Magic number or what?

Seeing that Amazon says HFQ has 608 pages, I checked some of the other books for comparison. MTaT and HFF both have exactly 608 pages in the hardback edition! (LAMA 672) Is this spookily precise writing? Or is 608 pages a standard size for hardbacks? And do Tor fiddle with the typesetting to make it work out to that number of pages?

As I understand it, most writers contracts stipulate the number of words a book should contain which has a direct relation to the number of pages needed to print those words.

And if you typeset and layout the hardcover first, you can use the same typeset and layout for the paperbacks and a shrunken version for pockets with no or small changes.

So it's not that strange that the books have more or less the same number of pages.

Also consider basic book binding: Pages are printed, typically, 16 at a time on a large sheet(8 per side), which is then folded and trimmed, creating a 16 page booklet which is called a signature. Signatures are then sewed and/or glued onto a backing...

So books are basically always multiples of 16 pages long. 608 is 38 signatures.

I believe, but don't have anything handy to count on except a book actually published in Sweden, that modern signatures are 32 pages, not 16. Since 608 is also a multiple of 32, that still works. 672, btw, is 21x32, so LAMA is 2 signatures longer - and thus ~10% more expensive to manufacture - than the others.

And yes, the typesetting is indeed fiddled with to fit everything into the chosen # of signatures. That's how typesetters make the big bucks - getting all those words to fit and still be readable. It's also why authors are expected to shell out to make edits once the book has been set: changing the length of a single word can mean having to reflow 2-3 chapters to get it shoehorned in. That takes time and money.

MTO wrote:Also consider basic book binding: Pages are printed, typically, 16 at a time on a large sheet(8 per side), which is then folded and trimmed, creating a 16 page booklet which is called a signature. Signatures are then sewed and/or glued onto a backing...

So books are basically always multiples of 16 pages long. 608 is 38 signatures.

Louis R wrote:I believe, but don't have anything handy to count on except a book actually published in Sweden, that modern signatures are 32 pages, not 16.

I presume that these days with roll-to-roll manufacturing, bigger is better so there are certainly times when they use larger sheets. I just did a quick check: I have some Larousse books (french publisher, printed in italy) that are 32 page signatures.
I have a cookbook printed in the USA that is only 8 page signatures.

I'm now tempted to go do a survey of hardcovers, but I'll hazard a guess: small print runs done by small publishers are probably on smaller signatures. Mass printings (like Weber's book, or the harry potter series) are probably on larger signatures.

Even with powered machinery you hit a point of diminishing returns on both sheet size and number of folds - and, believe it or not, few if any books are printed on roll stock. Not the last time I checked. I have to admit that it could have changed in the last decade or so.

MTO wrote:

Louis R wrote:I believe, but don't have anything handy to count on except a book actually published in Sweden, that modern signatures are 32 pages, not 16.

I presume that these days with roll-to-roll manufacturing, bigger is better so there are certainly times when they use larger sheets. I just did a quick check: I have some Larousse books (french publisher, printed in italy) that are 32 page signatures.
I have a cookbook printed in the USA that is only 8 page signatures.

I'm now tempted to go do a survey of hardcovers, but I'll hazard a guess: small print runs done by small publishers are probably on smaller signatures. Mass printings (like Weber's book, or the harry potter series) are probably on larger signatures.

Louis R wrote:Even with powered machinery you hit a point of diminishing returns on both sheet size and number of folds - and, believe it or not, few if any books are printed on roll stock. Not the last time I checked. I have to admit that it could have changed in the last decade or so.

The folds is a big limiting factor. As a quick check, 5 folds looked to me to not, well, *FOLD* very well, so 32 page signatures is probably the practical limit, especially when you're working with heavier papers like they often do in hardcovers.

As for roll-to-roll... I've never worked in the print industry... but I did think that offset printing was usually done roll-to-roll, wikipedia says it is used for books. OK, so wikipedia says 2 other things of interest:
1) they don't seem to use the term roll-to-roll, they call it web-feb.
2) usually only used for print runs greater than 10k. I'm not actually sure of the circulation, but I presume that RFC's books are run in sufficient quantity.

If you didn't use "web-fed", does that mean that you were dealing with massive piles of roughly A1-sized paper in a cut-sheet feeder? How often did that jam?!?

Sorry if I mislead you - I'm not a printer. I have spent a fair bit of time in printshops, but not for quite a long time. The only ones that used presses big enough for web-web to be worth while were newspapers, but I can see it happening for any very high vloume run. Offset is the normal process for any sort of largish-volume work; I have seen offset presses that could easily handle A0 sheets in 500-1000-sheet bundles. Paper handling is pretty sophisticated [by suction, IIRC] and probably jams a lot less often than all the desktop printers in your office put together. Mind you, print runs were typically only guaranteed to +/-10%, which gives you an idea of the number of points the process can go south on you.

MTO wrote:

Louis R wrote:Even with powered machinery you hit a point of diminishing returns on both sheet size and number of folds - and, believe it or not, few if any books are printed on roll stock. Not the last time I checked. I have to admit that it could have changed in the last decade or so.

The folds is a big limiting factor. As a quick check, 5 folds looked to me to not, well, *FOLD* very well, so 32 page signatures is probably the practical limit, especially when you're working with heavier papers like they often do in hardcovers.

As for roll-to-roll... I've never worked in the print industry... but I did think that offset printing was usually done roll-to-roll, wikipedia says it is used for books. OK, so wikipedia says 2 other things of interest:
1) they don't seem to use the term roll-to-roll, they call it web-feb.
2) usually only used for print runs greater than 10k. I'm not actually sure of the circulation, but I presume that RFC's books are run in sufficient quantity.

If you didn't use "web-fed", does that mean that you were dealing with massive piles of roughly A1-sized paper in a cut-sheet feeder? How often did that jam?!?

40 years ago, I spent most of a year running an offset printing press. I got pretty cocky with it, and would routinely run between 6,000 and 10,000 8-1/2x11 sheets through it per hour. Nothing the size of those signatures, but if I wanted to print something like that I'd slow the press down a lot.

The reason that I mentioned my experience is that people who do that kind of thing develop some non-obvious skills. As fast as that thing was running, I could grab a sheet out of the output stack, hold it up to the light to check for whether it was crooked, or whether the ink was uniformly dark across the page. And then slip it back into the stack as if I'd never removed it (albeit in a different place).

I'm pretty sure that if there were going to be changes to 2-3 pages, it would involve a few minutes of computer time to revise the layout, and a plate-printing charge to make a new plate for those pages. If the number of revisions was kept within reason, I wouldn't expect the costs to be very much.

Then again, my experience was 40 years ago, and I have no idea what the Charisians have come up with since then.

Signatures is the answer to the original question. My wife is a librarian, and reviews many ARCs (in a different genre). They often say 304 and 352 pages, which differ by 48. My guess is that the previous poster who suggested that different run sizes and different signature sizes correlate is spot on.