Revisiting knit graph papers, charting, row tracking, and more

DIY proportioned charts may be created using spreadsheet programs. I began to use Excel for charting in 2009 and continued to in nearly all the colored charts in my color separations for knits posts up to my latest computer upgrade. I now no longer have access to Office and work primarily in native Mac OS programs. I reviewed links for working with Excel for this post, and these are still live in 2024
http://marniemaclean.com/blog/tutorials.html#.Um_-wpFQY7I
http://www.chemknits.com/2010/01/how-to-make-knitting-chart-in-excel_9394.html
http://fleeglesblog.blogspot.com/search?q=charting+with+excel
http://anniebeeknits.wordpress.com/2011/06/06/charting-in-excel/
http://www.youtube.com/watch?v=T9dwuarghqE
https://colornotesyarn.com/create-knitted-design-chart-with-excel-spreadsheet/

one for google sheets:
http://fibercraftmama.blogspot.com/2014/10/making-knitting-charts-in-google-sheets.html

my early posts on working with numbers with some edits
https://alessandrina.com/2013/10/28/color-separations-using-mac-numbers-for-knitting/
https://alessandrina.com/2013/10/29/charting-knits-in-mac-numbers-program-2/
https://alessandrina.com/2014/09/03/creating-knit-graph-paper-on-mac-using-excel-and-numbers/

Punchcard design planning, new: these are not to scale but can serve as a base for planning repeats on paper, may print to size with some tweaking by splitting on more than one page. The first sheet is marked as a Brother factory blank single card, the second has the same markings as the roll I purchased as a “Brother one”, which has row # 1 markings for Studio kms, so Brother knitters need to begin knitting on row # 3 punchcards.

I have revised the nearly-to-scale template from a previous post; the first is not for a full card but is designed to fit on one sheet. There are 36 rows/ 180 mm high, by 108 mm wide, of prints to scale on my printer. It is bordered in blocks of 6, with no row numbers: scale no numbers.  A second document includes an added 6 rows, brother numbers row markings on the right, and a column for notes such as lace arrow marks, etc, prints to scale on my printer scale card extra2

It is sometimes helpful to have help in tracking information for actions to be taken on specific rows ie for increases or decreases during knitting on specific rows, or for lock/carriage setting changes in more complex double jacquard (DBJ). Going the printed paper route, this PDF  tracks single rows, every 2, every 4, every 6, and every 8 for tracking such instructions on individual pages. They have enough blank area left to write in pattern name, stitch, cam settings, stitch size, yarn, etc. It may be helpful to include short clippings of the fiber used. Each page may be printed separately, with printer adjustments made if required with your software: tracking rows

Some rib setup with room for notes, in both P and H settings

Ribber fabrics produced with 2 knit carriages selecting needles

There are a number of ribber fabrics that are produced by altering the settings on the ribber’s carriage to slip for an even number of rows in both directions. This requires manually changing the ribber setting from slip to knit and back for the length of the piece. In electronic machines, where the pattern advances with every pass of the carriage, there is another option. For example, to produce DBJ with the backing in one color, settings would be changed manually every 2 rows, using the color changer on the left

Some things to remember: when 2 carriages are selecting, each carriage needs to move far enough at the end of the needle bed so as not to be locked onto the belt. Extension rails are required. On the Brother ribber bed, there are stops that keep the combined carriages from going off the beds. There is one on each side (magenta arrow on left, the blue arrow on right), and to remove the ribber carriage off its bed, it, in turn, needs to be tilted forward prior to reaching the stops in order to clear them on either side.

Altering the KC sinker plates and arm: remove 6 small screws from the sinker plates, leaving only their arm

The carriage with the altered sinker plate in place, in turn, will then be used to replace rows that were to be knit with the ribber set to slip in both directions <– –>. In my sample, it operated from the right, with the combined carriages (KC2), from the left. 

For consistency, I am editing the original post and will continue to refer to the coupled carriages= KC2, the altered single bed one= KC1KC is the abbreviation commonly used in publications for Knit Carriage. The change knob, which affects end needle selection, is marked on the KC as for I (end needle selection, indicated by black arrow) and II (cancel end needle selection). Initials KR in publications are often used to refer to ribber carriage in setting discussions. My beginning swatches were knit using ayab software’s ribber setting, which matches the KRC (2-color double jacquard separation) function in the unaltered 910. With my first try, I made no effort to consider which color gets chosen first in the color separation (ayab = black, 910 = white). There is a limit as to how far the single carriage from the right (KC1) can travel on the needle bed to the left because the combined carriages on the left (KC2) are held in place by the pin.

