It's common today for people to use better grade mics for things like Skype & in-game chat, while the quality of online video for stuff like YouTube & video tutorials etc. is often at very professional levels. That can mean that the people listening to your voice in whatever setting have higher expectations than in the early days of YouTube.
The easiest, simplest way to upgrade your recorded &/or online voice is with a USB mic designed for podcasts etc. like the Yeti.
bluemic[.]com/products/yeti/
They're not the cheapest price-wise, & they're not the top quality that you can get [it's easy to spend thousands on mics], but they're popular, with loads of people using them, so you'll meet the expectations of anyone hearing your voice. It's possible to get away with spending a fraction of that money, but it's also a mileage-will-vary sort of thing.
IMHO the best cheap alternative is a good Dynamic mic, the sort used by bands & for Karaoke -- their simple design means that big name brands can be had for cheap [$10-$30 on sale], & they're usually pretty rugged, e.g. bands & roadies brag about using Shure dynamic mics for hammers. A big downside is that they can be Very sensitive to vibration, meaning that if you use a regular cheap mic stand on you desk [~$10], every time you move or set something down on your desk the mic may pick it up. So you need to use a boom stand, which aren't that much more expensive, or they have clamp-on arms, like the adjustable stand for a drafting-type desk lamp. You can also pick up a shock mount, which suspends the mic in a web of essentially rubber bands.
Headset mics... You can pay a little or a lot for headphones or earbuds with a mic, & the same goes for the similar mics you see people wear on stage, on TV etc. By design they're very small, & that small size means low performance &/or inferior sound unless you pay more money for products with the engineering & electronics to compensate.
Condenser mics... They're generally better all-around, & more natural sounding with better tone characteristics, but they use electronics that means they usually need extra [Phantom] power supplied. Different brands/models have different sound characteristics, & some have switches to change the nature of their sound, but one thing most share is that they're at best somewhat expensive. Nowadays you can get some pretty cheap condenser mics [roughly $25], & they may have a better range &/or more natural sound than comparably priced dynamic mics, but they won't come close to a dynamic mic's durability, nor will they sound as good as the average $75-$100+ condenser mics. However if you're careful -- the included shock mounts are almost fragile -- they can work.
Now once you've got a mic [of whatever type] you need to plug it into something... Since Vista & the [alleged?] death of XP, add-on sound cards have become rare. Most PCs & all laptops use an audio chipset -- *designed for listening* -- on the motherboard instead. How well the mic input works [& how much noise gets added] depends on what PC or laptop you've got, and your personal standards. The cheapest way to step up is with a USB recording or input box or mixer -- both start at ~$50, with the mixer usually giving you more controls. Note that when it comes to the electrical noise that shows up in PC/laptop recording, USB is not a cure-all, as the USB circuits themselves normally introduce electrical noise.
As above, condensor mics usually require phantom power, so if you decide you want or need a USB device to plug the mic into, then you can choose between a dynamic or condenser mic -- if you don't use one, a dynamic mic is likely your only choice. Both condenser & dynamic mics are usually set up for & include a XLR cable... it's got 3 pins where it plugs into the mic, while the plug at the other end of the cable varies. That's something to bear in mind when/if you're looking at mics &/or USB devices, since if the XLR cable that comes with the mic doesn't work with the USB device [or your PC/laptop], you've got to get another cable or some sort of adapter.
You might want to buy or make a spit screen -- more politely called a pop filter -- which is basically a frame [usually circular] with material similar to womens' nylon stocking stretched over it. Some people also make or buy a baffle setup to cut down on the amount of extra noise reaching the mic in a non-studio setting... essentially some sort of frame that blocks the back & sides, covered on one side in foam rubber to absorb noise vibrations, it can sit on your desk or attach to the mic stand or boom arm.
https://en.wikipedia.org/wiki/Microphone
Recording Software...
