Something else that confounds me ...

I’ve been noticing this and have done some google searches for information but can’t find answers. I’m probably not typing in the query correctly. But something I’ve noticed that I can’t figure out is this: when I listen to a playback of my recordings, I’ll notice that at the beginning and end of each track, there’s like a … sort of … fade in and out at the beginning and end. Almost like the microphone doesn’t start recording sounds until I speak - and then it’s like a quiet build up, or rush of sound just before I’m starting to speak. Like the volume is being turned up just in time to capture my voice. I’m not using any fade in or out feature, and my Audacity settings are’t set up for sound activation recording. It’s the weirdest thing. And quite annoying because I can’t figure out why it’s happening or how to stop it. You’d think when I press record, it would start actually, you know, recording the “silent” room noise at least. I don’t even know if I’m describing this correctly. I’ll try to upload a sample when I’m able, if I’m not explaining things in the right way.

It just occurred to me though … maybe that sound is the result of using compress dynamics and normalization. It’s fixable with noise reduction and deleting certain areas, but it’s still frustrating. Ugh. I have so much to learn.

What’s your operating system?
Which version of Audacity? (look in “Help > About Audacity”)

I’m guessing that you are on Windows and Windows Sound Enhancements are messing up your recording. See: http://manual.audacityteam.org/o/man/faq_recording_troubleshooting.html#enhancements

For future reference, please refer to the instructions in the pink box near the top of the page.

You’re good.

Are you on Windows? Do you have any of the Windows built-in processing running? Audacity doesn’t apply any filters or tools during recording (or at least none that can affect voice quality), so the first place to look when people post of problems happening “by themselves” is the Windows filters.

http://manual.audacityteam.org/o/man/faq_recording_troubleshooting.html#enhancements

Another more uncomfortable possibility is your computer is “helping you” listen. Does it sound that way when you listen to the same show on a different computer or sound device? I once got stuck with a computer that always sounded a little funny no matter what we did. That turned “Cathedral Effects” was running on top of each and every sound playback.

Chris’s Compressor can do something like that under certain conditions. You can get around that by compressing a longer show, say before you cut it to the right length. Compress the show with “heads” and “tails”. Alternately, paste heads and tails on the show, compress it and then remove them.

I’m not a Windows elf, so if it’s one of those problems, we should wait for one of them.

Koz

You can get some of these answers by listening to the show before you do anything to it. Does it sound normally live with no oddball fades or volume changes ? If it does, then it’s something you’re doing.

The gold standard is make a “flat” recording, change the volume to meet ACX conformance and out the door. You only need compressors and advanced filtering if there’s some shortcoming with the original shoot. I think Dana_Tucker has said something similar several times.

This was a raw sound test I shot in my quiet third bedroom in the evening when traffic quieted down. I’m down the block from a police station and hospital, so recording this way is an adventure.

http://kozco.com/tech/audacity/clips/20151105MicTest-ES58-Mixer.wav

If you analyze it with ACX-Check, it should pass. No filters. This is right out of the sound mixer into Audacity. Change the volume to meet ACX peaks.

The studio is not exciting. It’s the Mac version of this:

Change the computer and eliminate that little silver box in the middle.

Koz

STOP RECORDING INTO YOUR PC! It sound like you have a “Gate” mixed into your chain according to your description. I have no idea. Skip the “compress dynamics and normalization”. Start clean and finish lightly! If you set your input levels correctly, you are 99.9% finished, if you have followed the other rules for recording clean audio. Please give a sample when you can.

What? :confused:
Isn’t that one of the purposes of Audacity?

ABSOLUTELY NOT!!! The purpose of Audacity is to mix, master and render! Only a newbie would ever consider recording into a Pc or Mac, that is unless you have a $20,000 system to do so. Most of us do not.

At the moment I have nothing else to record into but my PC. When I do the initial record, the volume is so low, that’s why I have to use compress and then normalize (this brings it up to ACX specifications - but this is probably where the background noise of the room stillness gets amplified … or whatever it’s called … the fffff). Maybe if I put the mic inside my mouth then it would be loud enough without having to process it. :stuck_out_tongue: Heading downstairs to play around with this some more.

Not sure I understand this. What else would you use to run Audacity on?

It is not a matter of what system you run on, it matters on what you record with. I was in a 4.8 mil dollor studio last week. They have a $8,000 mac with $14,000 Avid Daws installed that ran a $90,000 control board. It is 100% different to what 99% of other personal recording studios do, be it at Home or at a low cost start up recording studio. Bottom line, 99% of people recording quality audio, in a home environment, DO NOT record into their machine. They record into a external digital recorder that can produce 24 bit, 44.1 k in a wav format. PERIOD! (In the Podcasting World at least)

I currently run Audacity on a Linux PC and a Mac Mini. Previously I ran Audacity on Windows PCs.

So the issue did appear to be the processing - the compress dynamics to be specific. Now I’m just trying to figure out how to consistently pass ACX without my voice sounding like a loud boom right up inside your ear. I still pretty much always have to run Normalization at -3.2 (sometimes I try Amplify at -3.2 but that doesn’t always give me a pass on the check). When I turn the mic sideways pointing at my mouth, my voice sounds … I dunno, not clear … kinda muddy. I don’t know if that’s the right way to describe it. I pass ACX but I don’t like how it sounds. Obviously I get the clearest sound when it’s positioned right in front of me, but if I’m too close, my voice is overpowering … kinda boomy; again I pass ACX, but once more I don’t like how it sounds in my ears. So I’m trying to find that sweet spot in front of the mic that will give, as I said, a consistent pass rate on the ACX check and sound OK on the ears at the same time.

This sounds like good progress :slight_smile: A short audio sample in WAV format (just a few seconds) would certainly help us to hear where you are at with this, and hopefully help us to ,make some useful suggestions. There’s step-by-step instructions for posting an audio sample to the forum here: https://forum.audacityteam.org/t/how-to-post-an-audio-sample/29851/1
What we need is an absolutely “raw” recording - no processing whatsoever - straight off the mic.
Also, if you could give a bit of background information about your set-up: What sort of mic, what sound card (model numbers are good(, brief description of your recording environment.

I just did some more playing around with things such as mic position, my own sitting position, etc, and I think I might have found the sweet spot. I do still have to run normalize, but that just helps boost my volume to pass ACX and doesn’t appear to mess with the quality. My voice is clear and loud enough, but not overpowering. I use a bit of noise reduction (12,6,6) and so far that setting doesn’t appear to be causing unpleasant artifact sounds (of course, a longer track might tell me a different story; right now I’m just doing tests that last about 15 seconds). I’ll be recording later tonight - going to start off fresh, now that I think I’ve nailed down my equipment squabbles - and I’ll upload an unprocessed sample then.

Record a sound clip and post it. We can tell a lot from that.

http://www.kozco.com/tech/audacity/TestClip/Record_A_Clip.html

The quick version is on top, a longer explanation follows with side trips that have more detail. Do Not Process it before you post it. Announce it, cut it if needed, export and post.

I use Normalize because I can tell it the volume to achieve. I use it like this and I always use it the same way.

– Select the whole clip or show by clicking just above MUTE.
– Effect > Normalize: [X]Remove DC, [X]Normalize to -3.2 > OK

12, 6, 6 is not light Noise Removal. It’s medium. 6, 6, 6 is delicate, light removal.

What else would you use to run Audacity on?

I’m with him. A majority of the forum postings are people having trouble recording in Audacity. Record to Something Else, transfer the file(s) and edit and post in Audacity. I have a Zoom H4 and a Olympus WS-823. They will both produce good quality, reasonable volume, 44100, 16-bit, Stereo voice recordings with no computer errors. I don’t think we’re all that alone, either.

Koz

Thanks for the clarification. I used to use 24 for the top number, so 12 seemed light in comparison. I tried the Beast numbers, then 12,6,3, but liked 12,6,6 better.

I’ll post an unprocessed sample after recording tonight.

I’m with him. A majority of the forum postings are people having trouble recording in Audacity. Record to Something Else, transfer the file(s) and edit and post in Audacity. I have a Zoom H4 and a Olympus WS-823. They will both produce good quality, reasonable volume, 44100, 16-bit, Stereo voice recordings with no computer errors. I don’t think we’re all that alone, either.

This is an intriguing option to be sure, and I may try that in the future. Can they produce 24-bit? I saw someone in a video recommending using 24-bit for original recordings … I think it had something to do with producing a high-quality WAV file if a client ever needed/wanted that instead of an MP3

24 for the top number

[Gasp] I need to sit down for a second. I’m surprised there was anything left after that.

24-bit for original recordings

No. 44100, 16-bit, Stereo is Music CD quality. 24-bit is what they use to record Adele. I’m not Adele and you’re probably not either. Let’s stick with plain, ordinary, normal sound recording. That’s hard enough.

Koz

Didn’t get nearly the amount done tonight that I wanted to do - primarily because while I thought I’d nailed down a solution, the results from earlier this afternoon weren’t consistently repeatable when I returned to record for the project. At any rate, here’s a small unprocessed sample of part of what I did tonight.

I know this problem. I’m wondering if the type of mic also plays a part, was I’m sure it does in my case. My setup is this:

Mac mini 6.2 - 2012, Intel Core i7,2.3 GHz, Memory:16 GB. El-Capitan 10.11. 2TB Storage made up of -
Drive 0:HGST HTS721010A9E630. Upper bay. Drive 1:ST1000LM024 HN-M101MBB. Lower Bay
Audacity 2.1.1 - Focusrite SOLO Mini Preamp - Shure55SH Series II (Dynamic) - BTSKY™BM-800 (Condenser)

I have found that the Shure55SH records a more pleasant sound for my voice. Richer tones, a good spread of frequency. I have to be pretty close to it to speak, and get sufficient base volume into the recording, and this results in “mouth noises” when I swallow etc. Which I can edit out easily enough. The BTSKY™BM-800 (Condenser) mic, while I can adjust the gain of this mic via the SOLO Mini, and sit further back from it, doesn’t give such a rich tone. It’s quite … hmmm. bright? might be the word. If I’m being impolite, it sounds a bit ‘tinny’, higher pitched. Makes me sound like a teen ager - which isn’t a bad thing.:slight_smile: It’s not unpleasant, but I don’t think I could listen to it for hours on end…

So what I’m saying is, have you tried other mics? The Shure55SH I have is really good quality, Dynamic. The BTSKY Condenser cost £20 - about US$40 - It looks nice, but… One day I’ll buy a good quality Condenser mic and check the difference.

So I have tracks that are running about 20 to 30 minutes each at the moment. Voice only. So far I only have to run Normalise, usually around -3.2 to get the ACX-check pass. I don’t use noise reduction, as I’ve spent weeks and weeks test my recording environment, and trying to achieve optimal - minimal - sound floor readings. When I’m good to go, it’s usually at around -75dB to -60dB
My recording voice shows peaks when I’m speaking up to about -6dB, but usually more like -11db. Normalise then brings that up a bit.

interesting stuff.