How do you EQ?

Post your questions & get help from friendly LibriVoxers
SpookyNoodle
Posts: 41
Joined: January 7th, 2016, 3:02 pm

Post by SpookyNoodle » January 4th, 2020, 5:18 pm

Equalization is obviously going to be different from narrator to narrator, microphone to microhone, and set up to set up. However, I just started recording in the studio I set up in my basement, for the first time in over a year, and going over this last recording I did, I can't help but feel the recording needs a little...something. I don't like it, but I can't pinpoint why I don't like it, or what I need to do to fix it.

In the end, I settled on a tutorial, and my EQ ended up looking something like this:

https://i.imgur.com/msOefQV.png
It looks much more aggressive than it actually is-- notice that the parameters only go to +6 and -6 dB--and it certainly seems pretty standard, but it's just...not enough. When I listen with my HyperX or desktop speakers, it sounds fine, but when I listen through some proper monitors, with a flatter frequency response? Somehow it sounds boxy, closed, muffled, tinny, and nasally all at once.

What are some things that you do to EQ your narration? I can't seem to get it right, no matter where I put those bands, and I'd love some advice from people who are a little more practiced.
The real question is: why not?

philchenevert
LibriVox Admin Team
Posts: 18260
Joined: October 17th, 2010, 9:23 pm
Location: With my feet in the Bayou
Contact:

Post by philchenevert » January 4th, 2020, 5:36 pm

Ha. Boxy? Tinny? muffled? all at once? wow. I think you have found it! Seriously, I don't have any advice. I don't use EQ. I use high pass filter and then sometimes the Bass and Treble effect which I have fiddled with. I do envy those people who know how to use EQ though and wish you well. someone will come along and help no doubt.


------------
Oh, and the real answer is : because i said so.
Phil Chenevert, The LibriVox Video Guy

Why was Pavlov's hair so soft? Cause he conditioned it of course.


Pssss! Some Tutorial Videos here. Take a peek.

tovarisch
Posts: 2884
Joined: February 24th, 2013, 7:14 am
Location: New Hampshire, USA

Post by tovarisch » January 4th, 2020, 5:44 pm

Noodle,

The graph you've shared (BTW, not a good idea to share graphics since it excludes visually impaired folks who rely on the messages to be read to them) is not a good EQ curve (based on my understanding of the purpose of EQ curve) since it is essentially focused around two specific frequencies (~120 and ~400 Hz). Why? What bothers you in those frequencies? The 120 Hz is a usual source of trouble because of a noisy computer -- it's one of those electronic synch frequencies on which you can have some interference (like a spinning hard drive's motor, e.g.)... Use a notch filter to get rid of it!

The EQ stands for EQualization, and in my understanding the idea is to make the recording sound as pleasing as possible. What can there be about your voice that needs such a weird "correction" function?

Well, from those notes you probably can see that my acquaintance with equalization is rather superficial. I don't use it myself, nor do I really believe that one needs it for a voice-over. In most cases all you need is to cut down any rumble and, if your microphone is so inferior that it does not register highs well, bring them up a touch, and that's what you'd use EQ for. Of course, if your recording environment is quiet enough and the mic is good enough, there should be no need beyond, maybe, the low cut filter (often a setting on the mic).

Everything else is taken care of by Noise Cleaning.

That's my story and I'm sticking to it. :)
tovarisch
  • reality prompts me to scale down my reading, sorry to say
    to PLers: do correct my pronunciation please

annise
LibriVox Admin Team
Posts: 32744
Joined: April 3rd, 2008, 3:55 am
Location: Melbourne,Australia

Post by annise » January 4th, 2020, 6:49 pm

Can I just add - very few people listen to LV audiobooks via high-class speakers. most listen in cars, commuting, at the Gym, vacuuming the house, etc so you are probably over-processing.

Anne

KevinS
Posts: 7021
Joined: April 7th, 2019, 8:32 am
Contact:

Post by KevinS » January 4th, 2020, 7:20 pm

I am not a professional nor do I have professional equipment, so I convince myself of things so that I can feel better.

