Hacker News new | ask | show | jobs
by nickjj 2322 days ago
You can spend $700+ on his high end audio recommendations but 90% of home office style recording environments (even with acoustic panels) are going to have a noticeable amount of background hiss (computer fans, etc.) get picked up and his set up doesn't account for that.

If you goto the OP's Youtube channel and listen to his recordings, for example https://www.youtube.com/watch?v=lh0sTxLs9VA, there is a HUGE amount of hiss / static in his audio, to the point where I would be uncomfortable listening to his voice for a long period of time at a medium volume. He's using the $400 Shure mic in that video too.

The $100 for the dedicated 2nd pre-amp with the Triton seems pretty wasteful. As an alternative, for $299 you can pick up the DBX 286s pre-amp[0] which also acts as a noise gate (helps filter hiss / static), de-esser, compressor and basic EQ together. It's an awesome piece of hardware and offers way more bang for your buck.

Also the AT2005[1] mic is a solid upgrade from the AT2100 that he recommends, and both are about the same price. IMO the AT2005 is worth using as an end-game microphone. The AT2005 is one of those crazy mics where it almost feels like it's mis-priced at $80 and the DBX 286 has enough knobs and buttons to make it sound good with a number of different types of voices.

Ultimately your real natural voice is going to play one of the biggest roles in how you sound in the end.

[0]: https://amzn.to/2VPbmMc

[1]: https://amzn.to/39hyQwI

(Yes these are affiliate links, I've been using both pieces of hardware for years)

11 comments

A noise gate does not remove hiss or static, it simply reduces the volume of the signal when it gets below a certain level.

As a professional audio engineer in a past life, I can say the SM7B + Triton Fethead is a great recommendation. The SM7B has very good off-axis rejection characteristics and will pick up _much_ less room noise than most alternatives.

> A noise gate does not remove hiss or static, it simply reduces the volume of the signal when it gets below a certain level.

But the end result is with a configured noise gate the audio coming out will not appear to have as much hiss or room noise.

Check out the Youtube video I linked of his. It's super noticeable. With Sony MDR-V6 headphones, I hear a huge amount of hiss in his audio. To the point where it's distracting and drowns out his voice.

I'm not trying to pick on you or him, but if I go to your most recent Youtube video at https://www.youtube.com/watch?v=7lf6dwipRYg, you can very much hear the same type of hiss within the first few seconds. I can't say for what what it is, but it sounds like your computer fan is spinning at 50% and it very much comes through. At about the 16:22 mark in the video, your fans appear to be spinning at 100% because it's much louder than before you started coding. It sounds like your computer is about to launch into space.

It looks like you have the same mic too.

I'm not saying the mic is bad, it's a really good mic. I just think the OP should have talked more about one of the biggest things that will kill a recording, especially when talking about "professional podcasting" and recommending high end gear.

For comparison, here's something I recorded with the AT2005 + DBX 286s acting as a noise gate: https://www.youtube.com/watch?v=-Xd84hlIjkI

There's no software processing here either and my room is not treated with acoustic panels. In fact, my office is almost a worst case scenario. It's a wide open empty room with hardwood floors, angled ceilings and almost no furniture or rugs. I have super loud computer fans too and I'm right next to a window.

Even at maximum volume (unrealistically loud) there's pretty much no hiss or room noise. At least nothing I can hear with the same Sony MDR-V6 headphones.

I'm sure I could get similar results with the Shure mic too (using the DBX), but I'm happy with the AT2005. If I ever upgrade mics in the future it would be to a shotgun mic so it's out of frame.

What you're hearing in that video is a space heater running in my office — I don't put a ton of effort into making everything perfect for ad hoc live streams.

A proper screencast is probably the better thing to judge:

https://www.youtube.com/watch?v=Ff_n_QClipQ

I'm sure you'll find something to nitpick over but I think it's pretty good for a programming tutorial recorded in my home office.

> I'm sure you'll find something to nitpick over but I think it's pretty good for a programming tutorial recorded in my home office.

It's not nit picking.

The video you linked sounds much much better and really really good. Besides turning off your heater did you do any other processing with software afterwards? Such as using your video editor's noise cancellation filters or a proper DAW?

