Experiments with design and method: side-fastening shoes with trompe l’oeil toes

It’s been a while since I’ve posted a photo-essay showing some of my methods. So I thought I’d chronicle the making of a pair in which I experimented with various aspects of design and technique. This is the end result:

A pair of wide-toed, side-fastening shoes in blue leather with red highlghts — The toe shape of these shoes is exactly the same as other lasted pairs I’ve made, but they may not look as wide, thanks to the strong red patterns, which draw the eye.

Design

I was trying several new things with this design.

De-emphasizing toe width

To let the toes move naturally, the toe area of the shoe should be as wide as, or even wider than the ball of the foot. However, virtually all shoes that one can buy (for many decades now!) have tapering toes: fashion and shoemakers’ convenience trumps human anatomy and sense.

How can I make my toe shape seem conventionally tapering, while still being unconventionally wide? I decided to try a combination of color and shapes to draw the eye to the center-line of the foot. To ensure this “trompe l’oeil” pattern remained more-or-less the same no matter how the foot was moving, I made the shoes fasten at the sides rather than down the center line of the forefoot as in my previous lasted pairs.

Side fastening

I have never before made footwear that has a tongue and eyelets on the side of the foot, rather than at the front. I decided to make the shoes lace up on the outside of each foot (rather than at the instep), and to have a highlight “collar” around the shoe just under the ankle, that would flow downwards to the sole either side of the tongue, again drawing the eye away from the toe area. This collar is not padded: it’s simply a visual highlight.

No reinforcing at toes and heels

In previous lasted shoes and boots, I have used vegetable-tanned leather to provide reinforcement in the toe area and around the heel. Nailed to the last while wet, this stretched leather dries to form a stiff cap over the toes and the cup of the heel. However, for the current pair, I wanted to de-emphasize the wide toe shape, so I decided to forego the extra layer of toe-cap reinforcement and simply rely on the lining (3oz chrome-tanned cow-hide) to provide durability. We will see how well this wears, compared with the footwear with stiff toes!

Other design elements

None of my previous pairs has had metal eyelets, so I thought I would try using some for the first time.
As in previous shoes, I wanted to use a discreet welt to fasten upper and lining to sole.
As previously, I wanted to try to keep the amount of leather wastage to a minimum, so I tried to make pattern pieces be parsimonious as possible for the shapes I needed to achieve. This meant, for example, cutting the toe pattern as three separate shapes to be sewn into a converging pattern, rather than as a single three-pronged shape. I also cut the midsole from the same piece used to cut the welt (photos below).

The pattern

Having conceived the design, I started drafting the pattern on the same last that I would later use to assemble the shoe.

Wooden last covered with masking tape on which a prototype shoe design is inked in colored pen — Having covered the last in masking tape in such a way that it would be possible to pull off in a single piece (in theory!), I drew my prototype design onto it using pencil and colored pens.

I used pens to color each design element that I thought would form a different part of the pattern. When satisfied with how this appeared on the last, I cut through the masking tape with a scalpel and removed it from the last in two pieces:

a small one used to form the pattern for the tongue
and a big piece on which the patterns for the remainder of the upper (and lining) are based.

Masking tape with a shoe design drawn on it being pulled off a wooden last — First, I trimmed away the tape above the orange line marking the top of the shoe, then I cut off all the tape that was attached on the sole. This left only the tape showing the upper design, ready to be peeled carefully off the last.

The masking tape would be impossible to flatten onto a piece of card without major wrinkling and distortion at the toe and heel, so I used scissors to cut slits into it. This enabled it to be flattened, while preserving the shape of the top of the shoe as faithfully as possible.

A masking-tape shoe proto-pattern lying alongside a pair of scissors — To be able to flatten the toe and heel area onto card, I needed to cut slits so that the masking tape in these areas could flatten in a fan shape. This shape can be gathered together when the shoe is made in leather.

Having cut these slits, I pressed the tape shape onto a piece of thin card (making sure that the straight lines down the front of the foot and down the heel of the foot remained as straight as possible).

Part of the pattern-making process: pressing tape removed from the last onto a piece of thin card — Note the distortion around the lower edge of the tape: it is fanned out at the toe and heel, and wrinkled together in other areas. If there’s too much wrinkling, the tape can be carefully peeled off the card for another attempt at flattening it.

Having flattened the tape with as little distortion as I could manage, I drew around it onto the card to form the pattern for most of the upper. I made sure to mark the center lines at toe and heel, as well as the points where the various colored highlights intersected each other. When I had the pattern for the main body of the upper marked out, I then used scissors to cut out the masking tape some more to make separate patterns for all the pieces of the highlighted parts (intended to be red leather).

