We did it! Just 2 hours after uploading this video, Suno emailed that they're bringing back V3.5 for all covers, personas, extends and Replace sections. And also fixing the "shimmer". Love Team Suno. 💙
@AIAutomationLabs The improved vocal quality and remastering features seem super promising. The workaround for the ‘Shimmer’ issue is a nice touch-props to the community for figuring that out! Excited to see how this tool evolves.
02:39 Remaster Feature Example I see it differently. At first glance, it’s clear: V4 delivers cleaner, more detailed vocals. But sometimes, they’re too clean. That level of perfection can strip away some of the human touch and emotional depth. What concerns me even more is that the playback in V4 significantly loses impact and expression. The energy and presence a track needs often feel subdued. On the other hand, V3.5, when paired with the right vocal tagging prompts, can produce incredible results. Vocals, rap, backings, and adlibs can sound more musical, emotional, and organic. It might require a bit more finesse in crafting prompts, but the outcome is often more powerful and expressive. So, the question remains: Do we prioritize technical perfection or the musical soul?
I agree. I think it comes down to the Genre of music. Some genres like EDM, Trap don't really need much of the 'soul'. Good audio quality will suffice. But in genres like Pop Ballad, Orchestra, Opera, Jazz, the expression or the 'soul' matters a lot. And yes, I've noticed that Remaster kinda removes that 'soul' from it while cleaning it up. I've also seen it sing off key on a critical note that messed up the beautiful song. Lets just hope Suno comes up with a better solution to this. Maybe a feature like Udio AI where you can set the slider to set the level of uniqueness Or similarity we want.
I accidentally listened to several songs on the main page and there is no noise and distortion! The quality is almost like that of Udio, even written in version 3
How the quality of STEM features right now ? UVR actually doing good job for real recording materials and several songs generated by UDIO , but always struggle to get usable results with every SUNO generated materials even for slow tempo ones ( which rarely using any sidechain technique to get more pumping fx)
I've tried remastering lots of songs up to v4, and most of them aren't as good as the original gen. One song in particular removes so many awesome instrumental parts of the original gen. No matter what I try it always removes the electric guitar rhythm and intensified drums in the repeated chorus at the end. Makes me sad every time I try (and it fails). So far, it just hasn't worked for me. But, I'm so thankful Suno continues to push forward!
Yeah it kinda takes away the soul of the song in some of the genres. Have you tried remixing the song on Udio AI? You can use the upload feature then select Remix and set the similarity slider 'similar' and remix the song. Udio does a great job. I'd recommend that you give it a try. I've been using this trick long before Remaster came out in Suno V4.
While I was able to successfully create Greek songs using the previous version v3.5, I have encountered significant issues with the new version v4. Specifically, v4 appears to lack Greek language recognition. This results in frequent misspellings of Greek lyrics by the singers in generated songs. This problem was infrequent in v3.5, but it is consistently occurring in v4. As a Premier Plan subscriber, I believe that v4 seems unsuitable for creating Greek songs. I will continue using v3.5 and hope that future updates to v4 will address the Greek language issue, preventing such lyric errors in the future.
I see. If you're on Discord, you can report this issue in the V4 Feedback channel. Everyone's writing their issues with V4 on there. But yeah, for now just continue with v3.5 until they improve v4.
The sound is cleaner and the new features are great, but creating more creative and hybrid music is not as easily achieved in version four. 3.5 was not as cliché and generic or Top 40 sounding.
If you're on any Paid Plan, you get commercial rights to all the songs you create with it. So you can sell, stream, distribute, monetize in any way you want.
We had The EAGLES, AC/DC, Pink Floyd, Van Halen, Ozzy Osborne, Lynyrd Skynyrd, and so many others. Sorry, but ai will never come from the heart or soul. It's pointless.
I wish we could upgrade to paid plan and extend to get the commercial rights to monetize. I was told unfortunately since it was created under the free plan it is suno ai property to claim and good news ! The extended portion is covered for monetization only
3.5v can have some weird vocal artifacts. One song ended with just vocals where the vocalists named the band and thanked the audience for coming out. I believe the band was Blazing Balls so if you need a band name, there you go. Another song ended with the vocalist acting like it was trying to sing the song without music. 4.0 still sounds sing songy in these examples. No scats, noo runs, no improve yells etc.
A few things. 1. Oddly, today, vocals weren't as clean as the day V4 launched. Traffic? Temp I hope! 2. Another issues was when I was replacing a section, that section was mastered differently or had different effects on the vocals. It sounded ridiculously different. However, you can remaster the whole resulting track to make it all consistent again. If you liked the original master, you will just have to keep remastering to find something similar. 3. Remasters sometimes change the melody or screw up the lyrics when the original was perfect. 4. Same when I was replacing a section. I was replacing a verse 1 and Suno was using some of the lyrics from verse 2. Over and over and over and over. - It's amazing how fast we become annoyed with what was maybe the coolest tech of our lives so far. :D Reminds me of Amazon Alexa in USA or Siri on iPhone. Man are they stupid now. Use ChatGPT in your tech or GTFO! ;) Nice channel Sub'd!
Lol. I agree. We are already complaining about the coolest tech, but we also want it to get even better. Oh Man, what a time to be alive. Suno's servers are working overtime to handle the traffic Lol. And If you have any verse or a section looping over and over, you can add a [Break] or [Interlude] while editing the lyrics.