It would be possible if needed to separate KC2 and push further out on the extension rail, but perhaps not practical, so there are some constraints on the fabric width able to be produced.

My first swatch has some manually created pintucks on the knit side (white only knitting extra rows, joined together by all knit rows with pink). The reverse, purl side, is knit by the ribber set to N/N, all in one color. The stitches held on the ribber while the white only knits on the top bed are visibly elongated (left swatch bottom). There is some color confusion on the knit side on the first couple of rows of DBJ, solved in the second swatch. My repeat for the planned width using Ayab:

Knitting on 910 a single repeat may be programmed, start knitting with color intended for white squares, which will also serve as the solid backing color. For more similarities and differences between the original and the altered 910 see ayab diary post.

To knit using Ayab: begin the knitting with the color intended to be used for areas marked in black squares, which will also serve as the solid backing color. Preselect the first row from left to right with ribber set to knit (N/N), it will remain set that way for the remainder of the process. When on right, set coupled carriages (KC2) to slip <– –>, knit one row to left, both carriages stay on left. Knit the next 2 rows for DBJ using KC1 with altered sinker plate operating from the right, using color intended for areas marked with white squares. The main bed only will knit. Set change knob to end needle select (KC I) to ensure first and last needles in use knit. Return the KC1 to the right, on the extension rail, and knit the next two rows with KC2 operating from left. Repeat, changing carriages and consequently colors every 2 rows.

Blisters or pintucks are created when one bed knits more rows than the other, whether as simple knitting or in the pattern. Periodically the knitting is sealed by at least one-row knitting across all stitches on both beds. In this version, sealing rows must occur in pairs to allow for color change. The first chart shows a tentative repeat, planning for black squares to create the blister shapes, drawn in 2-row blocks to allow for color changes every X even number of rowsThe image color inverted, so white areas will create the blisters in slip stitch (col 2) while black squares will knit (col 1)Most published patterns for these fabrics will also include an all knit rows to seal the shapes knit on the main bed only

By using 2 carriages to select needles, one (KC1) may be set to slip <– –>in order to knit X number of rows on the top bed only, while the pairs of carriages (KC2) are set to normal knit on both beds, its cam button set to select needles, KCI. Selection will continue, but no patterning occurs as a result. A proof of concept swatch knit on only 24 stitches, the pink knits for 4 rows, the white for 2; the pink yarn is a cotton, the white an acrylic/wool blend:

An expanded pattern repeat is planned for a larger test swatch. Here there are 2 black squares added at each end of the repeat to ensure that those stitches on the top bed are knit on all slip stitch rows. The new color is wool, this time knit for 6 rows in slip stitch prior to sealing with 2 passes of the combined carriages with the contrasting color.

Single bed slip stitch rows appear on the knit side in color 2, the reverse, purl side is in a single color (1), and formed by the all knit rows. Its stitches are in turn elongated, since they are held and not knit while the opposite bed knits for multiple rows. In the sample, the first and last stitch on each side were on the ribber, creating a single white slip stitch edging. One can adjust such details to suit. The first preselection row was made after cast on with both carriages(KC 2)set to knit, moving from left to right with color 1, where they stay. KC 1 with altered sinker plate was threaded with color 2, and begins from and returns to the left-hand side.

A detailed shot of the edge: note the white, single, elongated stitch uppermost, and the pairs of contrast color ones in the “border” Designs with the deliberate placement of white blocks representing each blister can be created. It is a good idea to test tolerance for each of the yarns involved as a hand tech or repeat such above before planning significantly longer repeats. All black rows are required at intervals if the fabric is to be knit changing ribber settings for all knit rows. The same rows could be left blank if using carriages in the above manner, and lack of needle selection would be an indicator for switching to the double carriages for the 2 sealing rows, while not having to track the count for the slipped ones.

