Starting anywhere else would be dishonest. So this piece starts by conceding the ground that has genuinely moved, and then sets out the five things that are decided while the recorder is running and cannot be revisited afterwards. Those five are what actually changes when you record professionally. The room is only one of them, and it is not the most important.
Post production now fixes problems that used to require a treated room
Adobe's Enhance Speech v2 states that it removes noise and echo from voice recordings, and lists reverb, chatter and background music among the things it takes out. It is free to use, handles files up to 1 GB, and allows up to four hours of enhancement a day. Whatever a studio blog published in 2023, that capability is real and most readers have already tried it.
Adobe's own guidance goes further than most studios would like. Its explainer says that you do not have to rent a recording studio just to get perfect audio and that you can fix it in post. Any comparison that ignores that sentence is arguing with a reader who knows better.
Adobe's own caution marks where the tool stops helping
The same page carries a warning that is easy to skim past. It notes that keeping the enhancement at full strength can make a recording sound unnaturally perfect, which is why the slider exists at all. That is a vendor telling you its processing is a reconstruction of your voice rather than a recovery of it, and that pushing it hard is audible.
So the accurate position is narrow rather than absolute. Cleanup handles what sits underneath and around a voice. It does not handle decisions that were baked into the recording itself, and those are a different category of problem.
Five recording decisions cannot be reversed after the session ends
Each of these is settled at the moment of capture. No processing, paid or free, reopens them.
Separate channels are a capture decision rather than an editing option
Two voices recorded onto one track are one signal. Nothing downstream separates them again, which means one person's cough, one person's level and one person's plosive are now everyone's. Recording each microphone to its own channel is what makes a conversation editable at all: levels balanced independently, a guest's phone buzz removed without touching the host, an interruption tidied on one track only.
This is the difference people notice in the finished episode and almost never attribute correctly. They hear a show that sounds controlled and assume it was a better room. It was usually a better channel count.
Camera coverage is fixed at the moment of recording
A single fixed camera yields a single angle for the length of the episode. There is no second angle to cut to, no reaction shot, and nothing to hide a jump cut behind, so an edit either lives with a visible cut or stretches an unbroken take. A second camera creates the option, and the option has to exist while the recording happens.
This one has become expensive to get wrong. Video is now where podcast discovery happens, and a back catalogue recorded on one webcam cannot be cut into clips that hold attention on a phone screen.
Lighting is recorded into the image and cannot be added later
Colour can be graded and exposure can be nudged. The direction light came from cannot. A face lit by a window behind it stays lit by a window behind it, and grading a dark subject lifts the noise with it. Lighting is the most visible difference between a home recording and a studio one, and it is the least discussed in comparisons like this.
A host who runs the room is not hosting the conversation
This is the difference that never appears in a specification list. Watching levels, checking that both cameras are still rolling and worrying about a red light all draw attention away from the person opposite, and it shows in the conversation. An operator in the room converts that attention back into the interview.
It also removes the failure that ends more sessions than any technical fault: discovering afterwards that something did not record. When a person whose only job is the capture chain is watching it, that class of error largely disappears.
Setup and troubleshooting time is spent, not recoverable
A home session includes getting the gear out, positioning it, setting levels, discovering a cable fault, and putting it all away. None of that is in the episode, and none of it comes back. The practical consequence is not cost, it is frequency. Sessions that take a whole evening happen less often, and inconsistent publishing is the most common reason a show quietly stops.
Broadcast standards put numbers on what a treated room actually means
Professional rooms are not a matter of taste, and the reference figures are published. EBU Tech 3276, the European Broadcasting Union specification for critical listening conditions, sets a nominal reverberation time between 0.2 and 0.4 seconds, scaled by room volume, and requires continuous background noise preferably below NR 10 and never above NR 15.
| Specification in EBU Tech 3276 | Published requirement | What it means in practice |
|---|---|---|
| Nominal reverberation time | Between 0.2 and 0.4 seconds | Speech stops almost the instant the speaker does. A domestic room with hard walls runs longer, which is what an ear hears as roominess |
| Reverberation scaling | Tm equals 0.25 multiplied by the cube root of volume divided by 100 cubic metres | Bigger rooms are allowed slightly longer decay, so the target moves with the space rather than being a single number |
| Continuous background noise | Preferably below NR 10, absolute maximum NR 15 | No audible air conditioning, fridge, traffic or computer fan under the recording |
| Noise character | No impulsive, cyclical or tonal content | A rattling vent or a ticking clock fails the requirement even at a low level |
| Room volume | Floor area at least 40 square metres, volume no more than 300 cubic metres | A purpose sized space, which is the part a spare bedroom cannot be argued into |
Read from the EBU Tech 3276 specification on 17 September 2026. The document covers listening conditions for the assessment of sound programme material, so it describes the reference environment broadcasters work to rather than a legal requirement for a podcast.
An isolation figure in decibels is a measurable claim, not an adjective
A room described as treated to minus 40 dB is making a specific claim. Because decibels are logarithmic, 40 dB of attenuation means outside sound arrives with one ten thousandth of its original power, which is a hundredfold reduction in sound pressure. A truck on the street outside becomes a level that sits under the noise floor rather than under the dialogue. Konnect Studios publishes minus 40 dB across its recording rooms, and a published number is something you can hold a studio to in a way that a word like soundproofed is not.
Delivery targets differ by platform and one of you has to hit them
Whatever the room, an episode is judged against published numbers when it reaches the apps. Two of the three largest platforms set different targets, and articles on this topic routinely quote one figure as though it covered both.
| Platform | Loudness target | True peak | Other stated requirements |
|---|---|---|---|
| Apple Podcasts | Around minus 16 dB LKFS, with a tolerance of plus or minus 1 dB | Must not exceed minus 1 dB FS | WAV or FLAC at 44.1 kHz minimum, 16 or 24 bit with 24 bit recommended. MP3 mono 96 to 128 kbps recommended, stereo 128 to 256 kbps |
| Spotify | Normalised to minus 14 dB LUFS, measured to the ITU 1770 standard | Below minus 1 dB true peak, and below minus 2 dB for masters louder than the target | Louder masters are turned down on playback rather than rejected |
Apple figures read from the Apple Podcasts audio requirements page and Spotify figures from the Spotify loudness normalisation article, both on 17 September 2026.
The two targets are 2 dB apart, which is why a single master cannot be optimal on both and why most producers deliver to the Apple figure and let Spotify normalise. The point for this comparison is simpler. Somebody has to measure loudness and true peak and correct them, every episode, forever. At home that person is you. In a studio session that includes editing, it is part of the deliverable.
What actually changes, line by line
Collecting the differences in one place, with the fixable ones marked as fixable.
| What it is | Home studio | Podcast studio | Fixable afterwards |
|---|---|---|---|
| Background noise and hum | Whatever the room and the street provide | Treated room with a published isolation figure | Largely yes, with AI cleanup |
| Room reverb | Depends entirely on the room | Designed to a short decay | Partly, and processing is audible when pushed |
| Channel separation | One track unless the interface supports more | Four microphones each on their own channel | No |
| Camera angles | Usually one, fixed | Two cameras, framed and colour matched | No |
| Lighting | Ambient or a single key light | Full rig, set before the session | No |
| Switching between angles | Cut manually in the edit, if a second angle exists | Cut live to a director's feed during the session | Only if the angles were recorded |
| Someone running the capture | The host, while hosting | A dedicated operator | No |
| Setup and pack down | Every session, unpaid | Done before you arrive | No |
| Loudness and true peak | Your job, every episode | Your job, or part of an edited session | Yes, and it must be done |
| Set and backdrop | One look, whatever the room offers | Two sets inside one booking | No |
A home studio is the better answer in four situations
This is not a close call in every case, and pretending otherwise would undermine everything above.
- You record solo. Most of the five irreversible decisions concern more than one person in a room. A single speaker with one microphone and a good AI cleanup pass gets very close to a studio result on audio alone.
- You record frequently and at unpredictable times. A room you can walk into at 11pm beats a booking, and consistency beats production value in the first year.
- Your show is audio only and will stay audio only. Four of the differences in the table above concern video. Remove video and the gap narrows sharply.
- You already own the gear and know how to use it. The five decisions are only advantages if someone is exercising them. A confident operator recording themselves is exercising most of them at home.
If any of those describe you, the useful next question is cost per finished episode rather than cost per hour, and what podcast studio hire actually costs in Sydney works that calculation through. How to start a podcast in Australia prices the equipment path in Australian dollars if you are leaning towards building your own.