I've seen that It usually happens when people write Verse 2 in the lyrics meta tags. Suno interprets it as singing twice. It happens rarely though. And it can be edited out with the replace section feature or in any DAW.
I got a few times, Yes. But I believe it's when I gave it very few lyrics, and Suno realized that it has to do something to reach the 4 minute limit or anywhere near. So it just kept singing the same thing over and over again. Also happens when there's no Outro or End or Fade out in the lyrics, to let Suno know the entire song context, so that it can decide when to end and Not rush through the lyrics or repeat the same thing over and over.
I have a question, probably a dumb one. Why is it when you make a song in V4, it sounds good but then when you remaster it to V4 again it sounds better in most cases? Shouldnt you making the song in V4 should have already made it cleaned up? I been contemplating testing out remastering already made v4 songs again further due to this. It seems atleast for me sometimes V4 will make a very dry mix, and when you remaster it...it will be more punchy but then the 2nd remastered sounds worse as if you are muffled or under water lol! Im strictly talking about V4 to V4. Not V3 or V3.5 to V4.
@@AIAutomationLabs Hip-Hop. Sometimes regular, others 1990's Hip-Hop. Usually those come out with a very degraded sound on purpose I think, like a old bootleg lol. I was wondering if anyone else felt like remastering a v4 made track was beneficial. I only attempted the remaster of a v4 track because I keep getting that slot machine style sound that everyone keeps complaining about. Watermark possibly?
Hmm could be. Suno's discord mods denied that it's a watermark. V3.5 had a lot of artifacts but they were not so clear, and were muffled. So it was like swept under the rug. But now with V4 with more fidelity, better audio quality, it's all coming out and is audible more clearly.
Been testing v4 extensively, it does a good job of cleaning and sounding clean but it removes the soul and the energy from the singer. I made comparison by remastering tracks from 3.5 to v4, v4 sounded clean but we lost the soul and the energy of the singer
I agree. It sounds flat in some genres when you remaster it. I've been using Udio to remaster Suno songs even before the remaster feature came out on Suno V4. It does a great job
I have paid version mu problem is it's not sing as "rap" .suno ai rap section also sing as normal song, sorry for my English.i wnat rap but it's sing as normal song my song structure is chorus and rap .
Up until V3.5, It's easy to tell with certain genres, because almost all songs have that same robotic voice and vibe. But with V4, it's hard to tell now.
Suno has an inaudible watermark to identify the Song made with Suno. It remains when you download the file. But if you recreate the song yourself, then there's no way to tell that the Song was made with Suno. Just don't set the Song to 'Public' status on Suno Platform. And if you're on any Paid Plan for Suno, you own the commercial rights to the song. You can sell it, stream it, distribute it, recreate it, whatever you want with it.
@@S.... As a udio user I can say the difference isn't that much suno v4 sound better the robotic voice decreases significantly. I can say udio is still better in terms of Voice but as a sound structure it still isn't that close to suno but I hope they fix it with udio V2 model.
@@Lost_Circle I'd argue that "out of the box", with minimal work that is true. But with good prompts and some work I made much better tracks, with incredible melodies and structures, than I would ever dream in Suno. And I could actually use them instead of just thinking to myself, "wow, what a great idea for a track, I should record it myself..."
I've noticed that the 'Remaster' enhances mostly the vocal quality. Sometimes doing that, it removes the soul of the song, the emotions etc. And for instrumentals, if your original had a lot of artifacts, the remaster enhances them even more.
For those that have used Udio religiously ever since its release, we don't understand all the fuss with Suno V4. Vocal quality is probably 90% of the the original Udio v1 model and Suno has been playing catchup with features and functions ever since Udio came on the scene. And unfortunately V4 Suno still sounds like Suno music, its still easy to spot, and unless you are prepared to work really hard in the DAW to manipulate and change stems is not production quality output. I really wanted V4 Suno to outshine Udio to force competition on quality, but its a real let down. A real bummer. A missed opportunity.
I use both. Both have their Pros and Cons. Yeah Suno sounds more Cliche/Top 40 modern day music. And although Udio has natural sounding vocals. It's a pain to get one good melody or workable sample to start. One song takes a lot of credits for a full song. While in Suno, you can get a full song that you can work in your DAW for any modifications. TLDR: Suno is quick & easy to use and Udio takes time, dedication and effort.
I started with Suno v1 and was there for a long time and paied a lot of money, until some bugs ruined the great V3_beta model and changed it to shitty V3 and V3.5 never to achieve the quality V3_beta had, and I just got angry and annoyed all the time... I tried Udio then, and boy I was shocked! Everything I dreamed about while using Suno was there, every problem Suno had was non existent in Udio or can be fixed, the quality was there, it followed instructions - it was night and day difference. Still is. Arguable Suno had better melodies "out of the box" but I could achieve even better ones with Udio if I just tired.
@@S.... 100% agree. I have tried, so many times, to enjoy Suno, but it's such a challenge to get over the lack of depth of the instrumentals, and the vocals. V4 doesn't go far enough to fix it. Udio is not perfect either, but the music I create is anything but generic. It's sophisticated, complex and rich in a way that Suno just isn't. I'm happy to be proven wrong BTW, hopefully one day soon.
This v4 creation and remastering is ruining some of my songs with added clicking noises. Really irritating that I’m burning credits to end up with terrible quality songs. V4 needs to go into beta and give free credits to test it out.