I am always interested in automating stitches to facilitate as many details as possible in creating fabrics that imitate hand techniques, without relying on row counts and a lot (in this instance) of “hooking” stitches up manually. The swatches below were part of a series of posts on ruching as a hand technique

The hand tech chart mirrored, as the springboard for my automated fabric Both led to my exploring the possibility of a cousin, working on the double bed, and using the 2 carriages. The process began with a proof of concept swatch. Two colors were used to highlight what the stitches knit with each of the 2 carriages are doing. Both KC1 and 2 were set to knit in both directions. I began with KC2 knitting the needle set up for the fabric in a dark color, on a small number of stitches

KC2 operated from left, KC1 from the right. With KC2 returned to left, KC1 knit 8 rows. It is best to work out the limit for how many rows will knit on the top bed without stitch problems prior to any automation of functions. *Two rows were knit with KC2 in the dark color, 8 rows with KC1, ribber racked to proper position**, * to** were repeated. The elongation of the stitches on the purl side results from the fact the ribber knits far fewer rows than the main bed, and in addition, the stitches on the row knit on its second pass in the pattern are pulled on across the bed at distances matching the racking positions. The plan is to automate the texture, knit it in one color, and find a way to track correct racking positions: cast on may be in any preferred method. With the pitch set to P, it is easier to transfer stitches between beds to the desired configuration. Every needle is in work on the top bed, for a multiple of 10+6ribber needles with stitches that will be moved with racking

placement of stitches on respective beds
change pitch after transfers, ribber moves slightly to right final configuration prior to patterning and racking sequence begins

The starting racking position is 5. Racking handle markings for Brother begin with 0 on left, 5 at the center, 10 on right. The ribber is set to half pitch since part of the needle bed will be knitting every needle rib. An often-overlooked clue as to what is happening or is about to is found in the arrows just below the racking position indicator. The red arrow indicates the direction in which the bed was racked on the last move. Since racking for my experiment will only be to 3 different positions, I began by choosing to use 5 pixels center, left, center, right for the full repeat sequence, but later amended the repeat to numbers of pixels equal to specific racking indicator number. The needles will be selected prior to the next row knit, a reminder of racking selection. Sequence will actually be 5, *0, 5, 10, 5,** 0, 5, etc. The racking indicator

My first Ayab repeat was planned for use with 2 carriages: KC1 (single carriage) is set to knit but to not select end needles (KCII), and will be producing the multiple rows gathered to create the blisters/ hems. KC2 (coupled carriages) are also set to select needles, both KC (knit carriage, cam button also on KCII) and KR (ribber carriage) are set for normal knit. They will create the sealed stitches joining up the blisters/ hems. The same color yarn is threaded in both. Blank squares will knit because a carriage set to knit overrides needle selection or lack of it in stitches on needles in the work position. My repeat is 46 stitches wide. Because the knitting is started with the ribber already in racking position 5, the first move in the pattern is turning the handle to the right, toward 0. The concept illustrated

set up and ready to go, racking position 5

4/19/18 As is true in any knitting, things can go quirky. I began to have a single needle on the main bed not knitting on rows knit with the combined carriages, then ran into dropped stitches in racked groups. The problem was initially not with the software but appeared to be a ribber issue, which after checking and balancing was resolved.

4/22/18 And today’s problem is the software, with persistent, intermittent selection errors. I did achieve a sample by manually pushing wrongly selected needles out to D by hand on problem rows, which tended to be 5 and 7 in all black areas. Some reminders and observations: ayab auto mirrors all images. If this repeat is entered (I added a single square as a marker for racking position 0), the software will actually be knitting this, the “image as it would appear on the knit side of the fabric”. What to program? anticipating the above getting mirrored by the software I entered this resulting in this

That said, remember that turning the racking handle to the left is toward increasing numbers on the indicator, to the right is toward decreasing numbers. For me that is counterintuitive. Mirroring the image again, and working with the repeat below can help with tracking racking movement even more. With the single dot on right, turn the handle toward it, to the right, and the movement will be towards 0. With the movement of marking row to left, turn racking handle to left, toward 5, and so on.

