I’m brand new not just to Audacity, but to any kind of recording & manipulation.
My wife used a Blue Yeti, through a usb hub and was about 12-18 inches from the microphone. The only editing I have done so far is to do a noise profile of a silent portion, noise reduction with that, and then silenced the non-speaking portions. All other settings are default.
Do we have a goal? Different goals have different production requirements.
The voice has a little “recording in the kitchen” sound. Was this a home office with bare walls? Even though the Yeti is a directional microphone, it’s not perfect and you will pick up reflections and echoes from the walls, ceiling, and floor if you have no carpet. I had a joke I could tell you the size of your recording space by analyzing the echoes. It’s permanent. There is no filter for that.
This is why some people outfit a closet for recording. I published a “Kitchen Table Sound Studio” for casual home recording.
Volume is good, if slightly low. Blue wave tips and peaks should reach up a bit more to around half-way (0.5). It’s OK as it is. The two extremes to avoid are having any part of the wave go all the way up—that’s permanent distortion, and having the waves too low. Then you struggle with background noise even more than you did.
Any reason you submitted in stereo—two blue waves? You can do that, but it doubles storage and transmission time for no good reason. The audiobook people recommend all voice work in mono—one wave. There are some odd exceptions. I used to submit stereo voice work to the video editors because I knew it was going to be folded into a stereo show. If I submitted in mono, I would be the odd duck. Still workable, just a bit odd. A hit to the professional reputation, etc, etc, etc.
You submitted in MP3. Any specific reason you did that? Never do production in MP3. Sometimes you have to deliver in MP3, but that’s up to the client. MP3 is something of a time bomb. MP3 gets its small, convenient files by carefully adding distortion. If the client needs to make their MP3 from your MP3, the distortion won’t hide any more. There was a forum poster who did a music review at home using download MP3s. He delivered to the station in MP3 and they couldn’t make the station podcast. Couldn’t do it. The music turned to trash.
Ask what the client wants or deliver multiple formats. In your studio, you should save raw performances and deliverable backups in WAV. You can make those into anything else. You can’t fix MP3, and you can’t make the distortion go away by folding a WAV in the middle. Once you make an MP3, you’re stuck.
Do you have a theater model? Anybody whose voice you like? I wish I had saved it, but there was a video of a woman in a very well padded home studio. She was the polished, professional “Please stay on the line. Your call is important to us” voice.
Your voice sounds like she’s carefully reading from a script instead of having an intimate, serious conversation with me. The pauses are a little off. Also avoid the bored instructor in front of a blackboard voice.
You posted in the Audiobook Production forum. Are you planning on publishing with ACX? We have tools, formats, and hints to help with that specific job.
Thanks for the great feedback!
Not only is this our first recording project on the computer, it’s my wife’s first voiceover. You’ve already told me enough to make huge improvements I’m sure. Here are answers for your questions:
• It is for an online video course that will be built with Camtasia.
• I built a PVC frame that goes around the desk, and hung shirts on hangers around it. A friend told me you don’t want it too closed off, but after looking at your kitchn link/post about a booth, I’ll be closing it off more, probably won’t even use the desk or have the computer in there.
• I gave no thought to stereo/mono. Should online course be mono?
• MP3 only for file size; sounds like I should use WAV for the video course?
• It’s for us, not a client.
• Yes, she was reading and I was glad to get that feedback. It’s my words and her voice. We’re going to do it now in shorter segments so she can read and digest my words and then tell it in her words in a more conversational manner. We actually recorded 20 minutes worth; I only saved the first minute as MP3 for posting on this forum. (Our internet sucks, so small file size is a good thing!)
• Sorry if I posted in the wrong forum; I am brand new to this forum and am not familiar with the protocol…seemed like the best fit at first glance.
Thanks again for the valuable feedback!
probably won’t even use the desk or have the computer in there
Home performers always fail noise. It’s practically a law. I know people that are trying to make echoey, kitcheny, bathroomy voice “normal,” but until they do, a quiet, clear voice is good.
The moving pad trick is interesting because while it’s only fair at rejecting household noises, it’s terrific at rejecting echoes because the performer’s voice has to go through the pads twice, losing effect each pass.
Anybody can put echo and reverb in later. You can’t take it out.
Don’t forget the blanket on the desk. Sound reflections from a desk can cause odd tonal distortions.
Should online course be mono?
That’s my opinion. Do you gain anything by having some sound from the left and some from the right? If not, mono is good. A Mono production automatically plays from both left and right speakers.
I should use WAV for the video course?
Use WAV for everything until you post for use. That’s when you make the MP3. MP3 is a delivery or listening medium. Not steps in the middle. There is no ‘open an MP3 and make it into something else.’ Another similar thought, if you have to make a correction to the voice or information, do it to the WAV edit master and then make a new MP3. DO NOT patch the MP3 and repost it.
