Saturday, February 6, 2021

Gauge: a mystery of knitting

Why is stitch gauge variable between knitters? It's a mystery of knitting encompassing many factors. Widely understood to have an influence are needle size, needle composition (material), tension in yarn supplied; and this post considers all these.  Yet, a factor equally important--if not more--so lies in the formation of the knit stitch itself. We'll start there. 

We can say there are two properties of a knit stitch: its width, and its height. In the simplest conception, width arises from the "tail" of the stitch ( red,  below) and height is measured from the "head" ( blue ) to the tail. 

The head is formed when your needle pulls through a loop, and the tail of the stitch when you move the formed loop along your knitting needle to get at the next loop. The tail here is shown split in half, but of course, the tail of one stitch butts against the next, making one continuous, shared, width. 

Tail or "'tweener strands" illustrated

Below is a photo-diagram of the edge of stockinette: this is a top view. The heads are not colorized, only the tails. 

Stockinette fabric, top view
"'tweener strands" of current row colorized red

As with the opening illustration, the tail is colorized   red.   Here, the tail is seen top-view with the purl side of the stockinette fabric in view. This little red tail connects one stitch to another along the row--in this case, the current row. Because this tail runs between stitches, we'll call it the 'tweener yarn

Why does tail yarn need a special name?  As it turns out, all 'tweener yarn is tail yarn, but not all tail yarn is 'tweener yarn.  As we go along in future posts, we'll need specific names for the different kinds.  So, to be specific, 'tweener yarn is that particular subset of tail yarn which connects stitches aligned in the same direction. Stated otherwise, 'tweener yarn connects adjacent stitches of the same kind (knit to knit, or purl to purl) along a row, as in stockinette.  (In a future post, we'll look at a different subset of tail-yarn which we'll give a different name, and that is the kind which connects different kinds of stitches: knit to purl or purl to knit.  When that post goes live, there will be a link placed here.) 

Here is a split-screen view of stockinette fabric, as seen from the front (knit) and back (purl) fabric faces, respectively.  Again, the 'tweener yarn of the current row is colorized   red. 

The most obvious difference between the two fabric faces is that the 'tweener yarn is unevenly distributed: far more appears on the back fabric face than the front. One consequence of this unbalanced distribution is behind the reason why stockinette curls so badly (discussed here).  But 'tweener yarn is also an important factor in gauge--our topic today.

Stitch gauge:  Differing lengths of 'tweener yarn supplied between individual stitches lead to differences in number of st/in.  In other words, it is the length of this little red strand laid down between stitches, which helps account for the difference in stitch gauge between two different knitters using the same yarn and the same needles. This makes intuitive sense because more distance between the stitches = fewer stitches per inch of fabric. 

Row gauge: The length of the 'tweener yarn is an important factor in STITCH gauge, but not the only factor in overall gauge-- there is also ROW gauge. 

In my analysis, it is the height of the stitches which adds a factor for difference in row gauge, and by height, I mean how tightly each loop of yarn conforms to the size of the barrel of the needle.  Once the stitch is off the needle, a stitch which lays looser around the needle barrel will lay taller in the fabric than a relatively tighter-wrapped stitch, so looser loops = taller fabric

Stitch + row gauge, theory: 

Putting together the two factors--length of 'tweener yarn + height of stitches--theory suggests...

4 samples, 4 different styles, 
same yarn, same number of rows & stitches,
same needles, same knitter

  • WIDE: If you pull your yarn tight over the barrel (smaller, tighter stitches) but leave plenty of room between stitches (longer 'tweener strand) you'll have--relatively speaking--a wide-ish stitch gauge (more st/in) and a short-ish row gauge (fewer rows/in). I could get this to occur consistently by operating on the upcoming stitch very near the holding needle tip (left needle), but drawing the new stitch well back  onto the working needle tip (right needle).  This near-the-left-tip, far-from-right-tip method produced the "wide" swatch illustration.
  • TALL: If your loop is loose over the barrel of the needle, but each loop is right up against the next loop (shorter 'tweener strand), then your row gauge appears relatively long (fewer rows/in) and your stitch gauge relatively narrow (more sts/in).  In my experiments, this style can result from stitch formation via drawing the two needle tips apart from one another like the arms of a scissors, with significant up and down motion between the needles, but the yarn held near both needle tips. I heavily "scissored" in knitting the tall sample. An alternative way of getting the same effect is by wrapping the stitch loosely, holding it somewhat open by tensioning the working (right) needle upwards, then giving a pretty good jerk to the running yarn.
  • TIGHT: If your 'tweener strand is short and the height of the stitch short, you're a tight knitter (more st/in and more rows/in).  In my experiments this was accomplished by knitting at the tippy-toppy-tips of both needles, drawing the running yarn so tight to the barrel of the needle they would hardly slide, and, of course, by heavily tensioning the yarn at all times. 
  • LOOSE: If your 'tweener yarn is long and the height of the stitch tall, you're a loose knitter (fewer sts/in, fewer rows/in).  This is how I usually knit.  Therefore, to get gauge, I have learned to drop needle size, usually two US sizes smaller than what is considered "standard."  For the below sample, I knit as loosely as possible, tensioned the yarn lightly, held each loop well away from the left and and right tips, drew a lot of yarn into each stitch and scissored each stitch wide open.
Another view. Wide (top L) and loose swatches (bottom L) are same width, but loose swatch is longer.  Tall (bottom R) and loose swatches are same height, but loose swatch is wider. 

 All these samples were knit using the same needles (Boye aluminum US size 5 dpn's) on the same day, by the same person (me), all 16 stitches and 20 rows, using yarn from the same ball of Patton's Roving.  I chose this yarn because it has very little twist and is airy.  Airy yarn like this which compresses but also fluffs up can be worked at many different gauges. Further, it is pure wool of a not-very-soft kind, and so, ultra-grabby. The only thing which varied between the samples was the height of the stitch, and the length of the tail yarn between the stitches--the 'tweener yarn. 

Now, theory is a very fine thing, but unless you're using an ultra-grabby yarn like mine, stitch height and 'tweener length aren't really in a fixed relationship.  Looking back at the structure of the stitch, you see that a stitch is a very fluid little creature. As yarn comes off the needles it is freed to wiggle around in the continuous and interconnected web which is knitted fabric. So, yarn can get pulled from the loop of the stitch into the tweener strand, or vice versa.

Via blocking, a taller stitch can simply become a wider stitch, just depending how the fabric is handled (when you block a fabric wide, it becomes shorter). Similarly, blocking can turn a wider fabric with more 'tweener yarn to become a taller fabric (blocking a fabric long makes it narrower). The more overall slack--whether in stitch height or 'tweener strand--then the more possible it is that in the blocking process, or even in just wearing the garment, the fabric can trade width for length and vice versa.  In other words, whether you made the stitches tall with a short length of tweener yarn, or you left a lot of 'tweener yarn stranded between the stitches but made the stitches tight, these variables don't matter as much as the overall amount of slack increases.  This factor is especially true if you are knitting in a slick yarn. 

In trading length for width, yarn slickness is huge.  Different yarns have different amounts of friction vs. slide.  The slicker the yarn, the easier the interchange between stitch height and tweener yarn length (And this is why it is so notoriously difficult to "get gauge" in  alpaca, acrylic, linen or silk, those slick and frictionless yarns.)  For my original cream colored samples, I chose high friction yarn which made it easy to show the relationship between stitch height and 'tweener yarn length.  But if you're knitting with a slick yarn, the story is completely different.  

The gold samples below are worked in some kind of acrylic yarn--the ball band went missing years ago. All I know is it is super slick (so slick, in fact, I usually use this yarn for COWYAK provisional cast-on: it slides out of the work perfectly). Both samples below were made within a few minutes of one another, with the same needles (those aluminum Boye size 5's again) same yarn, same hands. I knit both samples as I usually knit, which is loosely--long 'tweener yarn as well as tall stitches. The only construction difference is a few purl bumps on one so the two samples can be told apart. 

After the samples were knit, I didn't wet-block because blocking has no lasting effect on acrylic.  Instead, I simply tugged each sample repeatedly and strongly, in order to work the slack yarn either into the loop (tall) or into the 'tweener yarn (wide). The slick yarn simply slid around in the fabric, and here's the result stuck on a magnet board. Although both samples started off at the same size, tugging made the plain sample longer, and the bumped sample wider. 

Plain sample (L) tugged long, bumped sample (R) tugged wide

In this next photo, I left the plain sample alone, and tugged the bumped sample tall and long.  Now, both samples have the same length and width. 

Plain sample and bumped samples both pulled long

In the last iteration, I left the bumped sample long, but tugged the plain sample wide. This final experiment shows that tugging into a wide shape is not permanent in slick yarn

Plain sample (L) tugged wide, bumped sample (R) tugged long

These samples demo the influence of yarn friction (or in this case, lack of friction).  Given that gravity never sleeps, if I wanted to knit a sweater out of a slick yarn like this, I'd pull the swatch as long as possible, and measure gauge that way. That would better match the eventual lengthening of garment as it is worn over time. So, for alpaca, linen, cotton, acrylic, silk, and other similar fibers, consider tugging or blocking your swatches long to get the truer gauge. 

(And speaking of gravity, I might also add here that gravity affects ALL knitted garments to some extent, of whatever fiber they are worked.  So, if you're happy with a garment's current fit, do not store it hanging, no, not even on a wide-shouldered suit hanger.) 

Adding hardware to the analysis, adds more complications.  Needles play a role with needle size the most obvious influence. Bigger needles yield bigger loops and looser fabric, all other things held equal.  However, even among knitters using the same size needles and yarn, there is significant variation in fabric produced.  So, each knitter must learn by experience, her own correction from "standard."  As mentioned earlier, I generally start out swatching with two or even three sizes smaller needles than are recommended as standard on the yarn ball-band.

