CAD u37 Corrective EQ, Make it Sound Similar to Other Microphones!

CAD u37 Corrective EQ, Make it Sound Similar to Other Microphones!

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
_caustic_ submitted a new resource:

CAD u37 Corrective EQ, Make it Sound Similar to Other Microphones! - Using carefully crafted EQ curves to improve the sound of the CAD u37

The CAD u37 is a budget oriented microphone that is getting more and more popular in the UTAU community.
Unlike the Blue Snowball it has only one polar pattern and diaphragm and a basic design. Meaning more money went into the sound and not anything extra.
It's very clear and not muffled at all for it's price and is a wonderful budget option for the beginner.

However you get what you pay for and this microphone could still use a little help. That's where this resource comes in!

Here you...

Read more about this resource...
 

HoneyPai

Defoko's Slaves
Defender of Defoko
But isn't is best to have zero effects on the vocal when going into utau? From personal experience, I say yes, because most of the time, effects on the vocals can make an utau sound very strange, and even very boxy. Plus, you can always just add EQ in mix, it gives it a better sound. Raw vocals (in utau) are very easy to mix with, while a bank with effects on it, such as EQ, come with mixing limitations.
This is my experience
 

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
Thread starter
But isn't is best to have zero effects on the vocal when going into utau? From personal experience, I say yes, because most of the time, effects on the vocals can make an utau sound very strange, and even very boxy. Plus, you can always just add EQ in mix, it gives it a better sound. Raw vocals (in utau) are very easy to mix with, while a bank with effects on it, such as EQ, come with mixing limitations.
This is my experience

It's best to have as few effects as possible generally, yes. I'm a major promoter of this idea.

However these are made specifically for the CAD u37 to improve the sound without negative effects and are what I consider to be an exception to that rule.
As long as you know what you're doing you can process audio samples without ill-effects.
 

HoneyPai

Defoko's Slaves
Defender of Defoko
Have you tested it in utau?
I've heard a particular bank EQ'd by you, and it didn't sound very good, I'm not saying you don't know what you're doing, the bank just didn't sound good before EQ and didn't sound any better after EQ
 
  • Like
Reactions: Kitcat190

Kitcat190

Defoko's Slaves
Defender of Defoko
Not to just jump in here after another comment, but I saw this and had to say a couple things.

1. I don't really see the major point in this. Why would you EQ curve to match another microphone, the point of a high quality microphone is to get that range. I get the sense that for those who can only afford some types of microphones you want to aim for a better sound. However you've recommended the CAD numerous times despite its problems. And to be honest, its hard to match up every time seeing as every voice is different.

2. It doesn't seem to actually work. All you've done is raise the bass, which in turn, just made you sound like you got closer to the microphone. I can see it raising clarity, but again, getting closer with limited background noise works just as well. You've called out blue products for dumping money where it shouldn't go however I've used the different settings to my advantage every time I record. (not to mention the CAD and snowball are almost identical) Trying to mimic another microphone doesn't actually help the quality, especially when it's butchered through utau. All this is, is doing things that should be done to the exported files from UTAU when you mix.

3. Like I've said before, not all voices are the same, and therefore by changing the EQ curve to fit your voice to mimic the sound of the microphone, will not sound the same to another. That and many many many things can go wrong especially with just presets.

Personally I'm not a fan of EQing samples, unless it's very minimal with using gate noise. But curving to mimic just seems like a bad tactic. You're never going to have a perfectly clean recording, if that were the case, than EQ curving would work. But all you're doing is boosting frequencies that wouldn't hurt during the mix.

Not trying to call out or anything, I'm honestly curious as to what the exact point of this is, it's a bit naïve.
 

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
Thread starter
Have you tested it in utau?
I've heard a particular bank EQ'd by you, and it didn't sound very good, I'm not saying you don't know what you're doing, the bank just didn't sound good before EQ and didn't sound any better after EQ
I have tested it in Utau and the result was a-okay.

Not to just jump in here after another comment, but I saw this and had to say a couple things.

1. I don't really see the major point in this. Why would you EQ curve to match another microphone, the point of a high quality microphone is to get that range. I get the sense that for those who can only afford some types of microphones you want to aim for a better sound. However you've recommended the CAD numerous times despite its problems. And to be honest, its hard to match up every time seeing as every voice is different.