Konnect Studios is built around the five decisions rather than around the room
Every studio has a treated room. The five irreversible decisions are what separate a room from a session, and they are the design of this one. Konnect Studios sits in the Parramatta CBD, six minutes on foot from Parramatta station and the light rail, and teams record with four separate microphone channels in a Parramatta studio from $174 an hour on the Capture session.
| The irreversible decision | How a session at Konnect Studios settles it |
|---|---|
| Separate channels | Up to four Shure SM7dB microphones, each recorded on its own channel, so every voice stays independently editable after the session |
| Camera coverage | Two Sony FX30 cameras recording 4K, framed and colour matched before the session starts, so a second angle always exists |
| Lighting | A full lighting rig, set before you arrive. Low key and cinematic on The Vault, warm and natural on The Luxe |
| Someone running the capture | A dedicated studio operator sets the room up before you arrive and manages the technology for the whole session, so the host only has to host |
| Setup time | The room is built and tested. Sessions book in 1, 2, 4 and 8 hour blocks, and a full day batches several episodes without a single pack down |
| The room itself | Acoustic treatment to minus 40 dB isolation, which is a published figure rather than a description |
| Angle switching and the edit | Blackmagic ATEM live switching cuts angles during the recording, so the edit begins from a director's feed. Craft and Conquer sessions add the full episode edit, five reels and one teaser |
Session details, equipment and rates read from the Konnect Studios podcast studio hire page on 17 September 2026. Up to four guests, Monday to Saturday 7am to 9pm, after hours by arrangement.
The Shure SM7dB is worth one line of its own, because it is the microphone doing the work in most professional spoken word rooms. Shure publishes a 50 to 20,000 Hz frequency response, a cardioid pattern chosen for rear rejection, and a switchable preamp offering plus 18 dB or plus 28 dB of gain. Rear rejection is the specification that matters in a room with four people in it, because it is what keeps each microphone hearing its own speaker.
Three questions settle the decision in one sitting
- How many people are in the room? One person makes this close. Two or more makes channel separation the deciding factor, and channel separation is not something you can add later.
- Does the show need video? If yes, four of the ten differences apply immediately and none of them are fixable in post. If no, the gap narrows to the room and the operator.
- What stops you publishing? If the honest answer is the time around each recording rather than the recording itself, a booked session that starts when you walk in is solving the real problem.
If the answer points to a studio, how to choose a podcast studio in Sydney sets out the three fields that decide a booking and are missing from almost every listing.
Questions people ask when comparing a home setup with a studio
Is a podcast studio really better than a home studio?
On audio alone the gap has narrowed, because free AI cleanup now removes noise, hum and echo, and Adobe states plainly that you do not need a studio to get clean audio. The gap that remains is in the decisions made while recording: separate channels for each voice, a second camera angle, lighting, an operator watching the capture, and the setup time that never appears in the episode. None of those can be added afterwards.
What can you not fix in post production?
Five things. Two voices recorded onto one track cannot be separated. A single camera angle cannot become two. The direction light fell cannot be changed. A host distracted by running the equipment cannot be made attentive after the fact. And the hours spent setting up and packing down are spent. Noise, hum, echo and loudness are all fixable, which is why the argument should not be about them.
What loudness should a podcast be?
It depends on the platform, and the two largest differ. Apple Podcasts asks for around minus 16 dB LKFS with a tolerance of plus or minus 1 dB and a true peak no higher than minus 1 dB FS. Spotify normalises to minus 14 dB LUFS measured to the ITU 1770 standard and recommends staying below minus 1 dB true peak. Most producers master to the Apple figure and let Spotify normalise.
How quiet does a professional recording room have to be?
EBU Tech 3276, the European Broadcasting Union specification for critical listening conditions, asks for continuous background noise preferably below NR 10 and never above NR 15, with no impulsive, cyclical or tonal character. It also sets a nominal reverberation time between 0.2 and 0.4 seconds. Those are reference broadcast conditions rather than a rule for podcasts, but they are the numbers behind the word professional.
Does minus 40 dB isolation mean anything?
Yes, and it is measurable rather than descriptive. Decibels are logarithmic, so 40 dB of attenuation reduces incoming sound to one ten thousandth of its power, a hundredfold reduction in sound pressure. A studio publishing a decibel figure is making a claim you can test. A studio describing a room as soundproofed is not.
Can AI remove echo from a home podcast recording?
To a degree, and better than it could two years ago. Adobe's Enhance Speech states that it removes reverb along with noise and background music, free, on files up to 1 GB. Adobe also warns that running the enhancement at full strength can make a recording sound unnaturally perfect, which is why the tool has a strength slider. Treat it as a genuine repair with an audible ceiling rather than a substitute for a quiet room.
When is a home studio the better choice?
Four situations. When you record solo, because most of the irreversible decisions involve more than one person. When you record often and at unpredictable hours, because consistency beats production value early on. When the show is audio only and will stay that way, because four of the ten differences concern video. And when you already own the equipment and know how to run it.