Audio rejected for echo and mic technique
Hi!
I recently moved and have a new studio setup. My audio is getting rejected though due to echo and mic technique.
Here a link to my audio as well as pictures of my setup: https://drive.google.com/open?id=1OOcugxQjhgsSUbPLodg3pZ3LFudtNEO0
I'm recording in a padded closet, with the door opened about an inch. I use a Stellar X2 Large Capsule Condenser Microphone and am recording directly into a Zoom recorder.
The only processing I do is normalizing to -3 in Audacity.
I know my setup is in a really small space, but I have nowhere else to record in my apartment. Is there anything I can do?
Please sign in to leave a comment.
Hey Alex, a fellow VB voice here, so my two cents are just that.
First of all your post gives all the info needed for you to receive feedback.
Your recording space is indeed boxy and confined. This will create undesirable acoustics in your recordings but I think you can try a few different things. I'd be curious to hear the sound with the door fully open, the mic angle upward at 45 degrees making sure you speak off axis.
Naturally we want to avoid all plosives in the capsule by using correct technique and as for any boominess you could try a high-pass filter rolling off everything below 100-120hz.
Your tablet is very probably behaving as a hard and ugly reflective surface. Are you able to apply more acoustic treatment down the whole wall, perhaps some bass traps.
I highly recommend that you play with technique, volume, and mic position. LISTEN very carefully when monitoring and be your own harshest critic. Boxiness and boominess is easy to hear in a second, especially with any loud 'OH' sound
I do believe that your space is workable with the right treatment.
All the best
Hi James, thank you so much for your help!
So I tried recording with the door open and the mic at a 45-degree angle, but it sounded even worse.
I then recorded with the door open and the mic directly facing me. I normalized it to -3 in Audacity as well as added a High Pass Filter set to 100 Hz and -6db per octave rolloff: https://drive.google.com/open?id=1C2KmhSzdHbbW5quSSnGH3TFkuR5ZW8zZ
Do you think it sounds better? I'm having a hard time hearing changes in the audio.
Thanks!
The open door recording does sound pleasant in terms of a neutral, natural voice in my opinion but it isn't acoustically ideal and there is quite a loud noise/hum in the open door version. Does the adjoining room have loud electricals?
I liked the performance and mic technique this time. How are we with the door closed and using a roll off?
Another thing to listen for is ANY mouth noise related to dryness, saliva or general mechanics. Additionally short breath bursts before a line are completely undesirable unless the passage is impossibly long, even then we have editing. See what you can accomplish in single breaths and by decreasing those bursts in post. VB typically recommend a fade, personally I try to get it right at the source.
The water heater is directly across from the closet, so that's probably causing the hum. Couldn't I just you a noise reduction to take that out?
I recorded again, this time with the door closed: https://drive.google.com/open?id=1EkQg6JIYclgyBX6h3oHQR-pHoFaN4VIn
The roll off is definitely helping, but it still sounds a little boxy
As you know the noise floor is a crucial element in a professional recording. Noises 'in the gaps' can be cleaned up but that extra external layer will lay atop the whole recording so ultimately the goal is a noise floor of around -60db.
That said your 'close door high pass' recording is easily the best so far imo. Very close to what VB expects in my experience. As you say there is some boxiness and unwanted reflections but we are getting close.
Also great job technique-wise this time, aim for that in every recording!
Can you cover the remaining hard surfaces? And you can experiment with the roll off if you want to introduce some bass at 80-90hz
Ok, I placed foam pads all the way down to the floor.
I think this sounds the best so far: https://drive.google.com/open?id=1LjyF24lxtf_wjsddpJZKChdmbbuKENNX
What do you think?
Excellent, acoustics and technique are both now under control in my personal opinion. Two things to consider would be how you decide to manage high and low pass roll offs if any. Feel free to experiment with no high pass now or a lower one to introduce some bass and if things feel a bit high and sibilant cut some of the highs. And if possible experiment with other microphones too.
There is drastic improvement between your first and last recording. VB will most likely approve such audio I believe.
Good work.
It appears VB still isn't happy :/
I just submitted this audio for a speedy: https://drive.google.com/open?id=18-dsz8BZX4c62CKGenVnkYzcvvkglPPd
It was rejected for Room Echo and Processing.
Why do you think it's still being rejected for echo?