Yeah they messed it up. The vocals are clean but the instrumentals and backing tracks are bad. And yeah, the shimmer and clicking noises. We had these in V3.5 too but it wasn't clear enough to be audible. But when we remaster it, it's clear as day and messing up the track. It says Beta on the Pop-up when you see the "What's new" Section but they didn't advertise it as Beta nor gave us free credits to try. I hope they come up with a fix soon.
AI like Suno is crazy impressive, no doubt, but it’s more of a tool than a replacement. People still crave that human touch, the emotions, and the imperfections that only live musicians can bring. AI is great for experimenting and speeding up the creative process, but it’s never going to replace the magic of a real person pouring their soul into an instrument.
Just 2 hours after uploading this video, Suno emailed that they're bringing back 3.5 for all covers, personas, extends and Replace sections. And also fixing the "shimmer". Someone at Suno is watching our videos bro. GG!
@AIAutomationLabs It seems to be lacking emotion and too bland. The 3.5 has some slight noise quality but has more emotion and sounds more natural at singing.
You can copy the lyrics, paste it into ChatGPT and ask it to keep the same song but replace the word neon with anything that fits the song. Then come back to Suno and replace that word using the Replace Section feature.
It means that Suno tried to remaster your track, but it failed. So Suno refunded your credits because it failed. What's the length of your original track? Have you tried selecting the original and tried remastering it again?
@@AIAutomationLabs usually gives me 2,05-3:25 length, but i now i have a new problem: service create song with watermark: " This is a fake app, please go to (original site here)" . I really dont understand what is this. I send message to support to solve this problem, and waiting the answer:((. Did u ever heard about it?
There are so many fake apps using Suno's API illegally. So Suno has added this watermark, whenever Suno detects unusual activity or someone accesses Suno outside Suno's website or iOS App. If you've created the song on Suno's website and you're still getting that message, the support team will help you don't worry. Are you using VPN or something like that?
As for now the WAV format is no better than MP3 one, it was added because people automaticaly thinks WAV is better and loseless. It's like you got an mp3 and convert it to wav - you won't get better quality this way.
Well, I agree to some extent. But I'm still waiting for the day, Udio can make at least one workable song output in one go. Like Suno makes a full 4 minute song that you can rework in your DAW. That's what the people want. Ease of use. On the other hand, Udio takes multiple credits just to get a 30 second sample to start working with. Their new Model has been a downgrade in terms of coherence. It mumbles lyrics like crazy
@@AIAutomationLabs First of all, almost nothing out of Suno is usable in my opinion. And even ignoring that, I always had more luck making a 4 (or longer) minute song with Udio (even before v130) and extending those, than with Suno. And I usually make songs that are 6-12 minutes long - this is impossible with Suno since degradation gets worse to longer you go. Although I do agree that it's more a hassle with Udio and 4 minute mark would be great. And Suno is much easier for those that do not want to do anything themselves. (If Udio want's to mumble with your prompt/text just move the lyrics strength slider to 60-70% or add/substract some text if that's too much. Also adding slow/fast to style, to defeat that "too much lyrics/too little" and [Break] / [Instrumental Solo] to extend those empty spaces works like magic. I do not use the 30s model anymore, at all, not even for the beginnings - I usually start vocals in first second and than extend into, or I just use v130 and tell it to start singing at 15-20 second. It might eat more credits, but I still end a month with 100-400 left on Pro subscription and don't feel the need to save)
What genre of music are you making? I agree. Suno is for the users who want it easy. And it's horrible for extending songs more than the given 4 minute mark. Suno is for casual users/ to bounce off ideas/meme songs. While Udio is for people who would like more control over every bit of their songs. I'm waiting for Udio's new update to make a comparison video between Udio and Suno. That way we can bring in new users to Udio who have never even tried it.
@@AIAutomationLabs It's hard to get past 2 minute mark in Suno imo - even in a single 4 minute track, Suno tends to introduce more and more noise from about 1:30, instruments sound worse and worse, it forgets the track etc. I do all styles, I play a lot. From death metal, through orchestral, pop up to acapellas and operas. That's another thing with Udio - it knows more styles. New update from Udio will hopefully make Suno try to achieve more. Let them race, for our benefit! ;)
I see. Nice. Yeah me too, I love these genres. Lol yes, Suno vs Udio, at the end of the day, we get the best model which gives it's users what they need.
@@AIAutomationLabs Oha, the songs are practically all 3:14 long, and they end as if you'd taken the needle off the record. If you extend the songs, the collages are poorly done, you notice a flaw in the collage, which makes it impossible to extend more than 3:14. When they're shorter than 3:14 they always fade out, but that happens rarely. The quality of the voice has improved but the number of artifacts has increased, mainly clicks and resonances in the drum kit. The musical composition has worsened, since the melodic line was more elaborate with 3.5, now the music just repeats parts, it doesn't make a proper composition. For me, the only good thing that's happened is having a persona for each project. That's very positive, but for the rest, it's gotten a lot worse, both in terms of musical composition and melody, as well as in the clicks and resonances of the drums, and in extending the song, which has become almost impossible. It's a shame.
Hmm I agree. Yeah it was kinda like a great Vocal quality update but everything else messed up. I'm strictly using V4 for new songs and reworking them in DAW for now. Personas are working great for the genres and personas I had created with v3.5. Extensions and replace sections are a hit and miss. And yeah the 3 14 is annoying.