If this is a video project, then the video format will take the place of the MP3. MP4? The format generators have their own methods of compressing and processing the sound, but it’s always good to have the absolute best quality, uncompressed sound to start.
Sorry if I posted in the wrong forum
If you post in the wrong forum, we’ll move it. If that’s the worst thing that happens all week, I’m good. You are going to run into all the same problems that the audiobook people have. So it’s good where it is. There is a fuzzy rule that if you can pass audiobook standards, you can post anywhere else successfully. They’re pretty picky.
There are some interesting philosophies about screwing up. If you make a mistake, some readers like to use the Punch and Roll tools to correct it immediately. Stop, and you can tell audacity to back up from the mistake and play some of the good work jumping into record at the place you choose, usually just before the mistake. It takes some practice and setup but that was a welcome tool development.
Some readers don’t stop. They make a sound effect so they can find the error later, (clicker toy). Take a breath and re-read the whole last sentence through the correction, and then keep right on through the rest of the chapter. That one, of course, needs post production editing to remove the trash, but that also gives you a chance to carefully match the metre, rhythm and mark.
What you don’t want to do is make a mistake and try to go back next week and announce a correction then. That one is doomed.
Other oddities. Do Not Change Anything once you start reading until you finish a job. The forum is full of people who decided to do an Audacity update and the microphone dropped dead or the machine stopped recording.
I wouldn’t do any computer updates, either, but that can be controversial. There was a Mac “update” a bit ago that killed Audacity. That got everyone’s attention.
You may find as many have, that post production editing takes at least five times the length of the show. Plan on that.
I mastered the last post for audiobook submission and it failed for an odd reason. It’s too good.
Where did you get the silence to put between the words or sentences? In general you shouldn’t use Generate > Silence. That produces the Blackness of Space silence wherever it’s used. Also, what were the Noise Reduction settings you used?
If the noise tests turn up unnaturally quiet background sound, the submission could get rejected for overprocessing. However, you are the publisher, so if you like the way it sounds, that’s what counts. I’m going to slowly back away from this because the voice seems clear with no processing distortion. Who is “The Producer?” upper case intentional. That’s the person with final say. Talk to them.
As this is the new version of ACX-Check, I’ll clarify a detail in the test results.
Each of the three measurements produces a numeric result - the “dB” measurement.
The numeric result is followed by one of: “Pass, Fail, Warning”.
“Pass” means that it meets the specification published by ACX.
“Fail” means that it does not meet the specification published by ACX.
“Warning” is when ACX don’t publish a figure, but there is probably a problem.
“Fail” and “Warning” are usually followed by a hint about the problem.
Example: ACX specify a maximum peak level of -3.0 dB, but they don’t specify a minimum value for the peak level. However, if the peak level is below -6 dB, then ACX-Check will show a “warning” because there is probably a problem. It would normally be expected that the peak level will be close to -3 dB.
There are irrational answers, too. Peaks can’t be quieter or the same value as RMS. Noise can’t be louder than RMS.
A proper reading depends on you mastering or applying some corrections to your work. Nobody can reliably read directly into ACX. You have to do something in the middle. You may hit one performance by accident, but that’s no way to read all the chapters in a book.
But that’s assuming audiobook standards.
Did you settle on a volume to produce your work? That’s another interesting discussion. There are two competing standards: the audiobook/ACX one and LUFS
You didn’t think this would be easy, did you?
What did your Producer say about the background sound? Go with it?
Thanks again…I’m on the run today; will digest the rest of the feedback later today. But…it’s only for an online video course to be used with a combination of slides, panned & zoomed images, and some stock video. Oh, and I’m the owner/Producer/writer, etc. It’s all for our business consulting practice and a line of courses we’re setting out to do.
So…perhaps it’s ok for it to sound a little like a lecture rather than a conversation?
Also, if the bottom line is for understandability and also for the viewer to just not be turned off by the production - and not for a 300-page book - do you think it’s near acceptable?
You all are so generous with your feedback. Thanks!
Do you have one hat with all the labels on it, or do you change hats each time? Are they color coordinated? “I can’t talk to you right now. I’m wearing the wrong hat.”
it’s ok for it to sound a little like a lecture rather than a conversation?
It’s OK for it to sound however you like it to sound. Much of that is my opinion. Our job is to make the performance meet whatever the requirements are. In audiobooks it’s relatively easy. They wrote everything down. You’re something of a blank slate, but there are still theater considerations.
Do you have instructional videos you like? That’s another way to decide the sound and production quality. I’m surprised nobody ever says they want their show to sound just like someone else’s. It’s nice to sound fresh, new, and exciting, but it’s also a good way to make a lot of mistakes.