2. It doesn't seem to actually work. All you've done is raise the bass, which in turn, just made you sound like you got closer to the microphone. I can see it raising clarity, but again, getting closer with limited background noise works just as well. You've called out blue products for dumping money where it shouldn't go however I've used the different settings to my advantage every time I record. (not to mention the CAD and snowball are almost identical) Trying to mimic another microphone doesn't actually help the quality, especially when it's butchered through utau. All this is, is doing things that should be done to the exported files from UTAU when you mix.

3. Like I've said before, not all voices are the same, and therefore by changing the EQ curve to fit your voice to mimic the sound of the microphone, will not sound the same to another. That and many many many things can go wrong especially with just presets.

Personally I'm not a fan of EQing samples, unless it's very minimal with using gate noise. But curving to mimic just seems like a bad tactic. You're never going to have a perfectly clean recording, if that were the case, than EQ curving would work. But all you're doing is boosting frequencies that wouldn't hurt during the mix.

Not trying to call out or anything, I'm honestly curious as to what the exact point of this is, it's a bit naïve.

1. ALL cheap microphones have problems. Some have low recording levels, some have noise, etc. The CAD has an issue of some bass rolloff while otherwise being fairly adequate in every other aspect. I recommended it as it's still fairly good for the price paid.

2a. I actually didn't just raise the bass, I did frequency sweeps with both microphones and compared the differences. Actually getting close to the CAD doesn't result in much of a bass increase and overdoing it can cause more harm than good on any microphone.

2b. I still stand by all the things I've said about the Blue Snowball, it still has small diaphragms, unnecessary extras, strong high frequency peaks that are tamed with thick foam that muffles the sound, etc. Those cannot be fixed easily while the sound characteristics of the CAD u37 are at least manageable.

3. same as above. Since I used frequency sweeps these EQ curves will be universal to everyone and not just me.



I have already had several people test these with good success which is why I have decided to release these publicly as a useful tool. Matching to other mics isn't necessary but provided a good goal/baseline for me to aim for.
 

Kitcat190

Defoko's Slaves
Defender of Defoko
I have tested it in Utau and the result was a-okay.



1. ALL cheap microphones have problems. Some have low recording levels, some have noise, etc. The CAD has an issue of some bass rolloff while otherwise being fairly adequate in every other aspect. I recommended it as it's still fairly good for the price paid.

2a. I actually didn't just raise the bass, I did frequency sweeps with both microphones and compared the differences. Actually getting close to the CAD doesn't result in much of a bass increase and overdoing it can cause more harm than good.

2b. I still stand by all the things I've said about the Blue Snowball, it still has small diaphragms, unnecessary extras, strong high frequency peaks that are tamed with thick foam that muffles the sound, etc. Those cannot be fixed easily while the sound characteristics of the CAD u37 are at least manageable.

3. same as above. Since I used frequency sweeps these EQ curves will be universal to everyone and not just me.



I have already had several people test these with good success which is why I have decided to release these publicly as a useful tool. Matching to other mics isn't necessary but provided a good goal/baseline for me to aim for.




1. All USB microphones will not get you "studio" quality. The CAD is a microphone not recommended for singing. It distorts and catches background noise far more than other mics we've compared in the discussion. In almost every review, it's had negative feedback unless if it's used for podcasts, and even then there have been major complaints. The CAD is a large diaphragm microphone, which in turn picks up background noise a lot easier. It does not make them better.

2. I don't see how you can stand by what you've said about the blue snowball when you back up the CAD. I'm not even saying this because I have it and I'm "butt hurt" I know it's not a good mic, however, the CAD is far worse. The Blue isn't even a small diaphragm, it's 14 mm which is more of a medium size. Small diaphragms don't mean they're bad, the downside is the self noise, but that happens in all mics. Small diaphragms have wider frequency ranges, ironically meaning, you'd have better luck doing this on a blue snowball verses a CAD.

