Shelf Material

I certainly don't like MDF in practise - a very muddy sound. I don't have the ability to try such things as the BDR or Polycrystal shelves before buying (since noone stocks them here), and hence this post. But I quite like the BDR cones and so bought some of the BDR pucks - and they sounded awful. If they represent the BDR shelf then no thanks. I have actually got a lot of good ideas from the above posts and am in the process of trying or acquiring some of the less expensive suggestions. I am happy to splash out on the expensive stuff, but would prefer to see a concensus of more than one before doing so. It looks like the Butchers block qualifies. I wonder if Kevin has a point. Caterham and I tend to concur that you want light and rigid, release the energy quickly but damp the main resonances. But Kevin is suggesting the light and rigid shelf, and then deal with the resonances with some form of compliance between shelf and component. This may very well be right, but I tended to think of the rack as doing the light and rigid task and that you would begin to damp the resonance with the shelf. I don't like two forms of compliance in a system and so I have tried to get "dead" sounding shelves and use cones. My theory may be right, but the practise is possibly too dificult to achieve - we will see. Kevin may have a more practical solution. Am I missing something Kevin?

What about carbon or carbon fiber?

It seems to me that MDF would make a very good shelf material - low mass and high rigitidy. I use 3/4" instead of 1/2" with good results as a standard (non-suspended) platform. It also has no grain. Am I thinking correctly? I have a 5' x 2 1/2' x 2 1/2" maple glue-lam butcher block that I salvaged from a restaurant but would rather go with MDF based on the info in these posts.

As for Subaruguru's comment about the sitka spruce, if it does have a high Young's modulus, then it would be an excellent "shelf material". My original comment applied only to the selection of the shelf, as the title of this thread states. Once a shelf has been chosen, then a suspension system can be applied to decouple the component / shelf from the structure which it resides upon. The goal of the shelf should be to rigidly couple the energy from / to the equipment into the suspension system. The lower the mass (for the combination of the shelf and product), the lower the amplitude (and higher the frequency of the resonant system) assuming that the compliance remains constant. Another concept to consider is the "Q" of the suspension and the use of any damping (lossy) properties. A good field of study would be to take a look at what turntable manufacturers are doing. They have had to deal with suspension / decoupling systems and have a wealth of practical experience. Kevin Halverson

High Young's modulus and low mass.... hmmm...doesn't that describe sitka spruce, universally used for piano soundboards?...and guitars, etc.? Whereas I'm tending to agree with these requirements, the additional critical requirement is that this wonder-shelf also drain or absorb energy (as heat), rather than reradiate it like a musical instrument, no? Caterham's analysis and resultant highly-evolved product seems attractive. I just spoke with a friend (Ken Parker) who is a custom luthier, and currently produces carbon-fiber covered wooden-body guitars as well. His whole life is currently wrapped up in making 3 lb ultra-stiff $3000 musical instruments that "sing" loudly even before making use of their piezo pickups. I mentioned to him all these suspension/isolation/damping options, and his initial thoughts are to try clamping the hell out a CDP by mass-loading the top and then thin-skin binding it (sorbothane sheet or 1/8" gum rubber you can buy cheap) to a 3" granite base that you can buy from machine shops (who use them as precision flat surfaces, discarding them after they get nicked or chipped).......... Yet again that old bumper sticker "Think Globally, Act Locally" comes to mind: what about damping/tuning the interior and suspension parts of CDPs and transports...y'know, getting right into the heart of it? A guy in CT (Virtual Mode) uses a Rotel 855 he damps internally as his reference CDP for the design oh his well-respected passive attenuators!...claims most CDP problems are mechanical, not DAC related. I'm tempted to send him my 855 for tweaking before replacing it with a Bel Canto DAC/DVD setup, etc. Yet Caterham speaks of not upsetting the original voicing of a product by such internal damping in lieu of vetting its vibrations through his shelf, which, with additional hard cones, further isolates the component from externally-generated vibratory spuriae, as well. Christ, I don't really know where to start here...still using that 5 buck wheel barrow inner tube from Home Depot! Redkiwi--the deal with maple butcherblock is that the individual strips of wood are glued together in opposing-grain arrays to provide mechanical integrity--no warping, and controlled expansion/contraction. I suppose it won't ring like spruce, of course, but don't see why it would be better than cheap MDF in this regard. It's VERY hard, and will require carbide tools to cut, dulling them readily because of the glue in the joints, as well. It's beautiful and lasts forever, though, so you'll get furniture-grade shelves as a bonus...... Anyone familiar with Virtual Mode's $200 CDP mods? Thanks for these thoughtful and educational posts, guys. Ernie