But one truth that I think is universal and true---a true truth, if you will--is that the listener quickly adjusts to a recording's overall sound and the mind focuses on the meanings of the words rather than their timbre etc. etc.
New modem installation delayed until instructions are clear.

tovarisch
Posts: 2884
Joined: February 24th, 2013, 7:14 am
Location: New Hampshire, USA

Post by tovarisch » January 4th, 2020, 7:24 pm

... right, and that's why we concentrate on delivery and content quality. A boring book or a book badly read can sound wonderful, clean, dry, and yet be boring or badly read. No studio equipment or brilliant sound engineering can save one from inability to engage the listener...
tovarisch
  • reality prompts me to scale down my reading, sorry to say
    to PLers: do correct my pronunciation please

SpookyNoodle
Posts: 41
Joined: January 7th, 2016, 3:02 pm

Post by SpookyNoodle » January 4th, 2020, 11:12 pm

https://music.tutsplus.com/tutorials/a-master-guide-to-voice-equalization-how-to-apply-eq-to-voice-recordings--cms-25184

This is the article I consulted when I was doing my preliminary stuff, and while the before/after comparisons between this recording and the raw recording indicate that the effects make it sound better, it still sounds like it needs something more.

The headers that I specifically referenced were as follows:
  • Use a High Pass Filter to Cut Everything Below 80Hz
    Cut 100-300 Hz to Add Clarity
    Cut 300-400 Hz if the Voice Sounds ‘Muddy’
    A Wide Boost Between 2-6 kHz Can Improve Clarity
    Cut Around 3-5 kHz if the Voice Sounds too Sibilant
Now, these were just headings, and the article has many more details that explain why they suggest those frequencies.

Whenever I just try a boost or cut band, and slide it around to find a frequency, I can hear an obvious difference as the frequency changes, but nothing sounds truly right.

I'm recording on a Blue Bluebird run through a Scarlett 2i2, and the recording space is pretty well treated, so I'm not super worried about EMI, although the plumbing runs through the wall pretty close to the mic, so I try to be on the lookout for that.

And while I'm aware that most people won't be hearing a difference, from what I understand, EQ can make an unconscious but impactful difference on the listener. Sure they'll adjust after ~10 seconds of listening, but I want to make that adjustment as smooth as possible. I'd like to use the experience I gain from volunteering with LV to hopefully become competent enough to produce stuff for Audible, or other Audiobook services, so even if that effort won't mean much to the average listener, I want to make my 'product' the best it can be, you know?
The real question is: why not?

lurcherlover
Posts: 953
Joined: November 10th, 2016, 3:54 am
Location: LONDON UK

Post by lurcherlover » January 5th, 2020, 1:07 am

Good advice from many of the above, especially from tovarisch.

I take it that you are a male reader. I use very little EQ and even then very sparingly. Maybe a 3-4dB boost at around 100-200 Hz. But using good mics in a treated room with very clean mic pre's means I do not need any EQ.

Can you send an example of your voice in a short clip - maybe 20-30 seconds? If you are not speaking into the mic centrally and instead across the front at an angle (depending on cardioid, omni, hyper-cardioid etc) - that can degrade the sound quality.

"Cut 100-300 Hz to Add Clarity Cut 300-400 Hz if the Voice Sounds ‘Muddy’"

This sounds wrong to me. Cutting at those frequencies will add mud, not cut it. You need to leave it level, or even boost those frequencies. It should not be necessary to do much drastic EQ'ing. Or it may be that you have a voice which you don't like (others may like it) and no amount off messing around with EQ will make much difference.

With regard to Audible.com - don't have high expectations, unless you have a recent best seller to record, you will only make peanuts, and it's not worth the bother. Record for Librivox instead!

P S Listen on headphones and you may get a better idea of how your voice sounds. I rarely if ever use speakers as they (even the best) are poor reproducers of sound. I have to EQ my rubbish speakers a lot to get the same result as I get with any moderately good headphones. And speakers are effected by the monitoring environment i.e. the room - and so are not really accurate.

tovarisch
Posts: 2884
Joined: February 24th, 2013, 7:14 am
Location: New Hampshire, USA

Post by tovarisch » January 5th, 2020, 12:03 pm

Noodle,

What Peter wrote made me think that perhaps what you were trying to do was due to the quality of the speakers you were using. You fixed how your voice sounded there, then switched the speakers, and got a totally different result perhaps because the other set of speakers needed a different correction altogether themselves. It's great that you can hear the difference those three decibels make. Try using reference headphones (ones that are true to the sound, not boosted in any way in any part of the spectrum).
tovarisch
  • reality prompts me to scale down my reading, sorry to say
    to PLers: do correct my pronunciation please

SpookyNoodle
Posts: 41
Joined: January 7th, 2016, 3:02 pm

Post by SpookyNoodle » January 5th, 2020, 10:06 pm

Yes, I am EQing using a pair of flat (relatively) frequency response headphones. Commercial headphones sound fine, but they're not the ideal choice for processing.

