I-35 reconstruction through Norman, OK

[rte66man]

IMG_1495 by rte66man, on Flickr

IMG_1483 by rte66man, on Flickr

IMG_1486 by rte66man, on Flickr

IMG_1490 by rte66man, on Flickr

[Scott5114]

Great pictures. Check out the Lindsey sign in the first one, where instead of putting a MUTCD-style CLOSED banner on, they were trying to black everything out with electrical tape, but ran out of tape and ended up having to do it the correct way anyway. (There was a few days between when the tape and the CLOSED banner went up)

[rte66man]

Eastbound OK9 at NW 24th St:
IMG_1480 by rte66man, on Flickr

Westbound OK9 at I35.  You can see where the permanent northbound ramp will be:
IMG_1483 by rte66man, on Flickr

[Scott5114]

Some new signs have gone up, including this one:



(Pardon the bad image due to the rain and not expecting this sign to be up yet.) This appears to be Oklahoma's first attempt at an APL. Looks like a change order went through at some point here, since all of the rest of the signs in this project use Clearview. Of note here is that 1) the capital of Oklahoma is "Oklahoma C ity" now, 2) "EXIT ONLY" is apparently regulatory 3) there's no real need to make sure that the sign is level before you bolt it to the gantry.

[Plutonic Panda]

Incredibly annoying. I usually won't even look at the signs when I'm in Oklahoma or I get angry.

[Bobby5280]

This appears to be Oklahoma's first attempt at an APL. Looks like a change order went through at some point here, since all of the rest of the signs in this project use Clearview. Of note here is that 1) the capital of Oklahoma is "Oklahoma C ity" now, 2) "EXIT ONLY" is apparently regulatory 3) there's no real need to make sure that the sign is level before you bolt it to the gantry.

I'm guessing that the people doing the traffic sign design work and fabrication work (either for ODOT or for subcontractors) probably aren't getting paid worth a damn. They're obviously not using the two wet, round things on their faces to observe whether the sign's layout, fabrication and installation look correct. They're just punching in and punching out without any regard to actually doing a good job.

The "Oklahoma C ity" spacing is another hint that seems to confirm the sign fabricators are actually applying vinyl letters one at a time to the sign panel. Notice how badly off-center "Oklahoma C ity" is placed. It's flush to the right edge of the sign panel rather than centered in its portion of the layout. They obviously placed the I-35 shield and "North" later, centering both elements above the "Oklahoma C ity" letters. The "Exit Only" and arrow elements are hugged over to the right even farther rather than being centered with the other OKC elements on the sign panel. Maybe they did this odd-ball graphics placement as an attempt to center the copy and arrow over the right lane. Doing so would require a longer (and more expensive) sign panel for the display to look properly balanced. Yet it still disregards the effect of the middle lane. Overall it's a crappy layout.

But there's at least one positive thing to it: at least the lowercase letters weren't shrank to 75% of of their normal size. There's a bunch of horrible new signs bearing that glaring error on the first few miles of I-44 North of the Red River.

[Scott5114]

Actually, I think that this sort of error is more indicative of some sort of software problem encountered when fabricating the sign–bad kerning on the FHWA series font file is most likely, but it could be an issue with the program itself, or even an error like the sheeting slipping while being cut. Then someone just weeded the vinyl, slapped the transfer tape on, and stuck it to the panel. Even if they noticed the kerning error, what can you do? The vinyl's cut, it's too late to fix now, and they're not going to bother to re-cut it.

The tell-tale signs of manually-applied elements tend to look more like the subtle errors you find on button copy signs (which were usually manually assembled)–you might get slight vertical displacement of letters here or there, or rounded letters, especially 's', are sometimes rotated a few degrees. But the kerning is usually good–people have an intuitive instinct of what properly kerned text looks like. It's one of those things that's easy to do by eye but gets complicated when you try to work it out mathematically.

My theories about this being the result of a change order have been strengthened: the next panel (on the gantry behind this one) has gone up and it now references Ed Noble Parkway, which I explicitly remember the plans on the as-bid sheets not doing. My guess, entirely unsupported by any actual evidence, is that the city of Norman made a request to include Ed Noble on the sign, as the construction (along with some important but unrelated outside factors) has been absolutely slaughtering the businesses along Ed Noble Parkway, and the city is desperate to keep the area from becoming a long-term blight.

Another fun thing about this sign: it lies. You cannot reach I-35 from the left lane. That lane splits off to Ed Noble/Lindsey.

