The very best speaker cabinets are made from MDF.  This is because MDF is far superior to particle board, has very good internal damping, and machines nicely.  But MDF can be improved further by the addition of hardwood.  Hardwood is brutally stiff, and improves the appearance of the cabinet.  IMO, even very good veneers fall slightly short of the rich/deep hardwood appearance.

Most folks really like the look of hardwood on a speaker box.  My favorite happens to be striped walnut.  There are some problems with walnut, and all hardwood for that matter.  The first problem is resonance.  Resonance is good in an instrument, but bad in a speaker cabinet. The back of a piano isn’t covered in MDF. It is covered in solid wood. This is because solid wood (lumber) resonates very nicely. While this is good for instruments, this is not good for a speaker cabinet. Ideally the cabinet should be acoustically dead. There is room for debate about the box being part of the speaker’s sound, not something to eliminate sound. I agree with most that solid & dead is better than live & loose. Most excellent commercial cabinets are very heavy and strong.  A wimpy cabinet will produce warm, sloppy, and inaccurate bass.  This is because audible bass is emanating from the sloppy cabinet walls.   I prefer my bass tight, lean and accurate. This requires a strong cabinet.  Therefore I err on the side of solid and dead than on the side of resonant and sloppy. This means that I need to have a strong/dead internal structure and have lumber only on the outside. The inside skeleton of my cabinets are 3/4"+ MDF and the exterior is veneer on the sides and lumber on the baffle.  The lumber on the baffle is glued to the MDF.

Over time and changing environmental moisture, lumber will expand and contract. MDF doesn’t (relatively speaking). I consulted the Madisound discussion group and a local Cabinet shop owner concerning this issue.  A few offerings were conveyed. One was a sand gap. Another was a constrained layer with asphalt roofing paper. Another was to use some esoteric EAR dampening as a constrained layer material. The general consensus on gluing the oak to the MDF was that if the span across the grain was great enough the oak would split over a period of time. Byron Studer of St. Peter, MN glued solid wood to MDF with poor results.  The sides cracked.  He used good lumber from his brother's cabinet shop on a speaker cabinet and the sides (not the front) cracked in about one year.  I then asked Dave Baker, owner of Custom Cabinets, in Great Falls, MT about gluing oak to MDF and he said it wouldn’t split.  Hmmmm...

I attribute the difference of opinion to one of four variables. One, Bryon Studer used Silicon glue to allow for some movement. Maybe a more solid aliphatic resin glue would be better. Presumably this might hold the oak in place where it would form smaller cracks instead of a big crack after the Silicone allowed the oak to shift and squirm. Two, having grown up in Minnesota I know that most folks homes are very dry in the winter, and quite humid in the summer. Great Falls, MT is always very dry. This would explain why the cabinet shop owner said I would have no problems regardless of cross grain lumber span.  The varying home moisture content may play a role in causing Bryon Studer’s Minnesota lumber to crack. Because of the extreme hot/cold and resultant home humid/dry conditions (relative humidity), Minnesota furniture is placed under severe stress. Third, I don’t know how the other folks applied pressure to the material (MDF and lumber) to glue it together, but this may be an issue too. Bryon Studer probably did the glue-up in the cabinet shop, so this probably isn’t the causal variable. Fourth, the tangential distance across the material was likely a contributor. None of the Gentlemen who responded said that their baffle cracked. They said that the sides of their speaker cracked. The sides of their cabinets were substantially longer than the baffle width, the wider span across the grain was the most likely factor.  Hence, there is veneer across the sides of my speakers and solid wood on the baffle.  I am very comfortable with this.

The big advantage of hardwood is it's superior strength.  MOE (Modulus Of Elasticity) is a stress to strain ratio.  Higher number's indicate stronger materials.  The Young's modulus of elasticity (MOE) of red oak is about 1.55.  The MOE of MDF (medium density fiberboard) is about .55.  This means that red oak is approximately 3 times stronger than MDF.  When oak is glued to MDF it makes a very desirable baffle.  The baffle becomes incredibly strong (oak) and dead (MDF).  MDF comprises the other factor desirable in a speaker panel.  The MDF is very dampened.  It doesn't vibrate.  I think of MDF having little "shock absorbers" built in.  The rap test on the oak/MDF baffle results in, well... sore knuckles.

MDF and lumber will move (i.e. expand & contract) as they gain/lose moisture during seasonal change.  When gaining moisture they will expand, and when losing moisture they will contract.  During the winter months, most homes are relatively dry.  During the summer months most homes are relatively humid.  My home in Omaha Nebraska has @ 35% humidity in the winter and @ 50% humidity in the summer inside my home.  This humidity is different from the outside humidity.  During the winter months, using a furnace to heat very cold air will effectively reduce the relative humidity.  During the summer months, using central air conditioning (A/C) to cool the very warm (and humid) air will actually remove moisture from the air.   I have lived most of my life in the Midwest and currently live in Nebraska.   Other regions certainly have different variations. 

I will offer herein a comment that is slightly tangential.  This comment comes from an HVAC expert (my uncle Richard).  He cited that high efficiency (i.e. High SEER) Air Conditioners (A/C)  remove less humidity than low SEER A/C.  This is because the temperature of the in a low SEER evaporator is lower and more moisture will be drawn from the air when passing through the evaporator.  For this reason, he recommends folks in the Midwest use an A/C unit with a SEER rating of @ 12-13.  He conversely recommends in the Midwest avoid using extremely high efficiency/high SEER (i.e. SEER 16) A/C.  While a high SEER A/C is more efficient, it will also leave more humidity inside the home.

Given the lumber in used, the type of cut dramatically effects the effects of changing moisture content within that lumber.  Given the same imbuing change in moisture content, flat-sawn lumber moves @ twice the amount of quarter-sawn lumber.  Quarter-sawn lumber is very stable.  In the good-old days, the fancy quarter-sawn lumber was used for two reason.  First, it looks very nice.  Second, and more important, it is very stable when moisture content changes.  The interior environment of homes built before... @ 1940 . 

There is more information about the relative movement of quarter-sawn and flat-sawn lumber here.