Canvas quality

A brief heads-up for the tailors out there.

I've been seeing some of what I consider to be inferior quality canvas cropping up and if you're not accustomed to looking for this, you may not notice it. So next time you are shopping canvas, bring a loupe or a magnifying glass.

The canvas is woven from yarns which are twisted; the animal hair will be mixed with wool, cotton, and maybe other fibers and twisted to make a yarn. The springiness of the animal hair, when twisted, will cause the yarn, and thus the canvas, to want to curl. To counter this, quality canvas is woven using what we call S and Z yarns (the shape of the letters gives you an idea of what this means); the yarn is twisted in one direction to create the S yarn, and the Z yarn is made by twisting in the opposite direction. This balances the canvas, but is more expensive than making, stocking, and using only one kind of yarn so certain weavers, mainly located in parts of the world which are new to weaving canvas (ahem) are cheating by only using one type of yarn. If you look at the photo above you can see a bit of the chevron which indicates that the weave is balanced.

If you're making up some heavy tweed or English flannel it probably won't matter, but if you're cutting something softer or lighter I would definitely take a closer look at the canvas you are buying.

Huntsman, Ripped and smoothed, part two

Since my last post, a reader emailed me to tell me that the initials of the cutter would be found on the label, those initials being "TH". I assume this must mean Terry Haste, former MD of Huntsman, but was he there in 2000? I'll have to go back and read the book again. Or do some more homework. But in any case, we're not evaluating the cutting here, only the sewing. It's just fun to know a bit about the suit.

Another thing I have been thinking about; there was a mass exodus from Huntsman during that period so anything not quite up to par may be a result of them losing their best people. Or it may not. More to think about. Meanwhile, let's get ripping.

The finishing inside is quite neat; the felling is so neatly done that you have to look close to see it in spots. There is no pleat in the bottom of the lining and to tell the truth, I'm not sure one is really needed, looking at this suit. Hmm.



The neck is neatly done as well. The facing/collar seam has actually been opened and drawn on the whole way, rather than clipping and working the facing over the top collar around the neck which creates some bulk. Extra work, but I like it and will use it.



Two pockets on the left, one on the right, neither with any kind of closure.



While I was surfing Ebay listings I found the pockets odd, as they seemed to have been worked onto the lining, the way manufactured suits are done, and like this one-



I prefer to set the pockets onto a cloth facing which will carry the weight better. It also makes relining much easier- if the pocket has been cut through the lining and goes onto the facing, as in the picture, it's a bit of a beast to reline neatly since you have to work the new pocket exactly over the hole in the facing- margins of error no more than 1/32" or so. Not a job for your average alterations tailor. Well, I should have known better. I didn't notice that the lining was a bit crooked at the pocket- this would have given it away. The pocket is not actually worked through the lining (though the smaller one just below it is)- the pocket is worked through a cloth facing and the lining is felled to conceal the work. It is common for most Italian tailors to leave this facing exposed in a bit of a decorative manner, like in this shot from a Zegna Couture coat (notice also the hand-felled lining with a backstitch instead of a felling stitch- we also saw this on the Brioni dinner jacket I dissected)


but Huntsman has chosen to conceal it entirely; naturally, it has all been done by hand.







Now on to some of what gives the suit its shape.

The bridle has been taped with lining instead of the cotton tape which is more common, and the lapel has been sparingly pad stitched. Remember that a lot of the stitching poked through- it is probable that the padding was done by an apprentice. I personally feel that denser pad stitching on the lapel will give a richer roll, and the results that I get would seem to support that theory, but I haven't properly tested it yet so I can't state it as fact, only as received wisdom. Which can sometimes be false. The undercollar (not shown) has been padded by hand, something that I HAVE tested extensively and found that it is not necessary to do by hand- the machine does just as adequate a job, in my opinion.



The canvas front is a lovely, dense hair canvas which has not been singed. I need to find out where this canvas comes from. Huntsman has a reputation for being a firm coat so I expected to find more in the chest. Instead, there is only one piece of haircloth, on a partial bias, and a rather thick piece of chest felt.