3. Frequency sweeps are just running through every frequencies and mapping it out, how do you imply it gives universal practicality, cause thats not how it works to match a microphone to another one, it will still have a different result. Making EQ cuts and boosts do not match up to different voice types. How did you perform this, especially on a CAD...
Also, did you define success through utau or just talking through samples, because it's subjective. Again, the only thing that sounds clear through the plain ear is just the fact the sample provided was louder.

Again I'm not saying stuff just to say it, I'm honestly confused on how this works seeing as this is a bit far fetched.
 
  • Like
Reactions: Zoku and HoneyPai

Tomato Hentai

dont call me a veggie
Defender of Defoko
1. All USB microphones will not get you "studio" quality. The CAD is a microphone not recommended for singing. It distorts and catches background noise far more than other mics we've compared in the discussion. In almost every review, it's had negative feedback unless if it's used for podcasts, and even then there have been major complaints. The CAD is a large diaphragm microphone, which in turn picks up background noise a lot easier. It does not make them better.

2. I don't see how you can stand by what you've said about the blue snowball when you back up the CAD. I'm not even saying this because I have it and I'm "butt hurt" I know it's not a good mic, however, the CAD is far worse. The Blue isn't even a small diaphragm, it's 14 mm which is more of a medium size. Small diaphragms don't mean they're bad, the downside is the self noise, but that happens in all mics. Small diaphragms have wider frequency ranges, ironically meaning, you'd have better luck doing this on a blue snowball verses a CAD.

3. Frequency sweeps are just running through every frequencies and mapping it out, how do you imply it gives universal practicality, cause thats not how it works to match a microphone to another one, it will still have a different result. Making EQ cuts and boosts do not match up to different voice types. How did you perform this, especially on a CAD...
Also, did you define success through utau or just talking through samples, because it's subjective. Again, the only thing that sounds clear through the plain ear is just the fact the sample provided was louder.

Again I'm not saying stuff just to say it, I'm honestly confused on how this works seeing as this is a bit far fetched.
I'm pretty sure _caustic_ knows that you can't get studio quality from a USB mic, the whole point of this is just to make the CAD u37 sound clearer.
Every single usage example of Blue's Snowball that I've heard sounds extremely noisy in comparison to most uses of the CAD u37. I was honestly was originally considering maybe buying a Snowball until I heard how the u37 usually picks up a lot less noise.
 

Kitcat190

Defoko's Slaves
Defender of Defoko
I'm pretty sure _caustic_ knows that you can't get studio quality from a USB mic, the whole point of this is just to make the CAD u37 sound clearer.
Every single usage example of Blue's Snowball that I've heard sounds extremely noisy in comparison to most uses of the CAD u37. I was honestly was originally considering maybe buying a Snowball until I heard how the u37 usually picks up a lot less noise.
I appreciate you speaking on his behalf, but what he's doing is just eqing when you can do that when you mix. And also, here are my examples.
I never said blue were good, every usb is bad like I said before. But CAD has far more complaints, and ones that i've heard personally. I'm not exactly arguing on "no this one!" i'm arguing the logic behind this when the facts aren't there
 
  • Like
Reactions: Tomato Hentai

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
Thread starter
Everything that is being said here are things that I've said before.

1. All USB microphones will not get you "studio" quality. The CAD is a microphone not recommended for singing. It distorts and catches background noise far more than other mics we've compared in the discussion. In almost every review, it's had negative feedback unless if it's used for podcasts, and even then there have been major complaints. The CAD is a large diaphragm microphone, which in turn picks up background noise a lot easier. It does not make them better.

2. I don't see how you can stand by what you've said about the blue snowball when you back up the CAD. I'm not even saying this because I have it and I'm "butt hurt" I know it's not a good mic, however, the CAD is far worse. The Blue isn't even a small diaphragm, it's 14 mm which is more of a medium size. Small diaphragms don't mean they're bad, the downside is the self noise, but that happens in all mics. Small diaphragms have wider frequency ranges, ironically meaning, you'd have better luck doing this on a blue snowball verses a CAD.

3. Frequency sweeps are just running through every frequencies and mapping it out, how do you imply it gives universal practicality, cause thats not how it works to match a microphone to another one, it will still have a different result. Making EQ cuts and boosts do not match up to different voice types. How did you perform this, especially on a CAD...
Also, did you define success through utau or just talking through samples, because it's subjective. Again, the only thing that sounds clear through the plain ear is just the fact the sample provided was louder.

