The Clearview thread

[BigMattFromTexas]

To be honest, the lettering of Clearview I actually kind of like, its mostly the numbers that are so hideous that even makes a blind person's eyes bleed.

I used to but now that I see more and more of it around San Angelo and all around Texas I seem to like it ALOT less...
Im actually becoming more fond of highway gothic, which I know is the exact opposit of what I said in the very first post of this thread.
BigMatt

[myosh_tino]

There's ongoing construction on the Bronx River Parkway in southern Westchester County.

The County DPW is showing it's absolute brilliance with the new signs:


Not only are they using Clearview, but they're borrowing NYSDOT's "box the street name" technique. Double fail!

My eyes!  My eyes!  
I'm sorry but that sign looks absolutely hideous!

[Duke87]

Hey, if you think that's hideous, check out the gore point sign:

[vdeane]

I've seen some strange signs around NY but that one takes the cake!

[Brandon]

To be honest, the lettering of Clearview I actually kind of like, its mostly the numbers that are so hideous that even makes a blind person's eyes bleed.

Agreed.  The Clearview numbers are bad, especially the 6 and the 9.  The letters are tolerable, but the "g" needs a trim, and the "l" could do without the tail.  I don't mind the tail so much on the "y".  The capital Clearview letters though, IMHO, look pretty decent.

[exit322]

And a good reason Clearview is catching on:

My wife works in downtown Akron and takes the part of I-77 three days a week where the Clearview signs were posted.  When they first started going up, she said to me, "they're putting up new signs on 77, and they're a lot easier to read when driving through."

[mightyace]

"they're putting up new signs on 77, and they're a lot easier to read when driving through."

Yes, because they're so ugly you can't help but notice them! 

[Brandon]

"they're putting up new signs on 77, and they're a lot easier to read when driving through."

Yes, because they're so ugly you can't help but notice them! 

LOL!  As far as I can see it, the increased readability seems to be due to the sheeting, not the font.  I'm sure that FHwA font at the same size and reflectivity would be just as legible as Clearview.  Problem is, I've never seen the two ever put side-by-side at the same font size and reflectivity when one is promoting Clearview.

[jdb1234]

^^ Looks like ALDOT is replacing all older button-copy signage with the Clearview signs along US 31/Red Mountain Expressway.

You would be correct.  I noticed most of them 2 weeks ago.  There have been a few put up over the past few days.  An interesting thing is that the last sign I posted, the 8th Ave S is still in button copy. 

[Hellfighter]

I'm Speechless! 

[vdeane]

LOL!  As far as I can see it, the increased readability seems to be due to the sheeting, not the font.  I'm sure that FHwA font at the same size and reflectivity would be just as legible as Clearview.  Problem is, I've never seen the two ever put side-by-side at the same font size and reflectivity when one is promoting Clearview.

I don't know about reflectivity, but you can get same size on I-90 east at exit 37 - the 1 mile advance sign has not yet been made Clearview, but the next one has.

[Riverside Frwy]

There's ongoing construction on the Bronx River Parkway in southern Westchester County.

The County DPW is showing it's absolute brilliance with the new signs:


Not only are they using Clearview, but they're borrowing NYSDOT's "box the street name" technique. Double fail!

My eyes!  My eyes!   
I'm sorry but that sign looks absolutely hideous!

That's an understatement.This sign makes me want to rip the eyes from my sockets and throw it in the street so I won't have to look at it. 

[realjd]

LOL!  As far as I can see it, the increased readability seems to be due to the sheeting, not the font.  I'm sure that FHwA font at the same size and reflectivity would be just as legible as Clearview.  Problem is, I've never seen the two ever put side-by-side at the same font size and reflectivity when one is promoting Clearview.

You personally may not have seen it in a test like this, but keep in mind that there was extensive research done that did show a significant increase in legibility (at same font size and reflectivity), especially among older drivers. The font was designed by engineers for legibility, not necessarily aesthetics.

