Audition Feedback
Hi all,
I just had my audition rejected and was looking to get some feedback from the community on your thoughts on what I could do to improve my recording:
1) Room Echo - The acoustic conditions in your recording are not ideal. The \'sound\' of your room is audible in your deliverable even at its present amplitude levels.
2) Levels - Your audio levels are soft and uneven, not normalized to the -3dBFS Peak Level prescribed by Bunny Standards.
3) Processing - The muffled sound of your voice indicates that the waveform may have been treated to some degree with noise reduction/suppression, impacting its dynamics and true loudness.
I was a little surprised at the reasoning for a few reasons, I can't hear my room "sound" at all through my headphones. I have a 5'x6' fully foamed sound booth in my basement I've been using for years without issue. Regarding the levels and the processing, the peak level in the instructions said not to go above -6db and it said not to use any treatment or effects in post which I didn't. Which completely baffles me on the processing note.
Any thoughts on what I can do?
https://drive.google.com/file/d/1c1ci1KFwrvEpowD-y8A44COxTiB0TDrG/view?usp=sharing
Please sign in to leave a comment.
Hi Stephen,
Thank you for reaching out with this!
Before we move forward with this, is there a chance you can send us a RAW recording?
Also, would you mind sharing pictures of your recording space with us?
Below is a picture of my recording space. Fully treated to allow for minimal bounce or echo.
Here's a file I just recorded in a raw, pcm format. I did not normalize this or do anything to it. https://drive.google.com/file/d/14A_0p7WChPVO2Qy9JzFGNvaqkrHsKbqA/view?usp=sharing
If you have issues opening the pcm, here is a wav pcm of the same file.
https://drive.google.com/file/d/1lmpeBhJGA7vuttEJYzZvSjzZ7yWKB0Pj/view?usp=sharing
Hi, Gents!
First of all, Stephen, thank you for the great photo of your setup, it does provide a clear picture! Now, onto the audio...
I analyzed your sample and found a couple of issues with self-noise/mouth noise, plosives, breaths, and two very weird frequency spikes, one at the 3200 to 3400 Hz range (no biggie though) and the other around the 1200 to 1700 Hz range, again nothing that some smart EQ can't fix. As for the self-noise it is easily fixed as well. Let's detail each issue with a proper mitigation strategy:
1) Self-noise - The inherent noise in your mic and/or mic pres in your interface, along with the faint 'reading noises', they can be eliminated with a single pass of a noise gate at a threshold of 38 dB with the fastest attack and release settings. I used the stock plugin in Audition to achieve this.
2) Plosives and Funky Frequencies - They can call be solved with smart EQing. Setting a simple 10 band EQ (again a stock plugin in audition) to cut your 79 Hz frequencies and lower by 6 dB and the offending frequencies at a 1400 Hz and 3300 Hz by 1.5 dB should do the trick.
Now, performance-wise there are some things to consider. Some words with hard consonants at the end (like in 'result', 'support', etc.) tend to be very quiet, almost silent in some cases. This is a diction issue that should be addressed with practice in reading. A pro tip is to 'clone' sounds to save a take.
In this processed version of your audio, you'll notice that all hard consonants now sound better and clearer. I took the '-d' sound in 'asked' to splice it in where other similar-sounding consonants were lacking in presence. I'd highly recommend you flip that mic upside down and point it at your forehead/bridge of your nose so that you can 'pop' and enunciate at your heart's content. That should also help a lot with mouth noises and breaths, as the mic positioning will mitigate their natural amplitude (which is already very workable).
As you can tell from the file, the noise gate I applied practically eliminated all of them noises without me having to do a single mute/fade or spectral edit to remove them. Finally, I manually lowered the amplitude of some open vowels that represented the loudest peaks in your recording; again this is something that proper breathing (diaphragmatic) will help you solve in the long run, but saving a take by doing this before the application of normalization will ensure that this process works as it should and that the valleys in your speech are amplified along with the peaks at an even level. One could also go for automation if that works best or if it's more convenient.