Different colored pieces of masking tape removed from a last being used to make a shoe pattern by drawing around them onto thin card — Having drawn the pattern for the main body of the upper, and added a 20mm lasting margin (dotted line), I cut out and drew around the pattern for the parts of the design that would eventually end up being made from red leather.

The cloth prototype shows up design flaws

Once I had the various pieces for the upper pattern, I made a prototype shoe in heavy cotton cloth, with different colored cloth to stand in for the various “tapering toe” highlights.

A cream cotton prototype shoe with green and red highlights — This shoe is a prototype, to check fit, and check that the design is working as intended. The cloth is simply taped (using masking tape) to a stiff cardboard sole, to provide a “quick and easy” simulacrum of what the eventual leather shoe will look and feel like.

This prototype showed me that the way I had envisaged the colored pattern overlapping from the main part of the upper onto the tongue of the shoe would not work on a moving foot, and would also be obscured by lacing. So I decided to omit it, and to re-work the pattern for the tongue slightly.

Having modified the pattern for the upper based on what I had learned from the cloth prototype, I made a pattern for the lining. This was simply the same as the pattern for the upper, cut slightly oversize at the top of the shoe (so that it could be trimmed back after being sewn to the upper: much easier than making it exactly to fit, then having to line everything up precisely while sewing). For both upper and lining, I needed to mark the lines where the red leather would be sewn; I cut holes for these lines into the card using a scalpel, and used a narrow punch to mark their beginnings and ends.

Pieces of card, together with scissors, pens and scalpel

Since I wanted to use the tozafoot method of welting, I also made patterns for the insole, the midsole and the welt, based on a tracing of the base of the last.

Using a card pattern to mark stitch positions on undyed leather insoles. — Leather that will become insoles being marked with the line showing where the midsoles will need to be sewn on. The right-hand piece of card contains the pattern for the sewing line: the aperture in this card exactly matches the aperture for the midsole pattern (left-hand piece of card)

Making the shoe in leather

After making the final pattern in card, the next step was to lay the card onto leather to mark where to cut out the various pieces to assemble into a shoe.

Drawing around pieces of card on thick leather — Patterns for the welt (larger piece of card) and midsole (smaller piece of card) marked onto 6oz vegetable tanned, undyed cow hide. Note that the midsole fits almost exactly inside the welt, minimizing the amount of leather wastage.

Pieces of red leather lying alongside the card patterns used to mark out their edges and the scissors used to cut them out — The red leather highlights for both shoes were marked out from the pattern using silver pen, then cut out, ready to be stitched to the rest of the upper.

Having cut the leather pieces out, I marked stitch positions in pen.

Two pieces of undyed cowhid lying on a mat alongside the knife and scissors used to cut them out — After the pieces of welt, insole and mid sole are cut out, and stitch positions marked in pen, punches are used to make stitching holes (in the welt) or slits (in the midsole and insole).

I used biro to mark the sole pieces, since no one will see these in the assembled shoe, but silver pen (which can be removed using an eraser or water) to mark the upper and lining.

Four pieces of leather with drying white glue on parts of their surfaces, lying on a piece of newspaper next to the glue bottle and applicator — Before punching stitch positions, I glued the midsole to the insole (taking care to keep the glue within the stitching lines) using water-based glue. The glue alone would not be enough to hold the leather: the glue is just to form a temporary bond to prevent the pieces moving relative to one another while being stitched.

I then punched stitching holes…

A rubber mallet lying next to a white nylon board on which two leather pieces are lying, one of which has a chisel protruding from it. — The mid sole is stuck to the insole, and a chisel is used to punch stitch positions.

… and skived where necessary, to thin the leather where an overlap would otherwise result in too much thickness (for example, where one piece of red leather overlaid another along the edge of the upper).

Two pieces of blue leather lying next to the skiving tool — I skived the blue leather of the uppers along its top edge, where it would attach to the red leather.

A ring-shaped piece of leather lying on a glass board next to an implement containing a razor blade used for skiving — The interior of the welt is skived, since this will be overlapped by the combined upper and lining around the perimeter of the foot.

I also dyed the top surfaces of the welts to their final color at this stage, because dying them after they’re attached to the upper carries too much risk of dye staining the upper as well as the welt.

Two black hoops of leather with stitching holes punched in them — The dyed welts, with stitching holes punched, ready to attach.

Having cut out and skived the pieces of the upper, I sewed them together. As with the mid-sole and insole connection, I used a little water-based glue to hold the pieces of leather together temporarily until I could stitch them permanently using (0.8mm waxed nylon) thread.

Red and blue pieces of leather with silver lines marked on them showing where they need to be assembled together — It was rather fiddly to sew the red pieces to the main body of the upper. Before starting to assemble everything, I marked very carefully where each piece needed to fit (according to the pattern) using a silver pen.

