![]() So, could we get some guidance on how to achieve natural sounding voice when using the denoiser? It can remove bg noise really well, how do we achieve better separation of voice? I am not an audio professional, but as mentioned before, seem to be doing good with Bertom Denoiser and Davinci Resolve noise reducer, this allows me to conclude that the Denoiser I can't seem to handle well is the one in Lightworks. Imagine if you have an hour long edit of bunch of people talking where you used the noise reduction and you have to go through each voice audio clip and set the quality. Apparently on my system Lightworks export audio respecting the selected quality of the effect. In addition I thought that this quality setting was supposed to be for editing time purposes and exported video always take the best of the best audio effect quality approach to render everything. I was thinking that maybe quality settings will do some tricks, and they do have effect as some artifacts are gone in Very High setting versus High, but if I set it to offline quality, which I assume is the best quality possible, I don't hear noticeable difference from the very high quality setting. I am able to get similar background noise elimination inside Lightworks, however with a big trade off, I get bucket voice, and especially noticeable when the talent breathe in. Even the birds in the background were reduced really dramatically, and the talent sounds very natural. ![]() This allows you to denoise images while actually recovering. DeNoise AI uses a fundamentally new approach called deep learning: after a lengthy process of learning from millions of images, DeNoise AI learned to accurately distinguish between real image detail and noise. The resulting audio sounds very natural with no buckettery voice, even breathing in sounds good while background noise is really well removed. Noise reduction software has been the same for over a decade - until now. My latest encounter with noise reduction was in Davinci Resolve, where I was able to get desired background noise removal with a default preset with no modifications needed at all. I still am hoping that VST3 support will get some more development and enhancement to work better, during the early stages, even if not in a perfect way, I still was able to use plugins like Bertom Denoiser, which provided me better noise reduction results, I used to be able to reduce undesired noise while avoiding bucket sound really easily. So either I miss something in the setting and don't understand, or current tool has some limitations, while it does remove background noise really well, it still dramatically affect what the one would expect to be untouched. Especially noticeable when talent breathe in/ breathe out. ![]() No matter what settings I do, if I want to have undesired noise removed from the audio, typically I get a bucket sounding voice. By this I mean, usually noise reduction tool in Lightworks gives me a type of sound that where I live, we call it a bucket sound. I am a bit puzzled and not because it doesn t work, it works and I do tend to reduce noise quite alot. Since we don't have a manual, could someone provide with an explanation how it works and how to get the best possible results? I am having some questions regarding noise reduction available in Lightworks 2023.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |