FEEDBACK, PLEASE!
Hello! Can you please give feedback to my new setup?
-Rhode NT1A Mic
-Behringer Uphoria UCM202HD 2 channel Interface w/pre-amp
-Sony 7506 Headphones
-Audacity
-Room: I am in the corner of a room, with sheetrock walls.
-Booth: Clearwater Tall Sorber Baffle Panels, a moving blanket and a thick blanket completely closing me in, with a thick blanket over my head and a rug underneath me.
2 Samples:
https://drive.google.com/file/d/17b6nodeYzYxvuiAgtQi_6n0l5XdAObRi/view?usp=sharing
This Sample is recorded again, this time with the Gain up on the Interface a little more:
https://drive.google.com/file/d/1rwtBbH2tp1BHhwmmbmsIsMksCWk8uSRt/view?usp=sharing
How does this sound? THANKS!!
D
Please sign in to leave a comment.
Hi, Daisy!
Thank you so much for reaching out and asking for feedback, we love to help whenever possible =)
Here are my observations:
1) Gear
Your equipment seems more than adequate! We shouldn't have any problems in this regard.
2) Acoustic Conditions
Your echoes appear to be under complete control and your acoustic balance also seems very adequate. Let's keep all your acoustic treatment unchanged for the time being.
3) Audio Quality
That Røde is sounding quite sweet on your voice! However, we must address the elephant in the room: both submitted recordings are too quiet (peaking at around -22 dB) which will cause some issues if we try to normalize the waveform. I did just that and the result is this:
The spectral analysis of your waveform post-normalization (at -3 dBFS Peak Level) is that of very loud hiss (white noise) which you can see as the pink/purple noise all over the place. The present levels of this noise would make it impossible to completely remove to salvage the audio, as its presence is widespread through the audible spectrum (20 Hz to 20 kHz), so attempting noise reduction would degrade your voice too.
Suggestion Time!
1) Check your Mic and Interface
It's abnormal to receive such a low signal. The gain knob in that particular interface should be anywhere between 10:00 to 11:00 and never higher than 12:00. In your DAW, you should be receiving a signal of -9 to -6 dB, with the least amount of hiss possible. Double-check your mic to ensure no pads are active (they reduce the output of the mic). Also, make sure that phantom power is enabled in your interface.
2) Test for Gain Levels
With the previous indications in mind, test extensively until you get such a signal: -9 to -6 with the least amount of hiss possible.
3) Normalization
Normalize your waveform to -3 dBFS Peak Level (never use RMS). Use manual leveling or automation to even out your speech if necessary.
4) Edit and Export
Scrub the waveform looking for mouth clicks, noises, breaths and such. Use fades and mutes as per this guide if necessary. Should you notice audible hiss at this point, the use of a noise gate set at around -38 dB with the quickest attack and release times should help to eliminate it. Trim the start and finish of your waveform to .5 (point five) seconds of silence and export your work to the Bunny Standard for WAV files: 44.1 kHz/16-bit.
5) Listen Back!
Take a chance to listen as needed and try to be as critical of your work as possible: is it the best? Go ahead and show us. Is it still in need of improvement, show us too!
We'll be eagerly awaiting any developments and progress, so don't be shy and show us what you can do 😉
Cheers!
- Héctor Adolfo Ituarte (Bunny Studio QC Agent)
Hello!
I made the necessary changes. Can you let me know if this is now correct?
SAMPLE: https://drive.google.com/file/d/1R2Hx-kr4uDBbmt_zN7bjor1RQvmi1mfV/view?usp=sharing
Thanks!
D.
Hey, D!
Wow, you knocked it outta the park! This is a 97% improvement and I'm very proud of you! Here are my final three suggestions for you. These three tips will help you achieve pristine audio with world-class quality:
1) Apply a Noise Gate
Your interface and mic combo are silent as they come and the audio as you sent it would immediately qualify for any submission, but we can do better! A noise gate will automatically silence all of that almost inaudible noise/rumor that's inherent to all recording equipment, giving your audios that super professional sound and feel of a totally silent noise floor.