The direction of the hairline is surprising since this won't support the upper shoulder as well and there is no additional shoulder support piece.

The "vees" or shoulder cuts are left open rather than being fixed. I'm not sure how I feel about this yet. Readers who survived my pagoda shoulders will remember that these vees are opened and fixed in place to provide the shape to the shoulder and without them being fixed in place they can shift around so the shoulder shape will not be constant. This could possibly be a good thing since there is perhaps more movement allowed, but I think this would be very dangerous in lighter cloth- the hollow created by the open vees would make a very visible dent in the shoulder which would be very obvious in light cloth- the only thing covering them is the piece of striped lining on the top and bottom. I need to think about this for a while.



There is a moderate amount of wadding in the shoulder and two small pieces of collar linen. This is certainly a lot less than either Oxxford or Brioni put in their shoulders, again a surprise, since I expected a firmer shoulder from Huntsman.



Here is the pad removed



In the sleeve head, FOUR layers of heavy felt. This is a lot, even for me who likes a good, healthy rope. The effect is not bad, but its a bit chunky and lumpy since it's put in straight rather than shaped, which is now the norm (sometimes tradition should give way to innovation).


The back of the scye (as well as the underarm) has been padded with cotton and chest felt. This was common in the days of body coats (Sator will love it) but I haven't seen much of it lately; I'm not sure if this is common practice for them or if this gentleman had very hunched shoulders which needed some help. Judging by the, um, proportions of the suit, I am guessing the latter.



EDIT
I went plowing along without mentioning that the shoulder seam was sewn by hand and that the sleeve has been sewn in by hand. More on that later.


So far so good, the only major surprises being how LITTLE stuff there was inside. I was expecting much more. As some of my other garments begin to arrive, we'll have a better basis of comparison, though. Need to find an A&S......... But before that, the trousers, which seem to have changed little in the last 100 years.

Constructing the pagoda shoulder, part 1

Preparing the canvas

The shoulder is arguably the most discussed and the most controversial element of a tailored jacket; the whole garment hangs from the shoulder, and the shoulder receives much of the strain from movement. There are perhaps as many ways of constructing, or “expressing” a shoulder as there are tailors, but there area few general shapes, or silhouettes. Some of the names have become muddled throughout the years, but I am going to chiefly examine one of them. Known as the “spalla insellata” because of its resemblance to a horse saddle, it is also known as a “spalla incavata” (concave shoulder) or pagoda shoulder, for obvious reasons. In the 30s the Apparel Arts referred to a natural shoulder as having a concave line, not the round one most often referred to today as a natural shoulder. To understand why, we look at the human shoulder (well, a plastic one. This is a family show, after all)



It is clear that the shoulder line between the trapezius and the acromion forms a concave curve; the front of the shoulder area is more complex due to the curvature of the clavicle and the protrusion of the top of the humerus, the ball which fits into the socket of the shoulder joint. It is this joint which is all important here, because this is where the whole arm moves.

Observing all these hills and valleys we can appreciate the complexity of the shape of the front of the shoulder and perhaps understand why the shape and location of the shoulder seam is crucial to this area. The closer the seam comes to all these curves, the easier it will be to accurately contour them. It was once fashionable to slant the shoulder seam backward which was a way of introducing some fullness over the shoulder blade, much like the vault dart or shoulder dart; the yoke of the shirt conceals one of these darts, and on heavier figures they are often included on the scye or the shoulder seam. In my posting on the one piece back, we saw how the gentle curve of the blade could be accounted for with a bit of stretching, shrinking, and a bit of wadding (shoulder pad) to fill in the hollow, rendering a backward slant unnecessary. It can also have some unfortunate results should the garment encounter humidity and the fullness puff up, unsupported by the shoulder.