Suno can assist with ideas, melodies, but the soul of music-emotion, storytelling, and connection-remains a human touch. Many Pros are already using it in their workflow to enhance what Suno lacks.
So, as far as you know (or anyone else in the comments), is it okay to mention specific artist/band names in the prompt? I had somehow gotten the impression I wasn't allowed to do that, and it's probably been one of the biggest downsides for me. The same goes for song names...any idea if we're allowed to reference specific song titles in the prompts? I'm just always worried with these AI media generators, like Midjourney, for example, that I'm gonna get suspended, banned, or in some kind of trouble, when I'm just looking for certain influences that I like, not cuz I'm trying to copycat someone else's work. I want it to be my own. It's just difficult to get specific styles otherwise. Live your vids, btw! Thanks for all the work you put in! Really well done. 🫡👍
You can enter Artist names. Most of the time it will throw an error. If you enter any known artist names in the old model for lyrics or the 'Style of music' section, it will give an error. But this new lyrics model somehow allowed it and surprisingly I got similar vocal outputs too. I'm sure Suno will patch this or fix this to throw an error when someone uses artist names. But no, you won't get banned for using artist names.
@AIAutomationLabs Oh, I understand. That's pretty cool. I'll give it a try and see what happens. Thanks a lot for the reply and helpful info. You're making great resources out here. Really appreciate it. 🤠
We did it! Just 2 hours after uploading this video, Suno emailed that they're bringing back V3.5 for all covers, personas, extends and Replace sections. And also fixing the "shimmer". Love Team Suno. 💙
Awesome! I agree!
Awesome cos IMO v4 is terrible 😡
@AIAutomationLabs The improved vocal quality and remastering features seem super promising. The workaround for the ‘Shimmer’ issue is a nice touch-props to the community for figuring that out! Excited to see how this tool evolves.
Yes, it's awesome. More fun stuff to come in the future :)
Wooww I like that, keep publishing video like this
Thank you for the support!! :)
Thank you for this video.
I'm glad it was helpful :)
Thanks for the explaination at the end. Now i can minimize the shimmering and enjoy the songs more!
You're welcome. I'm glad it was helpful :)
Could you please tell me what you mean by shimmer in this case? Thanks.
Amazing video 🫡👍
Thank you! I'm glad you liked it :)
02:39 Remaster Feature Example
I see it differently. At first glance, it’s clear: V4 delivers cleaner, more detailed vocals. But sometimes, they’re too clean. That level of perfection can strip away some of the human touch and emotional depth. What concerns me even more is that the playback in V4 significantly loses impact and expression. The energy and presence a track needs often feel subdued.
On the other hand, V3.5, when paired with the right vocal tagging prompts, can produce incredible results. Vocals, rap, backings, and adlibs can sound more musical, emotional, and organic. It might require a bit more finesse in crafting prompts, but the outcome is often more powerful and expressive.
So, the question remains: Do we prioritize technical perfection or the musical soul?
I agree. I think it comes down to the Genre of music. Some genres like EDM, Trap don't really need much of the 'soul'. Good audio quality will suffice. But in genres like Pop Ballad, Orchestra, Opera, Jazz, the expression or the 'soul' matters a lot. And yes, I've noticed that Remaster kinda removes that 'soul' from it while cleaning it up. I've also seen it sing off key on a critical note that messed up the beautiful song. Lets just hope Suno comes up with a better solution to this. Maybe a feature like Udio AI where you can set the slider to set the level of uniqueness Or similarity we want.
I accidentally listened to several songs on the main page and there is no noise and distortion! The quality is almost like that of Udio, even written in version 3
Most of the songs on Suno's main Page are from Suno's admin team, discord mods and discord members.
Wow! Improved in deed! If ever this version turned to free, I'd probably use it over and over again. Keep up the good work! :)
I'm glad you liked it :) Yeahh V4 still needs some fixes to be made. But I'm sure they'll make it free someday.
Just pay for it.
How the quality of STEM features right now ? UVR actually doing good job for real recording materials and several songs generated by UDIO , but always struggle to get usable results with every SUNO generated materials even for slow tempo ones ( which rarely using any sidechain technique to get more pumping fx)
It still isn't very good compared to UVR. You can still hear a bit of vocals leaking into the instrumental stem.
Uvr can extract up to 14 channels, not perfectly tho or some parts still sound “mushed” compared to daw samples, but vocal has improved drastically ❤
I've tried remastering lots of songs up to v4, and most of them aren't as good as the original gen. One song in particular removes so many awesome instrumental parts of the original gen. No matter what I try it always removes the electric guitar rhythm and intensified drums in the repeated chorus at the end. Makes me sad every time I try (and it fails). So far, it just hasn't worked for me. But, I'm so thankful Suno continues to push forward!
Yeah it kinda takes away the soul of the song in some of the genres. Have you tried remixing the song on Udio AI? You can use the upload feature then select Remix and set the similarity slider 'similar' and remix the song. Udio does a great job. I'd recommend that you give it a try. I've been using this trick long before Remaster came out in Suno V4.
Awesome ❤
I'm glad you liked it :) ❤
Nice breath work with the Female Vocal
Finally someone commented on this!. I liked it too.!!
I'm trying to remaster one of my own songs but it will only let me upload 60 seconds. Is the any way to bypass this?
Upload limit is 60 seconds for Free users and 120 seconds for Paid users. You can speed up the song 1.5x and upload the sped up version.