1st off, if you're recording voice or music for whatever reason, one option is not to use a PC or laptop etc. at all, avoiding the noise that comes not just from any cooling fans, but from the electronics. I won't go so far as to say that they're cheap [IMO anyway], but recording devices can be surprisingly affordable -- the world's full of starving musicians, & it makes no sense for many companies to price them entirely out of their market. The only caveat I'm aware of is that recording audio this way doesn't always work with analog video capture, but recording the audio afterwards to match the captured video should be OK.
If you record yourself speaking while also recording video, say of in-game footage or for a tutorial, the software you use will integrate the audio with the recorded video. If you're doing a voice-over or narration to accompany video you've already created [e.g. recorded game footage or a tutorial], it's probably easiest to record the audio in a NLE [Nonlinear Video Editor] alongside that video. Not all NLEs are created equal however, & there are several audio apps that let you preview video as you record & edit your audio if you want or need to. Or you can just play the video in a separate player window. If you're not recording in a NLE, try to make sure that the video player & graphics hardware driver settings aren't set to change the fps [Frames per Second], e.g. dropping frames for smoother playback.
Normally you're better off to record audio as-is, adding any FX later, since that lets you change your mind, Undo etc. In that respect one app's as good as any other -- the quality, amount of noise etc. is hardware dependent. If you know your recording hardware's internal sample rate [e.g. 44.1 or 48], recording & later working with your audio at that sample rate may avoid some lossy conversions during recording, processing etc. -- use software with dithering if you need to change the sample rate when you're done with it.
Many recording apps have an ARM button you click before recording -- it makes all the connections & works just like when you're recording, but nothing's written to disk -- so that you can check & set levels etc. Watching the software's meters you want the level as high as possible without hitting the ceiling or upper limit -- nothing over that upper limit will be recorded [Clipping], so whenever the input volume exceeds it, you'll get a flat line of what sounds like distortion. The most obvious solution might be to set the input levels low enough that clipping would be impossible, & then boost the volume in editing software later -- the problem with that is that you boost the volume of everything, both the stuff you want [your recording], & the stuff you don't [noise].
Some software gives you the option of automatically limiting peak levels to avoid clipping... usually I think that's something you see in more entry level apps because it potentially adds latency [the lag before what's recorded is actually written to disk] while reducing the original dynamic range.
Better audio software models a real world recording studio in that you can route the audio in ways that mimic the various patch or jumper cables between separate pieces of equipment in the studio. This can seem pretty confusing -- "dry sound definition" [w/out quotes] gets 32 million + hits on Google, so if you're confused you're Not alone. Your audio is *Dry* if it's as recorded, no FX added -- once you add an effect, say reverb, it's *Wet* -- and you might route wet & dry audio to several virtual devices or filters, mix wet & dry audio and so on.
Sometimes the ability to manage that routing is handy, even when all you want to do is record audio... if/when you use a USB device as input, it'll likely have its own drivers. And that device might show up as input as well as output -- it might have headphone &/or speaker jacks so that you can monitor the input live, without any latency [delay] due to software, and it may be possible to send the audio from your software back out to those headphone &/or speaker jacks. But what if you want to use that USB device for input, & listen to the results on headphones or speakers connected to your PC or laptop? In some software it's necessary to select the USB device for recording, then select your PC or laptop speakers so you can listen to the results. If instead the software lets you select input & output devices separately that can mean less work -- the cost is that those apps require more learning.
Arguably one of the best free audio apps comes from PreSonus... you have to create an account to get it, & once you're signed in, you have access to the manual as well as the app itself. It can be worthwhile to get the manual 1st & have a look -- it explains how the routing that I talked about works in the software, giving you an idea if you are willing & want to go there [or to anything similar]. The software itself doesn't have much impact on Windows, but the optional media downloads add up to gigabytes -- those are best installed via the app itself, & there's no way I can see to have them install anywhere but the default Documents folder [though that folder itself can be moved]. Otherwise the popular favorites of Wavosaur & Audacity should work fine. Resolve can make for a nice, free, NLE, but again there's a learning curve.
presonus[.]com/products/Studio-One/Get-Studio-One
blackmagicdesign[.]com/products/davinciresolve