While I won't pretend to have an ability to attribute sonic characteristics to any of the proposed construction techniques; I will make the following comment. Anything that increases the mass that does not have a corresponding increase in stiffness (high Young’s modulus) will result in an increase in the amplitude of deflection. In simpler terms, as stiff as possible, without unnecessary mass. Kevin Halverson

What about the polycrystal and black diamond products in the audio market? Are these worthwhile?

Redkiwi: I should warn you that I designed and built all of the curtain rod/window treatment accessories for our apartment and ended up using blue rubber "dog balls" for the finials in the dining room. I do have a tendancy to cut corners.

Redkiwi: I do not know what sonic difference the solid core piano wire will offer. I suspect that any braided or elastic cable whether metal or nylon will stretch from the weight of the gear/platforms and that the amount of stretch will not be uniform per cable. This stretching would necessitate constant leveling of the platform on which the CD player rests. I was just trying to be "practical" with the suggestion. ROFL. And yes, spikes sound like a good way to go with plan B. The aluminum sheeting on the bottom of the shelves should hold up to them.

Caterham1700 - thanks for your generosity, if you will allow me to be a little generous in return. I will send you an email directly. Recres - I have heard similar comments about Corian, but am curious to try because I think it will ping less than perspex. On the other hand "slow" is what I fear most about it as a material, being so dense and heavy. Did you use up-turned spikes underneath the Corian? Great to hear confirmation of the butchers block idea, and given Albert's recommendation I have tried to source some locally. However, I have found that no one brings it in and that our native South Island Beach is used universally in this country. If I can get some I will try it - but I think I will now have to begin investigating bringing some Maple in from the US. Dekay, I definitely recommend you use up-turned spikes between lips and shelf for your plan B, over any compliant material. But so far plan A (ie. suspension) sounds better still. Apart from issues of stretching, is there any reason to believe that a taut wire like the piano wire you mention will sound better than a wire that has a little give, such as the nylon I have used? Oh yeah, Gthirteen, this really defines the freaks among us, but whether it is my room or my hearing, this area of vibration appears to me to be as utterly critical as room treatment, yet unlike room treatment in that there does not seem to be any conventional wisdom on the best strategy. As I indicated when starting this thread, this is the one area of this hobby where I feel I am still way off achieving decent results (let alone understanding). I did try plexiglass and found it to be better than glass but not as good as perspex, but I might try a thicker piece and see what happens.

I love it. You guys are all freaks. I should know, we can smell our own. I am following the thread closely, as I will be re-positioning my stuff soon enough. Something you may want to try is Lexan, which is similar to plexiglass. A thick piece of plexiglass is much lighter than glass, and shouldnt ring. Good luck!!!!

Redkiwi: Thanks for the update. I just screwed up my left shoulder and am going to have to wait on the project, but here are a few ideas that I have had. Using small maybe 1/64th piano wire with line tightners for the cables (they should not stretch like braided wire or plastic line)and should pretty much remain level after the frame settles in. If I can squeeze the extra height in my cabinet I would also like to try double shelves that layer as follows (MDF/Vibrapods/MDF/component with stock feet resting on top) and a sheet of aluminum on the bottom side of the bottom layer to help shield my amps from the power supply in my CD player wich will rest directly below the amps. I figure that if it doesn't work I can always reinstall the same shelves by placing lips on the inside sides of the cabinet and resting the shelves on them. Then just play with isolation teqniques between the shelf and the lip. In the meantime I am going to have to rest my amps in "cat territory" on the outside top of my cabinet.

I've tried Corian, Redkiwi. And I didn't like it. It ruined dynamics, sounded slow and muddy. The maple butcher blocks that Albert describe work MUCH better. You may want to call the local restaurant supply businesses close to you and compare prices. I got mine for a little less than the McMaster Carr price. Good luck!

I have a demo Neuance Alpha sized 18-1/2" x 14-1/2"(fits Mana Mini/Reference Table) that I would be happy to let you have for the cost of shipping.Let me know if this is something you'd like to investigate. Best, Ken GreaterRanges/Neuance greaterranges@msn.com www.neuanceaudio.com

OOPS - it was Onhwy61 that suggested the suspension idea - thanks Highway Man!