For these, and a few other reasons, the backward slant is generally only seen on garments from a few houses who like to maintain tradition and who may not have taken the time to fully examine its implications, feeling secure in the received wisdom without questioning it.

Going back to this complex formation, the astute pattern maker will recognize that, not only is the shoulder seam location important, but another seam or dart, or perhaps several, would be appropriate, even necessary. Throughout published literature, tailors have illustrated ways in which these additional shaping mechanisms could be introduced to the canvas, with three principle variations. These two seem the most logical;



Darts like this will introduce shortness in some areas and length in others, shortness being indicated by the red arrow, and length being indicated by the blue arrows. The first example makes sense, but it introduces some length to the neck area, which, in my opinion, is asking for trouble as extra length may give trouble in the form of a collar lifting from the neck. It also introduces a great weakness in the canvas along that seam by cutting through the hair line. Any shaping of the outer canvas must be carried through all layers and it may be difficult to maintain this shape while providing support.

The second example is also valid, but introduces fullness to the chest, which is fine in a draped style but perhaps not so for cleaner cut chests. It also creates weakness in this area, though it would be easier to make up for this with haircloth.

Another look at the shoulder anatomy suggests that neither of these is the best location of the cuts or seams- the ideal location and slant are indicated by the red and green lines.



The third example is the one generally seen today, though the illustration is somewhat incomplete and inaccurate.



The cuts, or vees, should correspond with the hollow indicated by the red line in the photo of mannequin just above. These will throw fullness over the shoulder point and create the little hollow and bulge, but as we saw in the soft shouldered coat, this alone will not create the pagoda shape. Most modern coats have these vees to give some forward pitch without creating fullness at the neck or over the chest. To get the pagoda shape, a few more cuts and manipulations are required.

As we look at my way of constructing this shoulder, rather than a slavish adherence to any formulae I may suggest, try to understand the concept behind it so you can adjust to suit your taste or figure, or to work out a totally different shape of your own.

Start with a cut about 3” deep, 1/3” of the way across the shoulder, with an inclination similar to the little red line on the mannequin. Using a bias piece of canvas, open this cut 5/8” to ¾”. When I cut my bias strip, I make a step the width of the desired opening to act as a guide. This will stick out the top but I will trim it off after.



This will provide forward pitch but to get the complete shape, a little stretching will be required along the armsye. Normally I leave it off to the end of the canvas construction, but to give an idea of the final shape, I have done it first.



Observe the hairline of the canvas, how it bows upward. The combination of the cut and the stretching create the same effect as the two cuts illustrated in the first diagram, combined, creating length along the shoulder and the armscye, but keeping both the chest and the neck edge short.

To cut the haircloth piece, strike a line along the straight grain of your canvas about 10” long to measure the inclination of the roll line. Let’s say that the short arm of this angle is 3”. Draw a straight line down the center of the haircloth long enough to extend from the shoulder to the desired length of the chest piece (another subject entirely), use the measurement obtained from the canvas, 10” by 3”, to draw the angle of the roll line, and cut along this line. Place the hair cloth pieces together along the roll lines, line up the canvas roll line with the cut edge and draw in the rough shape of the shoulder seam and the first cut location as well as the scye shape, then rough cut the chest piece, which will later be trimmed to fit the canvas more accurately.


It is vital that all shaping remain constant through all layers, but if all the cuts occur in the same place, there will be structural weakness and thickness which is not good. Treating the cuts as darts, which can be transferred anywhere around their apex (a subject covered at great length in regular patternmaking courses so I won’t go too much into here) the cut of the haircloth should be pivoted slightly so as not to occur in the exact same place as the canvas cut. A cut is then also required is in the scye area as show, since the stretching we will do on the canvas is difficult to achieve properly in haircloth. We will open the cut of the haircloth in the same manner as the canvas, but leave the armscye cut open. For now.