I'm just asking because taking a quick skim through your videos, your live stream videos all have some form of hiss but your shorter non-live screencasts do not which makes me think additional post-processing is being done on them.

But I think the OP's channel intro video and your live stream video is a good example of the importance of managing background noise, which isn't mentioned in the article at all but it's probably the most important thing to do and you can't just blindly throw money at the problem to fix it with no prior knowledge.

For a professional set up, it seems like a no brainer to want to filter out background and room noise before it even hits your computer. That's some of what the DBX does. This way the amount of effort it takes to do a live stream or screencast is the same when it comes to audio quality. You just turn it on and hit record, and you're done. It also works the same in OBS, your video recording tool of choice, zoom, hangouts, skype or any program that records your mic.

You mentioned being an audio engineer. Surely you know how important having a good audio source is, and how time consuming the editing process is when you need to fiddle with cleaning things up after the fact. Configuring all of this stuff to happen live is something I did after recording about 20 videos. Now 400+ videos later I couldn't imagine having to do post-processing on videos just to clean up audio.

I listened to the first clip expecting it to be much worse than it ended up being. I'm listening on HD280pros, the headphones I keep at work, although I use Sony MDR-7506s when I'm in front of my audio gear.

The noise floor sounded more to me like bad gain staging than it sounds like fans and room noise. And I honestly think that the majority of people who watch that video aren't going to notice. I think you're too close to the process. I would venture that almost no one else in this thread would watch that video and call that an all-caps HUGE amount of noise.

A long time ago, I was mixing down some rough recordings of my band, and as I was putting everything together, I kept hearing the squeak of my cheap kick pedal every time I hit the bass drum. It was incessant, and it didn't help that I typically played at upwards of 160bpm. I tried EQ, notch filtering, all kinds of things available to me back in 2002 to try and get rid of that squeak without destroying the other things coming through the room mic, and I just couldn't do it, so I had to leave it in. No one else in the band could hear what I was talking about.

On my way home, I put in a CD by the band Cake that I'd had for a couple of years at this point, and I had heard probably a hundred times before. And even though it had never stood out to me at all before, all of the sudden the only thing I could hear on this professionally recorded album was THE SQUEAK OF THE KICK PEDAL. It was absolutely maddening.

I went on an electronic music binge for a couple of months after that, just to cleanse my palette.

> I would venture that almost no one else in this thread would watch that video and call that an all-caps HUGE amount of noise.

I listened to the video and while I wouldn’t go all-caps, it’s way more noise that I like to listen to. So I downloaded the audio track and measured the RMS noise and signal levels, getting about a 30dB SNR. That’s a somewhat bad number, which matches my subjective experience.

A few years ago I converted some old audio cassettes with voice recordings of deceased family members to MP3. I heard all sorts of weird noises, but I spent a while getting rid of them with EQ and notch filters. I don’t know if anyone else in the family cared about the work I did cleaning up the audio, but I did.

Everyone has different standards, and it’s not quite fair to say that nobody in the audience cares about these things, just like it’s not quite fair to say that everyone in the audience cares.

You’ve probably tried it but OBS has a reasonable software noise cancellation filter you can apply to audio sources.
A more directional and less sensitive mic like the NTG-2 may work better with some dampening on the walls or by adding a carpet. Additionally if you can't fix the room you can use a plugin to help reduce noise in dialog, something like Izotope RX7 will do a good job of doing this, but fixing the room should always be the first step.

I've used both the SM7B and the NTG-2 and I would record vocals all day with the SM7B if the room was in a good condition to do so or I was recording actual singing vocals. Otherwise I'd just use the NTG-2 which with it's highly directional nature is pretty good at removing a lot of other background noise.

Does the NTG-2 work well in a desk setup with minimizing typing? I’m looking to find a better quality mic (blue yeti now) that does better at minimizing keystroke sound.
The problem with the NTG-2 is it sounds reasonable enough when you have nothing being compared to it, but when you put it side by side with a decent dynamic microphone where you can listen to each one in an A / B test then it falls apart due to sounding super muddy and thin. It's something you can pick out in a blind test 100 out of 100 times if you have decent headphones and there's nothing you can do in post-processing to fix it.

