Convexing thoughts

[tgbto]

So maybe there are weird things going on with your camera, but...

On pics 707 and 708, there are visible indents on the edge and on the black coating (I circled them in red). If they're visible on a low-res pic, they probably are quite obvious irl, and I think polishing would only increase the contrast.

I'd wager they come from the not-so-controlled up-down movement. The fact that it has to be there at all makes it extremely hard to have a smooth movement along the length of the edge. I'm sure you could get rid of those with a lot of practice, but still... If the process introduces an intrinsic weakness, maybe the tooling isn't adequate.

[Dutchman]

... snip
I'd wager they come from the not-so-controlled up-down movement. The fact that it has to be there at all ...

Remarkable observation, supported by the idea that the up-down grinding should cause this. In my opinion this is fatal for this method unless you practice and recover a lot.

[Thy Will Be Done]

So maybe there are weird things going on with your camera, but...

On pics 707 and 708, there are visible indents on the edge and on the black coating (I circled them in red). If they're visible on a low-res pic, they probably are quite obvious irl, and I think polishing would only increase the contrast.

I'd wager they come from the not-so-controlled up-down movement. The fact that it has to be there at all makes it extremely hard to have a smooth movement along the length of the edge. I'm sure you could get rid of those with a lot of practice, but still... If the process introduces an intrinsic weakness, maybe the tooling isn't adequate.

I have seen pixelation in photos uploaded to share online before which look exactly like those dips.  I don't believe the KJ-45 could make hard dips like that, it would likely be more liked rounded holes in the edge similar to half moons.

I have used the convex method with the KJ-45 and while your concerns are well warranted I do believe it's something that can be overcome with proper technique for the most part.  I certainly did not have that issue which the photo suggested.  What I did have was less than ideal blending in a few spots.

[tgbto]

AFAIC I ended up with some kind of a wavy pattern along the edge, which I believe were some high/low spots.

Just think about it: if you go up-down-shift-up-down-shift then you might ensure proper convexity, but legnthwise you're going to go shift-stop-shift-stop, which we know is bad technique.
If on the other hand you go up-down in a regular fashion but with a constant speed, you will create a wave pattern. So of course you can hope for a proper statistical distribution of the high and downs, but yeah, that means lots of passes and very good technique.

To be actually homogeneous, you would have to have a true triangular motion. Stop a bit on the stops and you overgrind compared to when you move. A sine wave is hardly better. But even reversing the motion means having the direction of the resulting force that moves with respect to the plane of symmetry of the blade, therefore you don't grind the same amount when going up as when going down. So a triangular motion with a different speed when going up or down. All in all this motion goes contrary to the consistency we try to build when grinding.

Again, I'm sure there are ways to improve, and maybe a few dedicated sharpeners will get a fairly decent result. Will it be close to what you get in 5 minutes with a 150$ belt sander and only a few hours under your belt ?

[cbwx34]

I think you could move the blue line upwards, so that it lies fully inside of the black lines.
This would illustrate the fact that the grinding wheel cannot add material to the knive.
Changing a flat grind to a convex grind with a fatter edge always reduces the height of the knife.
Grinding the flat grind to a thinner edge maintaines the knife hight, however increases the bevel hight - as you already explained.

You are correct... the diagram just shows the relationship, not the location of the actual grind.

[cbwx34]

So maybe there are weird things going on with your camera, but...
...

Must be it, as that is not visible or felt on the actual edge.  (FWIW It started as a digital zoomed pic, that was then reduced & rescaled...)  (Maybe look at 0711... I don't think it suffered the same abuse.) 

...
...Will it be close to what you get in 5 minutes with a 150$ belt sander and only a few hours under your belt ?

No, and like I said before, "there are better ways"... I'm not saying that it's the best convex grind.  You do get some of the "benefits" of a convex, for example a soft shoulder, bit of a convex shape, etc.  Of course there'll be some "defects" I said that before too... like you said a bit wavy, etc.  (The more you zoom, the more you see.)   But for what it is, I think it does OK, and I'm not sure they'd matter in use?

[tgbto]

You do get some of the "benefits" of a convex, for example a soft shoulder, bit of a convex shape, etc.  Of course there'll be some "defects" I said that before too... like you said a bit wavy, etc.

But then is it not better to just thin the knife by lowering the support bar one or two full turns before sharpening, get close enough to the edge and then sharpen at the desired bevel angle height ? That will soften the shoulders, result in a convexish shape, and be much faster and more consistent ?

[cbwx34]

But then is it not better to just thin the knife by lowering the support bar one or two full turns before sharpening, get close enough to the edge and then sharpen at the desired bevel angle height ? That will soften the shoulders, result in a convexish shape, and be much faster and more consistent ?

I think you'd end up with more high and low spots, (but realistically, in use I doubt it'd matter).  I don't think it'd be any faster or more consistent.  This went pretty quick for me.

Edit:  I took a quick video of the edge and a couple of slices (doesn't catch on anything)...

(for you to beat up on... )

(BTW, these were done on a T-4 198mm wheel.)

[Thy Will Be Done]

You do get some of the "benefits" of a convex, for example a soft shoulder, bit of a convex shape, etc.  Of course there'll be some "defects" I said that before too... like you said a bit wavy, etc.

But then is it not better to just thin the knife by lowering the support bar one or two full turns before sharpening, get close enough to the edge and then sharpen at the desired bevel angle height ? That will soften the shoulders, result in a convexish shape, and be much faster and more consistent ?

This is a nice idea but it's liable to have the same tendency to inaccuracies on a different level.  I think if you really care about high precision then it's best to use bench stones to finish.  I've not found a way to rival the precision I can get cleaning up bevels by hand on stones.  When I did KJ-45 convex for a few knives I always finished on stones to get a nice shaped convex.