Should I try processing it at 120 HZ? I'm afraid if I don't process it, they're just going to reject it for echo.
Thanks so much again for all your help! I really feel like we're almost there!
Hmm with my 770 pros I'm honestly not hearing any echo or reflections. No the solution isn't cutting more bass frequencies, that will just remove any sense of body from your voice.
At this point we need deep audio analyses from VB themselves. To speed up that process inform them of your issue as well as this thread at support@voicebunny.com
That speedy audio sounds clean to me in every way.
I downloaded your speedy audio and amplified it to 125% (far beyond realistic) to exaggerate every characteristic. What I hear is a faint hiss and what could be perceived as high end sizzle (processing). Now using another pair of headphones I'm still not detecting echo.
Typically VB are against any sort of aggressive gating, compression or eq, but I'm just not hearing aggressive processing either. We need some other ears in here Alex.
So I just emailed support and they got back to me saying they would forward my concern to Bunny Pro Management and I'd hear back within 48 hours. I'm curious to hear what they have to say.
Before when my audio had room echo, they would get back to me with a revision. Now they just outright reject it. Maybe my account is flagged or something because I've submitted so much poor audio?
Yeh possible, if that's the case your account could be in poor standing statistically. Let's wait for now, I expect you'll see a reply in less than 48 hours.
One thing is without question. Your audio improved.
Alexander and James - Thank you very much for participating here!
Alexander's audio quality indeed has improved! Much kudos to your effort ;-) While I can only observe audio frequency range presented in these audio samples, there are some frequencies that seem to have been emphasized possibly due to your voice booth structure... and here's what I suggest to bring down:
Please experiment... Thank you!! We look forward to collaborating with you and your improved audio!! :-)
Best regards,
Oki
Thanks for your response, Oki!
I'm a little confused as to what I should change.
In Audacity, under the High Pass Filter Effect, the two options to change are "Frequency (Hz)" and "Rolloff (db per octave)". I currently have it set to 100 Hz and 6db.
Are you saying I should try it at 120 Hz and 174 Hz?
Thanks!
oh no , not the high-pass EQ :-(((
I'm not familiar with Audacity, but there has to be a feature that you can pinpoint the frequency (again it's NOT roll-off)
I hope this helps... thank you!
Best regards,
Oki
What program are you using? What's the feature called in that program to pinpoint the frequency?
So I discovered an effect in Audacity called "Notch Filter". It allows me to change the Frequency and Q Factor.
Here are some samples -
Normalize -3: https://drive.google.com/open?id=1guTdaS0nV9n4GszNfRLkvUOb7-6GWDhr
Normalize -3, 120 Hz, Q Factor 6.0: https://drive.google.com/open?id=1TSy4L_GOZ8PNxAJP8tR18CFw8SSd3vqH
Normalize -3, 174 Hz, Q Factor 6.0: https://drive.google.com/open?id=17yitW0EAhoztXq6T5Qgp7sS5rQJzmJXZ
To my ear, the 174 Hz one sounds the best. What do you think?
Just submitted a speedie with Normalize -3, 174 Hz, and Q Factor 6.0: https://drive.google.com/open?id=16qclqZYYr07OOlB08o2jduVBULPOCWPJ
It was rejected for:
Your recording sounds undesirably over-processed, which is impacting the quality of the deliverable. It may be due to a noise reduction plugin, an EQ, or a noise gate, for example.
Your delivery sounds undesirably robotic, as if was computer generated. Our clients expect natural and engaging reads, unless specifically instructed otherwise.
This is getting ridiculous. Not only are they rejecting it for processing, but they're saying my delivery is robotic? The instructions called for "calmly, smoothly and with a relaxed tone".
I'm at a total loss. I don't know what to do.
Hmm.... :-( To me, it sounds a-OKay.
Let us see what exactly is happening.
In the meantime, I had a chance to once again download and use Audacity ... and it is UTTERLY HORRIBLE :-((( !!! I am by no means a poor craftsman blaming his tools (well... at least I'm making my living out of doing voiceovers), but I really don't think I can produce any presentable audio products with it :-(((
What do I use? I use Steinberg Wavelab Pro ... but there should be a more affordable version named Wavelab Elements.
Alex I see that one of your speedies passed QA. Good stuff!
Yeah!!! Congrats!!! ;-)