Shotgun mics with a very good dynamic range and a pleasant tone tend to be pretty expensive. Usually in the $1,400+ range. Compared to something like the AT2005 which is a dynamic mic and is $80, but now it sits in front of your face.

This video has a pretty decent comparison of a high quality dynamic microphone to a $300 shotgun mic: https://www.youtube.com/watch?v=LiYEX-omlFk. It has both talking and typing tests. It's not the NTG-2 shotgun mic tho, it's the Deity-S. But it does serve as a decent base line comparison of how much worse a shotgun mic sounds unless you go up to the very high end.

> But the end result is with a configured noise gate the audio coming out will not appear to have as much hiss or room noise.

> Check out the Youtube video I linked of his. It's super noticeable. With Sony MDR-V6 headphones, I hear a huge amount of hiss in his audio. To the point where it's distracting and drowns out his voice.

If the noise floor is high enough that it interferes with the actual signal, a gate is just going to make it more annoying to listen to (as you hear the noise cut in and out).

> a gate is just going to make it more annoying to listen to (as you hear the noise cut in and out)

my experience as well. its best to just use an EQ to dampen all high frequencies over a threshold so that its not distracting.

It would like this in the EQ: https://dt7v1i9vyp3mf.cloudfront.net/styles/news_large/s3/im...

I was a recording engineer at a TV studio before I got into programming. When we used gates it was very shallow, long compression so you couldn't notice abrupt changes.

What USB interface do you use with the DBX 286?
I have alot of experience with live audio, and from my understanding I can't see what the point of the Triton Fethead is? Is that used in addition to the pre-amp in your interface? If so why? Does the SM7B have super low output? It's not used much live, but when we did, we never had trouble with level, and I'm pretty sure an inline pre-amp would have added more noise.
Thank you for providing non-affiliate links for those of us who don't want someone else making money off our unrelated purchases in the next 24 hours.
I know, the idea that somebody could link to something useful and then be compensated for the value of their recommendation is so awful. It should be illegal for people to make money for doing things online. It should all be free. If we could get farmers and ranchers in on it, then municipalities and auto makers- heck- lets get the whole government on it- everything will be free and everyone will work for the common good. What could possibly go wrong!
Had you actually read my comment, you'd know I'm not complaining about this recommendation. I'm complaining about Amazon's affiliate structure specifically. If I buy a hammer on Amazon 23 hours after absent-mindedly clicking the GP's link, they get a cut of that. A cut that they don't deserve.
Why would you care? You would prefer that Amazon gets to keep an even higher percentage of the profits?
I would prefer HN be one of the few places online where I can trust links because of the products behind them, not because of someone trying to drive affiliate revenue.
I did throw a disclaimer in there that I've used these products in a professional environment for years.

I didn't wake up that day hoping to make affiliate sales off a product. I happened to have the links handy because they are listed on my site in a "tools I use" blog post[0], and honestly getting the links from there was faster than getting the non-affiliate links.

[0]: https://nickjanetakis.com/blog/the-tools-i-use

I'm going to disagree on a lot of this.

The Triton fethead is just a "plug it in and go" solution in a lot of setups. That dbx preamp has a lot of knobs that, for the most part, need to be set right in order for it to work at all - let alone help the situation.

Also worth noting that the fethead is tiny - it sits inline with your cable, whereas that dbx requires you to mount 19" rack gear at your desk, which is a big ask for a lot of people.

> The Triton fethead is just a "plug it in and go" solution in a lot of setups. That dbx preamp has a lot of knobs that, for the most part, need to be set right in order for it to work at all - let alone help the situation.

But that's the thing. The Triton is just a pre-amp, the Scarlett also has its own pre-amp, but maybe it's not strong enough to drive the Shure mic (it's strong enough for most other mics).

The DBX also has a pre-amp but it does many other things to improve the quality of your audio where as the Triton does nothing except boost the signal. I would classify the DBX as very much helping the situation. Way more than just having a pre-amp.