Again I'm not saying stuff just to say it, I'm honestly confused on how this works seeing as this is a bit far fetched.

1. there are good USB mics but certainly not for $50. That being said none of them would be considered "studio". I already know this and have said it myself on several occasions.. The CAD mic does not distort on singing unless you're clipping and that's as simple as lowing the input volume. When comparing the noise levels it catches to other large diaphragm condenser mics it is basically the same.

2. The blue is still way overpriced for the components that it contains, it's issues are unfixable in post while the CAD's only issue is simple an issue of the EQ applied by it's internal electronics (which all mics do to some extent) and is easily remedied. Larger diaphragms are optimal for voice recording. Smaller diaphragms have higher self noise than Larger diaphragms which is why it's a negative. Larger diaphragms capture low frequencies better than smaller diaphragms and easily capture sound well beyond our audible range which makes that a moot point.

3. Frequency sweeps are even and do not create room for error and bias unlike a voice recording. Frequency sweeps are used to tune rooms and even perform diagnostics during microphone manufacturing to make sure it's within spec (though this is to a much higher accuracy). Since these EQ curved were created specifically with the CAD u37 and Shure SM7b (which should all sound identical to other mics of the same model) this means that the result will be the same every time.
This also means that you can only use these EQ curves on recordings from a CAD u37. You technical can but the result won't be as expected.
I never said it would sound good with every voice, simply that it will work as expected with every voice/recording. If it doesn't fit a voice then they may not sound good with the real thing either. That's just random chance and not what I'm trying to achieve. I chose the SM7b for example simply because it's such a versatile microphone that tends to sound at least decent on most people.
If it doesn't fit then you can simply try a different EQ until something matches nicely or do minor tweaks (+-3db) to one of the EQs yourself if you want to.

My goal isn't to make miracles but to let people try different "flavors" to get a better overall sound.

I defined success both in terms of how the records sound on their own and through Utau.
 
Last edited:

HoneyPai

Defoko's Slaves
Defender of Defoko
Everything that is being said here are things that I've said before.



1. there are good USB mics but certainly not for $50. That being said none of them would be considered "studio". I already know this and have said it myself on several occasions.. The CAD mic does not distort on singing unless you're clipping and that's as simple as lowing the input volume. When comparing the noise levels it catches to other large diaphragm condenser mics it is basically the same.

2. The blue is still way overpriced for the components that it contains, it's issues are unfixable in post while the CAD's only issue is simple an issue of the EQ applied by it's internal electronics (which all mics do to some extent) and is easily remedied. Larger diaphragms are optimal for voice recording. Smaller diaphragms have higher self noise than Larger diaphragms which is why it's a negative. Larger diaphragms capture low frequencies better than smaller diaphragms and easily capture sound well beyond our audible range which makes that a moot point.

3. Frequency sweeps are even and do not create room for error and bias unlike a voice recording. Frequency sweeps are used to tune rooms and even perform diagnostics during microphone manufacturing to make sure it's within spec (though this is to a much higher accuracy). Since these EQ curved were created specifically with the CAD u37 and Shure SM7b (which should all sound identical to other mics of the same model) this means that the result will be the same every time.
This also means that you can only use these EQ curves on recordings from a CAD u37. You technical can but the result won't be as expected.
I never said it would sound good with every voice, simply that it will work as expected with every voice/recording. If it doesn't fit a voice then they may not sound good with the real thing either. That's just random chance and not what I'm trying to achieve. I chose the SM7b for example simply because it's such a versatile microphone that tends to sound at least decent on most people.
If it doesn't fit then you can simply try a different EQ until something matches nicely or do minor tweaks (+-3d:shades: to one of the EQs yourself if you want to.

My goal isn't to make miracles but to let people try different "flavors" to get a better overall sound.

I defined success both in terms of how the records sound on their own and through Utau.


I do not understand why you keep trying to say the CAD is better than the Snowball, price wise, it all depends on where you get it, I got my Snowball for $50, I am certainly willing to pay $10 more for something better than the CAD.