While I was able to successfully create Greek songs using the previous version v3.5, I have encountered significant issues with the new version v4. Specifically, v4 appears to lack Greek language recognition. This results in frequent misspellings of Greek lyrics by the singers in generated songs.
This problem was infrequent in v3.5, but it is consistently occurring in v4. As a Premier Plan subscriber, I believe that v4 seems unsuitable for creating Greek songs. I will continue using v3.5 and hope that future updates to v4 will address the Greek language issue, preventing such lyric errors in the future.
I see. If you're on Discord, you can report this issue in the V4 Feedback channel. Everyone's writing their issues with V4 on there. But yeah, for now just continue with v3.5 until they improve v4.
@@AIAutomationLabsHello, thank you for the information 😊 Have a great day ahead ‼
Thank you. You too :)
Yup. The shimmer is appearing almost always at the near end of the song.
Yeahh let's hope they fix it in the next update.
The sound is cleaner and the new features are great, but creating more creative and hybrid music is not as easily achieved in version four. 3.5 was not as cliché and generic or Top 40 sounding.
Hahaa I agree. "Top 40 sounding" is the best way to put it.
3 was much better at originality than 3.5. And 4 is significantly worse than 3.5. Udio’s v1.5 model suffers from the same thing.
nice
Thank you :)
Thanks. Do we have rights to sell any music we create?
If you're on any Paid Plan, you get commercial rights to all the songs you create with it. So you can sell, stream, distribute, monetize in any way you want.
We had The EAGLES, AC/DC, Pink Floyd, Van Halen, Ozzy Osborne, Lynyrd Skynyrd, and so many others. Sorry, but ai will never come from the heart or soul. It's pointless.
Good voice quality
Yes :)
Which site is the singing avatar from .The quality is great .Many thanks
It's Hedra AI for lip syncing And I used Ideogram AI for the Avatar image. :)
@@AIAutomationLabs Many thanks for the reply
"the first big change in v4 is the model v4" .uummm sir, was is supposed to be model v69 lol xD anyway thanks for the video!
Lol 😅 sorry to state the obvious. Thanks for watching :)
Por estos lados, todo eso que dices es para las versiones pagas...
Gracias de todas formas por traernos las novedades..!
Sí lo siento, espero que también esté disponible para las versiones gratuitas en futuro. Y Me alegro de que te haya gustado el vídeo :)
I wish we could upgrade to paid plan and extend to get the commercial rights to monetize. I was told unfortunately since it was created under the free plan it is suno ai property to claim and good news ! The extended portion is covered for monetization only
If you email Suno, they allow commercial rights to one free plan's song, if you're now on the Pro Plan.
How did you get the video of the girl singing?
Made the image with Ideogram AI and the talking head video with Hedra AI. Links are in the video description :)
@ thanks mate
Can you upload your own track and lyrics and have Suno create a vocal?
Yes you can upload it using the "Upload Audio" button
@@AIAutomationLabs Cool, thanks!
3.5v can have some weird vocal artifacts. One song ended with just vocals where the vocalists named the band and thanked the audience for coming out. I believe the band was Blazing Balls so if you need a band name, there you go. Another song ended with the vocalist acting like it was trying to sing the song without music.
4.0 still sounds sing songy in these examples. No scats, noo runs, no improve yells etc.
Yeah V3.5 pulls out random stuff from training data during song generations. 4.0 sounds polished but sometimes lacks emotion.
Awesome
I'm glad you liked it :)
So we got an AI voice talking about AI music 😮.....
Why is it news to you? Are you new to UA-cam? The channel is literally all about AI.
i also try with my three four songs it is not working with some genre and indian raags properly
Seems like the model hasn't been trained on a lot of music styles and genres.
A few things. 1. Oddly, today, vocals weren't as clean as the day V4 launched. Traffic? Temp I hope! 2. Another issues was when I was replacing a section, that section was mastered differently or had different effects on the vocals. It sounded ridiculously different. However, you can remaster the whole resulting track to make it all consistent again. If you liked the original master, you will just have to keep remastering to find something similar. 3. Remasters sometimes change the melody or screw up the lyrics when the original was perfect. 4. Same when I was replacing a section. I was replacing a verse 1 and Suno was using some of the lyrics from verse 2. Over and over and over and over. - It's amazing how fast we become annoyed with what was maybe the coolest tech of our lives so far. :D Reminds me of Amazon Alexa in USA or Siri on iPhone. Man are they stupid now. Use ChatGPT in your tech or GTFO! ;) Nice channel Sub'd!
Lol. I agree. We are already complaining about the coolest tech, but we also want it to get even better. Oh Man, what a time to be alive.
Suno's servers are working overtime to handle the traffic Lol. And If you have any verse or a section looping over and over, you can add a [Break] or [Interlude] while editing the lyrics.
They have things to fix before its perfect. Did they get rid of the same verse coming twice
I've seen that It usually happens when people write Verse 2 in the lyrics meta tags. Suno interprets it as singing twice. It happens rarely though. And it can be edited out with the replace section feature or in any DAW.
@AIAutomationLabs Yes I edit them away. But i seen many say the same. I always use [Verse 1] and [Verse 2] and so on. So you never got it in 3,5?