Next on the drawing board: a fabric using the same technique, but that I might like more. This image was a portion of a greyscale pin of a pattern book from a Russian pinI tested a concept for recreating it as a hand technique, trying to sort out how many rows I could knit before racking and the racking sequence. The best result was with a single sealing row, which in turn required changing the ribber slip setting for one row only,

so it’s back to the drawing board. I think any automation is best done using the ribber to do the all knit rows, the main bed to needle select racking positions. Results will be added to the post on combining KC patterning with racking

The present set up with 2 carriages may be used for solid color backed quilting. Using the altered KC1 operating from the right with no yarn in feeder should work to drop stitches in drop stitch lace where the repeat is altered to allow for the knit carriage with no yarn to do the stitch ditching while selecting needles as well. Related color separation and swatch may be found in the last segment of the post: revisiting-drop-release-stitch-lace/

 

Revisiting knitting with 2 carriages single bed, 910 vs Ayab so far

The question has often been asked as to why knit with 2 carriages. For me, it became a personal favorite for knitting fabrics requiring color changes every even number of rows. I tend to push the limit with materials and found the Brother single bed color changer to be quirky. When using 2 KCs there is no pushing the wrong button for the proper color in the sequence, or accidental choosing of an empty slot and dropping the knitting. Some of the loops at the color changer side edge or yarn getting caught in wheels or hooks are also eliminated. Each carriage can be set for a completely different function from the other with no manual cam button changes ie single color knitting vs FI, its own tension depending on desired effect or fiber (ie striped shadow pleats), etc. Floats along the vertical edges of the knit as a result of the striping happen on both sides rather than all on the color changer side. The principles apply when using the lace carriage and combining it with tuck, slip, or FI settings in the KC. If one is following a published punchcard pattern for such fabrics, frequently there are guidelines as to when to change colors, but no explanation as to why the cards are punched the way they are, so understanding what happens enables one to interpret directions for electronics. When going from published 2 carriage punchcard machine repeats to electronic, the repeat needs to be adjusted to compensate for the change to row by row selection. In punchcard machines, the needle selection does not advance when you switch to operating the carriages from the opposite sides, so the pattern for those rows knits twice.

Many of the stitch world books seemed more concerned with larger repeats than with some of the more labor-intensive, specialty fabrics. The punchcard books, now downloadable for free are worth having and studying no matter what your machine. They do need to be brand specific. Studio and Toyota ones may be used, but adjustments may have to be made in terms of starting rows or even flipping the card over vertically.  Download links for punchcard patterns, with symbols used and pattern samples, can be found in my previous post

I have added a new category to my sidebar to make this topic search easier: Patterning with 2 knit carriages.  Included are color separations for some fabrics suitable for this method of knitting.

I am presently knitting my samples on an altered 910, using an EMS kit purchased last December. I have had selection problems related to my hardware, but some of the 2 carriage operation or lack thereof has been evaluated and duplicated by others. My successful experiments with 2 single bed carriages selecting so far:

Lace and slip stitch: both carriages selecting in their usual positions 
2018/03/05/lace-edgings-on-brother-machines/
Start with 2 knit rows in a programmed pattern 
with KC on left, preselect to right, set to slip <€” €”>, knit first 2 all knit rows, ending with KC on right, proceed with selection with LC on left
Will not work if the pattern is adjusted so KC knits the first 2 rows and parks on left, with attempted use of LC from the right. LC will not select. 
Operating LC with KC plain knitting (no pattern selection) may be done with LC selecting from either side, with repeat adjustment 2018/03/05/lace-edgings-on-brother-machines/

Lace and weaving: 
2018/01/02/lace-punchcards-meet-ayab/
both carriages selecting in their usual positions 

Lace and tuck stitch: the fabric worked out for use with mylar sheet

Ayab actions for carriages, use LC on left, KC / tuck <– –> on right
a charted ayab repeat for my swatch 

I came across this image on a pinboard, and imagining the possibility of knitting it led me back to a previous post illustrating a similar stitch structure, knit on an unaltered 910. The pin:


working out the first repeat in the blog post

The swatch (sideways view) shows the repeat as illustrated in the upper segment of the above chart, and then lengthened X 2, with color changes every 4 rows (the length X2 is not presently an available action in Ayab)

In the past, when I tried to knit with 2 KCs, one on the left and one on the right, both selecting needles, I was unsuccessful, but in light of the above experiments, I decided to test the idea with the RKC (the one that would operate from the right) selecting and then knitting the first 2 rows of color, in turn resting on the right side. I then operated the second KC from the left, alternating carriages every 2 rows. Knitting at a slower speed and moving each carriage perilously far out on the extension rails did the trick.

To use the repeat in Ayab, in order to keep the proper color sequence, the last design row is shifted down to row 1 position. There is an error in the repeat, marked by arrows I did not notice until I had completed the swatch. The repeat may be used as is, and color reversed via action inverse in Ayab, or color reversed prior to loading the image, which I chose to do personally. 

This is grabbed from the ayab screen: the knitting is actually mirrored automatically by the software, so what is shown onscreen is what appears on the knit side of the finished piece. Something to keep in mind in situations where placement and direction might have some importance. 

Things to notice in swatch: as already mentioned, there is an error in the repeat in terms of my intention. Placement for the slip stitch blocks could certainly be planned for an asymmetrical effect, varied in size, and more. Slip stitch produces a significantly narrower knit than the stocking stitch. That would need to be allowed for were the stitch to be used in a garment or gauge dependent knit. 

Examining the purl side of the knit, which is facing as one machine knits: the intentionally dropped stitches at the start on both sides are there because I had cast on extra stitches that were outside the parameters of the programmed repeat width. The left side edge, next to LKC (left knit carriage) is visibly shorter than the right. The two yarns used were of different thicknesses but knit at the same tension. The LKC would have merited from a looser tension, and an edge weight on the very last stitch on that side would help with adjusting the length the white edge €œfloats€ from one color change to the other as stripes are created. The handling of side floats in striped knitting is dependent on the row height of such stripes.  
There are a great number of interesting ribber fabrics that can only be knit by manually changing the ribber settings to slip in both directions, so that the main bed only knits for X # of rows, with the ribber carriage periodically set to knit again. If the function happens for an even number of rows at a time and in both directions, the sinker plate for the KC may be altered for use with the ribber in the up position, to knit the main bed stitches only. Its change knob should be set to KC I to ensure that the first and last stitch is knit. I knit this sample using the altered KC selecting from the right, the KC engaged with the ribber sinker plate selecting from the left. I will address the altering process and steps involved in the knitting in a separate post. 

 

Fair isle variations

A review of links with associated hints and info:
Measuring gauge swatches, general information 
Matching patterns across sweater bodies and sleeves
Float control 
Scarf experiments
Design inspiration: binary alphabets
Adding hand techniques/ cables/ punchcard repeats
FI meets transfer lace on Brother machines 
Adding the ribber, FI on main bed Tubular machine knit fabrics: fair isle, Brother/Passap
Altered patterning using bleach discharge on knits

