Thanks so much Rainer. That’s why I have been making these videos. It may be a small number of people who have the question, but if one person asks a question it’s likely many others have wondered as well. So this is a way to address those questions and we can all learn together.
Digitally boosting a signal, like what you were doing with the razer mic, will not "degrade the signal to noise ratio" on its own. You may however end up clipping. What does actually worsen the signal to noise ratio is compressing the signal, which is typically followed by digitally boosting the signal - this is probably what you were thinking when you thought boosting would cause degradation. As for the degradation caused by digitally lowering the gain while recording, and then bringing it back up in post... Like you said, I think it may have something to do with the fact that it is recorded in 16-bit. p/s: I am assuming that the mic "gain" or "level" control in both windows and mac is essentially a digital fade/boost function, which can very well not be the case for certain USB mics with specific drivers (such as those that allow you to control the mic gain using a software).
Wanted to say the same things! My guess is that lowering the MacOS volume essentially means you're squeezing the signal's dynamic range within fewer bits, that's why stretching it back out made a lower-bit-sound. If these mics could be recorded in 32 bits, there would be no such issues, I would assume
It wasn’t meant as a “Call Out” or anything. It was meant as a “This comment made me second guess myself, so I tested it out only to realize I was correct, and I assume other people will have the questions so lets make a video explaining it and demoing it.” … Okay maybe there was a little bit of a call out because I could have done it without showing the comment, but that comment made my heart stop by making me think that I had completely f***ed up my video and tests and that I was putting out incorrect and misleading information. So if there’s a bit of callout, I’m okay with that.
@@Podcastage2 😂 I feel you dude. It really didn’t feel like a call out, I was just messing around. It was very respectful. I personally liked that you took the time to answer a comment this way and explain it. It was cool.
The theory says that in order to obtain the maximum signal-to-noise ratio, it is better to increase the gain of the first stage. So setting the mic gain to 100% is a great idea, as long as you don't overdrive subsequent stages.
The noise is generated in the preamp and the gain at 100% in software is 1:1 input. Any reduction in the software gain only reduces dynamic range / maximum loudness / SNR of the input, it normalizes the audio to a lower level. For example, he had to clip the preamp to hit -9dB in this very video when gain was reduced in software. Typically preamp SNR for USB microphones or USB powered audio interfaces is always significantly better at lower gain, that is the true first stage and the knob is on the physical microphone or preamp / interface.
@@BeepBeep2_ @kote315 If I right-click in the Windows settings and switch to dB there. The whole thing should then be at 0 dB. 100% was a full 30 db gain on my Rode. So thats not right. You have to look it up. Sometimes 100% is 0 db. sometimes 50% is 0 dB.
That explains why mine has this slightly distorted sound, just discovered that windows has this "percentage" and "decibel" view for the sound properties.
So yes. This is basic gain staging. There is a pre built into the mic. The noise floor is found here in the ADC. The computer end is attenuation only unless the software is directly controlling the head amp in the mic. Its not going through another audio (analog) gain stage in the computer. At that point it strictly digital and will just bring up the existing noise floor but should not add to it in any way. Love how internet heroes that have no experience are gonna make statements like that and make themselves look like asses. Great job in debunking it. The razor is a PoS for the record lol.
Thank you so much for the comment. That makes a lot of sense. I must have been thinking of pushing up the signal into a 0dB brick wall limiter. At that point the dynamic range would be compressed leading to the signal to noise ratio degrading. Also you’re exactly right about the computer level setting (unless there’s software/drivers that control the mics gain). This was a huge realization for me because I had called it gain in the past and someone called me out when I was discussing USB Soundcards. Those have a fixed input level, no amplification, and they just convert. I realized very recently that that would mean that USB mics without gain dials just have a set level, and you’re essentially only adjusting a fader in the computers audio settings. The razer isn’t great, but it’s like $30-$40 now and if I had to pick between the Seiren Mini or the Blue Snowball, I’d go Razer every time.
Thanks, really interesting test! What is the best way to connect a usb microphone to the camera's connection for audio in? Preferably with the possibility of monitoring maintained?
