In our catalog, editions credited to Macmillan Audio average 4.10โ for narration across 502 reviews, and 3.8% of those reviews carry an audio-issues tag. Editions credited to LibriVox average 3.18โ across 1,202 reviews, and 20.3% carry that tag. Same medium, same rating scale, and a gap wide enough to change which edition you choose.
I'm Kai Nakamura, a digital accessibility consultant in Seattle with moderate-to-severe bilateral hearing loss. I stream audiobooks straight into Bluetooth hearing aids, so a well-mastered recording is a joy and a poorly mastered one wears me out within minutes. Most listeners check the author and narrator before committing. I wanted to know whether the third name on the listing, the publisher imprint, tells us anything too.
A note on where these numbers come from, because it matters. Every rating and technical tag in this study comes from AudiobookSoul's AI-assisted review pipeline and is stored on each published review in our review database. Nobody, including me, has listened to all of these recordings end to end. What follows is an aggregate reading of our review collection: 10,764 publisher-credited reviews out of 10,771 published reviews that carry both a book score and a narration score.
The baseline: what "normal" looks like
Publisher & Studio Production Benchmark
Nearly the whole collection carries a publisher credit, so coverage is not the issue here: only 7 of 10,771 published reviews fall outside the analyzed set. Twenty-five imprints and distributors have 50 or more rated reviews, and those 25 make up the leaderboard below.
The block labels the 3.83โ figure a "commercial studio" average. In practice that means every publisher-credited review except LibriVox: 9,562 reviews. That set still includes open distribution platforms, so read it as "everything that isn't the volunteer catalog" rather than as a tidy studio benchmark. Within that same 9,562-review set, 13.5% of reviews (1,289) are tagged Great Production and 4.6% (436) are tagged Audio Issues. Keep those two figures in mind as the reference line for every row that follows.
The leaderboard: three bands, not twenty-five ranks
| 1 | Audible Studios | 86 | 4.14 | 4.18 | 10.5 | 8.1 | 1.2 | 15.2 |
| 2 | Macmillan Audio | 502 | 4.10 | 3.96 | 14.3 | 4.4 | 3.8 | 12.4 |
| 3 | Penguin Books LTD | 180 | 4.00 | 3.98 | 21.7 | 5.6 | 3.3 | 10.2 |
| 4 | Scholastic Inc. | 56 | 3.99 | 3.95 | 19.6 | 10.7 | 5.4 | 8.1 |
| 5 | HarperCollins | 82 | 3.98 | 3.88 | 22.0 | 8.5 | 6.1 | 7.1 |
| 6 | Blackstone Audiobooks | 357 | 3.96 | 3.95 | 13.2 | 9.0 | 2.2 | 12.4 |
| 7 | W.F. Howes | 64 | 3.95 | 3.97 | 3.1 | 3.1 | 4.7 | 12.6 |
| 8 | Recorded Books | 695 | 3.95 | 3.89 | 9.8 | 7.5 | 4.5 | 11.8 |
| 9 | Listening Library (Audio) | 245 | 3.95 | 3.98 | 16.7 | 9.0 | 4.1 | 8.8 |
| 10 | Random House (Audio) | 926 | 3.90 | 3.91 | 17.4 | 7.9 | 3.6 | 11.8 |
| 11 | Penguin Audio | 575 | 3.90 | 3.84 | 11.7 | 7.1 | 2.1 | 11.5 |
| 12 | HarperCollins Publishers UK | 76 | 3.88 | 3.86 | 13.2 | 7.9 | 5.3 | 12.3 |
| 13 | Hachette Book Group USA | 585 | 3.83 | 3.76 | 17.4 | 6.0 | 4.6 | 11.0 |
| 14 | Simon & Schuster Audio | 811 | 3.83 | 3.81 | 11.5 | 7.8 | 4.2 | 10.3 |
| 15 | HarperAudio | 673 | 3.82 | 3.84 | 10.4 | 6.2 | 3.1 | 10.4 |
| 16 | HighBridge Company | 133 | 3.82 | 3.90 | 10.5 | 7.5 | 6.0 | 9.6 |
| 17 | Dreamscape Media, LLC | 136 | 3.81 | 3.77 | 16.2 | 9.6 | 3.7 | 10.1 |
| 18 | Brilliance Audio | 572 | 3.75 | 3.73 | 10.5 | 5.6 | 4.5 | 12.6 |
| 19 | Thomas Nelson | 62 | 3.75 | 3.71 | 12.9 | 16.1 | 1.6 | 10.6 |
| 20 | Tantor Media | 888 | 3.71 | 3.74 | 9.7 | 7.5 | 6.6 | 10.7 |
| 21 | Thought Audio | 51 | 3.39 | 3.63 | 3.9 | 47.1 | 9.8 | 1.6 |
| 22 | Authors Republic | 59 | 3.39 | 3.27 | 15.3 | 27.1 | 3.4 | 7.5 |
| 23 | Ascent Audio | 156 | 3.36 | 3.57 | 21.2 | 28.8 | 9.6 | 7.5 |
| 24 | Findaway Voices | 179 | 3.33 | 3.38 | 16.2 | 17.9 | 15.1 | 6.3 |
| 25 | LibriVox | 1202 | 3.18 | 3.48 | 6.7 | 24.4 | 20.3 | 7.4 |
Twenty-five ranked rows invite you to read the table like a race. I'd rather read it as three bands, because several neighboring scores are only a hundredth of a star apart and some rest on fewer than 100 reviews.
