SA profile Row 4 Shift molds - Ready

[blastoid]

Despite the fact that "SHIF T" doesn't bother me that much, I'd be willing to subsidize the cost by buying smaller shifts I would never use.

Also, I saw that Carbon would be shipped with R3. Any word on Danger Zone?

Edit: Ignore me, Danger Zone is all-R3

[zslane]

7bit's Round 5a may end up being the first community GB to deliver R4 Shift keys. Fingers crossed.

[tofgerl]

Is carbon not counted as community-run?

[zslane]

Carbon isn't delivering R4 Shift keys, at least not to my knowledge. Those molds weren't available when Carbon when into production.

[tofgerl]

Actually they will be the first ones made. Or so someone said.

[Oobly]

Actually they will be the first ones made. Or so someone said.

...Carbon will get row 3 shifts first. Later they (the row 4 shifts) will be purchacable either on MassDrop or PMK....

Emphasis and bracket part added by me.

[zslane]

Ah, well chances are good then that SP will put R4 Shift keys for popular SA sets (not just Carbon) up on the PMK webstore before Round 5a goes into production. The race is on!

[swill]

Row 4 1.25 and 1.75 caps already exist in SA profile.

What 1.75u SA keycap is in row 4 already?  Isn't the only 1.75u key on the board the caps lock?  That key is in row 3 in all sets.  Are we sure that there are 1.75u row 4 shifts already?

[jbondeson]

Row 4 1.25 and 1.75 caps already exist in SA profile.

What 1.75u SA keycap is in row 4 already?  Isn't the only 1.75u key on the board the caps lock?  That key is in row 3 in all sets.  Are we sure that there are 1.75u row 4 shifts already?

1.75 row 4 shift exists, but most designers choose to not use them otherwise you could have a 2.25 left row3 with a 1.75 right row4 which would be funky. So everyone just does all shifts in row3

[Oobly]

Row 4 1.25 and 1.75 caps already exist in SA profile.

What 1.75u SA keycap is in row 4 already?  Isn't the only 1.75u key on the board the caps lock?  That key is in row 3 in all sets.  Are we sure that there are 1.75u row 4 shifts already?

1.75 row 4 shift exists, but most designers choose to not use them otherwise you could have a 2.25 left row3 with a 1.75 right row4 which would be funky. So everyone just does all shifts in row3

Hack'd by Geeks, Jukebox and Retro SA have R4 1.75u Shifts.

[tofgerl]

Actually they will be the first ones made. Or so someone said.

...Carbon will get row 3 shifts first. Later they (the row 4 shifts) will be purchacable either on MassDrop or PMK....

Emphasis and bracket part added by me.

Oh, I assumed they would be produced right afterwards, but I guess that was only inferred if you really wanted it to be inferred...

[KHAANNN]

I think that is their goal, yes. To offer R4 Shift keys in color combinations matching popular sets from the past, but I don't know any of the particulars there.

As for corrected legends, this was Melissa's reply this morning when I suggested that SP use my vector definitions for all new Shift keys going forward:

We are revising all of our SHIFT legends so that the T is closer to the F, so moving forward, anything that we are currently running or plan to run will have the adjusted SHIFT legend.

Make of that what you will.

This R4 Shifts thing is really good news, they should've done this long ago

I still think SP is too funky with their legends, especially their new custom ones, the ones for the Selectric sets etc., they all seem Comic Sans-ish to me, their SA legends are actually usually pretty good and consistent, yet their DSA ones have no standard, they mix and mash font-sizes and weights too much, that's why I'm still weary of SP sets

[zslane]

SP isn't a keyset designer, it is merely a keycap manufacturer. The variety of legend styles you see is all the responsibility of the custom keyset designers out there. SP sells generic dyesub PBT DSAs with a fairly heavy typeface, but that is the only full DSA set of their own "design" that they sell (that I know of).

It is especially easy (and inexpensive) to put almost anything on a keycap when using the dyesub process, and so dyesub PBT DSAs will have typefaces and icon legends that are all over the place. This can be a good thing (Granite) or a bad thing (Galaxy Class), depending on the taste of the designer. But in neither case does SP get any of the credit or blame.

It is far more expensive to create custom double-shot plates, and so you will find far more consistency between double-shot sets. Most SA keyset designers simply go with SP's standard Gorton Modified typeface and then pay for the few custom icon/graphic legend plates their sets need. 7bit's Space Cadet and APL keysets are rare specimens in the large number of custom legends they have. But again, SP has nothing to do with these designs.

[KHAANNN]

SP isn't a keyset designer, it is merely a keycap manufacturer. The variety of legend styles you see is all the responsibility of the custom keyset designers out there. SP sells generic dyesub PBT DSAs with a fairly heavy typeface, but that is the only full DSA set of their own "design" that they sell (that I know of).

