Is there a story for how AF addresses astigmatism?

I have a new itch to scratch… I think it’s accepted (here) that AF resolves myopia by acting as a stimulus for axial shortening. But is there a story for how it is able to reverse astigmatism ?

I happened to install an ‘astigmatism’ app on android. They seemed to be selling the story that their exercises would align the ciliary plumbing in such as way as to get the lens to compensate. Chimes with

So I went looking around, and found a few interesting refs. Apoligies if they’ve been discussed previously.

Review papers are always good places to start. I found A review of astigmatism and its possible genesis

leading to

All seems very interesting… (but all this extra screen time is doing me no good at all…)


Thanks for bringing this topic to the forefront again, as this is also an itch that I have been wanting to scratch. I have mentioned my own experience before in a few posts, but here it is again.

  1. During the course of not wearing correction for 30 years the axis of astigmatism in my left eye moved considerably, making the left lens in my 30-year old pair of glasses useless. That in my right eye had not moved.
  2. The shift in axis in my left eye turned my astigmatism into mirror image oblique astigmatism, and I found something online about that being less visually disturbing than parallel astigmatism. I find that in uncorrected binocular vision my astigmatism is hardly noticeable, whereas in moncular vision it is much more obvious.
  3. In the space of 6 months the axis of astigmatism in my left eye shifted again, but not to where it had been before. This has made all my reduced left lenses with cylinder correction useless. They turn the Snellen into a Dali painting. This was not too much of a disaster as I used them only for measuring and the very little driving I do. I now drive without glasses.
  4. With the reduction of myopia in both eyes, the degree of astigmatism has also reduced considerably. This makes me wonder what people are doing leaving the reduction in cylinder correction to a later stage.

Learning anything more about astigmatism has no practical value for me - I will keep on doing what I have been doing (zero diopter) - but it would certainly relieve my curiosity.

1 Like

this is very strange, if astigmatism is to do with cornea shape. My axis has been measured differing by max 10 degrees over the course of 20+ years

makes sense - lucky you! Unfortunately my axes are almost parallel so my two eyes cannot compensate for each other. Maybe this is why I have found starting my EM journey so hard - I tried to take Jake’s advice at face value (dropping 1 D from differentials’ CYL) and it stalled my progress for 4-5 months while fighting with constant eye strain even though the SPH value was easily high enough for my computer distance. Only when I went back to full CYL +0.5D did I start to make progress - so I found cyl especially annoying and inhibiting my progress.

reading a lot on here I’m starting to think there are 2 types of people (minimum) regarding astigmatism:
those who Jake talks about who had it introduced at some point with 0.25 D then kept increasing due to stronger lenses.
Those, like me, who had it in their first pair of glasses - my first pair had already over -1 D of cyl on both eyes - do I REALLY have a hope of reversing THAT?? :frowning:
In my case I find it prudent to do as Jake says and leave cyl for a while, especially after my bad experience with the first few diffs trying to reduce cyl…

That’s interesting… I had been trying to rationalise why mixing cylinder was okay. While I have only a small amount of cylinder on one eye, I was concerned that I tend to read in bed without correction. If the eyes really are trying to adjust themselves to correct for corneal imperfections, then I’m giving them mixed signals by having different cylinder corrections at different times. You’d have had the same, with different cylinder on diffs and norms.

The other thing I had been trying to rationalise was why, if resolving ghosts was entirely down to the visual cortex relearning, why it was necessary to go in small steps. One might intuit that if it’s having to relearn things, it might as well do it once, in one big bang. But if it’s actually structural changes in the eye, then incremental steps does make a lot more sense. (And the anecdotal evidence is that it is incremental steps work better.)

… and then there’s people like me, for whom it all started so long ago that they have no idea what their first set of glasses were. (Other than that they hated them.)

1 Like

Your experience has conrfirmed this. I don’t remember if I had cylinder correction with my first pair of glasses - too long ago - but I think it unlikely. I did have a diopter gap from the beginning, but it was small, and its increasing to a 3D gap over the years was mainly the fault of the optos.

that’s not really a 3rd type - that’s a “don’t know if I’m type 1 or 2” type :smiley:

hmm I never thought about this - just that cyl was reduced too much. But maybe this also plays a role, thanks for the idea

I was lucky, after I found EM I asked my parents and they still had my childhood prescriptions stashed in a box somewhere.

I guess this is the only thing I have going for me, visually :-\ that I don’t have much of a diopter gap

also kind of confused how much the usual prescription complexity reduction path (from Jake) applies to me - due to doubts over whether my astigmatism can be reduced since at LEAST 1D of it is not lens induced. Guess I will have to try and see, which might involve more experimentation and lost time


Hmm… I wonder what “lens-induced astigmatism” entails… is it that lenses have introduced imperfections which the eye has adjusted itself to compensate for ? If that is the case, then is there any reason that should be easier/harder to undo than genetic astigmatism ? Presumably lens-induced astigmatism isn’t anything happening to the cornea.

