks 😉

Come on SACEM, I’m waiting for you!

No, but concretely, I can totally understand SACEM’s positioning in relation to music generated by artificial intelligence… That said, it’s evolving quickly and if tomorrow AI offers an upload of tracks to serve as a basis for the reconstruction of a composition, I want to say that this will clearly change the rules of the game once again. I’m not saying that it’s a good thing or that it’s a bad thing; in reality I think that things are destined to change and if possible to evolve and if all the changes are not necessarily positive, this one could well break systems that were sometimes in the interests of the artists, sometimes in the interests of those who shamelessly exploited them… So there are positives and negatives but we’ll see in practice, in the meantime, to have your music protected by SACEM it’s dead but… There are undoubtedly other alternatives that I will have to explore but one thing at a time. For now, it is the beginning of June and after having gathered as much information as possible, I am launching a crowdfunding project.

Come on, it will transmute dry (and not wet)!

I had a good and a bad surprise with SUNO and the PRO subscription. I suggest we start with the bad, like that at least it’s done and you can take the measure of what is called “having your project ruined”. So, when I read the general conditions of use of SUNO, I had clearly said that everything that was generated while you have a PRO subscription is marketable even if you then switch to a BASIC subscription, therefore not paying… We therefore retain the possibility of marketing what was generated during this period, which is very good. I naively thought that there could be retroactivity on what had been generated with a BASIC subscription when you switch to PRO? Well know that no, it is clearly not possible. What does that mean at my level? Well, just that I have just spent about 3 months completely building 57 songs that I will not be able to market and that therefore they cannot appear in the current project. Worse, some absolutely brilliant versions can never be remade identically and so not only do I lose these compositions but I will also have to make new ones that may not be as successful as the ones I spent so much time making. On June 18, 2024, my project will therefore start from scratch and everything that was done previously can be put aside permanently. There you go.

Afterwards, it’s my fault. I had read the general conditions of use of SUNO but not the FAQ! However, this information is in the FAQ and I was not curious enough to go and check it. On the other hand, it must also be said that at SUNO, communication is sometimes something very vague and very obscure. You often have to go and do exploration and archaeology to find information that is not necessarily highlighted… I’ll give you one that will make you laugh: do you remember this story of 10 generations at the same time that was weighing down the realization of my project? Well actually nothing is explained about this but when I took charge of the PRO subscription, I had my 2500 credits and I launched my first generation. 2 versions. At 5 credits per version, that means I only burned 10 credits. Did I get lied to???

For those who still have the reference… Otherwise you will understand pretty quickly!

Now that’s a big surprise! So generations continue to be done in pairs and not in packs of 10? So what didn’t I understand? Well, what you had to understand is that in the BASIC version, you can only do two generations at a time and you can’t restart a generation right away: you have to wait until the generation is finished to restart one, whereas in PRO and PREMIER, when you’ve just launched a generation you can immediately launch another one and up to 5 at the same time if you want… But each time, each generation consumes 10 credits and produces 2 versions… SO, with 2500 credits I can actually make 250 pieces of songs and since I need 5 to make a song I can virtually type 50 songs per month with a PRO subscription and that changes everything!

No need to have 6 subscriptions, especially since what was generated on the BASIC subscriptions will not be recoverable or reusable anyway. With a single PRO subscription per year, I can easily create my album collection and maintain a sustained production rate for 4 to 6 months, which will allow me to stop the subscription as soon as I have completed what I have to do and that is truly very good news because it also drastically reduces the cost of my project. But above all, what totally changes the situation is that SUNO has gone from version 3.0 to version 3.5. What does that change? Pretty much everything! First of all, the generation time. In version 3.0, we generated a first track with a duration of 2 minutes and extensions of one minute… To make a 5-minute song, it took 4 to 6 generations, and a lot of assembly time. In version 3.5, we generate a first track of 4 minutes duration and extensions of 2 minutes; this means in plain language that it takes 2 to 3 generations to make a 5-minute song…

Inside I was screaming with joy. It was late, the kids were sleeping…

But that wasn’t the only surprise I had, far from it… Because SUNO has also implemented the ability to upload a music file that can be used as a basis for generating a song. Of course, it is clearly explained that the file must belong to us, that we must have the rights to it and that it must not be something known or already protected by copyright laws, otherwise… in the general conditions SUNO can clearly delete the account and everything that goes with it because we do not respect the conditions of use. So it is explicitly recommended not to mess around with this feature. There, it is said but is it effective and how does it work? Well to start with, you can only upload a music file of 6 to 60 seconds, which is a bit “light” for certain pieces with a complex structure. Then, this music file will serve as a basis for the musical composition and so if we only put the main riffs in it, it starts with a patchwork of riffs without head or tail. We can only use one file of this type per song and the following generations are made as extensions of this music file knowing that SUNO is honest enough to offer a first generation that is proportional to an original generation of 4 minutes minus the time of the music file; basically, if the music file is 20 seconds long, the first generation will be 3 minutes and 40 seconds long while the following ones will be 2 minutes long. It’s classy, ​​right?

So, the difficulty now lies in the fact that in 1 minute, we have to succeed in giving a “clean” introduction to the song and the main riffs that we want to build without it distorting the whole… Needless to say, it’s a pain and most of the tracks have almost a 60-second instrumental introduction. So yes, it may seem really rotten but as you learn to use the feature you manage to do things that are more and more sustainable with sometimes really surprising and interesting things. I remember having discarded a generation of a song made from an original music file by saying to myself “yeah no, now I can’t find the original song” and after a few days coming back to this version and saying to myself “damn, this is probably the best mix you got of this song with the original music file! Ok, it doesn’t sound at all like what you originally did but in concept it still matches what you originally wanted to do and even if it’s different, when you listen to it… well it rocks! Clearly you have a real hit there even if it doesn’t have the shape you were hoping for”… And that’s the hardest part in fact: managing to accept that another idea or another concept could be better than what was originally imagined.

When we still have a nail to hammer in somewhere and a message to get across…

So, you probably remember what I said about SACEM? So although I literally shit on the Society of Authors, Composers and Music Publishers by saying that they do not want to recognize the protection of authors who use AI, I probably really shit in the glue and I explain. In reality, SACEM specifies that an autonomous creation by AI cannot be protected, which is a reality since there would be no author… But at the same time, artificial intelligence does not generate autonomously if there is a user behind it and especially if he uses his own words or now his own music. So I can in principle have my music protected by SACEM since I meet the criteria… That said, I would actually take the time to contact SACEM first to be sure that this is indeed the case because it would still be a big pain in the ass to pay the €100 membership fee only to have them tell me “uh… actually no, we refuse your compositions generated by AI because we think that as an author-composer you are worthless”. You never know, to be continued… As is this story about generating a composition from a music file present since version 3.5 of SUNO and initially exclusively reserved for PRO and PREMIER accounts.

The last point that seems important to me is that now at least I am reassured about the “PrEvIoUsLy UnReLeAsEd” project because I know that it will be possible from a material point of view. It was useful to check it and with all the changes that it induces, it changes a little what had been imagined at the start of the project but not that much in the end. Let me explain: to create physical counterparts I would still need to own the commercial rights to the music that I will generate and there I will save a little on the cost of subscriptions since in the end I would not need 6 subscriptions but only one subscription and the joke is that the “trial month” that I was planning to do well… I have already transformed it into an annual subscription so I can immediately get to work on the project and generate the final versions of the compositions that I will put on my album collection, which I will present in the next article because well, my articles are already quite full, right?