If you're interested, the FHWA Interim Approval summarizes some of the research and the results, and provides citations if you're really motivated to dig up the papers:
http://mutcd.fhwa.dot.gov/res-ia_clearview_font.htm

The ClearviewHwy company includes a great writeup explaining the development of the font, including some of the testing that was done:
http://clearviewhwy.com/ResearchAndDesign/index.php

And this NY Times article is also a good read about the new font
http://www.nytimes.com/2007/08/12/magazine/12fonts-t.html

Again, I'm not saying it's necessarily a nice looking font, but it IS more readable, regardless of what your gut feeling may be.

[J N Winkler]

It is also worth noting that the type community does not see aesthetics the way we road enthusiasts do.  Formally Series E Modified, and the FHWA alphabet series in general, are what are called neo-grotesque typefaces:  i.e., relatively recently designed sans-serif typefaces where the stroke width is more or less uniform throughout the letter.  (Typefaces with this characteristic are also often called gothic, hence the common name "Highway Gothic" for the FHWA alphabet series.)  In contradistinction, the Clearview typefaces are examples of what are called humanist typefaces, which are sans-serif typefaces where the variation in stroke width through each letter is such that the typeface has an almost calligraphic appearance.

http://en.wikipedia.org/wiki/Sans-serif#Classification

The type community is full of snobs who prefer humanist typefaces over neo-grotesque and gothic typefaces, regardless of design role, because they think the greater geometric complexity equates to added sophistication and thus aesthetic appeal.  This is quite independent of the comparative field testing carried out by TTI which showed that Clearview has legibility advantages over Series E Modified which are marked for certain driver groups and for certain combinations of sheeting type.  The type snobs are not aware of this research and even if they were led to it, they would not consider it important or relevant to anything they are interested in.  From their point of view Clearview rules because it is a humanist typeface, full stop.  (While there is a case to be made for the involvement of type designers in the planning of a traffic signing system, type snobs tend to be myopic and to fail to realize that the real role of the type designer is not to design signs, but rather to develop a robust system which engineers and others not trained in typography can deploy to design signs which will both fit the context and look attractive.)

The generalization is made (including in the Wikipedia article linked to above) that humanist typefaces tend to be the most readable of the sans-serif typefaces.  This, however, is not the same as saying that a given humanist typeface is infallibly more readable than a given neo-grotesque typeface, much less that all letters in a given humanist typeface are more readable than all letters in a given neo-grotesque typeface.  Nevertheless, the type community has a monopolistic hold on the business of packaging prestige (at least for type) and tends to portray a preference for Series E Modified as showing a lack of sophistication.

Personally, I think it is entirely reasonable to prefer the look of Series E Modified, and I think there is a continuing role in traffic sign design for the FHWA alphabet series in general even if Clearview's legibility advantages are taken as a given.  But the field and tachistoscope testing of Clearview is hard to argue with (except where digits are concerned--as far as I can tell, the testing never looked at Clearview digits in detail), and the preference of the opinion-formers in the type community for humanist typefaces in general makes it difficult to argue that Clearview should not be used because it is ugly.  Right now I have an uneasy feeling that it is inertia, more than the ability to make a positive argument in favor of continuing to use it, that keeps Series E Modified appearing on signs in a majority of states.

[Scott5114]

But the field and tachistoscope testing of Clearview is hard to argue with (except where digits are concerned--as far as I can tell, the testing never looked at Clearview digits in detail), and the preference of the opinion-formers in the type community for humanist typefaces in general makes it difficult to argue that Clearview should not be used because it is ugly.

Except I think there has been a study showing either no benefit or even detriment to Clearview if used on dark-on-light installations, thus why the B series hasn't been approved by FHWA. If that is true, is there really much benefit to having the font, considering it would make the look and feel of positive-contrast signage completely different from negative-contrast? (Road signing is all about consistency, or should be, in any event.) 

I believe that more field testing should be done. I conducted some informal tests during my senior year of high school, where I had someone stand across the room and I held up two sheets of paper, one with Series E Mod, and one with the equivalent Clearview, and lo and behold, most people said E(M) was more legible.