It is especially easy (and inexpensive) to put almost anything on a keycap when using the dyesub process, and so dyesub PBT DSAs will have typefaces and icon legends that are all over the place. This can be a good thing (Granite) or a bad thing (Galaxy Class), depending on the taste of the designer. But in neither case does SP get any of the credit or blame.

It is far more expensive to create custom double-shot plates, and so you will find far more consistency between double-shot sets. Most SA keyset designers simply go with SP's standard Gorton Modified typeface and then pay for the few custom icon/graphic legend plates their sets need. 7bit's Space Cadet and APL keysets are rare specimens in the large number of custom legends they have. But again, SP has nothing to do with these designs.

Not entirely correct, far from it, there are a lot of things you can't change as a designer, those things that you can't change is why I don't like SP

Some legends are slightly bolder than others on SA, "Enter" and "Return" for example, almost all the custom icons are way outside the designer specs, it reminds me a lot of this: http://www.artistsandillustrators.co.uk/uploads/tinymce/images/ruined-spanish-jesus-fresco-painting.jpg

What I actually hate SP for is their DSA texts, many designers include some awesome keycap selections, but for me all of them were unusable because SP's funky font-size's, almost none of their "Shift" DSA keycaps have the same "Shift" font-sizing, the 2u, 1.75u ones all have slightly smaller font-sizes

Their 1.75 "Control" keycap has a lower font-size compared to the "Tab" and "Shift" keycaps, so If you intend to use the 1.75 Control, you have to live with that unnecessarily small "Control" phrase

I thought this was because they were incompetent, but after seeing this: https://geekhack.org/index.php?topic=76705.0 I think it might also be because some of their DSA keycaps have a smaller double-shottable area, tho this doesn't explain the font-weight issues they have with some SA legends/phrases

TL;DR: No it's not the designer's fault

[zslane]

If we're talking about the double-shot legend process, then variability is the result of two things: font choice (or in the case of icons/graphics, the correctness of the vector file sent to SP) and manufacturing tolerances.

SP only has three typeface sets, and 99.9% of their double-shots are made with their Gorton Modified typeface set. That set comes in three styles: normal, condensed, and wide (basically normal with extended Os and Qs and such). Each comes in numerous sizes. Designers will work with SP to come up with the font that will fit best on each key, but if a designer decides he wants condensed Gorton Modified on the SCROLL LOCK key, but wide Gorton Modified on the CAPS LOCK key, that inconsistency isn't SP's fault.

Of course, those fonts aren't digital, they are old-school metal blocks, with all the kerning issues that go with that. When you talk about things designers can't control, the shape and spacing of Gorton Modified letters is definitely one of them, but its the same for all designers, so for instance all SHIFT keys are SHIF T keys; its not like one keyset will have SHIFT while another will have SHIF T.

When it comes to icon/graphic legends, there has been a history of miscommunication/misunderstanding on the part of designers and SP alike. Some designers simply don't know how to send a proper vector file to SP, and SP simply doesn't know how to inform/educate designers so they get it right. Oobly is working diligently to improve this situation. We all stand to benefit from his hard work in this area.

Minute variations due to the nature of the plastic double-shot process are impossible to avoid, at least with the machines SP currently uses. Not much can be done about that without them spending big $$$ on new tooling.

P.S. If it's true that Modern Selectric was designed by SP themselves, then that is one example where we can lay the blame for its poor aesthetics squarely at their feet. But that is an isolated case, is it not?

[user 18]

P.S. If it's true that Modern Selectric was designed by SP themselves, then that is one example where we can lay the blame for its poor aesthetics squarely at their feet. But that is an isolated case, is it not?

Modern Selectric was not designed by SP, it was designed by user skullydazed and run through PMK.

See:
https://geekhack.org/index.php?topic=72254
https://geekhack.org/index.php?topic=71540

[zslane]

Okay, then my original assessment stands: SP is not a keyset designer, just a manufacturer.

Of course, history shows there are still plenty of ways for them to make mistakes. But mistakes are fixable and not something a keyset designer has to settle for.

[KHAANNN]

Thanks a lot for the insight zslane

It's also good to hear at least things are improving

I will probably wait a while myself before getting another SP set, also, interestingly, I find their cherry replica DCS sets to be the most consistent ones

[Oobly]

I understand your concerns, KHAANN, and am aware of the majority of issues with the current legends. As zslane says, most of the more obvious issues with SA legends are either due to designers providing incorrectly specified legends (some scaling issues due to software differences, choosing mismatched line thickness, etc.) or choosing mismatched font options, or the inherent problem of the "manually kerned" / "unkernable" legacy Gorton Modified font legends made way back when.

SP are cutting new, better kerned SHIFT legends now, to be used on all SA sets in future, since those were the most glaringly in need of work. zslane's efforts (http://deskthority.net/workshop-f7/fixing-signature-plastics-shift-key-legends-digitally-t11522.html) have no doubt played a part in inspiring them to make this improvement.

