Some tips I have picked up while messing around with the AI model. In the text prompt box, describe the essence of the song you want Suno to generate. So include - Genre (pop, hip hop, ambient, etc.) Overall mood/feeling (uplifting, scary, raw, filtered, happy, sad etc) Tempo/pace (express as bpm, so anything from 40 bpm to 180 bpm seems to work but from testing I'd stay in the 70 - 140 range) Instruments (specify what you want to include, I have not been able to exclude. e.g. Drums, Bass, Piano, Strings would include in varying degrees all of the instruments but may also add others as well) Song theme and topics (Literally e.g. a song about jack running up a hill to fetch a pail of water but falling down the hill and injuring himself, on a rainy day in London) Length (define a duration from 5 secs to 10 minutes (1 minute) on free package but they can be stitched to make complete duration) Time Signature (choose a time signature like 4/4 or 6/8) Key (pick a musical key like A Minor or C Major) Additionally in the lyrics box, try things like - [start] [Intro] [Verse] [Chorus] [Bridge] [Pre-chorus] (record scratching) *if you put (la-la-la-la) here it will be performed by backing singer(s) or (bird song) you'll get just that. [Drum Solo] [Guitar Solo] [Intense Guitar Outro] [solo vocal, outro] [Outro] [Fade Out] [End] These are just some that seem to work (usually) but I am sure there are many other prompts I have yet to try. P.S. Fadr is the perfect companion to Suno. Have fun 👍
@@allyourtechai Here is some more tips and tricks. Please feel free to use any and all of them in any forthcoming tutorials as tutorial videos are not the right fit for my channel and I am more than happy to share the information I have gathered. Lyrics Prompts: Suno can understand a variety of prompts in the lyrics input field, including (but not exclusively): Lyrical structure: You can use square brackets [ ] to indicate different sections of the song, such as [Intro], [Verse], [Hook], [Bridge] and [Outro]. This helps the AI model understand the overall structure of the song and generate lyrics accordingly or allows you to arrange your own lyrics as desired. Specific lyrics: You can directly input lyrics for each section, using natural language and punctuation. The AI model will try to match the style and tone of your prompts. If you are aiming for a particular accent I find it works best by writing the lyrics phonetically. As an example if making a rap song you type "I saw him coming, I started running", it will sound less human than if you type "I sorrim comin, I startid runnin', I hope that sense. Sound effects: You can insert sound effects within parentheses ( ) to enhance the musicality of your lyrics. For example, (drum beat) or (guitar riff) or (whistle) or (I assume thousands of options) at any stage within the lyrics field. Even mid sentence. Style of Music Prompts: Suno can understand prompts in the style of music field to influence the generated song's genre, tempo, vocal style, mood and instrumentation. Here are some examples: Genre: Use keywords like pop, rock, hip hop, electronic, or classical to specify the overall genre of the song. Tempo: Express the tempo using numbers or words like slow, medium, or fast. For example, 90BPM or medium tempo. Vocal style: Describe the desired vocal style, such as natural voice, female vocalist, male vocalist, or robotic voice. Mood: Angry, Happy, Intense, Sombre, Uplifting, etc etc. Here are some recommendations for formatting said prompts in the Style Of Music box: Use commas to separate the genre, tempo, and vocal style, etc: This provides a clear and structured way to convey the desired characteristics of the generated song. For example, "rap, 90 BPM, natural voice" is a clear and concise way to express your preferences. Consider using square brackets for genre: This helps distinguish genre from tempo and vocal style, which can be particularly helpful when using longer genre terms. For example, "[Rap] 90 BPM natural voice" provides a more organized structure. Parentheses for sound effects or additional instructions: If you want to include sound effects or specific instructions, use parentheses. For instance, "rap, 90 BPM, natural voice (drum beat, guitar riff)" combines genre, tempo, vocal style, and sound effects.
you can actually go way further and add musical descriptions into the brackets, like for example: music styles: Trap, Bass-Heavy Massive Sound, Electro-Trap [intro, sinister synths, building suspense] [trap beats, rapid-fire hi-hats, dark undertones] [pre-chorus] your lyrics... [build, swelling bassline, siren-like synths escalating] [verse] your lyrics... [drop, explosive future bass drop, mind-blowing synth chords] [verse] your lyrics... [pulsating electro house rhythms, distorted bass] [build, drum and bass break, accelerating tempo] [drop, massive big room house drop, broad sonic impact] [bridge, brief respite with ambient pads, the calm before the storm] [chorus] your lyrics... [final drop - unexpected switch to double-time trap, overwhelming bass] [outro, instrumental]
I been playing around with this as well, trying to get an instrumental song out and it produced this in the lyrics panel: [Verse] (Drum beats building up) (Horns blaring) [Chorus] (Orchestral crescendo) (Dramatic strings) (Epic percussion) (Choir vocals) I played further, adding in (flute) (piano) (violin). All had influence on the generation.
I first started with Suno and I started using the brackets and parentheses from the beginning. I will use parentheses in the brackets with "chorus" to add a horn section during the chorus. You can even use brackets: [1-bar Guitar Riff]
I have created hundreds of songs.. while it is not too amazing 90% of the time, 10% of songs are mind blowing hits. Listening to real songs on the radio or top 40, i recognize lyrics and certain melodies that have me questioning if those real radio songs were generated by ai… really crazy.
Playing with it for a day I was like meh for ai these aren't too bad then like you said occasionally it just makes a legit radio banger and I was blown away like you couldn't tell me this song couldn't have been played on the radio it's unreal!
@@tornatotato91 it really is a role of the dice! The general production quality also varies.. but all songs still have a very mushy quality about them. Good for using as a framework
You can create thousands, all it takes is a couple of clicks. But what you're going do with those tracks you created? Sell it? Or share it with friends who also have their own music created by Sonu. I play around with it for month and already got bored.
this is awesome thanks for walking us through btw, “bridge” generally later in the song, after the 2nd chorus. anything between verse and chorus is called “pre-chorus”. Maybe this helps for arrangements.
Absolutely Suno won me over. I can honestly say a person’s that really understands song writing this tool is a serious game changer. And to add to that is have a producer enter in the dials as prompt instructions, by god Suno works wonders. ❤ this is an awesome video by the way ✌🏼😎
my favorite thing i've ever added to a punk rock song was at the end i put [Music ends] (Man sighs) (man says: I just want some freaking sleep!) it took a few tries but turned out so cool!