Needle material also plays a role. Grabby needles like wood or bamboo deliver more yarn per stitch; smoother needles, fewer, and this is true even when all these needles are the exact same size.  Yet, it's even more complicated than that, because while needle composition does strongly affect stitch height (amount of yarn grabbed into each loop) it does not strongly affect the length of the 'tweener strand or tail: the loop is made over the grabby needle, but the tail is created mainly in the frictionless air. So if you get stitch gauge, but you feel your row gauge is too long or too short, consider a switch to needles of the same size but a different composition--smoother or grabbier--as required to correct.

Finally yarn size has quite an effect.  In unitary fabric (non-lace), thinner yarn gets up to fewer overall tricks because there is inherently less slack in smaller stitches knit on smaller needles. The bulkier the yarn you work with, the better will be the result if the yarn is grabby and not slick. So, if I wanted to work a garment in alpaca or silk or a synthetic (slick) yarn, I'd choose a pattern calling for a lighter weight yarn, so there would be less yarn to wiggle around in the first place.  For a bulky garment, I'd choose a grabby yarn, giving gravity a bigger challenge to dragging the garment long.  

In lace fabric knit with a lightweight yarn, the situation is exactly reversed because lace has inherently loose gauge due to its open nature. The yarn-overs in lace would tend to slide into the fabric and disappear if knit in a slick yarn. A grabby yarn holds the lace pattern much better. For that reason, yarn with a mohair content is a dream for lace--the "halo" of mohair fibers are like tentacles holding open the yo's. Shetland-type wools are a popular choice for lace for the same reason: they aren't at all slick and hold the lace pattern with little slide. 

Changing your gauge
I think that changing your gauge by changing the motion of your hands is one of the hardest things to do in all of knitting.  The way that your hands form the loops--whether tall or tight to the needle--becomes an automatic process after just a short time knitting.  Similarly, the amount of yarn your hands play out between stitches--the length of the 'tweener yarn--also becomes automatic. 

I don't say it can't be done--it can, and I have known knitters who did.  So, if you want to try, it helps that you know what factors to look for: the height of the stitches, the length of the 'tweener yarn, and how you got to those factors. As per  above, these include how far the loop is from the tips of each needle at the time of stitch formation, the tension on the running yarn, whether your needles "scissor" the stitches wide open at the time of formation, and (especially for English-style knitters, perhaps) whether you give the running yarn a jerk after each stitch is seated on the right (holding) needle. 

If you absolutely want to have a whole fresh start which corrects the "automatic" tendencies of your hands, consider experimenting with a different style of knitting.  If you knit English, try continental.  If you knit continental, try English-style.  I expect that even with this fairly nuclear-type option, though, your hands will soon develop habits in this alternate style--new and different habits, but equally automatic after a short time. 

An easier and more accessible way to change your gauge is to mess around with the interplay between the other factors identified in this post.  
  • If your swatches always lie, consider swatching and working your next project in a grabbier yarn--rustic wool, for one example, stays put once knit. Among synthetics, something with bumps, slubs, or other heavy texture would also have a better chance of staying truer to swatch gauge.
  • If determined to work in a slick fiber, then, as stated above, a swatch tugged or blocked long is a more accurate approximation of the final result (and the longer the garment is, the more true this is). 
  • If your final garment doesn't fit the way you would like, consider extreme blocking--remember than blocking can usually get fabric to trade height for width, at least somewhat. BTW? A high-end dry-cleaner can help with this too: they have commercial fabric-processing machines you simply don't have at home, especially steaming equipment. 
  • Match yarn to project: in non-lace projects, save slick fibers for thinner yarns knit on smaller needles to reduce overall slack; prefer grabby yarns for heavier yarns to increase friction and decrease sliding around. In lace, the situation is reversed: although lace is typically knit with thin yarn, grabby yarns give the best result in holding the holes of lace patterns open. 
As far as stitch and row gauge go...
  • If you get stitch gauge but your row gauge is too short (not enough rows) consider swatching with the same size needle, but a grabbier material like wood or bamboo, to drag more yarn into the height of each loop. 
  • If you get stitch gauge, but your row gauge is too long (too many rows) consider swatching with a slicker needle--non-coated metal of some kind, perhaps anodized aluminum.
  • If you get row gauge but not stitch gauge, consider changing needle size until your stitch gauge is correct, and then switching up your row gauge by changing needle material as above
  • Or maybe, just don't worry too much about row gauge in the first place.  See, for top-down or bottom-up garments (which is most of them) row gauge is less important than stitch gauge, because you can always alter the pattern by adding or subtracting a few rows.  You can usually even fudge sleeve caps this way, too, after a few runs at it.  For side-to-side sweaters, though, row gauge is important!
'til next time


Thanks to everyone who posted in an old 2008 Ravelry thread which really got my gears turning in thinking about gauge so many years ago, with a special thank you to Beanmama who started that thread, eightoclock who really ran with it and yarnspinner who dropped a crucial hint. 

The essential swatches

Many thanks also to the various knitters, who, over the years, contributed research swatches, labeled as to yarn & needle size and composition.  

DOWN THE RABBIT HOLE (related links):