I got a few times, Yes. But I believe it's when I gave it very few lyrics, and Suno realized that it has to do something to reach the 4 minute limit or anywhere near. So it just kept singing the same thing over and over again. Also happens when there's no Outro or End or Fade out in the lyrics, to let Suno know the entire song context, so that it can decide when to end and Not rush through the lyrics or repeat the same thing over and over.
@AIAutomationLabs Thanks, i had the some when they jumperd over v3 or v4 so the song was incomplete so got v2 twice
How to upload audio file from gallery for remastering...??
Click on Create, At the Top you will see an 'Upload Audio' button beside the "Custom" option.
@AIAutomationLabs it allows only 60 seconds
On the Free version, it allows 60 seconds. And on the Paid version, it allows 2 minutes.
I have a question, probably a dumb one. Why is it when you make a song in V4, it sounds good but then when you remaster it to V4 again it sounds better in most cases? Shouldnt you making the song in V4 should have already made it cleaned up?
I been contemplating testing out remastering already made v4 songs again further due to this. It seems atleast for me sometimes V4 will make a very dry mix, and when you remaster it...it will be more punchy but then the 2nd remastered sounds worse as if you are muffled or under water lol! Im strictly talking about V4 to V4. Not V3 or V3.5 to V4.
Lol that's interesting. What genre of music are you making?
@@AIAutomationLabs Hip-Hop. Sometimes regular, others 1990's Hip-Hop. Usually those come out with a very degraded sound on purpose I think, like a old bootleg lol. I was wondering if anyone else felt like remastering a v4 made track was beneficial. I only attempted the remaster of a v4 track because I keep getting that slot machine style sound that everyone keeps complaining about. Watermark possibly?
Hmm could be. Suno's discord mods denied that it's a watermark. V3.5 had a lot of artifacts but they were not so clear, and were muffled. So it was like swept under the rug. But now with V4 with more fidelity, better audio quality, it's all coming out and is audible more clearly.
I the remaster limit also 4 minutes?
No. It can be more than that too. The remaster depends on the original song length. I've tried for up to 7 minutes.
@@AIAutomationLabs thanks
😊💕❤️❤️❤️🌹🌹🌹
❤❤🙌😊
Sounds more clearer.
Yes the vocals are very good. The instrumentals not so much. But a great advancement from the previous version.
@@AIAutomationLabs And the generation is freaking fast one...
Yeah Lol true.
Very Based 👌🏻
Keeping it real 🤝. We all want Suno to grow and get even get better :)
@@AIAutomationLabs Really enjoying the content, keep up the great work man!
Thanks for the support. Will post more content soon!
Can you use your own melody?
Yes you can use your own melody by uploading a sample using the 'Upload Audio' option.
Been testing v4 extensively, it does a good job of cleaning and sounding clean but it removes the soul and the energy from the singer. I made comparison by remastering tracks from 3.5 to v4, v4 sounded clean but we lost the soul and the energy of the singer
I agree. It sounds flat in some genres when you remaster it. I've been using Udio to remaster Suno songs even before the remaster feature came out on Suno V4. It does a great job
I have paid version mu problem is it's not sing as "rap" .suno ai rap section also sing as normal song, sorry for my English.i wnat rap but it's sing as normal song my song structure is chorus and rap .
In the style of music write words like "Rap", " Gangsta Rap", "Fast Rap", " hip hop"
How do people know that the song is created by ”Suno”？
Up until V3.5, It's easy to tell with certain genres, because almost all songs have that same robotic voice and vibe. But with V4, it's hard to tell now.
@ if someone reproduce a “suno song” using their own instruments and vocal, could it avoid detection？
Suno has an inaudible watermark to identify the Song made with Suno. It remains when you download the file. But if you recreate the song yourself, then there's no way to tell that the Song was made with Suno. Just don't set the Song to 'Public' status on Suno Platform. And if you're on any Paid Plan for Suno, you own the commercial rights to the song. You can sell it, stream it, distribute it, recreate it, whatever you want with it.
I have the pro version. I got 50 songs done with 2.200 tries. I only have 300 credits left. I think it is not really effective.
What style of music are you creating? May i know what prompts you're using and Are you using Meta Tags to guide the song structure with your lyrics?
udio v 2.0 is also going to come out soon lets see how udio compare to suno if suno sound better than udio then I am going to switch to suno again
I'll do a comparison video when it's out! :)
Right now Udio is still better than Suno v4.
One can assume newer versions of Udio will make that difference more significant.
@@S.... As a udio user I can say the difference isn't that much suno v4 sound better the robotic voice decreases significantly. I can say udio is still better in terms of Voice but as a sound structure it still isn't that close to suno but I hope they fix it with udio V2 model.
@@Lost_Circle I'd argue that "out of the box", with minimal work that is true.
But with good prompts and some work I made much better tracks, with incredible melodies and structures, than I would ever dream in Suno. And I could actually use them instead of just thinking to myself, "wow, what a great idea for a track, I should record it myself..."
I tried remastering with a few of my songs and they came out far worse than the originals
I've noticed that the 'Remaster' enhances mostly the vocal quality. Sometimes doing that, it removes the soul of the song, the emotions etc. And for instrumentals, if your original had a lot of artifacts, the remaster enhances them even more.
For those that have used Udio religiously ever since its release, we don't understand all the fuss with Suno V4. Vocal quality is probably 90% of the the original Udio v1 model and Suno has been playing catchup with features and functions ever since Udio came on the scene. And unfortunately V4 Suno still sounds like Suno music, its still easy to spot, and unless you are prepared to work really hard in the DAW to manipulate and change stems is not production quality output.