Been using Suno for a few months, and ill catch myself scrolling through my playlists, and jamming out to a song, go to find the artist, panic, and then realize, oh, I wrote that banger, and AI brought my lyrics to life
Didn't go through comments see if mentioned. But if you want to add a pause in lyrics it seems to work when uses ellipses.... And then.... Continue.... This worked most of the time for me. Also it seemed to understand using (pause) or (silence) (music break down) (intsrumental) (instrumental prelude) (instrumental interlude) etc. Does a pretty great job understanding these things. I tried using (chorus) (verse 1) etc. It works, but I found it shuffles the composition of the song around more often then. I think it knows where to place those in the music so it tries to do that more often. So I leave it off now, if I want my lyrics exactly as I put them. Still shuffles around, but not as often. Particularly with lengthy lyrics.
thanks for showing us this, and I think it was really cool how you artistically represented the rain falling down with the record scratching cuz that's sort of like what it's like when it's raining; if you were to imitate the audio of rain, it's like record scratching.
AI critics, lawmakers and users need to realize that creating a successful song on AI involves a lot of work. Listening to countless false first song stems, altering lyrics for better structure, writing lyrics that flow to begin with. Currently, the popular misconception is you drop a few words into AI prompt box and out comes a hit song. But it can do amazing things.
I’d see it as a tool or a motivational device. I’m currently working on a book and made a rock ballad for my main character, and it just gets me in the mood to write. I’d see it being something cool you could post to youtube, but nothing that’s really worthy of, say, being put in the rock and roll hall of fame or anything of notable repute. Again, it’s a tool, and should be treated as such.
I think it is like asking the C student in your class to do your work. They are more likely to think outside the box on the topic, know just enough, but you don't want to rely on them for your final grade. But those C students and their friends bought that they were super smart and secretly that A student who you were. No AI song is a hit yet because there is no soul in it yet. It can go into motions programed with algorithms just like a psychopath can model emotions without feeling them because they know they are expected but you know there is something not right when you look in the eyes. The AI generated comes off as an immature attention seeking teen. It needs the adult to impart wisdom. Not all human users have wisdom though, and they flood that first touch of potential to the public with content creating repulsion. You might work really hard but the primary content publishers of AI content seeking money and attention are in the same infancy stages of morality and lacking quality control as the developing programs which they are exploiting.
I'm a lyricist I post some songs I write with songs I let ai write and it's a perfect tool for me. Not everyone wants to sell their lyrics or change them to someone's style. With ai I have full control, even if it's with a few minor generation problems I usually find EXACTLY what I want
My brother have you found any way to remix the same song and keep the same melody because sooner AI messes app pronunciation of words and I want to go back and change the lyrics but keep the same exact melody as if I'm the composer and I'm telling my singer to just re sing the lyrics and I just want the computer to do the same as a human and you would think when sooner came out realize that AI wouldn't mess app your song you would want to go back and fix the same song and not have to make a beeside of the same song Have you found any way my brother to remaster the song and fix the lyrics????
Thank You! Other than using something like RVC after the song is generated is there a way to keep a consistent singer between songs in Suno that can be specified as a prompt?
When writing lyrics in Suno Ai, how can we direct the vocal sound by placing periods, commas or other punctuation marks?..When writing lyrics in Suno Ai, how should we use punctuation marks?
ho inserito un testo e ho ricavato una canzone ma non so come scaricare la parte musicale separata da quella cantata. per cortesia, se è possibile, mi puoi aiutare? altra domanda, ma se io metto in rete questa canzone ed ha successo i diritti di autore compositore a chi vanno. come bisogna comportarsi? grazie mille
1:00 for exactly that im here, hope im getting answere, i tried something like [180 BPM] or [SPEED UP] because when i put something in () instead of [] it just spells it out like addlips
Any tips on how make a better "Continue from this song" ? With both prompt and the timestamp. Because I had a first generated song and it was good however it fell off around the chorus. What I wanted to do was: continue from song, set the time before the chorus hits, and make the model sing the same chorus but the lyrics are adjusted (basically same everything about the first generated) I tried to prompt it to be the "same chorus as source" and the likes, but it still generates a totally different chorus. Hope someone can help
i love suno ai you use the STYLE MUSICfor tags to bring out lyrics if you do cover music. :) example accoustic, piano power ballad, melancholic, male or female voice, rnb etc etc
You say : Some tips I have picked up while messing around with the AI model. In the text prompt box, describe the essence of the song you want Suno to generate. So include - Genre (pop, hip hop, ambient, etc.) Overall mood/feeling (uplifting, scary, raw, filtered, happy, sad etc) Tempo/pace In what box do use this? Under custom mode there are these options: first Lyrics text box, then style of music, and lastly Title. Should also every metatag include brackets? Can you show examples of code?
Hi! I just watched your video "Suno AI Advanced Generative AI Music Prompting Tips," and I found your advice extremely helpful for creating complete songs with Suno AI. I would love to know if you could create a similar video focused on creating instrumentals with Suno. It would be great to have specific tips and techniques for getting the most out of AI in instrumental music production. Thank you in advance for your consideration, and keep up the great work!
I know AI can created background music and complete songs but is there a way to create your song (complete musical song) with the chords you have written with AI and not have the vocal so you can record the your vocals later? Would be great if the song could also be created as stems so they can import them into your DAW.
I have a question there is not enough tools for edit my question I creat a song not finish yet but I want the Same style for the rest of the song how I can continuo and finish the song using the same style I am so frustrated please answer
I wont share my examples but if you dont add a genre and make it an instrumental, prompt a percussive type, and prompts with effects processor parameters or synth parameters and ask it to not include melodies or chords at all, you'll get some sick generations from it. So if you know how to patch a synth or play with drum machines, this os a cool exercise.