Also, as a free software/open-source enthusiast, and taxpayer to boot, I find it objectionable for the government to spend $600 per workstation on a copyrighted typeface when a public domain typeface exists that can do the same job and has for the past 50 years.

[J N Winkler]

Except I think there has been a study showing either no benefit or even detriment to Clearview if used on dark-on-light installations, thus why the B series hasn't been approved by FHWA. If that is true, is there really much benefit to having the font, considering it would make the look and feel of positive-contrast signage completely different from negative-contrast? (Road signing is all about consistency, or should be, in any event.)

I think I may have seen the same study too--wasn't it done by TTI also?

The requirement for consistency in traffic signing has more to do with things like use of the same colors for the same meanings, same sign for same application, etc. than it does with typography.  Clearview adopters are far from alone in using typefaces from different type families on traffic signs--for instance, Spain uses Autopista (looks like Series E Modified) and Carretera Convencional (looks like Transport Heavy), while in France L1/L2 and L3/L4 (which are from different type families--L1/L2 being neo-grotesque while L3/L4 is an italic humanist face) can appear on the same sign.  Personally, I think multiple type families on the same sign can look attractive as long as the functions of the respective type families are clearly delimited, as is the case with Clearview when it is used only for white on green legend and not in shields or on "EXIT ONLY" bottom panels.  Unfortunately I haven't seen a lot of consistency on the last point--when I saw a lot of pictures of Clearview signs posted here with "EXIT ONLY" in Clearview, I went through a few old TxDOT signing contracts and realized how hit and miss it was to have the bottom panel in Series E or Series E Modified instead of Clearview.

I believe that more field testing should be done. I conducted some informal tests during my senior year of high school, where I had someone stand across the room and I held up two sheets of paper, one with Series E Mod, and one with the equivalent Clearview, and lo and behold, most people said E(M) was more legible.

For characters printed in black on a nonreflective white surface, one would expect Series E Modified to be more legible.  Keep in mind also that you were working with young eyes, and probably some advance familiarity with the legend unless you went to a lot of trouble to hide your test cards, all of which are factors which had to be controlled in the various Clearview studies by expanding the sample size and using nonsense words (e.g. "Player" on the TTI test signs).

My own experience is more mixed.  I have tested Clearview several times by posting Series E Modified and Clearview signs to this forum, and then trying to read them on my laptop screen from across the room.  Obviously I know what these signs say in advance, having made them myself, so it is not a pure recognition test.  There is also some divergence between seeing the signs internally illuminated on a LCD screen and seeing real signs illuminated by retroreflective sheeting at night.  Nevertheless, when I am standing against the far wall, I can't read either Clearview or Series E Modified but Clearview does look less blobby.  When I start walking slowly toward the laptop, I feel like I can see what the Clearview legend says before I can decode the identical Series E Modified legend.  The real acid test, of course, is when I don't know what the signs say, and I can't do this myself as long as I make up the signs.

Also, as a free software/open-source enthusiast, and taxpayer to boot, I find it objectionable for the government to spend $600 per workstation on a copyrighted typeface when a public domain typeface exists that can do the same job and has for the past 50 years.

There isn't really any difference between Clearview and the FHWA alphabet series from this point of view.  In order to secure approval from FHWA, the Clearview designers had to renounce all trademark claims.  This means that anyone is free to generate his or her own versions of the Clearview typeface for use with traffic signing software, just as is the case with the FHWA alphabet series.  The fact that no-one has come forward to do so (aside from Michael Adams, who came up with Roadgeek versions of the Clearview typefaces which he licenses only for non-commercial applications) means, I think, nothing more than that first-mover advantage is durable for computer fonts.  It can be argued (and I have so argued in the past) that the fact that the Clearview type supplement on the MUTCD website has the glyphs as rasters, while the SHS book has the FHWA alphabet series as vectors, means that the Clearview designers are attempting a form of trade-secrets protection in the Clearview supplement.  On the other hand, the resolution is high enough that fonts compiled by third parties from the Clearview supplement would approximate the "real" Clearview fonts at least as closely as the various versions of Series E Modified that are available for use with sign design software.  (There is, for example, a Series E Modified font widely used with certain design packages where the C is just wrong.  At the high-end consumer section of the market, the 5 in Page Studio Graphics' Pixymbols version of Series E Modified is just wrong.)