> whereas that dbx requires you to mount 19" rack gear at your desk, which is a big ask for a lot of people.

I don't think having a piece of gear on your desk is a big ask for a personal recording studio. It's not like you're sometimes streaming or recording your podcast at Starbucks. You also don't need to mount it. It sits flush on a desk without straining any of the parts or sitting uneven.

It's been happily sitting on my desk for a long time now.

> But that's the thing. The Triton is just a pre-amp, the Scarlett also has its own pre-amp, but maybe it's not strong enough to drive the Shure mic (it's strong enough for most other mics).

An awkward reality is that while prosumer preamps like those in the Scarlett are very nice for what they are, you often find yourself needing more clean gain than they comfortably provide. The Scarlett Solo’s gain range is listed as 56dB, and while that seems like a lot, and it seems like it should be good enough, for voice applications at normal speaking volumes and with typical microphones, you want more gain. From what I understand, this is more or less the amount of gain you can get in a single gain stage without introducing tons of dirt.

I personally have a Scarlett and a small assortment of mics, including dynamics, small and large diaphragm condensers, and ribbon mics. It’s only for louder sources like drums and guitar amps that I feel comfortable plugging straight into the Scarlett. For everything else, I plug into an outboard preamp.

While the DBX is nice, everything but gain can be done in software later on. I’d go ahead and recommend the DBX to most people because it solves the problem “once and for all” rather than forcing you to configure every different piece of audio software you use, but for my own personal use I’m much happier doing all EQ, compression, gating, de-essing, etc. in software.

That stuff isn’t necessary in hardware, but you really do need the gain.

> While the DBX is nice, everything but gain can be done in software later on.

Totally, but...

Doing it in software can be a pain because ideally you just want to flip your recorder on, talk and be done with it. Especially when your day to day involves recording many videos.

I know OBS has VST support but if you ever record outside of OBS everything sounds different.

Or if you want system wide software processing, it gets really complex with audio redirects / patching where it becomes a ceremony just to begin recording, and you need to worry about xruns, pops and other weird abnormalities unless the software you use is 100% amazing.

The only software I ever found to work well for that was in Windows using ASIO Link Pro to patch the real-time output of REAPER as microphone input to any app, but the author literally died and his license key server went offline (I bought it like 4-5 years ago). That is what eventually lead me to use the DBX. Now I just turn it on and never think about it, because it works the same across all apps with no ceremony, and will work on Linux too (or any OS).

The title is "Remote Work, Podcast..." not focused on creating a personal recording studio. It's not really appropriate to have a rackmount pre-amp to just make you sound slightly better on a conference call.
The Scarlett absolutely is not enough pre-amp, on it's own, for that microphone.
> As an alternative, for $299 you can pick up the DBX 286s pre-amp[0] which also acts as a noise gate (helps filter hiss / static), de-esser, compressor and basic EQ together.

Do you know of any device with similar capabilities but in smaller form factor (even if more expensive)? Something of the size of Focusrite Scarlett 4i4 [0]. Or at least not meant to be mounted in a rack?

[0]: https://focusrite.com/en/usb-audio-interface/scarlett/scarle...

The 4i4 is just like the 2i2 (which is what I use). It basically acts as a way for you to plug in an XLR mic (or instrument) into your computer. It's classified as a USB audio interface.

It won't help you when it comes to background noise cancellation or any type of processing. For that you need something like the DBX or you can find other hardware that combos as a USB audio interface + pre-amp + processor (with comparable options to the DBX). I don't know of any offhand but maybe something exists.

Also, you can go the software route too to avoid needing any hardware (and just use a USB mic), but it gets a little more complicated. OBS supports VST plugins which is straight forward enough, but if you plan to record in other programs you won't have those effects applied.

Or you can delve into the wonderful world of software based audio redirecting where you can redirect the output of a DAW like REAPER into another app as input to capture and process audio in real time. I did this for years until the software I used stopped working due to the author dying and his license server went offline. Alternative software doesn't exist on Windows with the requirements I had so I decided to go with hardware.

The cool thing about the AT2005 mic I recommended from earlier is it has a USB and XLR connector, so you can try both out.

