I've been deactivated, please check my sample's audio quality.
I've been deactivated, please check my sample's audio quality.
I've uploaded three here:
https://www.dropbox.com/s/gcq3avh38ibjn62/Buzz%20Blackburn%202019-08-21%20Ski%20Video.wav?dl=0
https://www.dropbox.com/s/zbn0i6dc4pac1k8/Buzz%20Blackburn%202019-08-21%20drones%20OmniBus.wav?dl=0
Thank you!
Buzzblackburn
Please sign in to leave a comment.
Hi Buzz,
Fellow voice actor here. Great voice and like the deliveries.
Don't know what VB quality control will say, but they sound good to me. On the 2nd example (financial), thought I heard a plosive pop on a "p" word, but it was minimal.
What gear are you using? I am just curious to see what others on the platform are using.
Best regards,
Joe Brown
Hi Buzz,
Sounds good! Your profile was probably deactivated automatically due to samples being rejected for not following our recommendations: https://help.bunnystudio.com/hc/en-us/articles/360000253639-What-constitutes-an-effective-sample.
We've now reactivated your profile.
Hey Buzz!
Indeed, there are some issues with plosives and sibilance. This may be due to two frequently found issues:
1) Improper Gain Levels at Recording Stage
Keep in mind that high gain levels in our recording gear (usually the pre-amp or interface) will negatively impact our final result. Think of this in photographic terms: an overexposed picture. The camera captures too much light and no amount of editing will be capable of making the photograph usable. The same thing happens in audio: if we print a take that's too hot from the beginning, the natural sensitivity of the mic will be augmented by the amount of gain used, causing the noise floor to be raised. This will show as a wind-like sound or white noise, that will be raised if further processing (such as compression and normalization) is applied.
An instance of this can be found in your last sample (where white noise is audible in the silent regions) and also relates to some of the plosiveness in your consonant sounds since they are captured at a higher than ideal level.
2) Compression
We (as voice talents) love compression! It makes us sound bigger, fuller and leveled... more professional. The problem is that compression works by raising quiet sounds and lowering the louder ones. This is all fine and dandy, but depending on your settings, this can also impact your Ts, Bs making them "poppier" almost to a plosive level depending on your personal diction. This process will also impact your sibilants across the board but can make SH sounds super explosive and/or harsh. It's not a critical situation in your case, but if you take into account that we request voice-overs uncompressed and at -3dBFS Peak level, this suddenly becomes more of a problem: if a client tried to make the recording louder or try to process the audio any further, it would be super difficult and a bigger headache than it needs to be.
So, what can be done to avoid this?
First: be mindful of those sounds while recording. Sometimes we don't hear ourselves stressing our sibilants or plosives while reading and we forget how useful pop filters are; second: if those sounds have made their way into the recording, edit them out!
The best way to rid your audio of harsh sibilants is through careful de-essing or even better: through spectral editing. Most DAWs have a spectral editor that's easy enough to use and can help heaps in restoring, repairing and healing audio. I can't tell you the amount of time this has saved me, helping me fix takes I would've had to re-record otherwise.
This screenshot belongs to an actual revision, wherein Adobe Audition, I single out the frequencies for the pro to fix (6-8kHz range) with some 'SH' and 'T' sounds even reaching into the 10kHz territory. Using this process in Audition enables you to select the frequencies with a variety of tools, such as a brush that allows you to 'paint' over the unwanted frequencies (in this case the almost white-hot spots indicated) so that you can simply lower the volume by 2-4dB until you reach the desired mitigation.
This image shows the Spectral Edit pane in use, with the brush employed to select some of the 'hot' frequencies to be corrected, plus the volume tool used to do so. After selecting the portion to edit, simply lower or raise the volume to the best suiting level (in this case I applied a 3.3dB reduction on the selection).
Fig. 3
This image shows the offending frequencies corrected after repeating the process to achieve such a result.
Hopefully, you'll find these tips useful for your future projects! We're looking forward to hearing from you very soon.
- Héctor Adolfo Ituarte (Voice Bunny QC Agent)