I then sewed the linings onto the uppers, before deploying my new tool: an eyelet setter! I used eyelets that had a 1/4″ stalk and that fit a 3/16″ hole (punched with a round punch). This size fit my design perfectly. The eyelet stalks had to pierce and turn back on either three or four layers of 2-3oz leather (one or two layers of red detailing, plus the blue main body of the upper, plus the lining): approximately 8oz leather altogether.

An eyelet setter and a bag of eyelets lying next to a partly assembled shoe upper — Yay, we have eyelets! The setting tool is remarkably simple: an anvil on which the eyelet sits once it has been poked through the hole in the leather, and a solid cylindrical punch that is bashed with a mallet to flatten the end of the eyelet against the inside of the leather.

After installing the eyelets, I sewed each tongue onto the main body of the upper.

Two pieces of leather clipped together with a large bulldog clip; a silver pen lies alongside — Once the positions of leather are aligned in place (along the lines marked with silver pen), bulldog clips are extremely useful for keeping them in place while stitch positions are punched. Here, I’ve already sewn the lining onto the main body of the upper, and I am checking where to attach the tongue before starting to stitch it.

Since these shoes would be made on a last, and would need to be removed from that last after the welt had been attached, I made sure to only sew a small distance up the upper from the edge of the lasting margin when attaching the tongue: I wanted as much latitude as possible to be able to pull the assembled shoe off the last!

A completed upper lying next to its pattern — The tongue has been sewn in, and this upper is now ready to attach to a welt. Note how difficult it would be to punch eyelets at this stage: much easier before things start becoming three-dimensional.

Having assembled the uppers, and stitched midsoles to insoles, I was ready to nail everything to the last…

A leather upper nailed to the last — First, the insole with attached midsole is nailed to the sole of the last. Then, the fully-assembled upper is drawn tightly over the last and nailed to the insole around its perimeter. This is what the 20mm lasting margin on the pattern is for: to allow the lasting pincers enough leather to grasp so that the upper can be pulled tightly over the last before nailing.

… while checking from time to time that the upper was in the correct position.

A shoe with a last inside it — It’s important to check that there are no wrinkles in the upper, and that the pattern is correct relative to the toe and heel of the foot, before starting to trim anything. Here, everything looks good. Note that during lasting, the shoe is laced as it would be on the foot.

After I’d nailed the uppers (with attached lining) onto the last, the edges of the lasting margin had to be trimmed, so that the remainder would fit neatly under the welt, abutting the edge of the midsole.

A scalpel lying on the sole of a partly constructed shoe nailed to a last — A scalpel is a great trimming tool, but be careful of fingers, and don’t drop it!

Upper and lining of a shoe neatly nailed to a last — Trimming complete. Note how the edge of the trimmed upper + lining abuts the edge of the midsole (which has no holes punched in it yet).

For several previous pairs of shoes, I had pre-punched holes in the midsole ready to attach to the welt. However, it was very time-consuming trying to insert the curved needles through these pre-punched holes, especially around the tight curves of toe and heel. Therefore, for the current pair, I simply used a pen to mark the expected positions of the stitching holes onto the midsole, but I did not punch the holes before starting the welting process. Rather, I used an awl to punch them as I stitched.

Close-up view of white thread being used to stitch a welt to a mid sole — In-progress attachment of a welt. As usual, two needles are used. Here, I am using 1mm waxed nylon thread (“Tiger thread”), rather than the 0.8mm thread used on the uppers. Red biro marks the approximate positions where I would expect to pierce a hole for the thread on the midsole.

Furthermore, to reduce the chances of piercing the outgoing thread with a needle going in the opposite direction (all too easy when inserting a needle through a hole in the midsole that already contains a thread), I experimented with using one hole in the midsole from the midsole to the welt, and a second hole for the thread coming in from the welt to the midsole. This can be seen in the above photo: around the toe area, there are two small stitches in the midsole for each of the large stitches on the welt.

Side-on view of a welt being attached, with nails still in the unsewn portion of the upper. — Side-on view of the process of welt attachment. Note the silver lines on the upper, marking the position where it needs to meet the edge of the last (on the pattern, this was where the masking tape edge was). Nails are pulled out with pliers just ahead of each stitch.

Having stitched the welt the entire distance around (overlapping by 3 stitches on the instep, for strength), I pulled out the two nails holding the insole to the last. Then I cut a piece of 4oz goat leather (nice and springy) to the midsole pattern and glued it into the space bounded by the welt. I only needed a single infill piece here, unlike my previous lasted shoes, since I did not have an extra piece of vegetable tanned leather to add extra thickness between the upper and lining at toe and heel.

A dark green piece of leather filling the space inside an unbroken welt on the sole of a lasted shoe. — Here, I’ve used water-based glue to fill the space bounded by the welt with a piece of goat leather. The shape is that of the midsole, trimmed by a millimeter here and there to account for vagaries in the welt position. The edge of this filler piece touches the edge of the welt, but do not overlap it. The shoe is now ready to start making the outsole.