It needs to give you the stems for the songs. Without the stems you can't change the tempo, mix or master it your own way, or make it more of your own. It's great for non-musicians that's a neat gimmick. But musicians who use it might want Suno as a starting point or to add specific parts of songs (for example, if you need a singer). I don't see why they can't add the BPM, key, or even MIDI for each song. I'm not complaining; certain genres sound great, while others are extremely generic. But my biggest grievance is it adds stock drums even when you ask it not to. The drums sound cheap, like if you audio-to-midi a drum track. I'd like to use Suno to get ideas and use certain parts but not create the entire song. I personally don't feel like it's taking my job because it shows how formulaic and generic a lot of traditional music is. I can see creative musicians incorporating AI to make something innovative. But as of right now, there's almost no creative freedom, and I want my music to be my own. They should also let you elaborate on what you like or don't like so it can narrow down what you want. Still, this is the most impressive AI I've personally experienced yet. It's funny to use to make wacky songs and some really have good parts. I'm sure it's going to get better and better and cater more to music pros.
What are you talking about - there are no stems for them to give you. There is no MIDI. There aren't any individual parts to the songs it generates. It doesn't work like that. If you want a certain bpm, then put it in your prompt. Learn the tool you're using, that way you won't expect it to do things it isn't capable of doing..
@@SineEyed That's my point. Clearly to make it they generate parts separately. A full song is a series of generated stems together. If they generate notes and chords that's the midi information. I'm aware the tool as-is isn't capable of doing this but obviously the technology is capable. Therefore I made a suggestion that would drastically improve their service. Why are you so upset? Lol
@@Soundbuds Yes. I came back to this video thinking about a comment an artist who makes his own music but wanted vocals sang by the AI and I just found stems on Suno and came back here. Very useful. You can create stems, crop sections and replace sections of the song. These should give musicians a way to clip and rework their music "stems". I'm not a professional and didn't know what stem was til I clicked it. No idea how long that's been there. There is a video feature available for IoS and soon to android I'd like to try as well. There is definitely a line between what amateurs and professionals can accomplish using things like Suno (in conjuction with things like Pro Tools or Ableton LIve etc), but that line is blurring rapidly. This is new technology for all of us. In time that line will blur to where it barely exists.
I'm on v2, but in the last 4 to 5 days I can't get it to sing in the language I entered as text, before that it worked 99% of the time, the only thing is sometimes an instrumental without singing, but never until now in a language I didn't enter like the text, they messed everything up with the new version, almost all my credit goes to nothing ... when I get it right and when I want to continue, it gives me the second part in another language, terrible
I saw a global top that simply described everything and the AI did. She asked for metal, she added, soon after, a folk, pauses during the solo. At first he asked her for a radio with static reading a message and then a countdown. I'm still confused about using brackets or paranteses.
Has anyone figured out how to take a song that Suno has made and make it instrumental while not using external programs? I'm asking because some stem splitters can't handle metal vocals well or multiple voices at once without destroying the song itself. If it could re-render the song without the vocals, that would be quite useful.
Thanks to Suno... the AI generates fantastic music... top in almost all areas and super fast... in contrast to UDIO, which has other strengths... in any case, my SHizzzo radio works great with SUNO
I only use Suno for the music and the vocals. The lyrics are always 100% mine. I use the final product as a "demo" to show my entertainer friends. The problem I'm havimg is they are afraid to use my songs because of the issues with AI. I can see the issue if the entire song is AI, including the lyrics. If I take what I do, and whoever I give it to records the music with their own instruments, thrn records it with their voice, it's no longer AI. Right...? Yes, no, maybe? I would consider the song AI inspired.
Either way, according to Sunos agreement, as long as you pay for one of their plans, (Anything but the free version) you own the rights to all of the music it created.
I committed 1 day to suno. I'm still working on it but making custom songs is super easy. Though, not 100%. Of these songs are radio worthy, most are. It takes practice but you can easily make a radio worthy song with a detailed prompt, or custom lyrics.
Hello, I would like to create short musical pieces in SUNO as JINGLES for commercials. Is it possible to create them within a precise 30-second timeframe?
try entering the desired lyrics without any command. like [verse 1] "your german lyrics" [verse 2] "Suaheli" worked just fine with me. It would even pick up german hillbilly accents.
I tried to generate lyrics with 5 languages using chat gpt4 and make a song with it. I was surprised by the good result (the song is whispers of liberty in my page)
🎯 Key Takeaways for quick navigation: 00:00 *Suno AI allows the creation of amazing AI-generated music; however, there's a lack of official documentation.* 00:14 *Custom mode in Suno AI helps users generate random lyrics to start a song.* 00:43 *The lyrics provided by Suno AI have a structured format with specific annotations like verses and choruses within brackets and braces.* 01:10 *Analysis of top tracks reveals various structural elements like bridges and outros that can be used to guide the AI in music creation.* 02:04 *Brace annotations allow for the inclusion of backup singers or effects like record scratching in specific song parts.* 02:58 *The tutorial demonstrates how to structure a song using Suno AI, with examples of intro, verse, chorus, bridge, and outros.* 03:57 *Instances of structures repeating unexpectedly in the AI-generated music, indicating some unpredictability in the output.* 04:53 *Demonstrating the effectiveness of specific annotations like record scratching in enhancing the musical piece.* 06:11 *Encouragement to experiment with Suno AI's capabilities and share discoveries and creations with the community.* 06:24 *Announcement of a running document on Patreon to keep track of findings and tips related to Suno AI music prompting.* Made with HARPA AI
I had a song that I felt the lyric I used might be a "trigger word" , so for safety I put (edit) instead of the actual word, and Suno put the BLEEP in it. Yes, I self-censored my own lyrics.
i am trying to extend the song in suno but when i give it to extend to verse 5 to 10 it starts to extend perfect but it leaves some verse in it any solution?
can you make it make like a spoken section, like an intro before the singing where people are talking, or a song that has a skit before the actual song
[heavy riff] for guitar solo in Genre rock / hard Rock is great . What I realize is that the quality goes down as longer the song is. When u have a part 3 or 4 the quality is no more good as at start. But depends also on genre. English is the best output. There is a lot more of commands what works well as shown here
What do i type into suno to get the beat from (africa bambaata planet rock) with the distorted 909 bass kicks of frenchcore??? Because I wanna make iiiiiAGGRO-BASS!!!!!
Keep extending out from a certain time frame. Say you click extend at the bottom there is usually a 1:00 change it to 0:45 or whatever is the best part to extend from. If it sounds off or you don't like it. I usually go back to the custom prompt and extend it again til I get the result I want.
