AI Sound Effects Generator

Introduction 0:00 it's been a while since I've done a 0:01 Cling video and there have been a lot of 0:03 updates so let's get right into it first New Website Design 0:06 off the website has been completely 0:08 redesigned on the sidebar you'll now see 0:10 tabs for effects image generation video 0:14 generation the new sound generation tool 0:16 for creating sound effects the tryon 0:18 feature lip sync and extend let's check Effects 0:21 out the new effects there are six 0:23 available and the newest one is called 0:26 anime figure you've probably seen some 0:28 of these effects in viral Tik Toks or 0:30 YouTube videos here's the bloom bloom 0:35 effect dizzy 0:40 dizzy fuzzy 0:46 fuzzy mochi 0:52 mochi and boom 0:56 boom i uploaded this image to test it 0:59 out and I thought the dizzy effect would 1:01 be fun it costs the same number of 1:03 credits as a 5-second video in 1:05 professional mode and here's the 1:12 result very good and I didn't realize it 1:15 added sound too that's a nice bonus the 1:18 next update actually came out a few How to Create Consistent Characters with Subject Reference 1:20 months ago but it's definitely worth 1:22 mentioning it's the face reference tool 1:24 in the image generation tab here's the 1:26 updated interface now DeepSk is 1:29 integrated into all the prompt boxes 1:31 which helps improve your prompts 1:33 automatically right here is where you 1:35 can upload a reference image i'm going 1:37 to use this one I created in MidJourney 1:39 once uploaded Cling will try to detect a 1:42 face if it can't the image won't be 1:45 usable after that you'll need to choose 1:47 your reference focus the whole subject 1:50 just the face or the entire image i'll 1:53 start with subject let's say I want her 1:56 sitting at her kitchen table drinking 2:01 coffee i'll let DeepSeek generate a 2:04 prompt based on that this result looks 2:07 solid so I'll click use prompt there are 2:10 two key settings you can tweak here face 2:12 reference controls how closely the 2:13 generated face matches the original 2:16 including facial expressions then 2:18 there's subject reference which 2:19 influences how much the model sticks to 2:21 things like the subject's clothing 2:23 hairstyle and pose the higher the value 2:26 the more faithful the result i'll 2:28 increase that a little you can also 2:30 adjust the aspect ratio and choose how 2:32 many images to 2:34 generate here's what we got she looks 2:37 exactly like our original character in 2:40 one version she still has a hand in her 2:42 pocket and the other hand is holding the 2:44 cup and in the other image both hands 2:46 are holding the coffee her outfit's 2:49 almost identical but the pants are a 2:51 different color let's try maxing out 2:53 both 2:55 settings now you can really see the 2:58 difference the reference image has a 3:00 much stronger influence her clothes are 3:03 a perfect match and her pose is nearly 3:06 identical but in one image it kind of 3:08 looks like she was just pasted onto the 3:10 background without any blending next 3:12 let's try changing her clothes and 3:14 facial expression let's have her wearing 3:16 pajamas and smiling at the 3:18 camera i'm keeping both reference 3:21 settings maxed out for now just to see 3:23 what 3:23 happens as expected the prompt changes 3:27 didn't do much her clothes and 3:29 expressions stayed the same because the 3:30 reference settings were too strong i'll 3:33 lower both settings 3:42 this i better but her hair color is 3:43 different so I'll adjust the subject 3:44 value one more 3:49 time now her hair color returns to 3:52 normal but she's no longer wearing 3:53 pajamas maybe a better approach is to 3:56 focus just on the face so I'll switch to 3:58 the face 4:02 tab i'll leave the reference strength 4:04 the same and roll the prompt okay now 4:07 she's wearing pajamas but still no smile 4:09 and her hair is brown then I noticed the 4:12 prompt says her hair is chestnut which 4:14 is incorrect so I fixed that quick tip 4:18 whenever you're using DeepSk or any 4:20 prompt enhancing tool make sure to 4:22 actually read the prompt and make 4:24 corrections if needed now I'll lower the 4:27 reference strength a bit and try 4:31 again finally she smiles she looks 4:34 slightly different but this is closer to 4:36 what I wanted so depending on your goal 4:39 you'll probably need to experiment with 4:41 these settings a bit it's all about 4:43 finding that sweet spot you can now How to Use the Sound Generation Tool 4:45 create AI generated sounds directly in 4:48 Cling just head over to the sound 4:50 generation tab to get started all you 4:53 need to do is type a description of the 4:55 sound you want into the prompt box you 4:57 can also enhance your prompt with 4:59 DeepSeek and even save your favorite 5:02 prompts as presets cling also gives you 5:04 a few helpful hints to guide you along 5:06 the way you can generate standalone 5:09 sounds or you can generate sound 5:11 specifically for a video let's try 5:13 adding sound to a video first I'll New Video Generation Interface 5:16 generate a new video by going to the 5:17 video generation tab as you can see it 5:20 has a fresh new interface too i'm going 5:23 to use the image to video feature and 5:25 create a clip of a werewolf terrorizing 5:27 a city i'll use deepseek to improve my 5:30 prompt 5:32 i'll just enter people run and scream as 5:35 the werewolf runs in the city it only 5:38 takes a few seconds for Deepseek to 5:39 think once it's done you can actually 5:42 view its thought process by expanding 5:44 the thinking 5:45 section i'm going to use the prompt but 5:47 I'll make a few tweaks 6:00 the video turned out great so now it's 6:02 time to add sound underneath the video How to Create Sound Effects for a Video 6:05 there are options for lip sync and 6:06 extend but what we want is AI sound once 6:11 you click it you'll be brought back to 6:12 the sound generation tab the prompt box 6:15 will automatically fill in with a 6:16 description based on the video but you 6:19 can tweak it if you want the sound 6:21 length matches the video duration and 6:24 can't be changed it also looks like the 6:26 number of generated tracks is fixed at 6:28 four creating the sound costs 10 credits 6:32 once it is done the tracks will appear 6:34 below the video let's listen 7:06 wow these sound effects are incredible 7:08 they really bring the video to life 7:10 every sound lines up perfectly with the 7:13 action making the scene feel way more 7:15 engaging and polished it's impressive 7:17 how well everything fits together now 7:20 for each audio track you have a couple 7:22 of options you can download the video 7:24 with that specific track applied or just 7:26 download the track itself as an MP3 or 7:29 WAV file the video remains the same in 7:33 each case only the sound changes if 7:36 you're not happy with the results or 7:37 just want to experiment you can 7:39 regenerate the tracks but keep in mind 7:41 it'll cost credits each time one last How to Crop Images 7:44 new feature in the image to video tool i 7:46 almost forgot to mention under the 7:49 elements tab when you upload an image 7:51 you now have the option to crop it to 7:53 focus on a specific subject this is 7:55 super helpful if your image contains 7:57 multiple elements or if you want to 7:59 highlight just a person's face without 8:01 including their outfit for example Wrap-up 8:07 that was a lot of updates to go through 8:08 and there's still more on the way clling 8:11 is getting ready to release its 2.0 8:13 model very soon and I'll be covering it 8:15 in an upcoming video so make sure to 8:17 stay tuned for that if you found this 8:19 video helpful or learn something new 8:21 don't forget to give it a like leave a 8:23 comment to let me know which feature 8:25 you're most excited about and subscribe 8:27 to the channel so you don't miss what's 8:28 coming next your support really helps 8:31 the channel grow and I appreciate every 8:33 one of you thank you for watching