I've got problems with low mic volume. My sennheiser profile usb mic has the gain knob turned all the way up, the level in Windows 11 is 100% BUT even if I yell into the microphone the levels only go up to 20%. It was the same with my basic microphone. I'm new to high quality audio recordings
I have a cheap USB fifine 670, it has a distinct hissing when I turn its volume knob down, so I have to lower the volume on my pc. Fifine, what can I say
On my end I did hear the noise floor on all quarter percents you had tested and noticed the clipping on the mic gain, lower software volume/gain. To me the noise floor got worse at 75% and 4 o'clock gain than it was at 100/2 o'clock. 16.5% I heard it. 44% it improved. 72% is hardly noticeable. 100% I believe almost nothing to nothing I can notice. If this helps you in any way. I am currently using a K7, AutoEQ/Peace with FHS5 pro using FH5S EQ from AutoEQ (The Harman Target curve) For what it's worth I think you both are neither right or wrong. It's a mix of settings used and the setup another person is using and of course hearing. Volume at the "OS level" makes sense before adding addition gain on the mic. I'm no audio engineer so my two cents is probably worth dirt.
When you're setting your microphones gain you should not be boosting or attenuating the signal further down the line because that will skew everything. If you're attenuating, then you'll increase your gain and clip. If you're boosting, then you'll set your gain too low and your sound source will be recorded closer to the noise floor of your converters. So when you're setting your gain always make sure that your faders (which in this case is windows level) is set to 0%. Unless I"m completely mistaken about everything I believe (which has happened far too often than I care to admit), then this should also offer the optimal noise performance.
@Podcastage2 that does make sense. I've probably confused the verbiage or really overthinking it. I'm by no means an expert in any way. You have more experience than I. I don't have my mic at 100% in Windows. I believe it's at 75%. My gain at 12:00. My only recent experience with this is on Discord. A friend said I was a little quiet. I increased the gain, some of which they said introduced noise to which i put it back. I then increased it some in windows, which helped. Then, there is the possibility of discord settings messing with it. My mic is a Q2U with a UMC22. Also my understanding with audio (listening) is to have windows or os at 100% and then increase your dial to where the audio level is comfortable. Which I do have windows at 100% and my k7 at roughly 11 o'clock. Maybe it's analogous in some way. I suppose the audio converting is reversed then
I think this video may have been misunderstood. I'm not saying that every USB device should have the level set at 100%. I was simply demonstrating that setting the level for this mic at 100% was not making the noise floor worse. I have a video that I need to make about determining what UNITY LEVEL is, where you're not boosting or attenuating the signal. If your device has a gain control (i.e. mixer), that is where you should have your device set in my opinion.
It makes no sense that setting it at 100% in Windows will make for more noise as that's all digital, and what you're doing is is adjusting whatever has already been converted to digital, so the noise is embedded already. Keep all the volumes for things in your OS at 100% (unless you need them to be at different levels to each other) and adjust the volume with an analogue volume control that is controlling whatever you are listening to.
Try this with a yeti nano please. Mine FREAKS out if it's at 100%. Has to be around 60% so it doesn't sound like the microphone is in a fan. That's on windows at least.
That is COMPLETELY different. The Yeti Nano does not have a gain dial on it. All you do is set your level in your audio settings so you’re hitting the appropriate level. You should only set the yeti nano’s level at 100% if you need the level that loud (which it sounds like you don’t).
I can't use my Samson Q2U without having the level on my pc set at 100%. I'd have to increase the gain too much in OBS to be heard and that actually ruins my audio.
It was meant to be informative, so I’m thrilled you found it informative too. I learned something while testing stuff and I tried to include that in the video.
@@Podcastage2 pardon my dry sense of sarcasm. I know it doesn't translate, I don't know why I do it. I've been watching you for years, I thought i was subscribed but when I just watch this last video I found out that I was not, so I did so. I think your channel is invaluable I'm a 64 year old lifelong musician and I've learned so much from you. Thank you for your effort you are appreciated.
Hello. I am sorry for my English. This is an automatic translation. I have a Blue Yeti USB and a Macbook. As I understand it, the algorithm is like this: In MacOS I set the microphone input level to 100%, and the Gain on the Blue Yeti body is at a minimum? If the signal is weak - I increase the Gain on the Blue Yeti body? If the signal is strong - I reduce the scale in MacOS, so that the voice is -12 to -6 decibels. In reviews, bloggers recommend Gain on the Blue Yeti body at about 9:00 hours, and in the system MacOS 75%. But this is somehow not logical. Please tell me what is the right thing to do? Step-by-step. Thank you.