I know what both of those are, I was asking if there was anything with functionality of DBX but with a form-factor of Scarlett 4i4. It doesn't even have to have an integrated USB audio interface, although that would be cool.

Thanks anyway for the recommendation of DBX.

Oh sorry, I missed the part about being "the size of the 4i4". I thought you were thinking about using the 4i4 as a replacement to the DBX.

The DBX doesn't need to be mounted to a rack btw. It sits clean on any flat surface like your desk. The only thing that's not optimal is it doesn't have a power switch on it and you should keep it off when not using it.

Personally I have mine hooked up to a surge protector with an on / off switch, so I don't put any stress on the plug. You could also use a remote control surge protector so you don't need to fiddle with manually turning it on / off like a barbarian.

Don’t know of any small units that also include a noise gate but otherwise something like a http://www.joemeek.com/threeq.html could do the job. A compact channel strip. Plenty of these units and their predecessors on the used market too.

Running your signal through analog EQ and compression is super useful, because they’re intuitive to use (just twist knobs) and once you’ve got your settings dialed in, it will survive any OS update or kernel panic :) And you’d need some dynamics processing digital or analog if you want to approximate the kind of voice people associate with radio or voice over.

I think a noise gate is tricky though, if set up wrong it can make the audio sound choppy. If at all feasible, reduce the noise at the source. Especially if you use a compressor, because as great as it can sound, it will also amplify the noise. You might notice this less if the compressor has a very fast release time (i.e. its amplification drops out immediately after the audio drops below the level where it’s supposed to kick in) but also that risks making the audio sound choppy and unnatural.

In that linked YouTube video, I didn't hear hiss or find anything objectionable in the sound. But I found it really distracting the way the microphone was so big in the picture. It even looked like it could be easily adjusted so the mic was in the same place but the stand was more out of the way. But if it was me, I'd go for a smaller / less obtrusive one even if the sound wasn't quite as good.

Funny how different people are sensitive to different things.

Headphones play a big role here.

If I plug in my budget earbuds I can't notice anything but with my studio headphones (which I normally use for everything) it's super distracting.

Studio headphones ftw!

But yeah it's true, you hear more 'imperfections' when using higher-end audio gear. Wouldn't go back though.

I had the same experience way back, dithering on old CRT with composite videos is smoother since you create new colors given how the signal is transfered and converted compared to a high resolution computer display (800x600). Now days I'm pretty sure there are GPU filters for that.

Can this also be fixed by filters?

Ahh. Using speakers.
I'm somewhat guilty of it as well, but a lot of people geek out on the gear side and spend money to optimize a fairly small part of the overall audio experience.

Most of what I record for podcasts these days is on location anyway and I know I don't personally have a "radio voice." So I do what I can to get reasonable audio quality but don't sweat the details too much.

> I'm somewhat guilty of it as well, but a lot of people geek out on the gear side and spend money to optimize a fairly small part of the overall audio experience.

There's nothing wrong with buying gear.

It's just, he lists that set up as a "professional" podcast set up but forgot to mention the most important piece of hardware (or software) to get rid of continuous background noise.

If you're going to listen to someone talk for an hour, having a bunch of white noise hiss in the background is very tiring on your ears.

If you read that article without years of prior audio experience you might just blindly click those links and buy the top end set up, and then get upset that you just dropped $700 on a set up that sounds worse than what you could have gotten for $80 (+ free software) or half that in pure hardware but spending your budget on different hardware.

> home office style recording environments (even with acoustic panels) are going to have a noticeable amount of background hiss (computer fans, etc.) get picked up and his set up doesn't account for that.

You can definitely baffle a lot/all of that noise away, but you definitely need to be more mindful of your setup and not just have some wall-mounted panels and your tower right next to your mic.

I've been privy to some crafty home studio setups. It's workable—the first problem is paying attention/having an ear and the example you posted is so obvious I was surprised at it.

> the example you posted is so obvious I was surprised at it.

Yeah I was surprised too. Considering that clip is from the author of the post.

> you definitely need to be more mindful of your setup and not just have some wall-mounted panels and your tower right next to your mic.