I really like the “All The Stations” videos.
Hollywood has a thing called a Log Line. In one brief sentence, describe the show. I think they used to type it on the top of your file folder or something. All The Stations log line is:
“They get on a train and then they get off.”
A more dull and boring idea for a show you never heard, but it’s fascinating because of how they flesh it out—and they involve the audience…and they have guests.
So your job is to spice up Microsoft Spreadsheet Manager Lectures without being hokey or silly.
Do you have your T-shirts and coffee mugs designed?
But on a more serious note, what’s the production company’s name? Pick a good one.
“Boy, I really liked the instruction videos ahowes produced.” Flat. No pun or second meaning. Howe-To-Video? Intentional misspelling. That might be good.
Enjoying the thoughts
Yes, we are “spice it up” type folks…Right now, we’re just trying to get something out there as a MVP minimum viable product. My wife is a cut-up, so perhaps version 2 will incorporate more and be more riveting.
I knew for a fact though, that really bad sound would turn off people and we wouldn’t even have an MVP…that’s why I reached out on the forum initially. I beefed up with heavy blankets around the pvc frame, added carpet on top of the desk, and even got rid of the second monitor. I think we’re well on our way to an MVP now.
I’m anxious to get to version 2, though!
added carpet on top of the desk
Terrific idea. I’ve done that. It helps suppress that "talking into a wine glass’ voice. If the microphone is sitting on it, it also helps suppress floor and table rumble and vibration.
I’m anxious to get to version 2, though!
Post a sample when you get that far.
There are fuzzy rules for graphics, too. Simple cuts or fades are good. I had to talk down one of the technical supervisors giving a company talk. No, you don’t have to use every special effect in the arsenal. Nobody will be impressed by your 3D-Page Turn Transitions.
Keep them simple. Your graphics should not need a table of contents to figure out where everything is. Progressive reveals are good. Each heading fades in as it’s announced. Tuning the graphics to the voice is a lot easier than the other way to.
No coffee mugs?
Yep…I gave thousands of presentations in the corporate world and I kept my presentations bare-bones. If they’re being dazzled by graphics or having to read too much on the screen, they’re not paying attention to your voice anyway. But…I love the feedback.
I will have my wife Cri-cut a coffee mug for you!
BTW, since she has quite a sense of humor, we worked out what I think is a good solution to the lecture-y voice and the bit of sounds-as-if-it’s-being-read-from-a-script feeling. We’ll have the slightly boring, but stuffed with good information, corporate-ish video modules, but then at the end of the modules, have a 200-level set of modules that mirror the stuffier ones. We’ll call it Kelley’s Cliff Notes. The videos will be less than half as long, get the same point across but with fewer words and less detail, and allow her humor and bloopers inline. I’m sure she’ll have 1-1/2 glasses of wine first to loosen her up. It should be entertaining.
With the content we’re delivering, the more irreverent executives and managers will be the ones that thrive on the modules anyway. They say that you should niche down to make a good living, so my niche will be innovative and irreverent executives looking for a slick and easy way to gain an hour’s productivity per employee each day.
I kept my presentations bare-bones.
You probably don’t need us any more. I’m going to go make coffee.
But what are you going to pour your coffee into if you don’t have the mug yet?
Anyway…here’s another try at the wav. ~15 seconds or so. Any suggestions on tweeks to help the general sound? I have zero knowledge about sound; can you tell me what her voice recording needs…more/less bass…whatever…
This works out better if you actually include the sound file.
if you don’t have the mug yet
I’m using the mug somebody gave me when I got them out of trouble. That one was a little more involved. Someone supplied them with a European video to include in a US show. We had a video converter.
Well dang it. I suppose it does. I haven’t done that since leaving the corporate world 15 months ago
Do I detect some southern now that the voice is a little more relaxed?
I still got a lot of room resonance. Is this after all the blankets, PVC Pipes, and carpeting? I would expect not to hear the room at all.
Also slightly low volume. Blue waves on the raw recording should hit somewhere between 1/4 and 1/2. That’s -12dB to -6dB on the bouncing sound meter.
Are you using the Yeti right? Somebody got tired of people recording the wrong way and designed a graphic. I don’t think this appears in the formal instructions. I’ve seen people announcing into the rounded top. That’s not the sensitive part of the mic, and worse, it can point the mic to the room instead of the performer.
I don’t know we ever covered microphone spacing. Do you use a pop and blast shield? Then you can be as close as one power fist.
Doesn’t have to be straight in front, either. Slightly off can work. That can also make reading the script easier.
This illustration is of an end-fire microphone which is not yours. It’s just to illustrate position and spacing.