In regards to open-source software more generally, it is a sad fact that state DOTs use closed-source, platform-dependent software (always Windows, never Linux or any version of the Mac OS) almost exclusively, and the costs on a per-workstation basis are far in excess of the cost of a Clearview font license.  For instance, in order to do sign design you need to have specialist sign design software, all of which is closed-source and most of which is quite effectively dongle-protected.  SignCAD, for example, costs in the thousands of dollars (between $2000 and $3000 per workstation, I think).  Even a stripped-down viewer version of SignCAD which allows *.SGN files to be viewed and converted to other formats but not edited costs almost a thousand dollars.  GuidSIGN is comparable in price.  Integrating SignCAD output into construction plans takes a CAD package, and I think both MicroStation and AutoCAD are over a thousand dollars for a single-workstation license (not sure about volume discounting).  MicroStation and AutoCAD by themselves are not much use without specialist civil engineering design software, like Geopak or Inroads.  The costs don't stop there either.  Unlike consumer-grade programs like Inkscape or CorelDRAW, or even Google SketchUp, these programs are not designed to present a shallow learning curve to first-time users.  This means that anyone who trains a CAD operator, for example, has a significant training investment which a competing organization gets almost for free when it poaches the employee--usually it is the state DOT which loses out to a consultant because DOTs in general have to trade low pay for job security.

So, in short, a $600 license for Clearview is not that much when you have to buy something like $5000 worth of other software in order to get value from the $45,000 annually (not including your share of FICA, plus any other benefits like health insurance which you offer) you pay to cover the workstation seat.

It is possible to make a case for a reduction in design costs, including in design-related overheads such as the cost of software and font licenses, but I suspect that this is a relatively low priority for most state DOTs.  The mantra is to pick the low-hanging fruit first, and in the heavy civil construction sector (including highway construction) the construction phase affords the greatest scope for savings in out-turn cost.  Consultant fees for final design are typically around 5%-10% of the total construction cost, so the excess above estimate that can result from not getting competitive bids (which happens when the market for construction services shrinks to the extent that contractors are not "hungry," and bid on a "take it or leave it" basis) is often greater than the design fee.  State DOTs in general would love not to have to deal with a brain drain, but in reality this is possible only if the workload in design offices is consistent over a long period of time, which hasn't really been the case since the period of first Interstate construction.  Use of consultants tends to be on a higher percentage basis for large projects than for projects in general (in Kansas, for example, KDOT uses consultants for 70% of its design work overall, but for 100% of the major projects) because state DOTs are no longer equipped to handle step changes in workload.  Consultants can do it because they pay their employees more partly to accept a higher risk of relocation.  (Caltrans has a low level of consultant utilization because of a legal [constitutional?] requirement that plans for work on the California state highway system has to be sealed by Caltrans civil PEs, but it can be argued that the flipside is shown in chaotic project management, the high volumes of work done by local agencies and coordinated through Caltrans OSFP, large-scale project buybacks like the Bakersfield freeway network, and even Caltrans touting single-handed bridge design in its attempts to recruit new civil engineering graduates.)

