revision :
I slightly de-essed and removed some eq that was unnaturally boosting the highs, subtle changes without trying to alter my tone. I don’t have the best pronunciation of things so things are still a bit “essy” but its a bit better. I am indeed told that I am breathy but I tend to leave it be as it reflects some passion.
The voice sounds great. A couple spots where some pitchiness comes through, but easily something that could be fixed with pitch correction or multiple takes.
The processing on the vocals sounds too extreme for the song. It sounds like the microphone is right up inside your mouth, to the point where I can hear the clacking and shlacking. I think you can back off quite a bit on the compressor and/or the microphone and it will sound much more natural.
1 Like
you are absolutely correct, I practically ate the mic on this lol, I wanted the voice really upclose and considering my horrible mic technique and experience I butchered it. I will see what I can do to tweak the compressor. I have a sore throat so a retake is out of the question at this time.
1 Like
I agree with what a few others have said about reducing the amount of compression to reduce the sibilance. Like AJ said, you can always automate the vocal if the volume varies too much.
1 Like
revised. Link in original post
I hate to be a “Goldilocks” critic, but the de-essing is a little too much now - some “s” sounds are actually lisping (not all, though). De-essing is a little bit like low end - it’s tricky to get it to translate “just right”. It’s probably the main thing that I hear pros get wrong in “big boy” productions from time to time.
1 Like
Yes I was afraid of that. I ended up asking my mastering engineer who suggested to scale back the de-essing and adding some saturation to make the sound back to its former richness (which he liked) but without the harshness. I ended up scaling back a bit and added the saturation and liked the results. Thought not ideal just yet but I do like the closeness of the voice with the saturation.
Yeah, that’s better… but if I’m really picky, some “s” sounds are still slightly lispy.
At times when sibilants have proved to be particularly irksome, I’ve had to create a separate uncompressed track solely for the sibilant sounds and mult it with the rest of the vocal track… & in that case, it is sometimes necessary to automate or clip gain the sibilants to get some consistency…
…but (I would imagine) that’s probably a bridge to far your purposes here.
1 Like
Purpose is to learn. I am trying to learn to work with Vocals more, best way for me was to dive in and sing my own and learn the common pitfalls. (Like most people I am not a big fan of my own voice) If I can get it to a point where its pleasant enough for me, I will be happy lol.
decided to stop the revisions here, thanks everyone for their helpful input .
I have it to a point where its a clear and smooth listen while maintaining its closeness it was intended for. Had to tweak with the de-essing and cross fading the harshness where needed. Backed off the compressor and automated the vocals to keep them piercing just enough.
Next step is working on my mic technique!
1 Like
Nice! 
1 Like
It’s because it’s all done automatically, for example with a de-esser. The problem is that no two esses are the same. The only way to do it well is to manually automate, which obviously takes up too much time for the big boys.
Indeed. While I do have an automatic de-esser working around 8kish ( at a medium threshhold ) where most of my esss were located , I did have to manually automate and cross fade at many other parts.
I did learn a lot from working on my own vocals. In my experience, working on dry vocals is hard to do. There is not much delay and reverb to mask the pitfalls. It is what it is… raw.
I wanted to make dry vocals sound upclose, personal and rich. The tips in this thread from everyone are great and should be followed rigorously. The tips that helped me the most were to use the limiter sparingly and manually automate the vocals instead. The sheer clarity of the vocals that came out in the end compared to my original sound were a result of a lot of manual riding. That beats any kind of automated vocal rider or limiter. Unless in a rush, those kind of tools should be used with a grain of salt. The only thing remaining is micing technique and prononciations which no tools can fix in post without massive changes to tonal quality.
In the end I am pretty happy with the test results overall.
yeah, that’s workin 
Might be that you could just sing it over next time? Beautiful
1 Like
This thread has given me an idea for a plugin. Boooooz!
1 Like
I am not against making a good deesser. I’ve tried different methods, everything from standard de-essing to completely regenerating the S with a vocoder. I haven’t found a method that works perfectly for all sources yet.
I’ve got two ideas:
-
Sample replacement. The plugin scans for the sibilants, then replaces each one with a sample from its library. Detection threshold/frequency and of course, output level are user-definable, and users can add to the library if they wish.
-
The plugin scans as above, but simply automates each sibilant down to a user-defined level, as opposed to compressing it. (Would probably need to measure the loudness of each sibilant individually in order to know how much to automate down.)
I’ve got Drum Leveler, in my tiny inadequate mind I’m thinking I might be able to set it to perform as in 2.
Sorry for Hijacking your thread, Michelle 
2 Likes
Moving to a new thread.
1 Like
glad to be of help lol
Really enjoyed this song! You have a beautiful voice, and I think you did well with the production aspects.
Starliner
1 Like