Band one: 3.90โ and above (ranks 1โ11). Audible Studios leads at 4.14โ narration, but on only 86 reviews. Its 1.2% audio-issues rate works out to roughly one tagged review, so it is a promising signal rather than a settled verdict. Macmillan Audio is the more heavily evidenced top performer: 502 reviews at 4.10โ narration, against 3.96โ for the books themselves. That narration-over-book pattern is the kind of split we explored in When Narration and Book Scores Split.
The rest of the band is tightly packed. Penguin Books LTD (180 reviews, 4.00โ ) has a 21.7% Great Production rate, and HarperCollins (82 reviews, 3.98โ ) has the highest in the table at 22.0%. Scholastic Inc. sits at 3.99โ on just 56 reviews. Blackstone Audiobooks (357 reviews, 3.96โ ) and Penguin Audio (575 reviews, 3.90โ ) pair solid narration with low audio-issues rates of 2.2% and 2.1%. Random House (Audio), the largest imprint in this band at 926 reviews, scores 3.90โ with 17.4% Great Production.
From an audio-clarity standpoint: a displayed tie is not a measured tie. W.F. Howes (64 reviews), Recorded Books (695) and Listening Library (Audio) (245) all display 3.95โ . Random House (Audio) and Penguin Audio both display 3.90โ . The rounded figures cannot tell you which of them is ahead, and the larger counts don't make one imprint "better", only more broadly covered.
Band two: 3.71โ to 3.88โ (ranks 12โ20). This is where several high-volume names sit. Hachette Book Group USA (585 reviews) and Simon & Schuster Audio (811) both display 3.83โ , the same displayed value as the non-LibriVox average. Hachette's 17.4% Great Production rate equals Random House (Audio)'s, despite the lower narration average. HarperAudio (673 reviews, 3.82โ ), Brilliance Audio (572, 3.75โ ) and Tantor Media (888, 3.71โ ) round out the volume players. Tantor's 6.6% audio-issues rate is the highest among the 20 imprints at or above 3.71โ , though Thomas Nelson, at the same band's lower end with 62 reviews, posts just 1.6%.
Band three: 3.18โ to 3.39โ (ranks 21โ25). The drop from Tantor Media's 3.71โ to the next row's 3.39โ is 0.32 stars, the largest gap between adjacent rows in the table. Below it sit Thought Audio (51 reviews, 3.39โ ), Authors Republic (59, 3.39โ ), Ascent Audio (156, 3.36โ ), Findaway Voices (179, 3.33โ ) and LibriVox (1,202, 3.18โ ). Two of these are open distribution platforms. Our methodology notes that Findaway Voices and Authors Republic aggregate independent self-published productions alongside studio releases, so their rows pool very different kinds of recordings under one credit.
Here's the band view as a quick reference. Ranges come directly from the leaderboard rows in each band.
Band Ranks Narration โ range Audio Issues % range Great Prod % range 3.90โ and above 1โ11 3.90โ4.14 1.2โ6.1 3.1โ22.0 3.71โ to 3.88โ 12โ20 3.71โ3.88 1.6โ6.6 9.7โ17.4 3.18โ to 3.39โ 21โ25 3.18โ3.39 3.4โ20.3 3.9โ21.2Notice the last column. Great Production rates overlap almost completely across bands: Ascent Audio, near the bottom on narration, has a 21.2% Great Production rate, while W.F. Howes in the top band has 3.1%. That tag alone doesn't sort imprints by narration score, which is a useful non-finding. Look at it alongside the audio-issues column, not instead of it.