Is there somewhere to find other people Suno songs? The home page is kind of weird, it doesn't seem like there is much on it for songs, I have made several songs public and I don't think anybody is seeing them, I wish there was a place for feedback and sharing
Yeah the base song it generates are not necessarily trash but it misses the link between words, feelings and intent, 3 things humans commenly change the meaning of. I found it works best to daisy chain instructions and to seperate the individual instructions with brackets in order to give it more complex instructions. I sucessfully made a song with both male and female vocals and in the right order but it was a fight and a half. For some reason it kinda ignores bracket instructions in lyrics but it pays more attention to the style interface. If you can figure out the proper inputs you can get it to do most anything from specific generas on specific verses to changeing vocals and pitch, tone and pace depending on what you are doing. The biggest draw back i have found is when its programming wants to argue about what i want it to do vs. The conclusion it came to, i messed up once forgot to give it specific details on a dual song and ended up with vocals that alternated by verse and gender for the whole track. Its wonky for sure but with time it will improve. Vi can never replace ppl bc not only does it not have a heart it also can't understand what that means, its literally impossible for a VI. Singers have no fear, the average preformer stopped writing their own stuff along time ago anyway. I can make music for me and thats good enough, atleast its original and not rinse and repeat radio.
I try to write a duet flr my gf. With a male voice and a female voice. But seems like no code I use makes 100% sure that the app will do it in the way I want. Does someone have a tipp for me? I tried to write [man] and [woman] or [male] and [female] the same also with [male singer] or [male voice] doesnt worked out for me. Does someone have an idea?
Is there a way to tell it to repeat the chorus or something? I feel like sometimes it just writes a whole new melody for each section, and isn't able to distinguish the whole verse/chorus/verse/chorus aspect.
@@BJ_MUSICandMORE sorry, no, in suno, it is no possible to do that, but if you have knowlege in music, you can post edit your song and change the chords using Other program like Reaper
@@elliotcruz7821 i am a musician...writing and recording songs...but i'm not a singer... I am searching for a KI which is using my music an my lyrics an do the singing...do you know somthing like that?
hello-- I noticed when you are create music by yourself and put the letters in a different language such as Hebrew most of the time the pronunciation of the words are not correct my questions is if I can change the word until I found the best person pronunciation and to keep the same music so we will not lose this music that I like. second thing it is frustrated is that they put the rhythmic Library in a way that it's hard to find kind of rhythmic you looking for and they don't give you even to write down so you can search for the it
Some tips I have picked up while messing around with the AI model. In the text prompt box, describe the essence of the song you want Suno to generate. So include -
Genre (pop, hip hop, ambient, etc.)
Overall mood/feeling (uplifting, scary, raw, filtered, happy, sad etc)
Tempo/pace (express as bpm, so anything from 40 bpm to 180 bpm seems to work but from testing I'd stay in the 70 - 140 range)
Instruments (specify what you want to include, I have not been able to exclude. e.g. Drums, Bass, Piano, Strings would include in varying degrees all of the instruments but may also add others as well)
Song theme and topics (Literally e.g. a song about jack running up a hill to fetch a pail of water but falling down the hill and injuring himself, on a rainy day in London)
Length (define a duration from 5 secs to 10 minutes (1 minute) on free package but they can be stitched to make complete duration)
Time Signature (choose a time signature like 4/4 or 6/8)
Key (pick a musical key like A Minor or C Major)
Additionally in the lyrics box, try things like -
[start]
[Intro]
[Verse]
[Chorus]
[Bridge]
[Pre-chorus]
(record scratching) *if you put (la-la-la-la) here it will be performed by backing singer(s) or (bird song) you'll get just that.
[Drum Solo]
[Guitar Solo]
[Intense Guitar Outro]
[solo vocal, outro]
[Outro]
[Fade Out]
[End]
These are just some that seem to work (usually) but I am sure there are many other prompts I have yet to try.
P.S. Fadr is the perfect companion to Suno.
Have fun 👍
Fantastic tips, thank you for sharing. I'm going to pin your post to the top so others can benefit as well.
@@allyourtechai You're very welcome 🤝
@@allyourtechai Here is some more tips and tricks. Please feel free to use any and all of them in any forthcoming tutorials as tutorial videos are not the right fit for my channel and I am more than happy to share the information I have gathered.
Lyrics Prompts:
Suno can understand a variety of prompts in the lyrics input field, including (but not exclusively):
Lyrical structure: You can use square brackets [ ] to indicate different sections of the song, such as [Intro], [Verse], [Hook], [Bridge] and [Outro]. This helps the AI model understand the overall structure of the song and generate lyrics accordingly or allows you to arrange your own lyrics as desired.
Specific lyrics: You can directly input lyrics for each section, using natural language and punctuation. The AI model will try to match the style and tone of your prompts. If you are aiming for a particular accent I find it works best by writing the lyrics phonetically. As an example if making a rap song you type "I saw him coming, I started running", it will sound less human than if you type "I sorrim comin, I startid runnin', I hope that sense.
Sound effects: You can insert sound effects within parentheses ( ) to enhance the musicality of your lyrics. For example, (drum beat) or (guitar riff) or (whistle) or (I assume thousands of options) at any stage within the lyrics field. Even mid sentence.
Style of Music Prompts:
Suno can understand prompts in the style of music field to influence the generated song's genre, tempo, vocal style, mood and instrumentation. Here are some examples:
Genre: Use keywords like pop, rock, hip hop, electronic, or classical to specify the overall genre of the song.
Tempo: Express the tempo using numbers or words like slow, medium, or fast. For example, 90BPM or medium tempo.
Vocal style: Describe the desired vocal style, such as natural voice, female vocalist, male vocalist, or robotic voice.
Mood: Angry, Happy, Intense, Sombre, Uplifting, etc etc.
Here are some recommendations for formatting said prompts in the Style Of Music box:
Use commas to separate the genre, tempo, and vocal style, etc: This provides a clear and structured way to convey the desired characteristics of the generated song. For example, "rap, 90 BPM, natural voice" is a clear and concise way to express your preferences.