When this “filler” piece was glued in place, I could trace around the outside of the last onto the 10oz cowhide that would form the outsole. Then I removed the shoe from the last: a little bit of a struggle, but thanks to my foresight in not stitching the tongue far up the upper, entirely do-able even though the front of the shoe was well encased.

A thick piece of leather with the shape of a sole traced on it in silver pen, with a knife nearby. — There is less leather wastage if the shape of the outsole is traced after attaching the welt, rather than before it is attached (the welt changes shape very slightly as it is sewn on, so one can’t simply use the welt pattern for the outsole too).

In previous pairs, I had prepunched the holes to attach the welt to the outsole, but for the current pair, I marked the hole positions after I had attached the welt. First, I used a pen to make marks 8mm apart along the curved edges of a piece of stiff card: one set of marks on a concave edge and one on a convex edge. I then used this template to mark stitch positions on the top of the welt as far under the edge of the shoe as I could reach, starting at the instep. I then punched holes in these positions with a 00 punch.

An in-progress shoe with bulldog clips holding its welt to the piece of leather that will be the outsole. — Bulldog clips hold the welt firmly to the outsole, ready for the stitch positions to be marked onto the outsole, through the holes punched in the welt.

I was then able to transfer these stitch positions to the outsole, punch them, and then carve a shallow groove connecting these holes in the bottom of the outsole. The thread connecting the welt to the outsole will sit in this groove, so that the entire outsole remains flat enough to glue a rubber sole to the bottom of it.

Rubber mallet, grooving tool and 00 punch lying near a prepared outsole — The outsole ready to stitch. Notice the “bump” in the groove on the outside of this sole: this is where the red leather detailing bulged out over the welt more than the nearby blue leather did. I anticipated this might happen, which is why I didn’t pre-punch holes in the welt or outsole, but marked hole positions after the welt had been attached to the upper.

All that remained was to sew the outsole to the welt with 1mm waxed nylon thread. Again, I used curved needles, to avoid stabbing the upper as much as possible, and bulldog clips to hold everything in alignment.

A thread, pliers and stitched shoe upside down on a table — The hole diameter is fairly large compared to to the thread, so pliers were rarely needed to pull the needles through. Stitching started and ended at the instep, with the final few stitches overlapping the first few, before being tied off. Thread and knots are hammered as flat as possible before the rubber sole is glued on.

Now the outsole was firmly attached, it and the welt could be trimmed to make the edges as neat and close to the stitching line as possible.

A shoe and a knife on a cutting board, with a trail of cut leather leading off the edge of the sole — I really should invest in a proper lip knife to do this: using my trim knife or a scalpel can be a bit hairy: there’s a non-negligible risk of slicing into the upper. Not something one wants to do after tens of hours of work!

Trimming complete, I sanded the sole and welt edges with 400 grit sandpaper, then dyed them black (Fieblings leather dye) to match the top of the welt. I also cut a “sock” from 6oz undyed, vegetable-tanned cowhide to go inside each shoe on top of the insole. There is no need to glue it in: it is slightly larger than the insole, so is held in place by friction.

Two shoes on a piece of newspaper together with a pot of dye, a dye applicator, and a pair of latex gloves — Dye is messy stuff, and it helped to have the shoes off the lasts for this step: the leather of the uppers could be held firmly as far away as possible from the sole edges while the applicator was being used to dab dye onto the soles.

Onto the final step! And the stinkiest. Gluing on Vibram soles is something that I hate doing, but it’s necessary if I want a durable and grippy sole on the bottom of the shoe. As I have done for previous pairs, I attached “Newporter” large men’s blank soles using barge cement, and left them to cure overnight with everything clamped together using large bulldog clips. I then trimmed off the excess rubber using a scalpel (wearing leather gloves in case the scalpel slipped). After inserting some shoe laces, the shoes were ready to wear!

A pair of blue leather shoes with red highlights, at the end of fawn-trousered legs — Beam me up Scotty? I don’t think anyone else has a pair of shoes identical to these! (And because they are handmade, the left shoe is not 100% identical to the right).

As can be seen in the above photo, the lack of rigid toe cap means that the toes are slightly visible within the shoe, despite the two (or three) layers of 3oz leather between them and the air. But the whole shoe is marvelously flexible, and also lighter than my previous pairs: there is less leather at heel and toe, and one less layer of filler between the midsole and outsole. Also, I almost don’t need to do up the laces to keep the shoe on the foot: the shape of the forefoot keeps it fairly well attached even without the lace. I look forward to wearing them!

Thanks for reading if you got this far. If you want to get in touch about these shoes or anything else on this site, you can contact me on Twitter @tozafoot.