Someone has gone by and installed yellow patches over the black-and-white EXIT ONLY patches, though.

[Bobby5280]

Actually, I think that this sort of error is more indicative of some sort of software problem encountered when fabricating the sign–bad kerning on the FHWA series font file is most likely, but it could be an issue with the program itself, or even an error like the sheeting slipping while being cut.

Given my 20+ years of experience dealing with vinyl graphics, I say they're applying the letters one at a time. If the vinyl plotter is so badly out of whack that vinyl material is slipping while a cut job is in progress you will lose the entire job. Letters and other graphics elements won't get cut completely. You'll lose any sense of the baseline. As far as glitchy fonts are concerned, I don't know what software ODOT and its subcontractors use for creating traffic sign panels, but getting the tracking right is a very basic thing for any vector graphics application. I've used a lot of different traffic sign fonts since the early 1990's and never saw kerning errors on the magnitude of that "Oklahoma C ity" legend. That's just an example of a "designer" sleep walking through his job, or the likelihood fabricators are slapping traffic sign legends onto sign panels one letter at a time and doing a crappy job at it.

I also say they're doing the one letter at a time application approach given some of the odd-ball sign monstrosities here in the Lawton area. We're talking sign panels with letters of different sizes applied to the same legend. And not all of those letters are applied perfectly vertical either. You're not going to do that by accident. They're eye-ball placed on the panel, and badly at that. I don't care how much someone hates Clearview Highway, that kind of error is absolutely impossible for a font to deliver by itself. Even a bad piece of design software will at least keep the text object letters all the same height and same vertical level. It's like the ODOT sign makers are pulling individual letters out of a cardboard box and just sticking them to the sign in what ever manner they see fit.

[Scott5114]

I believe the software package OkDOT uses is called GuidSIGN, and if I am remembering right it is actually a plugin to AutoCAD, rather than a vector graphics program of its own. They definitely don't use anything like Inkscape or Illustrator.

Actually, the plans as produced by OkDOT include a table that lists each letter along with its X-axis position. If there's some sort of manual translation of those coordinates into the plotter, kerning errors like this could simply be the result of a typo or transposed digits.

[Bobby5280]

I believe the software package OkDOT uses is called GuidSIGN, and if I am remembering right it is actually a plugin to AutoCAD, rather than a vector graphics program of its own. They definitely don't use anything like Inkscape or Illustrator.

Inkscape has its own primitive issues. But Adobe Illustrator has far more advanced type handling than most other professional-level design applications. Adobe InDesign goes a little farther in some respects; Illustrator has some of its own exclusive bells and whistles as well. Sign industry specific design applications have lagged behind in areas like full support of OpenType features. Adobe has been way ahead in this regard for over a decade. CorelDRAW only finally started fully supporting OpenType in 2012 at version X6.

Regardless of the type technology used, the designer can always convert the type object to outlines, then check and fine tune kerning before the job is output to a vinyl cutting plotter or large format printer. It's not difficult to get the stuff right in the computer if the "designer" cares about those kinds of details.

I usually convert all the type in a completed sign layout to outlines to avoid font loading & substitution problems if a certain layout is pulled up years later to be reproduced again. There are many different "flavors" of the same named typeface, yet each build will have subtle differences in how characters are drawn and kerned. Character sets may be radically different. For example, the version of Times New Roman bundled in with Windows 10 is not the same Times New Roman included with Windows XP. Programs like CorelDRAW or FlexiSign Pro can deliver unexpected results with type when opening archived files made with a much older version.

Actually, the plans as produced by OkDOT include a table that lists each letter along with its X-axis position. If there's some sort of manual translation of those coordinates into the plotter, kerning errors like this could simply be the result of a typo or transposed digits.

When a cut job is sent to a vinyl plotter the plotter itself doesn't see any fonts. It's all just shapes to cut, usually in some form of HPGL code. The same goes for sending jobs to computerized routing tables, except that's often G-code being sent to the machine. If the position of glyphs is out of whack then it's an issue in the design software (and maybe a bug in the font itself). But it's an issue that can be overcome by inspecting the layout. When routing letters out of aluminum or other materials those shapes are "nested" in routing software to get the most letters onto one sheet of material. Things like plotted paper patterns are used to get routed letters correctly positioned on the final surface.

[Scott5114]

I think we may be arguing the same thing from different ends. My impression of what is happening is that there's a bug in the software, or possibly operator error, causing kerning issues, and nobody is doing the QA to notice or correct the errors. Therefore, when it's sent to the plotter to produce the paper pattern (or the spacing table), it contains errors that are preserved if the assembler faithfully follows it.

