Slightly edited, here is an IM exchange with a very nice and knowledgeable Patagonia representative from a few minutes ago:

You: Curious that the size specified for the new, current version of the MLC doesn't quite calculate. 20.5 x 14 x 7 doesn't really equal 2746 cu.in. I am by the way an owner of the original MLC, my bag now about 20 years old.
Peter: Okay the dimensions given for the bag are as if it was a solid sided bag , but as you know the MLC is a soft bag and will expand out past those measurements given, which accounts for part of that discrepancy
Peter: Volume for bags is not actually a length xmidth x height measurement they fill the bag with voulume sheres and calculate the volume from that.

The rest of the interchange basically confirmed what one can learn about the new, and apparently fourth version of this bag; I have the first version, now being replaced by an Air Boss, whose volume on their website, by the way, is listed as 2340 cu.in., even though its nominal dimensions calculate to 2184 cu.in. So, the Redoxx calculation is 7% higher than that suggested by the dimensions, but the Patagonia upward adjustment is 37!

VOLUME SPHERES. Hmm. I am NOT sure how this corresponds to real life packing. Perhaps the webmaster of OBOW will know. Perhaps a more real-life measure might be MTSE's, aka "Men's T-Shirt Equivalents." Six of my XL versions fit into an Eagle Creek 2" x 10" x 14" cube, a bit protruberant, so an MTSE would be around 50 cu.in., and should be representative of the sort of soft goods one packs for a real trip.

Measurement of carry-on capacity may be a subset of Einstein's Theory of Relativity it appears. I may measure my Air Boss when it arrives by the MTSE method....as I think I have close to forty-seven T-shirts, which, of course, my wife advises are TOO many. Or maybe...NOT!

As a footnote, a minute or so on a scientific calculator does show that for the circumference of a rectangle, 2:1 ratio, a CYLINDER will have +43.2% greater area, and its diameter only 4.5% less. So that MLC theoretically overpacked and deformed into a cylinder could hold the advertised volume, but I think that's an extreme and improbable case. I would buy +15 to +20%, for a soft bag loaded into a modified oval cross-section. It might also be interesting to try and load, say, a semi-framed wheeled 24" case, with those "volume spheres," and see how the measurement compares with what is advertised, and with a well-designed 21-inch semi-soft bag, like the Air Boss. My guess is that the VSV, the "volume sphere volume" would be pretty close one to the other.

Finally, presuming that "volume spheres" are after their being used to measure a pack or item of luggage,are then themselves measured by being placed into a calibrated cylinder or the like, EUREKA, we should credit the inventor of this method, who used of course, not small styrene balls or the like, but molecules of water, that being the Greek philosopher and scientist....ARCHIMEDES.

Yep, we've had that discussion in the shoulder bag shoot-out comments. So if I understand you correctly, Alan, than a cylinder will hold the same volume than a cube with a lesser surface area, i.e. exterior dimensions. Is that right?

That would be an argument for duffels and carpet bags or even those basically cylninder shaped satchels.

I think in real life the Airboss will actually hold more than the new MLC. But I also think that considerations of volume are secondary if they are close. Maxing out volume is probably not a good idea anyways in terms of weight and stress. How something packs and whether it has the right dimensions for what you need to pack seems more important.

For example one bag might have the higher overall volume but the compartments might be too small for your shoes. Or its shape will not allow you to put a laptop or to put files without them getting bent.

Before turning to Einstein and breaking out the scientific calculator, you could make tennis balls the standard measure of volume for soft-side luggage. "Bag A holds 72 tennis balls. Bag B holds 84 tennis balls." You could keep a table of the results as you review bags. You would still list your normal volume and dimensions, but the balls would give us a degree of relativity. They're cheap, uniform, and will last forever for this purpose. Even the small boutique manufacturers can afford a garbage can full of balls to calculate their own volume. It would also make it easier to calculate all the pockets. To measure, you would just pack it until it is absolutely full but will still close, then count the balls.