Using the haircloth piece as a guide, we will cut what we call the perk, or shoulder support piece, also out of haircloth. This piece is very important to a clean shoulder and should not be cut on the straight grain but on the bias. Some makers prefer a perfect bias, I prefer a partial bias; this bias not only supports the hollow created in the shoulder shown by the diagonal green line above, but also makes stretching easier. To get my grain line, I measure an angle 3” long and then 1” toward the neck point, then join the points for my new grain line. Use the chest piece as a pattern, using this new grain line. The perk should stop about 1” below the cut on the chest piece, and should go straight across. In the image I have drawn in the hairline, not the grainline, so you can see how the perk will support the area.



Trim the perk so that it is about 3/8” inside the roll line and shoulder edges to avoid thick ridges. Halfway down the roll line portion of the perk, make a horizontal cut which will almost reach the shoulder cut of the chest piece. Instead of using a bias strip on this portion, we will use the chest piece as an anchor to open up the cut, but only 3/8” this time. Notice how all the cuts radiate around the same point? This is extremely important to get everything to gel together correctly when finished. Notice also how there is some weakness which creates a ripple of length along the roll line? This will be privoted toward the shoulder soon.



Now baste the chest piece on to the canvas, being sure to keep the roll line straight, which will force all the length created by the cuts over to the shoulder point. VERY IMPORTANT.


A straight cut piece of canvas should be used to cover the front edge of the chest piece, where the hair often pokes through and bothers the wearer; this will also help to stabilize the roll line and keep the length over the shoulder. You may want to use a strip of pocketing to cover the lower edge of the chest piece to prevent poking through as well.



Despite the staggered cuts, the canvas should be coming together like this, with all the shaping being sent toward the shoulder point.



Place a piece of felt, flannel or domette as you like, and pad stitch the canvas the way you normally would (another subject) but leaving the shoulder free. Make sure to keep the area around the neck straight, and pad this area a little more heavily to stabilize that weakness in the perk we saw earlier. It is now that we should use the iron to stretch the armsye area; the canvas will stretch, the perk will stretch a bit but will also shift, and the scye cut in the chest piece will open up, but still be loose. Commercial canvas front makers like Interforme have big pressing machines which have shapes which support the chest while stretching the shoulder all in one shot. Once we have basted the canvas to the front and inserted the shoulder pad, we will then pad stitch through all those layers- the cut will be open but supported by the perk underneath which is longer for that reason and the shape will be fairly permanent and won’t “break down” with time. Ahem.


Now the canvas goes in water overnight (never never never skip this step, no matter what canvas makers say about being preshrunk) and while it is soaking and drying we can make the shoulder pad. The canvas will get a torough pressing when it dries out, and another little stretch in the scye. And now I wait for the flannel to arrive so I can cut the cloth. EDIT** CLOTH JUST CAME!*** But while we wait, go back to the posting on the canvas for the soft shoulder- one pice of canvas, one little cut. That's it. Compare the two, knowing we have only completed the first of several steps; some people contend that the soft shoulder is the greater art form. I beg to differ.

And meanwhile an amusing little article about Bill Fioravanti, who also favors the concave shoulder. I have never met him, but Chris Despos has only good words for him, and if I am not mistaken, called him one of the best fitters he knows. High praise indeed!

In response to Dukemati's question, my canvas and haircloth are both from Rovagnati Vincenzo spa; the canvas is the lana 180g (sometimes I use 200g) and the chest piece is made from 21 pick haircloth- nice, light and resilient. If you are in North America, their distributor is Interforme Interlinings, esturino at interforme dot com; Emilio Rovagnati will be here next week, I will ask him about worldwide distributors.

What, exactly, is pad stitching?

We hear the terms `pad stitching` and `padded lapels` frequently but not everybody understands the significance.