I really wanted V4 Suno to outshine Udio to force competition on quality, but its a real let down. A real bummer. A missed opportunity.
I use both. Both have their Pros and Cons. Yeah Suno sounds more Cliche/Top 40 modern day music. And although Udio has natural sounding vocals. It's a pain to get one good melody or workable sample to start. One song takes a lot of credits for a full song. While in Suno, you can get a full song that you can work in your DAW for any modifications.
TLDR: Suno is quick & easy to use and Udio takes time, dedication and effort.
@ don't disagree there. Udio has a higher barrier to entry and a lot of patience but when you master it…
Lol yes. I use Udio on my weekends when I get some free time for myself. On busy workdays it's Suno all the way :)
I started with Suno v1 and was there for a long time and paied a lot of money, until some bugs ruined the great V3_beta model and changed it to shitty V3 and V3.5 never to achieve the quality V3_beta had, and I just got angry and annoyed all the time...
I tried Udio then, and boy I was shocked!
Everything I dreamed about while using Suno was there, every problem Suno had was non existent in Udio or can be fixed, the quality was there, it followed instructions - it was night and day difference. Still is. Arguable Suno had better melodies "out of the box" but I could achieve even better ones with Udio if I just tired.
@@S.... 100% agree. I have tried, so many times, to enjoy Suno, but it's such a challenge to get over the lack of depth of the instrumentals, and the vocals. V4 doesn't go far enough to fix it. Udio is not perfect either, but the music I create is anything but generic. It's sophisticated, complex and rich in a way that Suno just isn't. I'm happy to be proven wrong BTW, hopefully one day soon.
This v4 creation and remastering is ruining some of my songs with added clicking noises. Really irritating that I’m burning credits to end up with terrible quality songs. V4 needs to go into beta and give free credits to test it out.
Yeah they messed it up. The vocals are clean but the instrumentals and backing tracks are bad. And yeah, the shimmer and clicking noises. We had these in V3.5 too but it wasn't clear enough to be audible. But when we remaster it, it's clear as day and messing up the track. It says Beta on the Pop-up when you see the "What's new" Section but they didn't advertise it as Beta nor gave us free credits to try. I hope they come up with a fix soon.
Human instrumental music is gone. I Think ia a fact. Musicians will be live playing only, same happened with translators a year ago.
AI like Suno is crazy impressive, no doubt, but it’s more of a tool than a replacement. People still crave that human touch, the emotions, and the imperfections that only live musicians can bring. AI is great for experimenting and speeding up the creative process, but it’s never going to replace the magic of a real person pouring their soul into an instrument.
They should at least give paid users the option to go back to v3.5 😢
Just 2 hours after uploading this video, Suno emailed that they're bringing back 3.5 for all covers, personas, extends and Replace sections. And also fixing the "shimmer". Someone at Suno is watching our videos bro. GG!
I'M not liking the feel and sound of V4 just seams a little off. So I'm just sticking to using 3.5 until it has been sorted.
Umm like what?
@AIAutomationLabs It seems to be lacking emotion and too bland. The 3.5 has some slight noise quality but has more emotion and sounds more natural at singing.
Ahh I see. Yes I agree with you on that.
They should give negative prompts for everything Neon. It is so annoying.
You can copy the lyrics, paste it into ChatGPT and ask it to keep the same song but replace the word neon with anything that fits the song. Then come back to Suno and replace that word using the Replace Section feature.
Unfortunately no Google pay in Europe
Ohh I personally use my Card for subscription.
Their song lyric writing model, version 4, seems less effective now because it no longer incorporates references from the Bible in creating lyrics.
I see. But you can still use the Classic old model. It still works.
@@AIAutomationLabs Looks like it doesn't work in the custom mode. But does in the non-custom mode.
In the custom mode, there's a dropdown to choose the classic model or the new ReMi lyrics model.
@@AIAutomationLabs Try typing in "Psalm 100, or any other Psalm #, or even somthing like Joshua 1:9 and see if the lyrics relate.
@@AIAutomationLabs I pulled lyrics for Psalms 80 - Psalm 150 before the update and now the classic model gives lyrics that don't relate.
So, i wait v5 when isntrument quality will be more clean and powerful 😂
Hahaa.. Yeah the instrumentals aren't the best at the moment. But still an improvement from previous versions.
@AIAutomationLabs I m trying to remaster my old songs, but I get 0:02 second track parts , and message: "credits refunded" what is that mean ?
It means that Suno tried to remaster your track, but it failed. So Suno refunded your credits because it failed. What's the length of your original track? Have you tried selecting the original and tried remastering it again?
@@AIAutomationLabs usually gives me 2,05-3:25 length, but i now i have a new problem: service create song with watermark: " This is a fake app, please go to (original site here)" . I really dont understand what is this. I send message to support to solve this problem, and waiting the answer:((. Did u ever heard about it?
There are so many fake apps using Suno's API illegally. So Suno has added this watermark, whenever Suno detects unusual activity or someone accesses Suno outside Suno's website or iOS App. If you've created the song on Suno's website and you're still getting that message, the support team will help you don't worry. Are you using VPN or something like that?
Please one video in Hindi full tutorial in Indian playback singer tutorial female
I'm making different UA-cam channels for other languages bro. I'll post it there.