So in short:
1) Change the position of your mic to have the capsule point at your forehead/bridge of your nose (at a distance of 3 to 4 inches max).
2) Keep those tracking gain levels of yours, they are working great!
3) Once you've completed the tracking (recording process) first thing to do is to apply the noise gate at the 38 dB threshold level.
4) Glance at your waveform and identify the loudest peaks. Lower their amplitude manually or through automation.
5) Apply the EQ settings: cut frequencies below 79 Hz by at least 6 dB, reduce the 1300, and 3300 Hz frequency ranges by 1.5 dB.
6) Listen back to your audio and try to find the 'minutiae' such as pops, missing sounds, or other quirks. If everything went fine, this step should return no incidences.
7) Proceed to normalize the audio at our prescribed -3 dBFS Peak Level (stay away from RMS Normalization as if were COVID). Export your file at the set resolution parameters and make sure to enable Triangular Dither (especially if you recorded at a higher sample/bit rate as one always should).
8) Have a brew, a cup of coffee, or your celebratory drink of choice. If your resulting waveform sounds like the reference file I attached (or better) you've graduated VO Recording 101!
I hope these tips and strategies help you out, I'm eager to hear your audios!
Cheers!
- Héctor Adolfo Ituarte (Bunny Studio QC Agent)
Hector,
Here is the problem. Everything you just said, on a recording that I quickly recorded because your colleague asked I send over a raw audio format, has ZERO to do with the feedback I was given as to why I was rejected. I know what noise gates are, I know what EQ settings I need to be at, and I don't need to be told I'm passing "VO Recording 101" when your own site's directions said not to make any post-production treatments to the audio during the audition. I've been doing this for a long time and know how to treat my audio, but for you and whoever graded my audio to tell me that there's that much wrong with it, when I'm just following directions says more about Voicebunny than it does about my background in audio. I'll be damned if I'm going to be talked to like I'm some rookie.
Hey, Stephen!
I totally understand where you're coming from both as a VO pro and a QC agent. The tips and strategies offered were never meant to treat you like a rookie nor to demean your experience in this field, but only as friendly advice on how to get the particular raw sample up to what we could approve as a successful audition.
One thing to consider is that pros in our platform receive constant feedback during revision requests and that it often includes detailed guides like this one, in which we advise them punctually on how to solve the particular issues detected during the review of their submissions, so being able to abide by the recommendations issued in a QC revision is of the utmost importance, regardless of our level of expertise.
When we dispense advice in these forums that might 'contradict the no processing request' found in our project instructions, it's only to help out our pros and aspiring talents, for we know that no home studio is perfect and that pre-vetted, carefully tailored suggestions (with examples of the results) can be the recipe for success when working or applying to work with Bunny Studio. Believe me that even when the most seasoned pros experience a change in their recording setup or their processing choices (because everyone does dabble in the post-processing arts a bit and we know it), we sure let them know that said changes do not work, as we keep a close eye to quality-over-time and their past work and so is done by the clients who issue the projects.
Now, let me reiterate what our invitation to join the forums states:
"We trust our feedback will help you improve the quality of your work and achieve professional results... we invite you to visit our community page. This is a safe space where you can get advice from the Bunny Studio QC Team, in which other members and pros are sure to provide you with constructive commentary."
I can assure you that we don't tolerate nor engage in demeaning, derogatory or insulting behaviors and that all of what we try to do for you and other pros, is done with the best of intentions. If you feel that my thoughtfully crafted advice is uncalled for or that my choice of words was meant to ridicule you, I'm sorry but that was never the intention. I can tell you that as a pro I took me more than my fair share of tries to be approved in the platform and that what I've learned with Bunny Studio helped me grow out of preconceptions and improve upon what I had learned from almost 20 years of radio experience and that was a tough pill to swallow at the time, so please believe me when I say that everything in my previous response was said in the friendliest, kindest way and in the best of spirits.
Respectfully,
Héctor Adolfo Ituarte (Bunny Studio QC Agent)