2) Normalize in Sections
The waveform complies with the normalization guideline (-3 dBFS Peak Level), but the peak in question is located in the phrase "I'm loving it!" (where the emphasis lies). This leaves the valley of your recording (the main part of the script) a little low on amplitude. The trick here is simple: if you select only the valley portion and choose to normalize that instead of the whole file, the non-peaky portion of your audio will rise to the same level of the peaks. Here's a before and after of how your waveform would look like by applying this easy technique.
3) Filter Unnecessary Frequencies
Røde mics are infamous for that rich low-frequency richness that can cause plosives in many instances. Your audio is not popping, but it has extra info in the sub 100 Hz range that is just wasted energy in any final mix or future post-production effort. Applying a Parametric EQ with these parameters will greatly help in cleaning up that useless information from your waveform and help reduce the thumping/hard quality of your hard consonants (and even help you salvage a great take that might be affected by a full-fledged plosive!). You'll notice from listening to your file after these techniques have been applied, that the sound is much more balanced and neat.
If you have de-activated profile or need to reapply to become a pro, let me know so that we can involve our Bunny Managers in this thread, as you're pretty much ready for primetime. Please come back to this conversation one more time after you've read, applied and mastered the tricks described! We'd love your feedback about the support received and we'd love it even more if you could record one last test after incorporating these pointers into your workflow.
Once again, congratulations on the fantastic work!
Hi Hector!
THANKS! I made 2 of the changes you spoke about above (normalizing the specific section and adding the noise gate). However, I am confused about the EQ. I am using Audacity, so my EQ looks like this: https://drive.google.com/file/d/1ZH0_d2dSIQR1TR1WvdlAbyyo_w2KWXgN/view?usp=sharing
Here is my new sample. Can you please let me know if these changes are better?
SAMPLE:https://drive.google.com/file/d/1ES4ez_ieriBwQux98luEDjd9wKrCzo6-/view?usp=sharing
THANKS!!
D.
Hi, D!
Yes, the audio sounds much more balanced with the selective normalization! Cheers on that!
Now, the waveform still needs some fine-tuning, especially with those two processes I've outlined. I'd highly recommend you changing DAWs, as Audacity is good enough for a certain level of work especially if you're in a pinch, but more professional options as Adobe Audition or Reaper would serve you very well in the long run. I use them indistinctly for all tasks, but the processes are pretty much the same in both.
Reaper is offering a very good trial (extended due to COVID relief) and a super affordable license for those of us who don't make more than $200K/year working in this trade, so I'd suggest you download and install it in your system. As an incentive to do so, I'll leave you with a preset FX Chain that you can simply add to your project and it'll take care of the problematic frequencies we've been working with. Click here to download a video on how to do this with ease (and this one to see a basic handling of a project in Reaper.)
Adobe Audition has a more limited trial, but it offers a month-to-month subscription model that could also result in a comfy purchase should you be interested (it's what I used to fix your original file). Sharing presets with Audition is kind of a pain, hence my previous recommendation as a first option.
Please, come back to this post and share more samples! I'll be thrilled to hear what you come up with!
Hi Hector!
Thanks for all the feedback. I'd love to reapply for the studio to become a Pro. Can you resend me the info so I can reapply?
Daisy
Hi Daisy,
We'd highly recommend sending more samples as requested by Hector.
That being said, here's the link to our Application Process: Application Process
Best of luck with your application!
Hello!
I have made another sample, and applied the techniques Hector asked for above. I did Noise reduction, Compression, EQing and Normalizing.
SAMPLE: https://drive.google.com/file/d/1Y_64AOWP68lGOSmTvEVzKyAWgcAfHXUM/view?usp=sharing
How is this? I want to make sure everything is good before I re-apply. Thanks!
D.
Hi, D!
I'm glad to hear more from you! This sample is much more balanced, but there's still muddiness associated with excess information in the low-frequency range. This excess information plays against the intelligibility of your audio too.
Here's the same file, processed with the low-cut filter previously suggested and normalized to -3 dBFS (which was also missing from your waveform). If you can produce an audition sounding like this file, you should be totally fine with the admission project!
Let me know if you need more info on how to process your audio to eliminate these sonic blemishes, I'm looking forward to your results =)
Cheers!