The Corian hasn't arrived yet so a report on that will have to wait, and it sounds like I should go directly for the Maple rather than try a native NZ wood - thanks Albert. BUT - I bought some 100lb strain guage nylon from the local sports shop today and suspended my source and preamp components (three of them, but each suspended separately) by merely tying the nylon to the steel frame of my rack - ie. I removed the shelf above the component and tied the nylon to the frame that had supported that shelf and then looped the nylon under the component below and tied it to the other side. I used two pieces of nylon for each component, going side to side, and used a trusted fishing knot each end, rigged up a twist arrangement to deal with fine levelling and sat back to listen. I have only listened to two CDs so far, but I am excited. Resolution is better, soundstaging is remarkable, bass is more extended and more punchy, and the sound is slightly more forward, but with oodles of depth. The sound is definitely more natural - applause sounding much more like hands clapping than rice crispies. There is none of the swimminess that the bladder products have, and there is no part of the spectrum that has any apparent resonance or suck-out effect (all shelves seem to suffer the former and bladder products suffer the latter). Before I get carried away however, while listening I could objectively say it was better than before, but I had a slightly nagging doubt whether it was in fact more musical than before. I will have to get more acquainted with the sound, because the source of that feeling may reveal itself given more time - on the other hand it may have been because I had to turn the equipment off and on, and it needs to settle again. While it was quick and cheap to do, it was very fiddly and if I decide to follow this up further I will need to come to a better arrangement - perhaps involving shelves (that Corian and/or Maple may be useful after all) - taking components in and out of the rack at present would be quite painful and probably require three or four hands. I will report on this again when I have listened some more, and played around with some variations on the theme. Dekay and Sound_decisions, thanks for the idea and recommendations - the results are very intriguing at this point.

Redkiwi, Hard Maple butcher block is commonly used in the food industry in the USA, but I don't know availability in your native NZ. Maple, although relatively light (compared to MDF or particle board), is dense, and when large solid lumber strips are glued together, the combined grain structures, glue bonds and naturally occurring differences in density in the various pieces, makes for a unique shelf material. You can visit McMaster Carr via the internet to look for yourself. Go to: WWW:MCMASTER.COM. The shelf material to look up to begin with is #4882T52. It is two and a quarter inches thick, thirty inched deep, and six feet long. The price is $271.10 (US) and will make multiple shelves for that price. I had three six foot pieces of this material shipped to me, half way across the USA for about $75.00. It is also possible that they have distributors that deliver in NZ. Their site says "worldwide," so who knows? Long distance air shipment is not always out of the question, I once had Flying Tigers ship a four hundred pound, 8 foot long item from Holland (Amsterdam) to DFW Airport (Dallas) for only about $135.00. Granted it took almost two weeks, as I bought the most economical (stand by) service, but it certainly made it affordable. Obviously, if this or a similar item is available nearby, this would be ideal. In any case, I wish you the best, and hope that some of the information I have provided will give you alternatives.

I tried hanging equipment back in 95 with very good results.I even hung the cables.Wild looking to say the lest.But remember the celing moves more than the floor.I am working on a full multi-suspension system for CES,with the rack of stainless steel and the shelves of Corrian.Very George Jetsonish

Redkiwi: If you try hanging or suspending a component, please post your results. I just realized that in my setup I would have to include small line tighteners in order to level the platforms. Can't think of a better way.

I will look for some Maple butchers block Albert but I doubt I can get it here. I have seen such blocks made out of NZ native timbers, mainly in Rimu - how hard/soft is Maple? I have always heard good things about the Vibraplane, and the item at machinadynamica.com sounds like a similarly impressive product. Not only are they expensive, but the freight to NZ will add even more, and hence I need to take a biggish risk going down that path - but maybe a risk I will have to take. Suspending with string is novel (at least to me). That is something I can try very quickly and will have a go at it soon. I have also located some off-cuts of Corian, so will try those in the next week as well. Strangely I have never liked Vibrapods in any application, and similarly do not like the Sorbothane or Sorbogel stuff. A friend of mine says I am on the wrong track and insists that I have to damp the components directly first, and he may have a point. Are there any decent damping sheets available? I notice that the Sonic Frontiers stuff has damping pads inside, and I have found their gear to be relatively immune to the effects of cones, shelves etc.

check out www.machinadynamica.com it may help!