Mine is on the floor, under my desk, which has a large thick piece of wood (the desktop) as a barrier and it still comes through without a noise gate. My mic is on a boom arm like 5 feet above my tower (I use a standing desk) and the mic is positioned opposite of my computer to further reduce it picking it up.

Also, a lot of people use laptops as their main computer which is even worse. Chances are you'll end up with your laptop on the top of your desk instead of below it.

That's why I was so taken back by him creating this article without mentioning noise cancellation.

Fortunately I do have the DBX and the noise floor is almost silent without clamping down too hard on natural frequencies. I've recorded about 400 videos, some with the DXB and some without and there's such a huge difference.

A single wooden panel won't do much to help you in that situation regardless. It's a great reflector and wouldn't absorb much of the noise, especially of the sides are wide open.

Most people, especially someone just recording dev screencasts or video conferencing don't at all need a $300-350 preamp. It just seems like so much overkill. You could get perfectly usable noise reduction baked into Audacity, or with $50 plugin if the noise is existing.

FWIW, I record (music) with a laptop on my desk, sometimes multi channel simultaneously and with the right mics[1] and sufficient ambient noise reduction I've never had an issue with fan hiss or hum.

[0] https://www.waves.com/plugins/z-noise

[1] https://www.shure.eu/musicians/discover/educational/polar-pa...

> Most people, especially someone just recording dev screencasts or video conferencing don't at all need a $300-350 preamp. It just seems like so much overkill. You could get perfectly usable noise reduction baked into Audacity, or with $50 plugin if the noise is existing.

I do record dev screencasts. Using Audacity is ok but it's extremely tedious if you record a lot of videos.

That means for every video you create you need to split and export your audio, import it into Audacity, get a noise sample, filter it out, export the new audio, import it back into your video editor and then edit as planned.

That work flow will drain your soul if you're trying to record a 150 video course, or you put out new videos every day.

Some video editors like Camtasia, Screenflow and Resolve have decent enough noise cancellation filters where you can press 1 button and wait 10 seconds to avoid having to do that with Audacity, but it also means if you ever do live streaming you'll need to set up a different type of VST with OBS, otherwise your live streams will sound bad.

But you're right, for starting out I wouldn't bother with any of that. That's why I think the AT2005 is a great mic for $80 bucks. It works over USB on its own but it also has XLR support so you can "grow" into it if you decide to use hardware to save time later. The DBX is a lot more than a $300 pre-amp btw.

> Ultimately your real natural voice is going to play one of the biggest roles in how you sound in the end.

Therein lies the main problem for me personally.

Is there a way to train your voice to become more comfortable for others to listen to?

If you're serious about it, I'd look for a voice coach. It's likely easier to find someone who trains singers than who focuses only on speaking. That is fine. Vocal technique is largely the same.

Additionally, I'd look into something like Toastmasters. This will give you a great place to practice outside of formal lessons with a voice coach.

Being comfortable to listen to is about both your vocal technique - the mechanics of how you form sound - and about your content and delivery.

> Is there a way to train your voice to become more comfortable for others to listen to?

This is something I struggle with too honestly.

I've recorded 400+ videos and I still think my voice is -ultra- cringe and I feel like it gets worn out after talking for only maybe an hour straight.

So what I usually do is record for about 45 minutes, then edit for a few hours and then go back to recording when my voice is fresh. I drink plenty of water while editing too.

I don't have any professional advice since I'm just a dude who records screencasts. But I think my biggest problem related to getting worn out is I tend to talk slightly louder than normal when I'm recording because I still internally link loudness with tone variance and "energy". So maybe talking less loud, raising the gain on my mic and learning how to have more tone variance without talking louder would be something to look into. If I were to guess this is probably a common problem for many folks.

I use the Shure SM7B + DBX 286s combo and it's _sublime_.
If you don’t have a really good quality preamp in your interface, the fethead makes making up the gain really well that the sm7b needs. Sooo much gain.
Partly as a reminder to self: how hard could it be to build a noise cancelling/phased array setup for streaming?
If you're only concerned about OBS, it's trivial. You can use a VST (software plugin) without needing hardware or anything fancy.