I hope Suno adds FLAC so we can use it instead of the (wasteful) WAV format. :/
When it was Version V3, we only had MP3. And then we got the WAV in v3.5, It's a good progress. I hope we get other formats too, in the future.
As for now the WAV format is no better than MP3 one, it was added because people automaticaly thinks WAV is better and loseless.
It's like you got an mp3 and convert it to wav - you won't get better quality this way.
@@S.... Good to know. Thanks!
Isme kuchh to khas h
Lame. Suno is still way behind Udio, even after update.
At least we can fix some of the older ones with remix.
Well, I agree to some extent. But I'm still waiting for the day, Udio can make at least one workable song output in one go. Like Suno makes a full 4 minute song that you can rework in your DAW. That's what the people want. Ease of use. On the other hand, Udio takes multiple credits just to get a 30 second sample to start working with. Their new Model has been a downgrade in terms of coherence. It mumbles lyrics like crazy
@@AIAutomationLabs First of all, almost nothing out of Suno is usable in my opinion.
And even ignoring that, I always had more luck making a 4 (or longer) minute song with Udio (even before v130) and extending those, than with Suno. And I usually make songs that are 6-12 minutes long - this is impossible with Suno since degradation gets worse to longer you go.
Although I do agree that it's more a hassle with Udio and 4 minute mark would be great. And Suno is much easier for those that do not want to do anything themselves.
(If Udio want's to mumble with your prompt/text just move the lyrics strength slider to 60-70% or add/substract some text if that's too much. Also adding slow/fast to style, to defeat that "too much lyrics/too little" and [Break] / [Instrumental Solo] to extend those empty spaces works like magic. I do not use the 30s model anymore, at all, not even for the beginnings - I usually start vocals in first second and than extend into, or I just use v130 and tell it to start singing at 15-20 second. It might eat more credits, but I still end a month with 100-400 left on Pro subscription and don't feel the need to save)
What genre of music are you making? I agree. Suno is for the users who want it easy. And it's horrible for extending songs more than the given 4 minute mark. Suno is for casual users/ to bounce off ideas/meme songs. While Udio is for people who would like more control over every bit of their songs. I'm waiting for Udio's new update to make a comparison video between Udio and Suno. That way we can bring in new users to Udio who have never even tried it.
@@AIAutomationLabs It's hard to get past 2 minute mark in Suno imo - even in a single 4 minute track, Suno tends to introduce more and more noise from about 1:30, instruments sound worse and worse, it forgets the track etc.
I do all styles, I play a lot. From death metal, through orchestral, pop up to acapellas and operas. That's another thing with Udio - it knows more styles.
New update from Udio will hopefully make Suno try to achieve more. Let them race, for our benefit! ;)
I see. Nice. Yeah me too, I love these genres. Lol yes, Suno vs Udio, at the end of the day, we get the best model which gives it's users what they need.
downgrade
Ow man, why?
@@AIAutomationLabs Oha, the songs are practically all 3:14 long, and they end as if you'd taken the needle off the record. If you extend the songs, the collages are poorly done, you notice a flaw in the collage, which makes it impossible to extend more than 3:14. When they're shorter than 3:14 they always fade out, but that happens rarely. The quality of the voice has improved but the number of artifacts has increased, mainly clicks and resonances in the drum kit. The musical composition has worsened, since the melodic line was more elaborate with 3.5, now the music just repeats parts, it doesn't make a proper composition. For me, the only good thing that's happened is having a persona for each project. That's very positive, but for the rest, it's gotten a lot worse, both in terms of musical composition and melody, as well as in the clicks and resonances of the drums, and in extending the song, which has become almost impossible. It's a shame.
Hmm I agree. Yeah it was kinda like a great Vocal quality update but everything else messed up. I'm strictly using V4 for new songs and reworking them in DAW for now. Personas are working great for the genres and personas I had created with v3.5. Extensions and replace sections are a hit and miss. And yeah the 3 14 is annoying.
Bye human artist
AI vocals lacks soul, and any pros can notice it instantly .
Suno can assist with ideas, melodies, but the soul of music-emotion, storytelling, and connection-remains a human touch. Many Pros are already using it in their workflow to enhance what Suno lacks.
So, as far as you know (or anyone else in the comments), is it okay to mention specific artist/band names in the prompt?
I had somehow gotten the impression I wasn't allowed to do that, and it's probably been one of the biggest downsides for me.
The same goes for song names...any idea if we're allowed to reference specific song titles in the prompts?
I'm just always worried with these AI media generators, like Midjourney, for example, that I'm gonna get suspended, banned, or in some kind of trouble, when I'm just looking for certain influences that I like, not cuz I'm trying to copycat someone else's work. I want it to be my own. It's just difficult to get specific styles otherwise.
Live your vids, btw! Thanks for all the work you put in! Really well done. 🫡👍
You can enter Artist names. Most of the time it will throw an error. If you enter any known artist names in the old model for lyrics or the 'Style of music' section, it will give an error. But this new lyrics model somehow allowed it and surprisingly I got similar vocal outputs too. I'm sure Suno will patch this or fix this to throw an error when someone uses artist names. But no, you won't get banned for using artist names.
@AIAutomationLabs
Oh, I understand. That's pretty cool. I'll give it a try and see what happens. Thanks a lot for the reply and helpful info. You're making great resources out here. Really appreciate it. 🤠
Sure. Have fun!! Thanks for supporting the channel:)