Interesting post Onhwy61. I have never thought of hanging/suspending components other than speakers (which I have done in the past with good results. I am in a postition to try this on my CD source and amplifier(s) that are in an enclosed piece of furniture with a cap or top that would support the cables. I would have to make holes in the top shelf to allow the support cables to run through to the bottom shelf, but other than proper alignment I do not see this as a problem. I am in the process of making templates for the top shelf (which with this feature would require two shelves - top and bottem) and may include this feature in the design. Guess that I would still use Vibrapods between the shelf and the gear. Has anyone "hung" their equipment, and is this a sound idea?

I'm currently using 1/2 thick glass slabs with either BDR cones or Vibrapods. I'm very satisfied with the combination, however, nothing is perfect and I don't believe any single material works in every situation. Consider this -- the best shelf material may be no material at all. Instead of placing components on a shelf, have the equipment suspended by small diameter cables. The energy storage capabilites of the suspending cables is practically zero, hence they should have negligible negative effect on the sound of the component. This method will work only if the suspended component has an extremely rigid case. A few years back a product like this was manufactured (can't remember by who), but I never had any first hand experience with it, but the concept does sound intriguing.

I have not heard it mentioned so far, and I have not tried yet, but maple butcher block is supposed to work great. I have tried the sandboxes, homemade and bought and the seismic sinks too, as well as a lot of diffrent footers. Hidiously expensive and heavy, though in terms of the improvment brought out a great deal is the Vibraplane. It is hands down the best isolation device I have ever used. It is sort of similar to the seismic sink exept that it has seperate chambers so that the air can move freely between the foot and the storage chanber, it also weighs 150lbs. The improvment it brings out is undeniably huge. You can order them factory direct from Sounds of Silence, Steve is a great guy to work with.

Hi Redkiwi. I just went through a similar situation in my own system. The sound that I found worked best for me was suggested by a turntable manufacturer. There is a food grade solid hard rock Maple butcher block shelf, that is factory fitted and glued. In the USA, McMaster Carr supplies this and is available in two and a half inch thick material. This comes in both 24" and 36" depth, and is available in a variety of lengths. I ordered mine in 6 foot, and had a local woodworker rip and sand this into the shelves for my system. I also had him laminate the front and top surfaces with gray Formica. The Formica seems to have no affect on the sound, it merely provided the finish I wanted. Please post your results, especially if you choose to try the suggestions provided by Caterham1700. I would like to know if it is worth considering starting over on my own project.

Only the Townshend are available in this part of the world, and they are not light compared with a light rigid shelf. I hear them as slowing things up. I am not familiar with the 855 so it is difficult to comment further on the problem you are experiencing. Currently, I use a very light and rigid chip-board shelf under my transport - the sort that is small chips of wood glued together and which includes a lot of air. It is much lighter and more rigid than particle board. It is clad with a very hard surface of compressed wood and so cones and spikes make no impression on it. I then use BDR #4 cones under the transport. In general I find that the right cone depends on the shelf being used, and that this combination works quite well. However, the reason for this post is that, although the result is fine, it is still a trade-off. It is definitely still more resonant than using a bladder product, but I prefer its more accurate pacing.

Hi Redkiwi. Found your opening remarks very interesting. I'm trying to struggle through with an old Rotel 855 while mulling new digital front end options, and am trying to reduce high frequency roughness now apparent in my new otherwise-highly-resolving system. I'm using an air bladder suspension under the CDP, and am not aware of any change in PRT, as you suggest is possible/probable. Before I try reusing my old cones, or sorbothane feet, etc., and struggle to hear changes, do any of you guys have suggestions for improving CDP isolation (esp HF grain) WITHOUT compromising rhythm?...and indeed is it true that non-rigid setups (like airbladders) risk timing anomolies? Thanks. Ernie