I've almost finalised my detailed legend design specification document for designers to use as a reference (which will go up on the DT wiki and I'll post links when it's up) and will be introducing a new, properly kerned, modern font for SA keycaps which we want to get funded and cut (will start a thread on this by end of November) as an alternative to Gorton Modified.

An interesting point about the Cherry Replica legends is that that many of them appear to have been generated from scans of actual doubleshot caps, so some of the slight deformities / inconsistencies that I thought were due to molding by SP are actually present in the legend SVG files already and are cut and molded very accurately by SP. This bodes well for new legend designs and I think the new font will look very nice

[KHAANNN]

Just checked zslane's analysis, it's indeed very impressive

One wishes that keycap companies themselves were this thorough

[zslane]

Oobly, I wasn't aware that you were endeavoring to create an entirely new, digital, typeface for SP. What do you plan to call it? Is it going to be designed-by-community?
Strike that. I went back and re-read the Digital SHIFT thread on DT and saw all your posts on this subject. I had totally forgotten about that, though I confess I probably just scanned past it back when you first posted it because it seemed tangential at the time and I was buried in trying to get my center-line vectors all tidied up.

It's worth mentioning, I think, that any new legend typeface will be a "rounded" typeface whether we like it or not, simply due to the cutting method. I don't have any problem with that, but I'm sure there are folks out there who don't like characters and symbols with rounded ends.

[tofgerl]

By ends, do you mean where the serifs would be?

[zslane]

SP can not double-shot the mitred ends of Helvetica, and any attempt to duplicate its letter forms will always result in something resembling Helvetica Rounded instead.

Actually, to be more precise: the method used to produce plates for text legends employs a round cutting bit not unlike a drill bit. As such, the ends of any line it "carves" are as round as the cutting bit itself.

[Niomosy]

Thanks, Oobly, for all the hard work on that.  Glad to see this come to fruition.

[tofgerl]

Sure, the only way they could fix that would be to use tiny tiny drillbits to make the rounded edge-corners smaller and the edge look flatter. But of course I'm thinking of this as a DIY thing, which is pretty far from how things work in plastics. Not to mention the drillbits are already pretty damn small!

[zlittell]

The reason they a rounded is that you need to have a certain percentage of draft on the part so that the mold releases from the plastic correctly.  Sharp edges and minimal draft means plastic not flowing into corners completely and parts not releasing from tooling correctly.

[Oobly]

SP can not double-shot the mitred ends of Helvetica, and any attempt to duplicate its letter forms will always result in something resembling Helvetica Rounded instead.

Actually, to be more precise: the method used to produce plates for text legends employs a round cutting bit not unlike a drill bit. As such, the ends of any line it "carves" are as round as the cutting bit itself.

Yes, they use a round bit, but the diameter is just 0.012", so "rounded" does not have to equal "round"...



They (SP) have two methods for cutting legend molds, one using just a centre line and line thickness, which is how most of the Gorton legends are made and it results in round ends with diameter the same as the line thickness, and one where you specify an "area" to be cut away, which has very specific requirements, but allows you to have finer detail and practically any shape as long as the largest outside curve is less than 0.012", the gaps and islands are smaller than a certain size and the legend doesn't extend too far out over the first shot support wall... This is all inculded in the document I'll be posting soon to the DT wiki.

Hold your horses.. I'll get the font info out there soon enough. It takes a LOT of work to create a full set of SVG files for all the legends needed, and there's a lot of going back to do things over again when I discover some particular legend doesn't meet the specs, so I have to change sizes or kerning for a bunch of caps again, but there's slow progress being made.

[zslane]

SP seems to steer every custom text legend project towards the center-line method, which results in rounded ends. I don't think there's ever been a third-party legend set (i.e., not a novelty set)  that used the in-fill method. If yours is going to be in-fill, Oobly, then I think it will be the first of its kind. And expensive. And probably tricky for SP to get right. But I look forward to seeing the end result.

[CQ_Cumbers]

SP seems to steer every custom text legend project towards the center-line method, which results in rounded ends. I don't think there's ever been a third-party legend set (i.e., not a novelty set)  that used the in-fill method. If yours is going to be in-fill, Oobly, then I think it will be the first of its kind. And expensive. And probably tricky for SP to get right. But I look forward to seeing the end result.

Signature Plastics said that both methods cost the same, and I believe the Cherry Replica Legends used the in-fill method for a full set before. I think it is primarily the effort required, for designing so many keys consistently and within SP's restrictions, that has been prohibitive for all but the most ambitious projects. It remains to be seen how well SP executes the design, so plans are to check the final molded keys for errors and try to fix them if possible. Oobly really has been doing a lot of work.