I'm not sure how to upload samples for general purposes on the forum, and I obviously don't want to use some poor MC's project, so what would be the best way to get a vocal clip to you?
The real question is: why not?

Availle
LibriVox Admin Team
Posts: 17631
Joined: August 1st, 2009, 11:30 pm
Contact:

Post by Availle » January 5th, 2020, 10:12 pm

You can use the "tests" folder, it is cleared out regularly.

Keep your sample short though, please :wink:
Cheers,
Ava.

--
AvailleAudio.com

SpookyNoodle
Posts: 41
Joined: January 7th, 2016, 3:02 pm

Post by SpookyNoodle » January 6th, 2020, 9:52 pm

https://librivox.org/uploads/tests/lib_raw_test_1_5_2020.mp3

Excellent, thank you!

(This is me just rambling on, not actually annoyed at anyone in the thread, I couldn't be more pleased :D )
The real question is: why not?

lurcherlover
Posts: 953
Joined: November 10th, 2016, 3:54 am
Location: LONDON UK

Post by lurcherlover » January 7th, 2020, 1:53 am

SpookyNoodle wrote:
January 6th, 2020, 9:52 pm
https://librivox.org/uploads/tests/lib_raw_test_1_5_2020.mp3

Excellent, thank you!

(This is me just rambling on, not actually annoyed at anyone in the thread, I couldn't be more pleased :D )
Well, I cant find too much wrong with that clip at all! No distortion, plenty of gain, silent noise floor.

I think you may not like your voice, but I would imagine many people will find it very good and clear with good diction. It then comes down to how good you are at expressing yourself with reference to the written content be it fiction or whatever.

(Not many people get all those things right, even after some time). Excellent voice and technical quality of the recording.

PS I put the file into "checker" the program we use to make sure the file is valid for LV. It came up with the message that the file appeared not to be an MP3 file. Can you check that?

Also the checker app is excellent so it is worth downloading.

realisticspeakers
Posts: 1237
Joined: December 6th, 2010, 5:15 pm

Post by realisticspeakers » January 7th, 2020, 2:14 am

SpookyNoodle wrote:
January 6th, 2020, 9:52 pm
https://librivox.org/uploads/tests/lib_raw_test_1_5_2020.mp3

Excellent, thank you!

(This is me just rambling on, not actually annoyed at anyone in the thread, I couldn't be more pleased :D )
This sounds great to me. I suspect this is processed and not flat. It's very smooth across the spectrum. Looking at the eq response graph I don't see any major humps. Personally, I think you could boost about 150Hz another 3dB and get away with it.

You are probably your worst critic and totally losing it with details. Sorry, I cannot relate to that at all...not one bit.
You should explore multi-band compression. It will give you that "oomph" you're looking for that eq can't do.
Truth exists for the wise, Beauty for a feeling heart: They belong to each other. - Beethoven
Disclaimer:
"Kind reader, if this our performance doth in aught fall short of promise, blame not our good intent, but our unperfect wit."

SpookyNoodle
Posts: 41
Joined: January 7th, 2016, 3:02 pm

Post by SpookyNoodle » January 7th, 2020, 7:06 pm

That's super strange, I exported this file at 128 kbps, MP3, 44.1 kHz, all the typically requirements that I see for LV stuff. I'll be sure to download the checker program and check it out, perhaps the LAME codec is interacting with it strangely? I'll do some tests.
realisticspeakers wrote:
January 7th, 2020, 2:14 am
I suspect this is processed and not flat.
😳 well thank you, I shall take that as a credit to my equipment. I recorded this raw into Reaper, then went straight to export, no FX whatsoever.

I guess if nothing sticks out as obviously wrong, then that's something I'll have to worry about when I'm harassing Audible with my demo reel 😅
The real question is: why not?

Post Reply