Below are 2 Samples: One Raw file with no editing, and One File WITH Complete editing:
Studio Sample (RAW): https://drive.google.com/file/d/1SYv4K2OezAbFzfDqaceDlnD_7sZ2U921/view?usp=sharing
Studio Sample (WITH EDITING): https://drive.google.com/file/d/19QYHp3i1dgn2AK0d2qlkHQuLpdfO_5qR/view?usp=sharing
I have attached photos so you can see exactly what I am doing to edit and my numbers:
Noise Reduction Settings: https://drive.google.com/file/d/1k2TfWXaCIvOzz54D2CnUl-ZY8ezrEJEv/view?usp=sharing
EQ Settings: https://drive.google.com/file/d/1uj1EMmYveS0bM939qlWNsV-F-0RgrZLT/view?usp=sharing
Compression Settings: https://drive.google.com/file/d/1zJqQ80HK_Uvns6VXQ3v1tMPo8iyz35Vk/view?usp=sharing
I normalized to (-3dB), but there were still some small sections that peaked out of the -6dB to -9dB range, so I normalized to (-6dB) in those specific spots to even everything out.
Is this up to par? Thanks!
Hey, D!
Thank you very much for the detailed information of your workflow and the samples as well. I gotta say, your audio and your magnificent voice do not need that much processing at all. Your tone is so lush and rich that applying all that editing is just overkill.
Let's break it down:
1) Noise Reduction - I don't mean to diss Audacity, but I wouldn't use their plugins even if my life depended on it. Their processing is just not up to par with professional results. I know we recommend it as a free option to normalize files that need that done, but I wouldn't really keep on using it any further, much less use their 'automated' noise reduction. For different pieces of software, noise can be interpreted as a whole lot of things within a broad or narrow frequency spectrum (most usually the first of these two). That will undoubtedly eliminate some echoes or background noises or hiss. But you don't really need it and I wouldn't use it.
2) EQ - the EQ curves you're using are doing the complete opposite of what they're supposed to do. In your case, as I've suggested many times, the only thing you need is a high-pass (or low-cut) filter to help clean up the muddiness present in your sound, which most of the self-produced talents need to use really, as our recording environments are far from perfect (acoustically speaking).
3) Compression - This is definitely something you don't need. Bunny Studio seldom requests talents to engineer the sound, quite the opposite. Most of the audio we receive from talents is produced down the line and compression is one of those things that can really screw up one's work without the option of taking it out later, as compressing the waveform will alter its dynamics forever. Hence, Bunny Studio enforces a zero-tolerance policy with compression on waveforms both in paid projects but especially in application projects as a general rule.
Now, what can you do to make your audio compliant with our standards? Easy:
1) Apply a noise gate - This cleans up the noise floor and leaves quiet silences all across the board as we've previously discussed.
2) Normalize and level your audio - I've already shown you how to do this so that you can bring levels up or down and then use a general peak-level normalization to ensure the whole waveform doesn't exceed our prescribed -3 dBFS Peak Level.
3) Apply a low-cut filter - This is paramount in your case, in particular, your audio does need the clean up of these frequencies. I've also included examples of this in previous posts.
4) Learn to do spectral edits - Some mouth clicks and noises of yours (and everyone else for that matter) can only be removed through spectral edits, where one can select a very focused part of the waveform and eliminate it as if we were deleting blotches of color in a piece of artwork.
Will all of this work? Hear for yourself! This is the raw sample with all that simple processing, nothing less, nothing more and it sounds bigger, cleaner, crisper than with all the effects you're applying (after the slate I left your processed file for reference). I really suggest you make the jump into the realms of Reaper or Audition, you won't regret it. By the way, all the processing done to your sample was made with stock plugins in Audition, nothing third party or fancy aftermarket purchases. It's as simple as it gets =)
Hopefully, you'll be able to tell the difference (as we're using the same headphones). Please, refer to the already dispensed advice on how to process your files and they will sound as good or if not, better than this one.
Cheers!
Thanks! I purchased Adobe and made the changes.
How does this sound?
https://drive.google.com/file/d/1TfJ01m9mVNcYe825kVASXu81u7rvTTmQ/view?usp=sharing
Hi, D!
Kudos on the acquisition! This sample is sounding great, however, it's a stereo file with a full mix. Would you happen to have the unprocessed voice for us to evaluate?
Thanks again for your hard work! =)