These are random FI samples from my collection, most from my teaching days. None of them were ever intended for use in the finished product. They were knit to illustrate some of the possibilities for the different techniques using each of the cam button combinations. Some were knit during class demos. The colors made them easily identifiable as mine, knit using a personal yarn stash. The contrast helped identify how stitches were formed.
In this swatch, marking for measuring stitches per inch is done by leaving a needle out of work. The width between the resulting ladders should be checked at various points after the swatch is treated in the way you plan to treat (block) the finished fabric. Adding a third color per row would require altering the pattern to a color-separated slip stitch one, or one may add that color with duplicate stitching. The spots in this test are colored in with a permanent fabric marker. At the height of the art to wear movement one artist, in particular, was producing limited edition knitwear by knitting the same design in black and white, and in turn over-dyeing the white for different effects in each piece in the series. Eyelets at the bottom of the swatch are tension markings for the piece. The vertical line created by end needle selection (normally used in FI to avoid separation of colors et vertical edges) is interrupted in rows that are knit in only one color. Recommended maximum width for floats is usually 5 stitches.  How much the floats droop and cause potential “problems” on the purl side depend on fiber content. Sometimes such floats are intentionally created and worn on the outside of the garment as planned design features. The longer blue floats are seen below in the areas of the ladders where only the yellow is knitting, creating a wider span of the alternate color. These repeats are very simple. They are commonly associated with card #1 and card #2 in basic factory packs supplied with knitting machines purchases. Card #2 is reproducible by using card #1 elongated X2. A reminder: if using either repeat in pieces of garments ie baby leggings, etc. take note of which yarn feeder each color is in. Even if the repeat is correct and placed properly, the surface of the knit will appear different to the eye if the color placement is reversed in alternate pieces. The repeats may be used as backgrounds for a variety of other more complex fabrics in DIY designing. Here stainless 32 gauge wire is used as the second “color”, making the piece moldable and shape-retaining.   Color may be added or “taken away” as seen in the post on bleach discharge on knits Another factory-supplied punchcard is used. Thinner yarns in lighter colors may have noticeable bleed-through of darker colors traveling behind them, as seen on the left, not an issue with the thicker wool on the right. Forgetting to set the card to advance can result in vertical lines, which may alternately be planned as a design feature. The longer floats seem manageable in these yarns, there is a bit of hooking up on the bottom right. The yarn traveling up the swatch on the right is an alternative way to mark for gauge measurements. A previous post provides some information on float control.  Varying the colors, fiber content, and considering complementary borders is worth exploring thoroughly at the swatch level, before committing to a larger piece. Truly contrasting yarn used at the bottom and top of the area to be measured for row gauge makes the process easier. As attractive and quick as single bed FI can be, keep in mind that long pieces knit in yarns with “memory” such as wool, will tend to roll to the purl side vertically even after blocking, and certainly with wearing of pieces such as scarves or shawls.  Tone on tone chenille and all rayon, with “color reverse” by switching yarn positions in feeder less effective with a flat yarn as the alternative to the chenille Using the same card:  every needle, 4.5 mm electronic machine.  Transferring stitches to every other needle, odd needles in work on one side, even-numbered needles on other using worsted weight (2 needles in the center in work side by side.   The motif is now used twice as wide with every other needle in use across the fabric width It is possible to vary designs by using the 3 functions of the card reader: locked, normal rotation, and elongation. Designs with long vertical features tend to separate at the edges where the 2 colors meet. Lining the fabric with a fusible makes the knit lose stretch, but it may be an option for stabilization, float control, and offers an opportunity for mock quilting by inserting some stuffing under floats before it is ironed on. High contrast colors are best for sorting out how stitches are formed. Embroidery alters the “step ladder” effect outlining the shapes. Hooked-up floats are not just for float control; note puckering on the knit side where they have been hung up in groups.  These swatches were worked from the bottom up, starting with positive/ negative comparison, sorting out the possible placement of the ladder with the intent of adding ladder lace details. Cancel end needle selection because of needles out of work, but bring needles into D or E position to avoid separation of colors and/ or dropped stitches at side edges.  From the bottom up, transitioning from a ladder resulting from a single NOOW (needle out of work) to 2 NOOW, hooking up floats on opposite sides, ending in “lace” pattern alone  Combined with transfer lace  Hand techniques (in this case cables) can be combined with FI. In Brother, it helps to be familiar with the pattern, as needle selection may have to be manually restored after the technique is performed to stay in the correct pattern.
With cables: some punchcard repeats
With using a sewing machine: there is a vertical, single stitch line due to end needle selection in the contrast color formed on either side of any needle(s) out of work which provides a visual guide for altering the fabric. This swatch was knit with wide NOOW spaces, then sewing machine stitching joined the contrasting vertical lines to form a 2 color “fringe” on the knit side (left) and purl side (right)Variations with fibers for exploring surface textures: wool with raffia on the bottom, fishing line on top The same swatch continued on, using 3M elastic as the second color  The same repeat in a rayon chainette and wool, followed by some felting. The rayon “bubbles” more visibly when the wool creates the wider floats  reversing color positions
The punchcard is limited to varying the vertical repeat automatically in 3 ways: locking the card, normal rotation, and double length. Repeat width is fixed. Felting can produce interesting surfaces if one yarn is capable of being felted (green), and the other not (blue). The stitches knit with the latter will create puckers/ blisters. Since the knit will shrink in both width and height, the repeats here were used at double length. Note the added drooping of the blue floats on the purl side.  A punchcard can be further manipulated by masking areas with tape. It is not a good solution for production knitting, but adequate for testing out ideas before committing to punching a full, new card. The surface blisters here are much more dramatic. The green floats do not felt as much as in the previous swatch, and are considerably wider. On the right, far side you can see some of them were latched up, creating yet another design detail.
The reverse of both swatches shows the resulting difference in relative width.  The contrast using a factory-supplied punchcard pattern with short floats, also felted. The fringe is created by ending on one side (in this case on the right) with a group of needles out of work and the outermost 2 needles in work, essentially producing a large “ladder”.  Decreases and increases on needles close to the edge of the knit were brought in and out of work to create the “zig-zag”. The two edge stitches of the ladder may be trimmed before felting. The knit side is shown on the left, the purl side on the right, no clearly visible, separate floats, its surface is fairly flat.