To return to sign design, in principle you could use Inkscape or even CorelDRAW instead of, say, SignCAD to produce sign designs for insertion in construction plans.  There have been reports of CorelDRAW being used for sign design in Canadian design offices, where the production environment and culture is somewhat different from that prevailing in US state DOT and consultant offices.  The problem, however, is that US transportation agencies use, and require their consultants to use, expensive CAD packages because they want drawings produced to true scale, so that what is shown on the plan sheet, and what is notated in dimensioning callouts, are always in a consistent relation to each other (even for sign designs which are often notated "NTS" or "Not to scale," meaning in this case that the sheet is not necessarily drawn to an uniform scale but that the elements in each sign design are in correct proportion to each other).  This is very difficult to accomplish with commercial vector graphics packages, even when they are enhanced with custom scripts, while in a CAD program it is easier to tell when you are using dimensioning callouts which are inconsistent with something that can be built in the real world.  (In both you can make drawings that tell barefaced lies--e.g. by specifying an inner diameter for a pipe which is greater than the outer diameter--but the CAD environment is more likely to keep you honest, partly because it is easier to take measurements in a CAD program, and partly also because the major elements of most CAD drawings are actually produced in overlay programs which are specifically engineered not to produce output which is inconsistent with reality.)

I apologize for the windiness of this post and for roaming onto more general "Plan Production 101" issues, but I hope the foregoing gives some idea why I think issues like the cost of Clearview workstation licenses are oversold as reasons not to use Clearview.

[Scott5114]

I believe that more field testing should be done. I conducted some informal tests during my senior year of high school, where I had someone stand across the room and I held up two sheets of paper, one with Series E Mod, and one with the equivalent Clearview, and lo and behold, most people said E(M) was more legible.

For characters printed in black on a nonreflective white surface, one would expect Series E Modified to be more legible.

Having a laser printer nearby with no personal consideration for ink costs, I did the test cards in white on green background.

Keep in mind also that you were working with young eyes, and probably some advance familiarity with the legend unless you went to a lot of trouble to hide your test cards, all of which are factors which had to be controlled in the various Clearview studies by expanding the sample size and using nonsense words (e.g. "Player" on the TTI test signs).

If I remember correctly, I was using random control cities from I-40.

I do have to wonder though what you mean by "young" eyes. I personally find the FHWA Series typefaces easier to read. However, though I am young, I am far from possessing 20/20 vision; in fact, at work I often find I cannot read the credit meter on a slot machine when standing one bank of machines away (about 4-6 feet or so).

The real acid test, of course, is when I don't know what the signs say, and I can't do this myself as long as I make up the signs.

Could you script your graphics programs to produce random text output of random length and capitalize the first letter? It would be for the most part unpronounceable gibberish, but you would have the benefit of not seeing the legend beforehand. If you wanted to spend some time on it you could probably create a simple algorithm for creating pronounceable nonsense words.

There isn't really any difference between Clearview and the FHWA alphabet series from this point of view....

That may be how the sausage is made, but I don't have to like it. I would like government at all levels to use more open technologies, as they in general are cheaper, better, and less prone to obsolescence than their copyrighted counterparts. Massachusetts was a leader in this regard, specifying OpenDocument as the standard for office documents in state government, but Microsoft lobbied until their policy was amended to add their botched Open Office XML format as an acceptable standard. Unfortunately, I think that corporate voices carry too much power with government for change to ever be realized in this matter.

[J N Winkler]

I do have to wonder though what you mean by "young" eyes. I personally find the FHWA Series typefaces easier to read. However, though I am young, I am far from possessing 20/20 vision; in fact, at work I often find I cannot read the credit meter on a slot machine when standing one bank of machines away (about 4-6 feet or so).

Where Clearview is concerned, by "young eyes" I mean eyes with good light sensitivity, the ability to accommodate quickly to sudden changes in lighting, and a very high tolerance for nighttime glare.

Personally, I am in my mid-thirties, so I am not (yet) middle-aged but my vision, particularly in high-demand situations like at night, is noticeably worse than it was when I was in my early twenties.  I have more trouble with glare and I won't drive in pitch-black dark if I can realistically avoid it.  And even when I was twenty, I still preferred to drive under clear skies and a full moon when I had to drive at night.  My vision has not been 20/20 since I was seven but it has long been 20/20 with corrective lenses (glasses from age 9, contact lenses from age 14, and a mixture of glasses and contact lenses since age 23).  My glasses prescription is somewhat out of date but it is still close enough that my glasses keep me ahead of the 20/40 night vision requirement for driving in Kansas.