One of the important principles of tailoring, and which distinguishes tailoring from ordinary `flat`sewing, has to do with the fact that concentric arcs or circles become longer as they radiate from the centre. In easier terms, picture the starting line of a circular race track- the starting point of each lane is in a different spot because the outer lanes are longer than the inner lanes as they go around a bend. Tailored clothing consists of many layers of fabric that bend around each other in order to create shape. The layers of fabric and canvas in a lapel, for example, are like the lanes of a race track; the outer ones are necessarily longer than the outer ones; unlike a race track, whose lanes are staggered to maintain a consistent length, the tailor will cut the layers of fabric to different lengths so that when bent into shape they all meet together at the seam.

We have seen in a previous post that canvas contains horse hair which gives it a lofty roll. This roll contributes to the roll of the lapel but is not solely responsible for it. Once the tailor has basted the canvas and the cloth together, he will roll the lapel over his hand; the canvas being the 'outer lane' of the race track is thus a little longer than the 'inner lane' of fabric. He will sew a row of pad stitches parallel to the roll line, just catching the cloth underneath so the stitch is not visible under the lapel. At the end of the row he will roll the lapel over his hand again and sew another row of stitches, and continue until the whole lapel has been done. When it is finished, the canvas and cloth now roll as one; the canvas lends loft to the roll, but the additional length of canvas has been stitched permanently in place and so the two want to maintain the shape they were given. Notice in the photo above that the lapel stays curled- it does not fall flat to the table as it would have done had it not been pad stitched. Not only does this pad stitching create a beautiful roll in the lapel, a sign of a good-quality garment, but also helps the lapel to hug the chest tightly and not pull away, another indicator of quality.

Machines now exist which can do this job automatically, using special sensors so that the machine knows when it has reached the end of the lapel, and it will return to its starting point, advance a row, and sew the next one. The machines are quite fabulous, but also fabulously expensive; a pair (one for the right, one for the left) can cost upwards of $80,000 so are within the reach of mass-producers only. The machine takes only a few minutes to do a garment, and can do the two sides simultaneously. The garment shown above has a wide, peak lapel and took me about half an hour per lapel to do. This is a photo of a pad-stitch machine.



This is a half-canvas garment which has been pad stitched by machine



This is the same garment, showing the degree of roll.



The machine stitch is very consistent, however I may want to roll certain parts of a lapel a little more tighly, such as the very tip of a peak lapel, which can not be done by machine but can easily be done by hand. You will notice that the garment which has been stitched my machine does not roll as much as the garment at the top of the page, which was paddded by hand. This does not mean that it will not roll as well; the canvas will still provide a great deal of loft to the roll. It may not, however, hug the chest quite as tightly as a hand-padded lapel will. By varying the amount of canvas worked over the cloth and by stitching closer together or further apart I can vary the degree of roll in the lapel. This is a very fine point and I am not sure that it makes a very great difference in a finished, pressed garment. My unsubstantiated opinion is that it is better done by hand, but to really investigate further I would have to make two identical suits using identical trims and machine pad one of them, and do the other by hand. Maybe when I have retired and have nothing better to do.

The same principles apply when making the canvas front- on the CANVAS page you see a full front which has been pad stitched by hand. Again, there are several layers of material which have been worked into a particular shape and stitched together so that they will hold this shape. As I stated on that page, while I enjoy making my own fronts, I do not think there are any substantial benefits to doing them by hand over doing them by machine. I have done both ways for a long time; one is generally hoping for softness in the chest rather than stiffness and the machine does an admirable job. The amount of shaping and density of stitching can be controlled by the operator, unlike when machine-padding the lapels so I think it really is a toss-up on this issue.

It is important to note that if one were to press the lapel or chest flat you would lose almost all of the shaping which is why we generally advise against having things pressed by dry cleaners who may not be trained in pressing tailored garments (an art and a post unto itself).

So to recap, pad stitching is used to hold several layers of fabric together permanenlty in a particular shape. It gives the chest its shape, and gives the lapel its roll, helping to prevent it from pulling away from the chest- a fused lapel will never roll like a padded lapel and makes judging a garment`s quality easier. While I think it is a little better to pad a lapel by hand, I have seen no real proof of it in a finished garment; this is a matter of great contention and I do not pretend to have the answer.