But, as Alice said, this is not about which one is better, we all know USB mics are not very good, the point of her statements were to get you to directly answer what the logic was behind this, why spend time going through samples one by one, adding EQ when you can just add EQ when mixing. This seems reasonable for podcasts, but not so much utau. I have also heard the CAD clips badly, a friend of mine owned one and it was just pretty bad. But do not focus on that statement, because once again, this is not about which mic is better. But to answer "the snowball is overpriced for the components " it is not over priced if you utilize them.

I, like Alice, would just like to know what the logic is, behind this, and why you continue to recommend a microphone that you yourself said was bad, when there are better budget mics. You just have to look for them and simply not have the mind set that you can only get a specific mic, for a specific price, again, the snowball is around $60, I got mine for $50. The CAD u37 is around $40, I've seen it for $20. I bought a Blue Spark XLR, it's price is $175, I got mine for $98
You really just have to look around.
 

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
Thread starter
The CAD and is almost consistently cheaper. The u37 trends around $40 to $47 while the snowball trends $50 to $70. I wouldn't say the Snowball is at all better in any way either for the reasons I've also stated previously.

I have stated several times the logic behind this.
But I'll try to satiate you.
This is easier for people as the bulk of the work has already been done. This is much different than EQing for a song as the result will be different.
Also, EQ applied before Utau is going to be different than EQ applied after Utau. It will sound different as it's supposed to. This also changes the base state of the output render which makes for better presentation (better perceived quality) and can also make it easier to mix later since less fiddling may need to be done.

Less work, better presentation, avoiding making changes after the render.
 

HoneyPai

Defoko's Slaves
Defender of Defoko
Because it saves time and effort for the user of a voicebank?

Adding EQ to each recording of a bank doesn't seem like much of a time or effort saver, what if you're recording a vcv multipitch? Or a vccv English? When you could easily just put a base EQ on the rendered vocals you're mixing and proceed to tweak it. If you really want to save time, you might as well download Virtual Audio Cable
 
  • Like
Reactions: Tomato Hentai

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
Thread starter
Adding EQ to each recording of a bank doesn't seem like much of a time or effort saver, what if you're recording a vcv multipitch? Or a vccv English? When you could easily just put a base EQ on the rendered vocals you're mixing and proceed to tweak it. If you really want to save time, you might as well download Virtual Audio Cable
You can do this to whole folders at once and then it's set and done and you never need to worry about it. Rather than having to do it individually on every single situation.




Sound quality is subjective, if you don't like this and what people are doing then don't do it yourself.
It's completely unnecessary for you to go and harass me over it and then proceed to post messages elsewhere about this.
 
Last edited:

Zoku

making doper vocaloid music than the rest
Defender of Defoko
@_castic_ tl;dr on cad issue: CAD hs better components than a blue yeti/snowball and consistently better quality than a blue yeti/snowball with a cheaper price point than a blue yeti/snowball


also iirc blue XLR mics are top notch
 

Tomato Hentai

dont call me a veggie
Defender of Defoko
Adding EQ to each recording of a bank doesn't seem like much of a time or effort saver, what if you're recording a vcv multipitch? Or a vccv English? When you could easily just put a base EQ on the rendered vocals you're mixing and proceed to tweak it. If you really want to save time, you might as well download Virtual Audio Cable
Even then, if you had to do it one sample at a time, it would still save more time for the USER of the voicebank.
Part of what I want to do with my voicebanks is make them all as easy as they possibly can be to use, so if I had to alter samples one at a time to make them sound nice and save the USER's mixing time, you bet I would.
 

HoneyPai

Defoko's Slaves
Defender of Defoko
You can do this to whole folders at once and then it's set and done and you never need to worry about it. Rather than having to do it individually on every single situation.

As I said to tomato, one might as well use Virtual Audio Cable since you can actually assign it to your mic, also stating that for users that do actually use this method, but do not have Audacity
 

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
Thread starter
This is meant to have an easy point of entry.
Most people don't like installing new software.
While many utau users will not have VAC installed nearly everyone has Audacity. Also VAC is not free software.

That being said, VAC would make more sense since you're doing it on the fly as you record. If that's what you want to do then that's you're choice, it's basically the same but different workflow.
I'll add a VAC version of the EQ curves for people who choose to do that instead and credit you.
 
Last edited: