I searched but did not find much on my question: I know about audacity and kdenlive, etc, and have seen comments on microphones and hardware, and podcast apps for phones and listening, but I have not seen much on podcasting “procedures / methods / sofware” used to MAKE the podcasts.
I can tell that you guys on the Destination Linux shows are probably on the cutting edge of putting together podcasts, and I would like to try to tap some of your knowledge about procedures. Obviously the technical quality is important, but what I am really talking about here are tools that help organize and pull together the “content.”
Have you guys ever posted an article or thread on how you assemble your weekly podcasts? I know that Mike Tunnell is doing most of the heavy editing, but to what extent are you guys scripting each episode ahead of time? Do you use particular software to develop and agenda and then type it up to use to prepare the show? I might have heard comments to the effect that you use a shared editing tool that is NOT google docs. Do you have any recommendations on that?
Also, as to actually putting together the final product (not the live stream, but the edited version) do you use something more sophisticated that Kdenlive and/or Audacity? I can see that it would be extremely useful to have some kind of program where you could organize audio/video clips, label them, move them around, and then assemble them into a final timeline, but I don’t see Audacity or Kdenlive as particularly friendly for doing that final assembly. Do you have any recommendations on that?
Obviously I (and I bet many others) are thinking about doing my own podcasting in different areas, and if you guys have any suggestions on the best software / procedures /methods to use I would be extremely appreciative.
Podcasting is relatively easier to get into (you just need a good mic) but harder to get a passable quality. You just have audio and its easy to nitpick on it.
Maybe your recording room is noisy and is capturing the entire environmental sounds: chatter in the next room, loud vehicles passing by, neighbors dog is barking. In this case you need heavy soundproofing in the form of heavy curtains and maybe you can do foams and egg trays to do a hacky sound proofing.
Mic quality is also important. I guess the cheapest basic equiptment you need may be a Blue Yeti (or Yeti Blue?). Audio equiptment is a deep deep rabbithole in which you can spend ungodly amounts of money and still not get it perfectly.
Audio cleanup and processing is also done. Joe Resington of JB/Late Night Linux advocates that you do certain adjustments like cut off certain frequencies so that your voice is bearable listening into. I think he had a conference talk somewhere in the internet that discusses exactly how to do this.
You may also need to get something creative commons or public domain music to do your intro. All podcasts has a bit of a music at the start. Make sure to source it properly, otherwise you will get litigated to hell.
Finally testing: While having studio monitor audio device is nice, do check the listenability with various devices like different cellphones, headphones, speakers, etc. Get someone to critique the technical quality of your audio.
Oh and of course content. You better have good writing skills to get a good topic to release weekly, and time to edit the episodes, of course. There is a reason why sometimes Micheal releases late: you really have to pay attention and edit things.
But wait there’s more: Your personal vocal quality is also scrutinized. Dont talk fast like Noah. Have a slightly lower pitch than Michael. (Love you guys lol). Talk like Zeb, in my opinion, deliberate and with class. That takes practice if you are not a Brit.
Thanks for the comments:
I remember an episode of ask Noah when Noah intentionally spoke much more slowly and methodically.
I am SO glad he want back to his normal machine-gun delivery!!!
As to the suggestions, yes I know that audio quality is very important. My question however is really more about organization techniques for pulling together content from a variety of people, such as collaborative software to assist in scripting, executing, and then editing the final product from various audio chunks.
Any ideas on that part?
@Compass on mintCast we generally record with the mic open but not talking for between 20-30 seconds of background audio. This will give you a base line for using the noise reduction tool in Audacity and this can help a lot. Then when you are editing your audio later you highlight this section of the track that you recorded background noise, select noise reduction in the effects menu, use the ‘Get Noise profile’ then highlight your whole track and apply noise reduction using 12db, 6 sensitivity and 3 for frequency smoothing. Obviously omitting as much background noise as possible during recording is a must but despite this there will always be background noise so this is a great step as part of the audio editing process.
I’m currently not aware of any group edit ability of audio files.
For online script/content collaboration. You can go with the usual suspects Google and Office 365 or you can go open source via Nextcloud/Collabora or Cryptpad.
I wouldn’t consider myself an expert on any of this. For DLN Xtend, @EricAdams and I record in Audacity, we use PulseEffects for gating and compression, though I only gate right now and use the noise reduction function in Audacity to further remove the noise in the recording. I have a decent USB microphone on a boom that does isolate some of the noise. We each record our own audio so that there aren’t any “Internet Artifacts”. The editing of that podcast is done in Audacity where each of the track recordings are edited in parallel so we can get rid of any talking over one another and remove a lot of the excessive “ums” and pauses.
For my own separate, solo podcast, I record in Audacity, edit it down and import it into Kdenlive as it works as a non-destructive audio clip editor too. I can assemble and line up the music clips to the speaking and what not.
That is my two cents.
I was a radio DJ for about 6 years and probably 50% if my shows were pre-recorded. Keep in mind that recording and editing will take at least twice as long as the actual content you produce and that’s on top of whatever prep you do before hand. Personally I use Reaper. It is the most flexible Digital Audio Workstation that I know of. You can create your own scripts in 3 different languages to extend the functionality and there are already hundreds of functions you can string together into custom actions that take a single click to activate. I used it for live/studio music recording for 10 years so feel free to ask me anything.
@CubicleNate - So you start the audio editing in Audacity and finish it in Kdenlive, even though it is just audio, or are you using Kdenlive to add video or graphics? I have listened to mp3s of your podcasts but not tried to watch them. Does kdenlive make it easier to rearrange the audio?
I record and do most of the editing of my podcast in Audacity. I do just the Audio editing in a non-destructive manner for the final assembly of it in Kdenlive. I like using a music bed, force of habit, beneath my talking. The way I can overlay things in Kdenlive makes it a LOT easier than with Audacity.
Here’s the process for DLN Xtend
- We have a show prep document that we collaborate on to capture the content for the show and add notes to. It is not a script, just sections and bullet points for each section of the show.
- We communicate via voice on Mumble. I also record this audio just in case something goes wrong with one of our own recordings. You have the option to record each participant as a single file or to downmix it into a single recording. These files are medium-quality WAV files.
- Each of us records our own audio on our end. We happen to like and use Audacity but anything will work. These recording are exported as high-quality flac files.
- We sync our recordings using Syncthing so each of us has the source files.
- My editing process is all done in Audacity and it is fantastic. Once you learn the keyboard shortcuts you can become pretty efficient at editing and the toolset is very powerful.
Audacity editing process
- I import both source tracks and then will usually run some filters to get the volume and waveforms to match as closely as I can.
- I align the tracks so they are in sync.
- I do a first-pass edit to remove any large bits of silence (Truncate Silence is AWESOME) or any background noise (Noise Reduction). We are also open mic the entire time so there tends to be a lot of extra conversation that isn’t part of the segments so that gets removed.
- Now I do a more careful edit to remove weaker or redundant parts of the conversation. As we don’t script anything, some of our thoughts tend to evolve as we converse and in the end I try to get to strongest parts of what we discussed. I am also trying hard to keep the length to 30-45 minutes and sometimes it requires some pretty tough decisions on what stays and what gets cut. I will also cut entire segments if necessary (it usually is) to keep to that time limit. Some of these get posted to our YouTube channel as Xtras.
- Now I add in the segment labels and music track, marking the sections so I have names and timestamps for chapters in the podcast and adding the intro, outro and bumper music where appropriate.
- I listen to it again, sometimes just checking the edit points and sometimes listening to the entire thing. If I’m happy, I export as a flac file.
Post Production and Publishing
- I create a show notes document that has the title, short description, full description, keywords, show links, chapters and timestamps.
- I use Auphonic for post production which levels the audio and does some other audio magic which generally makes it sound very nice. I upload the flac file from Audacity so the source file is of high quality. Auphonic also lets me embed the show metadata like name, description, logo and so on. I get a nice 112 kbps mp3 file as the end result.
- I create a new episode on Fireside and fill in all the details, upload the mp3 file, create the chapters, add the links and so on.
- Once it has been published, I take to social media and post links on forums, chat rooms, Twitter, Mastodon and so on.
I am just going to go out on a limb and say that this might be overkill for a lot of people but I am trying to put out the best, most professional podcast possible, learning as I go. I will also say that if this were a video podcast, all of this would be about 1000x more difficult. Michael has spent years honing his process so he doesn’t spend all week editing a single show. We are only 10 episodes in and right now it’s taking between 15 to 30 hours to produce each episode (when I do it at least). Maybe we find ways to make this easier and faster but for now, it takes the time it takes to get the result we want. Not easy but we are happy with the results and feel like it’s getting better with each show. Hopefully other agree!
THANK YOU Eric Adams! (and of course prior responses) but THANKS for that detail and for the graphic! That is exactly what I was wondering about. Apparently I have a lot more learning to in Audacity because one of the major issues I had was visualizing the segments, and I see your third track is some kind of labeling feature that I have not yet discovered how to do. I did not know Audacity could do that so I will have to go looking what that is called. Duh I need to tell myself to RTFM: https://manual.audacityteam.org/man/creating_and_selecting_labels.html
That response is just filled with tips that I needed. Thank you again!
Also: You guys are using a service (Fireside?) to host your podcasts? I was thinking that perhaps the DLN team was self-hosting … and of course that was another question, if you set up a server such as on DigitalOcean, what software do you use to host and serve the podcast?
Although this might be a topic for a different thread, as I really want to focus on methodology, but I am certainly interested in suggestions for podcast hosting options like Fireside (which may be the most professional?) vs Anchor (which is free?) and anywhere in between. Like I said I was thinking that self-hosting would be the way to go but you guys are the experts and if you have investigated and you don’t self-host, then that largely answers my question.
Not mentioned in the title of this post is podcasting HARDWARE. One of my favourite hardware companies, Audient, just announced something new that I thought I should share with everyone here.
There’s not a lot of information available yet since these are not even released yet but it is iOS compatible, which leads me to believe it is USB class compliant, which means plug and play Linux operation. It is marketed toward podcasters/artists of all skill levels so it’s a device that could grow with you.
Anyways, in my opinion, a lot of people have expectations that exceed what the Blue Yeti mic can do and I see that mic recommended a lot as a beginner tool to get started. I think it’s quickly outgrown. A device like the Evo, matched with a half decent microphone is going to serve you better in the long run.
I have high hopes for these units, especially the Evo 8. So long as the headphone outputs are not mirrors of the 4 main outputs we could have a very compact and inexpensive surround sound setup on our hands.
Huh, editing the audio in Kdenlive…
Now I’m wondering if I could use this method to create a “pseudo-video” version of my show. Perhaps just with images/graphics and a few relevant video clips of what I’m talking about.
Puts on thinking cap
Fair point. I struggled with hardware for a long time. USB audio is okay I guess but I could never get it to be as good as I was looking for.
I agree 100% that people have unrealistic expectations of mics like the Yeti. Condenser microphones can be tough to deal with if you don’t have a quiet environment and can’t position it properly. I use a $50 Samson Q2U USB/XLR Dynamic Microphone which I find to be much better than the condenser mic I had been using. The really nice part, apart from being inexpensive, is that it is dual-purpose with USB and XLR connections. I have it on a $20 arm so I can position it where it needs to be.
The USB connection is fine but I struggled with pulse audio to get the sound I wanted which led me to trying the XLR interface. I initially tried a USB interface like the EVO but wasn’t blown away by it. I tried a cheapo mixer that ended up arriving broken so I returned it and went with the Behringer Xenyx 302USB. It’s a nice little mixer. Now I can mix the mix levels on the analog side and send that to my PC the way I want it.
There are a ton of great inexpensive options like the ones you describe. The trouble with Linux is that no manufacturers, except for RME, make drivers for Linux and the drivers have a software interface for controlling elements of the hardware. In many cases the functionality is seriously reduced on Linux for this reason. Your option works well because you have all the control over the hardware independent of software. It’s a more complicated experience for beginners to choose hardware and definitely leads to a lot of disappointment for new Linux users.
That would be cool. I was going to do some research prior to replying but decided I would live in the dream-space between my ears about how fantastic that would be.