Thanks for your kind words and advice Cornfedboy. The sad reality is that the only racks/shelves that reach these shores are of the basic variety, with the exceptions being Mana Acoustics and Townshend (and I own examples of each - most of which are in "the closet"). There are importers of cones from BDR and Polycrystal, but noone brings in the shelves. And noone brings in Zoethecus so far as I know. For example, I wanted two amp stands for my monoblocks and called around all of the better dealers in New Zealand - or at least that was what I intended, but truthfully I gave up after the first six because I was sick of explaining what an amp stand was... if you get the picture. In the end I did what every other audiofile does here - got them made. I would indeed prefer to just try and buy, but I am much less keen on importing one, only for it to end up in "the closet" - I am sure we all have one of those. Having come to this point with shelves (ie. needing to splash out and hope), I posted here hoping optimistically that there might be a concensus as to the best shelf material/product so that the chances of success could be improved. I think Albert Porter has done a lot of experimenting with vibration control, but his posts indicate he agrees pretty much with Dekay ie. that there is no universal answer. Perhaps I am being too optimistic. Anyway, I will investigate Zoethecus from here, and perhaps that is what I may end up trying. I hope you do get to New Zealand one day. I have known many Americans who have come here - some stay, some get homesick and go back. The ones that stay are adamant there is no place better. Personally I don't think it is a place to spend all one's life, but it is a wonderful place when you are a kid, when you have kids, and when you want to live a quieter life later on. In between, most of us Kiwis travel a lot and I have been forunate to visit your lucky country a number of times - picking up audio gear every time of course.

redwiki: you've already received some excellent advice, which i have no intention of undercutting or disparaging. in my younger days, i, as you, had a compulsion to create for myself bits of my audio system that seemed within my means, ability and understanding to produce. among these were shelves and racks for my equipment. as i've grown older and, unexpectantly, have less leisure time than i did in my younger days, i've relied less on my own abilities and more on those i judge as "experts" in whatever field is necessary to my ends at that juncture. hence, i've looked toward commercial developers of racks and the shelves that go with them. the company i've come to trust in for this aspect of my audio system is zoethecus. their shelves are top-notch and manifest the qualities your last 2 posters recommend. they are not, however, what i would call lightweight (actually, not figuretively). you can, tho, come close to replicating them. as with others, they comprise a "sandwich" of several materials. the idea is to use different "dead" materials of varying intrinsic resonances, all of very low hz., to cancel out the resonace of any single layer. i do not know the actual formulation used by zoethecus in their "z-slabs," but suspect that they include alternating layers of mdf-like pieces and absortive layers resembling sorbothane. these layers, topped off with a metallic uppermost platform that can withstand the weight of heavy "spiked" equipment, are wrapped with what i expect is a heat-activated edging that makes a tidy, handsome package. you can find these shelves alone on audiogon every so often. you might wish to buy one (or "borrow" a copy from your nz dealer) to investigate their properties. i've found nothing better for all equipment, from tt's to amps (my main amp weighs in at > 150#). in any event, i wish you good hunting. your posts are invariably among the most useful, interesting and polite among all of the "regulars" on audiogon. i hope someday to visit your island that, in my mind's eye, must be a heaven on earth. cheers.

My comment that some resonance may improve the sound quality was based on the assumption that many products have been "voiced" when under the influence of a certain amount of resonance. To remove all or more realistically much/most of the resonance may in fact degrade the sound. This is hypothetical.

Wow, those last two posts sound on the right track to me - I will check them out. Many thanks to all that posted.

Redkiwi, Try checking out products that meet exactly the rigid/low mass/fast dissipation isolation philosophies you are looking for at www.neuanceaudio.com(my company) or www.cetech.co.uk(a respected competitor)as well as the RATA/Torlyte of Russ Andrews,UK(the man that wrote the book on low mass supports).Each has an unique manner in which resonant energies are controlled but all do so very effectively without destroying pace, rhythm and timing.Bladder devices have the disadvantage of poor location in space on a microscopic level and thus present new problems for turntables and digital reading mechanisms such as transports and cdp's. Cetech uses a carbon graphite/aluminum honeycomb composite for fast evacuation, while Torlyte is of low energy storage and internally damped.Neuance is a low mass decoupled monocoque of via constrained layer principles with a high ceramic content and internally damped.I firmly believe that any one of these products will satisfy your needs and tastes. Best, Ken Lyon GreaterRanges/Neuance

If you can get two pieces of perspex made up similar to a double-glazed window unit with the air gap evacuated and replaced with an inert gas, this may give you the rigidity and damping/energy release properties you are looking for. If you use Corian then in our experience there is no need to "layer" this material, but use some other material to frame the Corian to vary the damping/energy release properties (you can experiment with large section rubber "o-ring" material set into grooves in the Corian) but this gets expensive! Richard, vantageaudio.com

