Is it just me... or are the Bunny QA team on crack?
I am really tired of inconsistent evaluations of my audio by the Bunny QA team. I literally use the SAME processing chain everytime I record and sometimes it is accepted and sometimes it isn't. It really feels like they can just reject recordings willy-nilly, which goes against the whole principles of this platform.
When its not accepted they always say the same thing:
Here are the samples of the audio they rejected today for reference: https://1drv.ms/u/s!AiU1hP8RhYzUnLo8U0kh0U1K9kGRfw?e=zIuOmI
This is vastly frustrating for me! I have better things to do then record for an extremely low budget platform that can't get it straight what is acceptable for them to hear or not. The only reason I am willing to accept the ghetto ass price model is because it is based on the principle of paid to audition not paid for the gig, but if my auditions are sometimes being rejected and sometimes being accepted despite having the EXACT SAME PROCESSING this completely goes against this point.
Anyone else experienced the same?
Please sign in to leave a comment.
Hi Nathan, fellow VB Pro here. I've got to go with the Quality Control team on this one. The samples you provided for reference are WAY over compressed and limited especially since they request that you use little to no processing on your tracks when you submit. Additionally, there are peaks over the -3db threshold. Yes, it can be frustrating to use the same setup and have some tracks accepted and others rejected, but if this is representative of all of your submissions, then by VB's stated quality standards the big question is why any of them are being accepted. Try submitting your next tracks without any compression or limiting, and normalize to -3db to see if that helps. You've got a pleasing voice with excellent enunciation, but it's getting lost in all of the processing.
That's exactly it Chuck! "... by VB's stated quality standards the big question is why any of them are being accepted." It is inconsistent. Its easy enough to change my processing chain to adjust to the requirements of the QA team, as I did to create this processing chain originally, but if the requirements keep on changing from what is written, their rejections seem more aligned to 'we already reached quota for this and need a reason to reject this person' rather than 'they have violated our standards'.
For goodness sake, I even had a speedie rejected in the past for DIRECTIONAL reasons, which completely goes against the principle of the platform. When I mentioned I should be able to charge as the directional change they made results in a completely different approach on the read they said "There shouldn't be any additional payments on this occasion since it is a correction so that the entire recording sounds better." Think about this, a correction so the entire recording sounds 'better'. What the heck does better mean!?
It all just seems a bit whacky to me to be honest!
For one, what was the direction they gave you to result in the voice over "sounding better?" Usually you're expected to not take a fee if the mistake is on your end. If they change the direction for the read, say, "we've decided that we want you to read it faster" then yes, you should be paid for that revision.
The only advice I can give you is to forget any sort of processing chain for now, and just submit your natural voice, normalized to -3db, and see how that goes.
No mistake on my end, a change in direction on the client's end. Another example of inconsistency!
I wish it was as simple as normalized to -3db equals acceptance, but I have had recordings rejected with those standards as well, that being said, I will perhaps return to that.
Would be nice to have an actual QA comment on this though instead of us just speculating, although I do appreciate your speculation! :D
But here's the thing; I'm not speculating about the quality of the clip that you submitted. It DEFINITELY should not have been accepted, as it breaks 3 of the quality control rules: compression, limiting, and level.
If recordings of that quality were automatically being accepted, that's a problem on VB's end. That doesn't mean you should be submitting clips of that quality, though.
But, we are speculating about why they in some instances are being accepted and in other instances not.
I understand that what I am doing breaks their rules on paper, however I have been told in multiple instances by QA that EQ, Limiting, and Compression is acceptable in instances that it doesn't impact the audio quality heavily. And I have had this process chain accepted in gigs on the platform be they speedies, competitions, or direct requests. Despite the chain changing how the audio looks, it is really a soft touch processing chain IMO.
All I feel, in this regard, is inconsistency. And when time is money for us voice actors, and the pay is so low on this platform, it makes it difficult for me to want to invest time in it when there are other opportunities to chase.
I'll do the next few opps exactly as you say, and we'll test out what happens, but I honestly expect that they are going to be rejected on similiar grounds despite me only normalizing. Let's test it and I'll update it here later!