Consider using square brackets for genre: This helps distinguish genre from tempo and vocal style, which can be particularly helpful when using longer genre terms. For example, "[Rap] 90 BPM natural voice" provides a more organized structure.
Parentheses for sound effects or additional instructions: If you want to include sound effects or specific instructions, use parentheses. For instance, "rap, 90 BPM, natural voice (drum beat, guitar riff)" combines genre, tempo, vocal style, and sound effects.
thank you
@@BKKnews 🤝👍
you can actually go way further and add musical descriptions into the brackets, like for example:
music styles: Trap, Bass-Heavy Massive Sound, Electro-Trap
[intro, sinister synths, building suspense]
[trap beats, rapid-fire hi-hats, dark undertones]
[pre-chorus]
your lyrics...
[build, swelling bassline, siren-like synths escalating]
[verse]
your lyrics...
[drop, explosive future bass drop, mind-blowing synth chords]
[verse]
your lyrics...
[pulsating electro house rhythms, distorted bass]
[build, drum and bass break, accelerating tempo]
[drop, massive big room house drop, broad sonic impact]
[bridge, brief respite with ambient pads, the calm before the storm]
[chorus]
your lyrics...
[final drop - unexpected switch to double-time trap, overwhelming bass]
[outro, instrumental]
Fantastic! Learning more from you all every day. Thanks for sharing!
Nice
I been playing around with this as well, trying to get an instrumental song out and it produced this in the lyrics panel:
[Verse]
(Drum beats building up)
(Horns blaring)
[Chorus]
(Orchestral crescendo)
(Dramatic strings)
(Epic percussion)
(Choir vocals)
I played further, adding in (flute) (piano) (violin). All had influence on the generation.
Do you do this in a single prompt on suno?
@@GaLzZy. in custom mode, not in prompt mode
I've already created a few songs with Suno, and they turned out great. The results have been impressive!
Based on your explanation, the following may apply to sonu ai
**Music description word classification:**
* **Mood:** Fast-paced, quiet, mysterious, scary, startling, happy, relaxed, depressed, depressed
* **Music style:** Blues, Tango, Samba, Salsa, Cumbia, Reggaeton, folk, school folk, Korean Gayo, Electronic, New wave, Hip hop, Jazz, rap, disco, 80s, Gospel, soft rock, lite rock,
* **Others:** Strong sense of rhythm
**Vocal Type:**
*female vocal
*Soprano
*mezzo-soprano
*alto
*tenor
* baritone
* male bass
**Musical Structure:**
* [cappella]
* [Duet]
*[choir]
*[start]
*[Intro]
*[Verse]
*[Chorus]
*[Bridge]
*[Pre-chorus]
*[Drum Solo]
* [Guitar Solo]
*[Intense Guitar Outro]
* [guitar, outro]
* [outro]
*[End]
*[instrumental]
(record scratching)
* [solo vocal, outro]
* [Fade Out]
* [latin harmony]
Suno doesn't give us enough space to do this. I only have 120 characters. How am I supposed to give it a structure?
@@Xgil2Play some music style can be added in your lyrics ( use [music style] )
@@arshan2008 So I don't have to exclusively use the lyrics section for lyrics? 🤔 I'll try it for sure.
@@Xgil2Play To be honest, even if there are many styles of control instructions, Suno AI will often ignore our instructions.
@@Xgil2Playuse custom
I've learned also that adding powerful or quiet depending on what effect your looking for can change the chorus etc.
I like that tip, gonna be using this for sure
creating music is my new addiction... thanks for the tips!
I've had some success using [Spoken] followed by words that will be spoken instead of sung.
I first started with Suno and I started using the brackets and parentheses from the beginning. I will use parentheses in the brackets with "chorus" to add a horn section during the chorus. You can even use brackets: [1-bar Guitar Riff]
I have created hundreds of songs.. while it is not too amazing 90% of the time, 10% of songs are mind blowing hits. Listening to real songs on the radio or top 40, i recognize lyrics and certain melodies that have me questioning if those real radio songs were generated by ai… really crazy.
Playing with it for a day I was like meh for ai these aren't too bad then like you said occasionally it just makes a legit radio banger and I was blown away like you couldn't tell me this song couldn't have been played on the radio it's unreal!
@@tornatotato91 it really is a role of the dice! The general production quality also varies.. but all songs still have a very mushy quality about them. Good for using as a framework
You can create thousands, all it takes is a couple of clicks. But what you're going do with those tracks you created? Sell it? Or share it with friends who also have their own music created by Sonu. I play around with it for month and already got bored.
Was thinking the same thing my man 👍🏻
@@Livekernaleventbecause you're not even try to sell it yet
this is awesome thanks for walking us through
btw, “bridge” generally later in the song, after the 2nd chorus. anything between verse and chorus is called “pre-chorus”. Maybe this helps for arrangements.
Great tip, thank you!
Thank you for giving me new ideas with SUNO AI!
Absolutely Suno won me over. I can honestly say a person’s that really understands song writing this tool is a serious game changer. And to add to that is have a producer enter in the dials as prompt instructions, by god Suno works wonders. ❤ this is an awesome video by the way ✌🏼😎
Great video. We create many of our music channel's song using Suno. It's an incredible AI tool for sure!
It really is!
my favorite thing i've ever added to a punk rock song was at the end i put
[Music ends]
(Man sighs)
(man says: I just want some freaking sleep!)
it took a few tries but turned out so cool!
Been using Suno for a few months, and ill catch myself scrolling through my playlists, and jamming out to a song, go to find the artist, panic, and then realize, oh, I wrote that banger, and AI brought my lyrics to life
Didn't go through comments see if mentioned. But if you want to add a pause in lyrics it seems to work when uses ellipses....
And then....
Continue....
This worked most of the time for me. Also it seemed to understand using (pause) or (silence) (music break down) (intsrumental) (instrumental prelude) (instrumental interlude) etc.
Does a pretty great job understanding these things.
I tried using (chorus) (verse 1) etc. It works, but I found it shuffles the composition of the song around more often then. I think it knows where to place those in the music so it tries to do that more often. So I leave it off now, if I want my lyrics exactly as I put them. Still shuffles around, but not as often. Particularly with lengthy lyrics.
thanks for showing us this, and I think it was really cool how you artistically represented the rain falling down with the record scratching cuz that's sort of like what it's like when it's raining; if you were to imitate the audio of rain, it's like record scratching.