My thoughts are that you want a shelf to be dead rather than resonant, but that it released its energy very quickly. Dead means the least effect on tonality - ie. no resonant peaks. Releasing its energy quickly means no perceived slowing of the music. I suspect that engineering-wise, getting these two things at the same time is very difficult, if not impossible, hence the difficulty I am having getting something that sounds right. I concur with all of the above posts to some degree, but think the most fruitful approach may be to get a shelf that releases its energy quickly - ie. is light and rigid - and use some form of damping that does not add weight, and hence energy storage. This aligns with the suggestions from Dekay and Vantageaudio. So I will work on that avenue I think. Perhaps two layers of 6mm perspex bonded together - but perpexes rigidity:weight ratio is not great. Perhaps Corian is worth investigating. Any ideas? Of course Dekay is right that the exact answer each time is equipment dependent, but the general strategy ought to have some consistency in my mind. Thanks for the ideas so far.

The same shelf and devices will not have the same sound on different components. There are a lot of variables such as stock feet, cones, pods , etc., as well as the shelf (or layers of shelves and isolators) and also included into the equation is the specific component that you are isolating. My layer for my CD player is 3/4" Persimmon (the cabinet itself), Vibrapods, 3/4" MDF, stock feet on CAL Icon II and it sounds great. The cabinet construction as well as it's location from the speakers must influence the sound as well. I have also used 1" thick marble in the equation and found it to be too bright though a friend preferred that sound, probably because the setup sounded more like his Marantz 67 Mk II player which was his reference and taste. I once added stick on floor tiles to the bottom of an extended "plastic" shelf with good results (to me) and the tiles were not visible unless you were laying on the floor. I think that it is all up to individual tailoring for each component, and when taking into consideration individual taste, that there is no answer. I can even imagine in some cases that added resonance may even improve the sound depending on what the sound is that one is looking for.

I JUST BUILT A TURNTABLE STAND WITH INCREDIBLE RESULTS. HERE GOES.. 1- 18X18X2 PATIO BLOCK ON THE FLOOR.5-16X16X8 CONCRETE CHIMNEY BLOCKS. CENTER 1 CHIMNEY BLOCK ON THE 18X18 AND ATTACH WITH LIQUID NAILS.USE THE BLUE LABEL HEAVY DUTY CONSTRUCTION IN CAULKING GUN.THEN STACK THE REMAINING 4 CHIMNEY BLOCKS ON TOP USING THE LIQIUD NAILS BETWEEN EACH BLOCK. VERY IMPORTANT!!! MAKE SURE BOTH SIDES OF THE CHIMNEY BLOCKS HAVE NO BURRS OR HIGH SPOTS. ON MINE ONE SIDE NEEDED TO BE HIT ALL AROUND WITH A HAMMER TO GET A FLAT SURFACE.SINCE THE CHIMNEY BLOCKS HAVE AN 8X8 HOLLOW IN THE CENTER YOU CAN FILL THE WHOLE THING WITH SAND IF YOU LIKE.I THEN USED ANOTHER 18X18X2 PATIO BLOCK FOR THE TOP!YOU CAN THEN FINISH THE SIDES OF THE CHIMNEY BLOCK WITH CERAMIC TILE OR WHATEVER.THE WHOLE THING COST ME $52 AND I BUILT IT IN LESS TIME THAN IT TOOK ME TO TYPE THIS!! I PUT MY BASIS TURNTALE ON TOP AND COULD NOT STOP LAUGHING.THIS IS TRULY A ROCK SOLID STAND ANN IT IS PERPECTLY LEVEL. ENYOY!!

I've tried just about everything too, I use solid 1" thick oak, seems to give me the best results so far. The Brightstar Audio stuff seems to sound better in my system than the Seismic sinks do. As always, your mileage may vary...Jeff

Try and get two thin pieces (ie two pieces that combined make up the dimensions of one solid piece) of whatever material you select and bond them with a suitable adhesive/sealant (Dow Corning RTV or clear silicone), aligning the two pieces at 90 deg to each other. Use a spiral pattern of adhesive/sealant starting from the center, this way you will get an air/sealant cushion. Good luck! Richard.