I'd be interested to see what setting you're using for your processing chain. Because to my ear, your sample is overly-processed. what compression and limiting settings are you using?
I utilize UAD Precision Limiter and Maximizer. I increase the volume via the limiter by 10-16'ish dB (this will depend on the volume recorded at). I then utilize a Maximizer to cap the volume out at -3dB.
With a limiter, the main concern isn't by how many db's you're boosting, it's the number of db's your signal is being limited... how much gain reduction is being applied. By the looks of your waveform, it's a lot. And in truth, with voiceover, you shouldn't need any limiting of your voice.
Interesting find there. I just went in and looked at this recording specifically and the gain reduction was quite high on the settings I had, went as high as -14db reduction at certain points in the limiter which would definitely squish the audio uncomfortably. I don't set the limiter reduction level myself, it automatically happens in my Limiter all I do is control the gain. I'm going to adjust these settings on my end to reduce the gain on the limiter, it will definitely make it sound subtley more natural.
Yeah, 14db reduction is a lot. It sounds like your limiter has a fixed threshold. So just increase you gain until you're getting about -3bd reduction max. All you're looking to do is to tame any rogue bursts of volume. For example, here's a voice over I did for a client this morning. It contains no limiting whatsoever... simply normalized to -3db.
Great observation! Thanks for your ears, I tend to not judge my waveform visually but by ear, and the audio I bounce sounds great on my monitors, but I think pulling it back will definitely keep in more of the range of the delivery and reduce some of the crunch feeling at moments. Great find, I'll def incorporate that in my auditions and work outside the platform immediately.
Now the question is, do I utilize that philosophy on my next Bunny opportunity or do I just Normalize it to -3DB as per their on paper standards. :D
Using your ears is great but for voice over it's easier to hear using headphones rather than monitors. When you listen to monitors, you're hearing too much of your room sound, which can cover up the actual deficiencies in the sound of your track, especially if you have a fairly reverberent room. The idea is clean, dry audio, and you can only realistically judge that in a clean, dry environment (ie, headphones.)
For your next Bunny job, I would leave your tack as clean as possible--no processing whatsoever--so that Quality Control can accurately judge your track, then go from there.
Whoops sorry. I mean headphones and studio monitors, where I come from monitors means both interchangably!
On my next Bunny job what I'll do is submit the Normalized one only, and I'll then email the QA team both directly and see which they prefer and then update this note.
We'll see if the normalized one gets accepted.... :D
Hi Nathan,
Apologise for the late response we have been inundated this past month and are still playing catch up responding to all the posts here in the community.
Completely understand your frustrations about all this.
A few things to mention, for bookings we are a lot more flexible in our assessments as long as the audio quality is the same quality the clients expect. Back-up contests are the same in this sense.
Other factors which come into play when assessing submissions are the client's demands too. Sometimes clients have very short deadlines and need a recording as soon as, and aren't as strict on quality requirements. Other clients are very strict on the quality they require, therefore with these ones we have to be more thorough in assessing them... Nevertheless, I digress.
But rest assured, there are many things to consider when we are QCing the submissions and without context, I can completely understand that this would be very frustrating.
Don't get me wrong, there are times when it is inconsistency too, which we are constantly trying to improve upon and align the team.
If you want to post the links to the submissions which have been rejected by QC, I'll happily have a look into it and see what happened?
Regarding the client wanting a new read - Again, if you want to post the link to that project I can have a look to see what happened there?
But from the sounds of it, it sounds like it was out of QC's hands for this one. After QC has accepted your submission to be forwarded to the client. It no longer is assessed by QC. So it sounds like the client was requesting for a 'better read'. Whole heartily agree this can be annoying at times, I've dealt with some rather peculiar requests and directional comments from clients in my time.
Thank you Chuck for your fantastic response!
I see Chuck has addressed the problems that I also heard in your audio, and offered some valuable help!
With his kind assessment, I believe you have the know-how to implement the changes.
Feel free to attach just a normalized audio file here if you wish, always happy to lend an ear and offer advice.
Thank you chaps, it's been great to see the discussion here on the community page as this is what it is intended for.
Hope you both have a great weekend! :)
Kind regards,
Samantha