AI critics, lawmakers and users need to realize that creating a successful song on AI involves a lot of work. Listening to countless false first song stems, altering lyrics for better structure, writing lyrics that flow to begin with. Currently, the popular misconception is you drop a few words into AI prompt box and out comes a hit song. But it can do amazing things.
Spot on!
Idk, I prompted it to make a blues song in b-minor about being constipated and giving up on life and it brought a tear to my eye soo..
Hopefully V4 will have some improvements, but give it a year or 2 it just came out in December
I’d see it as a tool or a motivational device. I’m currently working on a book and made a rock ballad for my main character, and it just gets me in the mood to write. I’d see it being something cool you could post to youtube, but nothing that’s really worthy of, say, being put in the rock and roll hall of fame or anything of notable repute. Again, it’s a tool, and should be treated as such.
I think it is like asking the C student in your class to do your work. They are more likely to think outside the box on the topic, know just enough, but you don't want to rely on them for your final grade. But those C students and their friends bought that they were super smart and secretly that A student who you were. No AI song is a hit yet because there is no soul in it yet. It can go into motions programed with algorithms just like a psychopath can model emotions without feeling them because they know they are expected but you know there is something not right when you look in the eyes. The AI generated comes off as an immature attention seeking teen. It needs the adult to impart wisdom. Not all human users have wisdom though, and they flood that first touch of potential to the public with content creating repulsion. You might work really hard but the primary content publishers of AI content seeking money and attention are in the same infancy stages of morality and lacking quality control as the developing programs which they are exploiting.
I'm a lyricist I post some songs I write with songs I let ai write and it's a perfect tool for me. Not everyone wants to sell their lyrics or change them to someone's style. With ai I have full control, even if it's with a few minor generation problems I usually find EXACTLY what I want
My brother have you found any way to remix the same song and keep the same melody because sooner AI messes app pronunciation of words and I want to go back and change the lyrics but keep the same exact melody as if I'm the composer and I'm telling my singer to just re sing the lyrics and I just want the computer to do the same as a human and you would think when sooner came out realize that AI wouldn't mess app your song you would want to go back and fix the same song and not have to make a beeside of the same song Have you found any way my brother to remaster the song and fix the lyrics????
There is, (cover song) @@Shirtcobain79
@@murkywatter where do you put this (cover song)? Is it on the style of music?
I've found that if you want an instrumental break, you have to use \[instrumental break\].
You can't just use [instrumental break] or [interlude]
Thank You! Other than using something like RVC after the song is generated is there a way to keep a consistent singer between songs in Suno that can be specified as a prompt?
Thank you! Made some ridiculous raps thanks to you. Cannot stop smiling
Haha, great to hear. It’s pretty addicting
This was REALLY helpful. Thank you!!!
When writing lyrics in Suno Ai, how can we direct the vocal sound by placing periods, commas or other punctuation marks?..When writing lyrics in Suno Ai, how should we use punctuation marks?
ho inserito un testo e ho ricavato una canzone ma non so come scaricare la parte musicale separata da quella cantata. per cortesia, se è possibile, mi puoi aiutare? altra domanda, ma se io metto in rete questa canzone ed ha successo i diritti di autore compositore a chi vanno. come bisogna comportarsi? grazie mille
1:00 for exactly that im here, hope im getting answere, i tried something like [180 BPM] or [SPEED UP] because when i put something in () instead of [] it just spells it out like addlips
which metatage do you use if you want one verse to sing in "Opera" and the next verse in let say "Rap" is that possible.
Any tips on how make a better "Continue from this song" ? With both prompt and the timestamp.
Because I had a first generated song and it was good however it fell off around the chorus.
What I wanted to do was: continue from song, set the time before the chorus hits, and make the model sing the same chorus but the lyrics are adjusted
(basically same everything about the first generated)
I tried to prompt it to be the "same chorus as source" and the likes, but it still generates a totally different chorus.
Hope someone can help
i love suno ai you use the STYLE MUSICfor tags to bring out lyrics if you do cover music. :) example accoustic, piano power ballad, melancholic, male or female voice, rnb etc etc
Can you edit the lyrics of a song that it has made? In other words, keep the same song musically but just with lyrics slightly edited.
That's the next big step it needs, to be able to edit songs
Yes you can.
Cover song
I made a banger by complete accident but what happens if you get popular catchy lyrics and melodies
You say : Some tips I have picked up while messing around with the AI model. In the text prompt box, describe the essence of the song you want Suno to generate. So include -
Genre (pop, hip hop, ambient, etc.)
Overall mood/feeling (uplifting, scary, raw, filtered, happy, sad etc)
Tempo/pace
In what box do use this? Under custom mode there are these options: first Lyrics text box, then style of music, and lastly Title. Should also every metatag include brackets? Can you show examples of code?
What style of beat did you use in your first song?
Amazing video! Thanks!
Thank you!
Hi! I just watched your video "Suno AI Advanced Generative AI Music Prompting Tips," and I found your advice extremely helpful for creating complete songs with Suno AI. I would love to know if you could create a similar video focused on creating instrumentals with Suno. It would be great to have specific tips and techniques for getting the most out of AI in instrumental music production. Thank you in advance for your consideration, and keep up the great work!
I know AI can created background music and complete songs but is there a way to create your song (complete musical song) with the chords you have written with AI and not have the vocal so you can record the your vocals later? Would be great if the song could also be created as stems so they can import them into your DAW.
I have a question there is not enough tools for edit my question I creat a song not finish yet but I want the Same style for the rest of the song how I can continuo and finish the song using the same style I am so frustrated please answer
I wont share my examples but if you dont add a genre and make it an instrumental, prompt a percussive type, and prompts with effects processor parameters or synth parameters and ask it to not include melodies or chords at all, you'll get some sick generations from it. So if you know how to patch a synth or play with drum machines, this os a cool exercise.
It needs to give you the stems for the songs. Without the stems you can't change the tempo, mix or master it your own way, or make it more of your own. It's great for non-musicians that's a neat gimmick. But musicians who use it might want Suno as a starting point or to add specific parts of songs (for example, if you need a singer).