That's a great idea for the main compartments but not so good for the often flat side compartments. How about using ping-pong balls instead? Even lighter and cheaper, less wasted space, meaning the gradation of measure will be finer. They will also be able to judge an outer compartment. They are standardized to 40mm diameter. Golf balls (42.6mm) would work, too, but are more expensive and heavier.
In comparison a tennis ball is around 65mm.

Still your basic idea is brilliant because it is so simple, feasible and easily reproducible. Bravo!

Actually the point of my post and addenda was just to point out some of the vagaries of volume measurement, and what sounded to be a bit of hyperbole from the folks at Patagonia. Very clearly several other factors figure into choice of a bag. In fact my final choice went to the Air Boss in large part because it will allow me to separate those irregularly shaped shoes from easier to pack items, and its center compartment additionally provides protection for more fragile items.

As for measuring with TENNIS balls, good idea, probably a good measure for larger main compartments, but likely to underestimate the size of small side compartments substantially, as their diameter is similar to the depth of many bag compartments. A typicall soft carry-on bag the size of an Air Boss, theoretically should hold about 227 tennis balls, but practically, rather fewer than that. That would be a might heavy, at 28.8 lbs.

GOLF balls? These are likely to more accurately measure smaller compartments in particular, but, since nominally an RAB could hold 926, which would weigh 94.8 lbs., this would end as more of a seam torture test!

Mostl likely the good folks at Patagonia use something like a PING PONG ball, save a bit sturdier, and the test load would come out to around 6 lbs. Besides that, what better sport for someone with carry-on luggage than table tennis? Surely better than...golf.

Back to square one...wonder how many tee-shirts I REALLY have? My wife really wants to know how many I can toss....

The dimensions and volume are very difficult to comparewhen there is no "standard" to work with - and of course they don't tell you how they came up with the values in the marketing literature. Some seem to take the greatest external dimensions and then give a "displacement" voume, it how much room the entire bag takes up. Others will give an internal volume capacity of the bag, which is the available packing space - as the OP found. Having said that, I can't understand how "bulging" would increase the internal space - an increase in one dimension would cause a decrease in another, unless the fabric had some stretch in it.

I got badly caught out in my last bag purchase. The bags I own (with marketed measurements) are The North Face Amira (19x13x7", 26L), The North Face Sweeper (20x13x8, 30L) and High Sierra Passport (HSP) (22x14x8.5, 40L?). I bought a Tom Bihn Smart Alec (18.5x11.75x7.75, 26L) thinking it would be about right. However the Smart Alec is seriously huge in terms of packable volume - twice the size of the Amira, about 1.5 times the size of the Sweeper, and only slightly smaller than the Passport. Makes comparing bags online very difficult!!

You are correct. Ping pong balls would be a much better measure than tennis balls. They are cheaper, lighter, and allow much finer measurement of pockets. They are so cheap no manufacturer could justify not using them if reviewers started using them. I only gave it about two seconds thought when I sent you the suggestion about tennis balls. Well done.

You can buy used golf balls in bulk, and they could be useful for stress tests on the seams since they are heavier as Alan Birmbaum pointed out. You could fill a bag full of golf balls and hang it for 28 days for an effortless stress test.

They could use rice grains, too, for that matter. But ping-pong balls is probably very good. On the other hand, rice is easy to imagine if translated in weight. At least I know more or less what kind of volume I have to imagine when we speak of 2lbs of rice. When we speak of 234 ping pong balls, I have no clue.

Stress test would have to include some movement with varying directions and degrees of acceleration. Afaik, Eagle Creek actually does that kind of stuff. Otherwise, most of the high-end bags we talk about here will hold up to the rigors of normal travel for quite some time.

To a certain degree one can also deduct that from technical construction detail: what kind of yarn, double or triple stitched, etc. Andiamo and Redoxx as well as some other boutique makers are fairly specific and upcoming about these things.

I think the folks at Tom Bihn use "packing peanuts" to work out the volume of their bags. Must be in a forum post somewhere...

Hm, certainly possible but I think ping pong balls or rice would be better. Packing peanuts are irregular in shape. Depending on how they fall in place they will take more or less volume. They are also easily compressed.

Rice would be exact, easy to get, easy to imagine, and quite durable as long as it doesn't get wet. Also easier to take out of the bag than peanuts that will get all static and cling to the nylon and each other.