So if lens-induced astig. is evidence that the eye internal optics can compensate for weirdness on the incoming signal, then just letting it do its thing can compensate for any genetic oddities due to corneal imperfections.

1 Like

let’s hope you’re right.

and introducing the different cyl levels for norms and diffs is the standard EM/Jake method anyway…

It seems that astigmatism is not a fixed entity, unless perhaps for genetic astigmatism, and that keeping it ‘captive’ with lens correction can keep it in place, or even make it worse by not allowing the visual cortex to deal with it in the most efficient way. My wandering axis of astigmatism confirms that it is not fixed. Jake had an interview with an opto some time ago in which she confirms the variability in degree and axis of astigmatism, even throughout the course of a day.
Even now, with zero diopter, I see slight changes in the degree and axis of astigmatism in my left eye from day to day, whereas the spherical measurement remains fairly constant.


I wouldn’t discount the possibility. There are nerves in the cornea, it could be stimulated to change. Even fingernails can be stimulated to change if you provide stimulus to the nail bed. Any tissue with cell turnover has the potential to remodel. That action would be much slower than anything the ciliary muscle can do.

The fact though that axis of astigmatism can wander over mere hours says that it is the faster acting system at work there.


never experienced this - any studies backing this up?

Wandering Axis of Astigmatisim

Astigmatism gone missing

This one addresses repeatability of an autorefractor, no lag time was introduced between measurements, they found around 5 degrees of variability:

This one found changes of 10 degrees (by autorefractor) between visits 2 weeks apart:

This one found different variability with different devices:

yeah OK within 10 degrees, I have seen
I mean the complete change that Hannie experienced…sounds strange

Over very long periods, complete change is normal. The very young and very old tend to have axis opposite of school-working aged people.

That’s interesting, are you saying that one should have a different CYL correction in the differentials v. the one in the normalized?

In all of the astigmatism-related posts and videos I’ve seen on EM so far, I haven’t come across that particular suggestion. I was under the impression that the only difference between normalized and differentials should be in the SPH correction.

it’s not stated explicitly, but it’s implicit when he says “drop the cyl or reduce by up to 1D for differentials”

1 Like

been thinking about this some more… from a purely image processing perspective

normal spherical blur results in distorted information reaching the retina. This actually must result in LESS USEFUL information content of the light rays for the brain hence the brain cannot hope to reconstruct the missing information purely by image processing. This is why, I think there is some consensus on this forum that normal AF MUST be doing something physically with the cornea/lens/front of the eye - many people can even feel something happening there - and this allows the slight improvement in the path of the light rays hence improving the physical information quality reaching the brain in the first place.

At the same time, astigmatism often results in double vision - ok it is still a kind of angle-specific short-sightedness modifier - so you DO get some loss of useful information, but it is only partial compared to spherical blur - e.g. where you get (assuming sph blur is corrected) the correct information on some meridians. This COULD allow the brain to use the correct information in 2 ways: some meridians provide good light signals/information and if it manifests as double vision - you just get 2 of the same thing - this type of error COULD be eliminated or compensated for by image processing in the brain - it has enough information physically (“I know what the thing looks like, just gotta filter out one copy of it” vs. “I don’t even have enough information to know what it looks like”).

This would imply that it is not AF that helps astigmatism, but as suggested on this forum, also by e.g. @kem, that practice or training of the brain to fuse the double images is what helps it to learn this kind of image processing to improve on the negative effect of astigmatism on images.

Probably forgetting a ton of things here, but I don’t see how AF, as we experience it helping spherical blur, could compensate for astigmatism unless your ciliary muscle could bend the lens only in certain meridians, which I doubt…

…and this would mean that when practicing AF, it is not that which is improving cyl, but your focus/engaging your brain at the same time as AF-ing, teaching it to fuse doubled lines

I am here to get information about astigmatism. My .5 cyl is causing blur adaption and robbing spherical improvement.

@francois and @Lajos
I’m not 100% sure, but I think the reason for making cylindrical reduction in differential lenses is because within the “edge of blur” you can see without a corrective lens. When you use differential lenses, you need less correction in order to get positive stimulation because you are viewing closer to your edge of blur. For example my spherical is -1.50 with 71 cm to the edge of blur horizontally, but an additional -.5 cyl is needed to compensate for 53 cm edge of blur vertically. Within 53 cm I can see perfectly without correction. After that I need increasing cylindrical correction until 71 cm where the need for spherical correction begins.

To calculate your edge of blur measurements into diopter numbers use @jakey 's calculator