I don't see why they can't add the BPM, key, or even MIDI for each song. I'm not complaining; certain genres sound great, while others are extremely generic. But my biggest grievance is it adds stock drums even when you ask it not to. The drums sound cheap, like if you audio-to-midi a drum track. I'd like to use Suno to get ideas and use certain parts but not create the entire song.
I personally don't feel like it's taking my job because it shows how formulaic and generic a lot of traditional music is. I can see creative musicians incorporating AI to make something innovative. But as of right now, there's almost no creative freedom, and I want my music to be my own.
They should also let you elaborate on what you like or don't like so it can narrow down what you want. Still, this is the most impressive AI I've personally experienced yet. It's funny to use to make wacky songs and some really have good parts. I'm sure it's going to get better and better and cater more to music pros.
I agree on everything!!
What are you talking about - there are no stems for them to give you. There is no MIDI. There aren't any individual parts to the songs it generates. It doesn't work like that.
If you want a certain bpm, then put it in your prompt. Learn the tool you're using, that way you won't expect it to do things it isn't capable of doing..
@@SineEyed That's my point. Clearly to make it they generate parts separately. A full song is a series of generated stems together.
If they generate notes and chords that's the midi information.
I'm aware the tool as-is isn't capable of doing this but obviously the technology is capable. Therefore I made a suggestion that would drastically improve their service.
Why are you so upset? Lol
I see they have the downloadable stems now!
@@Soundbuds Yes. I came back to this video thinking about a comment an artist who makes his own music but wanted vocals sang by the AI and I just found stems on Suno and came back here. Very useful. You can create stems, crop sections and replace sections of the song. These should give musicians a way to clip and rework their music "stems". I'm not a professional and didn't know what stem was til I clicked it. No idea how long that's been there. There is a video feature available for IoS and soon to android I'd like to try as well.
There is definitely a line between what amateurs and professionals can accomplish using things like Suno (in conjuction with things like Pro Tools or Ableton LIve etc), but that line is blurring rapidly. This is new technology for all of us. In time that line will blur to where it barely exists.
you can also put {cut & scratches} only put few words under neither...
How do I create commands in style or lyrics for a song in a Major key?
Suno sometimes singing my prompt - how to avoid it?
I wish they made this software a little more flexible. Mode options. Like put my own chords, tempo and patterns.
I'm on v2, but in the last 4 to 5 days I can't get it to sing in the language I entered as text, before that it worked 99% of the time, the only thing is sometimes an instrumental without singing, but never until now in a language I didn't enter like the text, they messed everything up with the new version, almost all my credit goes to nothing ... when I get it right and when I want to continue, it gives me the second part in another language, terrible
3:10 Its soooo Amazing vibe :3
what to do about ai voice? about replacing or correcting a word in it?
I saw a global top that simply described everything and the AI did.
She asked for metal, she added, soon after, a folk, pauses during the solo.
At first he asked her for a radio with static reading a message and then a countdown.
I'm still confused about using brackets or paranteses.
how do you add you own instrumentals and add a voice over it
Has anyone figured out how to take a song that Suno has made and make it instrumental while not using external programs? I'm asking because some stem splitters can't handle metal vocals well or multiple voices at once without destroying the song itself. If it could re-render the song without the vocals, that would be quite useful.
Thanks to Suno... the AI generates fantastic music... top in almost all areas and super fast... in contrast to UDIO, which has other strengths... in any case, my SHizzzo radio works great with SUNO
Sweet upload
I have problems with the accent. Is there a way to replace/correct the vocals with the wrong accent?
Tell Suno that we need a full set of Cheat Sheets
I only use Suno for the music and the vocals. The lyrics are always 100% mine. I use the final product as a "demo" to show my entertainer friends. The problem I'm havimg is they are afraid to use my songs because of the issues with AI. I can see the issue if the entire song is AI, including the lyrics. If I take what I do, and whoever I give it to records the music with their own instruments, thrn records it with their voice, it's no longer AI. Right...? Yes, no, maybe? I would consider the song AI inspired.
Yes 👍🏻 you are recreating an ai skeleton using a real live human band
Either way, according to Sunos agreement, as long as you pay for one of their plans, (Anything but the free version) you own the rights to all of the music it created.
ai is ai, regardless. you still had human input at some level, even with ai generated lyrics.
Robots aren't artists and you don't have any actual talent.
I committed 1 day to suno. I'm still working on it but making custom songs is super easy. Though, not 100%. Of these songs are radio worthy, most are. It takes practice but you can easily make a radio worthy song with a detailed prompt, or custom lyrics.
Also use the home songs for help with prompts or lyrics. Very useful to learn quickly.
I made a mistake in the lyrics. Is it possible to edit my lyrics and regenerate it? Cause I dont see it in the “Edit Song” option.
No, you would need to regenerate as far as I’m aware
Yes it is possible.
@@Aresmusic.official how?
Hello, I would like to create short musical pieces in SUNO as JINGLES for commercials. Is it possible to create them within a precise 30-second timeframe?
No
You should try giving a prompt for a 30 second jingle segment
Is it possible to switch language inside the same song? Anybody knows any command to do so?
try entering the desired lyrics without any command. like [verse 1] "your german lyrics" [verse 2] "Suaheli" worked just fine with me. It would even pick up german hillbilly accents.
I tried to generate lyrics with 5 languages using chat gpt4 and make a song with it. I was surprised by the good result (the song is whispers of liberty in my page)
🎯 Key Takeaways for quick navigation:
00:00 *Suno AI allows the creation of amazing AI-generated music; however, there's a lack of official documentation.*
00:14 *Custom mode in Suno AI helps users generate random lyrics to start a song.*
00:43 *The lyrics provided by Suno AI have a structured format with specific annotations like verses and choruses within brackets and braces.*
01:10 *Analysis of top tracks reveals various structural elements like bridges and outros that can be used to guide the AI in music creation.*
02:04 *Brace annotations allow for the inclusion of backup singers or effects like record scratching in specific song parts.*
02:58 *The tutorial demonstrates how to structure a song using Suno AI, with examples of intro, verse, chorus, bridge, and outros.*
03:57 *Instances of structures repeating unexpectedly in the AI-generated music, indicating some unpredictability in the output.*
04:53 *Demonstrating the effectiveness of specific annotations like record scratching in enhancing the musical piece.*
06:11 *Encouragement to experiment with Suno AI's capabilities and share discoveries and creations with the community.*
06:24 *Announcement of a running document on Patreon to keep track of findings and tips related to Suno AI music prompting.*
Made with HARPA AI
Can you help me. I used suno many songs but now I cant make custom what that mean😢
thank you very much!