How would you measure the rice? "Bag A holds 10 lbs of rice," or "Bag B holds 45 liters of rice?" Neither one of those really mean anything to me. A lot of manufacturers list their volume in liters, so measuring the rice in liters wouldn't really be any advantage.

Also, if you are suggesting loose rice, you'll never get it all out of your test bag. It's farfetched, but that might cause customs dogs to alert on your bag. I think the ping pong balls are simpler.

The intention of my suggestion was to get away from these standard measures and to add a little mirith to the descriptions. Ping pong balls are more mirthful than dry rice.

Well, the test bag wouldn't be sold anyway. And I haven't heard of rice sniffing dogs. :)

If you go to an Asian supermarket or even just your ordinary supermarket you will see various sizes of rice containers. From there it is easy to remember and imagine how much volume a bag has. You know how big a 1lb bag of rice is. Sometimes they even give you volume measures on the packaging.

For chutzpe ping pong balls are better indeed. But I have a hard time imagining the volume of x munber of ping pong balls even when I hold one in my hand. I can easily imagine what 45 packages of rice look like even if I don't hold one in my hand.

The beauty of the rice method is that you don't need to measure it in liters but you can. Then you can tell the customer this bag holds 45.23 liters equaling 42lb of rice.

Maybe you're right and rice packages are better, but "42 packages of rice" will never sound as good or dramatic as "937 ping pong balls." Sigh... disappointment clouds my 9061 liter cubicle.

Rice seems like it has the potential to be very messy... and very heavy. If you're doing a capacity-only test, something like ping-pong balls seems more practical. But whatever will conform the best to the shape of the bag is obviously the most accurate way to measure the capacity.

Just for the fun of it I figured this one out; took two minutes.

A lot of the bags discussed here have a volume in the 40-45l range. To make calculations simple I used 200 US cups as a volume measure. That's 47.3 liters. 200 cups of long rice come to exactly 40kg according to this converter:
http://www.traditionaloven.com/conversions_of_measures/rice_amounts_converter.html

So 1kg of rice equals around 1.2l, almost like water for the calculations.

You are right that it might be a little dusty but one can do it outside. And it only needs to be done once for each bag.

2.2 cups weigh 1lb.

My Redoxx AIR BOSS just arrived. Anyone who has one of these knows what they are like as far as construction, which is A+. These also include a very nice thick CLEAR VINYL LUGGAGE TAG, which has already had my business card inserted, attached by a thick vinyl leash.

The WORLD'S FANCIEST PRICE TAG, an OVAL METAL DOG TAG. also comes with the bag, attached by a thin cable I do wish however, they had left off the price. A more useful touch might have instead been to put the new owner's NAME on that dog tag, plus a bag serial number (but NOT obviously, the owner's Social Security number.)

As to its measured VOLUME of the RAB, well, I'm not playing golf, tennis or ping pong recently, and gee, my wife and I couldn't eat that much RICE in ten years!

Possibly an easier way to measure a bag's comparative volume might be to simply load it with bath or beach TOWELS of various sizes, which are readily available in most residences, won't bounce across the floor if dropped, and which have a DENSITY likely similar to clothing, which is, after all, is what bags are about.

Sure, some items one actually packs are somewhat heavier, others lighter, but towels ought to be close enough to the density of travel items. Of course, this way one measures EMPTY weight then LOADED WEIGHT, the net load likely being enough to at least be able to COMPARE one bag to another. With a BALANZZA or similar scale, this is a very quick determination.

The bag I just received likely has a capacity of about 2340 cu.in., which is 10.1 gallons. If that were water, such would weigh about 84 lbs. Actually, clothing (or towels) probably has a SPECIFIC GRAVITY of about 0.30, so capacity likely should be about 25 lbs., i.e., fully loaded, around 29 lbs. I'll post later what the actual weight proves to be.

Nope. TOWELS are not a good idea. The RAB holds just over 8 lbs. of various bath and beach towels, yet clearly folks load this particular bag with over twice that weight in clothes and accessories. Reboot towels as a useful measure.