Where audio issues concentrate
- LibriVox20.3% (244/1202)
- Findaway Voices15.1% (27/179)
- Thought Audio9.8% (5/51)
- Ascent Audio9.6% (15/156)
- Tantor Media6.6% (59/888)
- HarperCollins6.1% (5/82)
- HighBridge Company6.0% (8/133)
- Scholastic Inc.5.4% (3/56)
- HarperCollins Publishers UK5.3% (4/76)
- W.F. Howes4.7% (3/64)
- Hachette Book Group USA4.6% (27/585)
- Recorded Books4.5% (31/695)
- Brilliance Audio4.5% (26/572)
- Simon & Schuster Audio4.2% (34/811)
The chart shows the 14 imprints with the highest audio-issues rates, from highest to lowest. The cleanest rows, such as Audible Studios at 1.2% or Penguin Audio at 2.1%, are in the leaderboard rather than the chart.
Two rows stand well apart. LibriVox has 244 of 1,202 reviews tagged with audio issues (20.3%), and Findaway Voices has 27 of 179 (15.1%). Thought Audio (5 of 51, 9.8%) and Ascent Audio (15 of 156, 9.6%) come next. After that the chart flattens into a 4โ7% cluster, from Tantor Media's 6.6% (59 of 888) down to Simon & Schuster Audio's 4.2% (34 of 811).
Why would distributors and volunteer catalogs sit higher? Our data can't answer that directly, since it doesn't record how, where, or by whom each recording was made. One possible reading is that a credit covering many independent or volunteer productions pools a wider mix of recording setups than an imprint line. That's a hypothesis, not a finding. Note that Authors Republic, also a distribution platform, posts just 3.4%, so "distributor" is not a reliable shorthand for "issues."
A second caution concerns the Clean Audio column in the leaderboard. Thought Audio leads it at 47.1%, and LibriVox's 24.4% is well above Macmillan Audio's 4.4%, even though LibriVox has the highest audio-issues rate. Clean Audio and Audio Issues are separate tags, not opposite ends of one scale, and Thought Audio's average runtime of 1.6 hours describes very different material from Audible Studios' 15.2. I would not read Clean Audio as a ranking on its own.
If you rely on crisp consonant separation, as I do, pay attention to the imprint. Then confirm on the edition's own review, because an imprint average is a pooled number across many different books.
Flagship productions worth opening
This table is a showcase, not evidence about imprints. Each row is a single review with a Great Production tag, a Must-Listen verdict and a 5.0โ narration score, and one excellent edition doesn't lift a whole catalog.
The list shows how wide the format range is. Dungeon Crawler Carl, read by Jeff Hays, is the Audible Studios entry at 13.5 hours. None of This is True, read by Alix Dunmore, represents Simon & Schuster Audio with 5.0โ for both book and narration. Brutal Telling, read by Ralph Cosham, does the same for Macmillan Audio.
Blackstone Audiobooks' entry is Story of a New Name with Hillary Huber, at 19.1 hours. Random House (Audio) is represented by Angels in America, read by Andrew Garfield and the shortest flagship at 6.9 hours. Lincoln Highway, read by Dion Graham, carries the Penguin Audio line, and Belle Sauvage, read by Michael Sheen, represents Penguin Books LTD.
At the long end, Gone With the Wind, read by Linda Stephens for Recorded Books, runs 49.2 hours. Brilliance Audio's pick is Peter and the Starcatchers with Jim Dale. Brilliance sits in the middle band at 3.75โ , a reminder that standout editions turn up across the whole table.
The engineering here is clean, even when the story gets chaotic. At least, that's how our pipeline tagged these specific editions. Each review page shows the tags for that one recording, and that's the level I'd base a decision on.
A listener checklist for publisher credits
Here is how I'd use this study, especially if you listen through hearing aids, cochlear processors, or anything else that makes clarity matter more than usual:
Before you choose What to look at Why Compare two editions Each imprint's Narration โ and review count 4.14โ on 86 reviews and 4.10โ on 502 are not equally evidenced Clarity is a priority Audio Issues % in the leaderboard Rates run from 1.2% to 20.3% across the 25 imprints Edition is from LibriVox or an open distributor The specific edition's review and tags These rows pool many independent or volunteer productions Scores look tied Treat rounded ties as ties Three imprints display 3.95โ ; rounding hides the orderThen open the edition's review page and check its technical tags, and visit the narrator's profile to see how their other recordings scored. If where you listen also affects what you hear, Audiobook Listening Contexts Compared is a good next read.
How we got these numbers
These are observational averages across different books, genres and editions, so they are not a controlled comparison of imprints. Scores and tags are corpus evaluations from our AI-assisted review pipeline, not acoustic lab measurements, and imprint averages say nothing about why a given recording sounds the way it does. Treat the publisher credit as a reason to look closer, then let the edition's own review decide.