I had a song that I felt the lyric I used might be a "trigger word" , so for safety I put (edit) instead of the actual word, and Suno put the BLEEP in it. Yes, I self-censored my own lyrics.
My problem is with the hooks trying to get the voices to hold a note
I search more for the promts Male/female voice and Autrotune off. The most voices have y
great video thank you !
cool , cheers from Portugal
i am trying to extend the song in suno but when i give it to extend to verse 5 to 10 it starts to extend perfect but it leaves some verse in it any solution?
can some one please tell how i get the best beat like free again been struggling for hours to get it is not the text I think but the beat
can you make it make like a spoken section, like an intro before the singing where people are talking, or a song that has a skit before the actual song
Looking for this. Please let me know if you got it.
[heavy riff] for guitar solo in Genre rock / hard Rock is great . What I realize is that the quality goes down as longer the song is. When u have a part 3 or 4 the quality is no more good as at start. But depends also on genre. English is the best output. There is a lot more of commands what works well as shown here
How do I merge two instrumental travks
What do i type into suno to get the beat from (africa bambaata planet rock) with the distorted 909 bass kicks of frenchcore???
Because I wanna make iiiiiAGGRO-BASS!!!!!
Thank you. Very helpful!😀
Glad it was helpful!
hi, it's have other commands to use like echo or delay or pitch shift like DJ thanks
Not that I know of, but they seem to be adding new features regularly
Thank you for your sharing!
do you have any tips how i can make the robot keep the same singer tune for each extend part ? and not change it
Keep extending out from a certain time frame. Say you click extend at the bottom there is usually a 1:00 change it to 0:45 or whatever is the best part to extend from. If it sounds off or you don't like it. I usually go back to the custom prompt and extend it again til I get the result I want.
cannyou put a singer who fit on your own beat without changing the beat?
What about harmonies???
Awesome! Thanks!
Is there somewhere to find other people Suno songs? The home page is kind of weird, it doesn't seem like there is much on it for songs, I have made several songs public and I don't think anybody is seeing them, I wish there was a place for feedback and sharing
Does anyone know how to get phrasing correct, with the accent on the right beat?
Yeah the base song it generates are not necessarily trash but it misses the link between words, feelings and intent, 3 things humans commenly change the meaning of.
I found it works best to daisy chain instructions and to seperate the individual instructions with brackets in order to give it more complex instructions. I sucessfully made a song with both male and female vocals and in the right order but it was a fight and a half.
For some reason it kinda ignores bracket instructions in lyrics but it pays more attention to the style interface. If you can figure out the proper inputs you can get it to do most anything from specific generas on specific verses to changeing vocals and pitch, tone and pace depending on what you are doing.
The biggest draw back i have found is when its programming wants to argue about what i want it to do vs. The conclusion it came to, i messed up once forgot to give it specific details on a dual song and ended up with vocals that alternated by verse and gender for the whole track.
Its wonky for sure but with time it will improve. Vi can never replace ppl bc not only does it not have a heart it also can't understand what that means, its literally impossible for a VI. Singers have no fear, the average preformer stopped writing their own stuff along time ago anyway.
I can make music for me and thats good enough, atleast its original and not rinse and repeat radio.
How do you guve them high notes?
Are you guys happy with the sound quality of Suno ai generated music? Is there a tip to improve the quality of the sound?
how can i create ai lyric generator (that can generate farsi lyrics)
When screen sharing you should adjust DPI so it is readable even at lower resolutions.
Good video, thanks! Can you tell, please, how to edit a song, after Suno finished it? Can’t find this option yet…
I don’t think you can yet
Download the stems. Then recreate the part .. eg vox guitar 🎸 with humans
Is there any API support provided by the suno?
Not yet. There's a couple open source third party APIs but they consider that a TOS violation. Eventually their own API should get released.
Is it possible to deley the piano?
would i be able to add an intro / some pre- chorus to an already made song?
I try to write a duet flr my gf. With a male voice and a female voice. But seems like no code I use makes 100% sure that the app will do it in the way I want. Does someone have a tipp for me?
I tried to write [man] and [woman] or [male] and [female] the same also with [male singer] or [male voice] doesnt worked out for me. Does someone have an idea?
Is there a way to tell it to repeat the chorus or something? I feel like sometimes it just writes a whole new melody for each section, and isn't able to distinguish the whole verse/chorus/verse/chorus aspect.
No seas perezozo. Escribe tu propia letra.
I'm here because DJ AUDITS channel latest video uses an Outro Rap Track that is an absolute Banger!! I was stunned! 😮
The problem i have is actually get the current set to stop playing even with an [end] prompt.
Can you fix your existing songs?
Thanks 👍🏼
Anytime!
Hi, Is it possible to tell suno the chords that I want it to use in my lyrics so that it can generate a song with those chords?
I haven’t tried that personally, but will see if I can get it to work
it will be excellent🤩🤩🤩@@allyourtechai
Have you found someting to do with chords? 🤔
@@BJ_MUSICandMORE sorry, no, in suno, it is no possible to do that, but if you have knowlege in music, you can post edit your song and change the chords using Other program like Reaper
@@elliotcruz7821 i am a musician...writing and recording songs...but i'm not a singer...
I am searching for a KI which is using my music an my lyrics an do the singing...do you know somthing like that?
hello-- I noticed when you are create music by yourself and put the letters in a different language such as Hebrew most of the time the pronunciation of the words are not correct my questions is if I can change the word until I found the best person pronunciation and to keep the same music so we will not lose this music that I like.
second thing it is frustrated is that they put the rhythmic Library in a way that it's hard to find kind of rhythmic you looking for and they don't give you even to write down so you can search for the it
Love everyone here . We’re ahead
How do you get clips longer then 25 seconds?
How To Create FULL Length AI Songs With Suno AI!
ua-cam.com/video/NmdsSnzNblM/v-deo.html