Then again, the "bag of letters" theory has some merit to it: there was a sign briefly posted in OKC for the "kilpatricK" turnpike. Someone switched the Ks, which is unlikely to happen by any other way than if they were just grabbing letters and nailing them down willy-nilly. (They later sent someone out with a bucket truck to swap the letters.)

Of course, it could be one thing in one situation and another on another panel. The root cause is carelessness or incompetence at OkDOT; the exact mechanism by which it manifests itself is not so important.

[rte66man]

Hey Scott (or anyone in the Norman area), is the bridge deck on Lindsey St finished yet?  I seem to remember they were nearly done when ODOT had the press conference about funding shortfalls.

[Scott5114]

I don't go over there much during daylight hours, but if I do tomorrow I'll try to remember to swing by there.

Last I remember is that some concrete had been poured, but the sites of the barricades makes it difficult to get a really good look at what's going on.

[rte66man]

My brother drives through there on a regular basis. He said all lanes in both direction are now open.  Not sure if that means 3x3 or just 2x2 but it's at least progress.

[jaysmc]

I just heard that the Lindsey overpass and all ramps will be open in the late afternoon/evening of Friday, July 21st.

[Scott5114]

Was just by there today. Definitely sounds believable–the deck looks totally done and the signals on the bridge are in place (though off, obviously).

Mainline I-35 is now striped for three lanes but every time I've been through there one or more lanes has been closed for whatever reason. They were still doing some kind of work on the underside of the bridge Monday night. There is no median lighting in place south of Lindsey yet.

[jaysmc]

There's an opening event Friday morning. Breea Clark, Ward 6, posted on Facebook: "Join us in celebration of the much anticipated opening of the Lindsey Street Bridge over I-35. Grab your bike, rollerblades, skateboard, or walking shoes and join Mayor Lynne Miller and your Norman neighbors in the first crossing of the Lindsey Street Bridge over I-35! All ages and any form of non-motorized transportation, with the exception of motorized wheelchairs, are welcome. The bridge and on ramps will formally open to motor traffic on the evening of July 21st. Parking available at Lander's Chevrolet of Norman, please meet on the west side of the Lindsey Street Bridge over I-35. 9:00am-10:00am, July 21st."

[Plutonic Panda]

If anyone could grab an pics that would be appreciated!

[jaysmc]

Just drove over the new Lindsey I-35 overpass. It's open.

[Scott5114]

When they took off the electrical tape from the Lindsey Street sign, the tape ended up damaging it, so now it's barely reflective at night 

[Scott5114]

Did you know Dallas is west of Norman now?

[rte66man]

Did you know Dallas is west of Norman now?

<sigh>  Oh, ODOT, how could you miss that one?

[rte66man]

If anyone could grab an pics that would be appreciated!

https://www.ok.gov/odot/I-35_SH-9_Lindsey_Street_Project.html

http://www.koco.com/article/crowds-gather-to-celebrate-reopening-of-lindsey-street-bridge-in-norman/10340230

[Bobby5280]

That's just a crappy sign all the way around. Even if they avoided the mistake with cardinal directions over the OK-9 and I-35 shields the whole layout looks poorly composed. The legends for Ed Noble Parkway and Lindsey Street are tiny compared to the Dallas legend on the right half of the panel. The I-35 shield appears to be using E-Modified for the numerals (rather than Series D). The shield is needlessly neutered. The OK state highway sign design continues to stink; I'd rather they go back to the generic white circle with numerals inside.

Overall this sign was about saving money. A larger, wider, overhead display would have been more effective, but cost more money to build & install. I guess people in Norman should be happy ODOT at least installed a ground mounted sign of substantial size. Here in Lawton at the I-44 & Gore intersection a large ground mounted sign like that was taken out by a vehicle. ODOT replaced it with a tiny sign more appropriate for a junction of 2-lane roads. I guess that's what we get with our "anti big government" state government and how they've totally wrecked the state's budget.

[Scott5114]

Since OK9 is East-West, I am going to assume the contractor mistakenly switched the directionals.

They actually switched the shields–I-35 is supposed to be listed first, after all (and as of Monday the shields have indeed been swapped to their proper location).

Overall this sign was about saving money. A larger, wider, overhead display would have been more effective, but cost more money to build & install.

There's an overhead sign further back containing the same information, but with "left lane" and "right lane" instead of the arrows.