Older drivers are susceptible to problems like dry macular degeneration which cut visual acuity.  Even if this and other causes of lost visual acuity are left out of consideration, older drivers still have to deal with a range of problems connected with normal degradation of vision, such as clouding of the lenses over time (which cuts the light that reaches the retina and can eventually require cataract surgery), a slowing in pupil reaction to change of light (the pupils in older drivers are slower to open up again when the lighting level drops), etc.  This, combined with the results of TTI's Clearview tests, leads me to think that there is something of a generation gap in Clearview acceptance--you hear lots of relatively young people who have no use for Clearview, but very few complaints about the changeover to Clearview from older people.  If I had to guess, I would say the median age of members on this board is probably in the mid- to late twenties, so I don't find it surprising that a solid majority of people on here don't care for Clearview.

Could you script your graphics programs to produce random text output of random length and capitalize the first letter? It would be for the most part unpronounceable gibberish, but you would have the benefit of not seeing the legend beforehand. If you wanted to spend some time on it you could probably create a simple algorithm for creating pronounceable nonsense words.

I could try that, except I am not sure there is a random-number generator.  I would also have to write a script to export the results to PNG to remove the effects of text rendering within the graphics program (the Roadgeek Clearview fonts are not particularly well-hinted and are best worked with at high zoom).  As it happens, I have to work on different parts of this problem, since at present I have no automatic mechanism for exporting my images to PNG.  This may be a holiday project.

That may be how the sausage is made, but I don't have to like it.

I don't like it either.  I particularly dislike the steep learning curve connected with anything CAD.  I just spent an hour this afternoon trying to get SignPC (a free TxDOT bolt-on program which is designed to allow you to put together basic signs in MicroStation 8.05) and I was never able to get it to produce complete text strings instead of widely spaced individual letters with little edge guides for spacing, or to figure out how to put together the different parts of the sign cell to finish the sign around the legend.  The help file was full of phrases like, "Do a fence move," "Snap to points," "Turn on level 2," etc.  Nowadays the impulse is to laugh at anyone who talks about buying a "X for dummies" book to get started with a consumer-grade program, but I am pretty sure that rather old-fashioned approach still obtains for major CAD programs because you have to have a lot of basic conceptual furniture in place just to understand what the help file is trying to tell you.

I would like government at all levels to use more open technologies, as they in general are cheaper, better, and less prone to obsolescence than their copyrighted counterparts.

The trouble with that position (which I tend to agree with in principle) is that there has to be an open-source option which offers the required functionality.  That is the case for standard "office productivity" software packages like word processors, spreadsheets, and database programs, but I do not know of a single open-source CAD program.  Also, once the agency has committed to a particular program, whether it is open- or closed-source, the agency has a lot of work (i.e., paid-for value) embedded in formats characteristic to the program in question.  This "legacy" material leads to switching costs if the standard program is changed, along with the running costs of keeping the material readable to later versions of the same program.  It is not typically possible to avoid these problems of cost and commitment unless the open-source software and open-source formats are around at the time the agency has to commit to a particular program for that application.  These have to be traded off against the significant costs in forgone efficiencies which result when an agency refuses to adopt a program for an application where no open-source alternative exists.  For instance, any state DOT which refused to adopt a closed-source CAD program would still be paying an army of draftsmen to draw construction plans by hand.

State DOTs have a mixed record in adopting open-source alternatives.  PDF, which many use for construction plans, is an open format, but it is designed to accept elements (such as images) in closed formats like TIFF.  I am personally not aware of any lossless open-source formats for bitonal image data (such as scanned construction plan sheets) which matches TIFF with CCITT Group IV compression for efficiency.  (JBIG--which I hate--and DjVu are not lossless.)  CAD programs are strictly closed-source, but I think the file formats are at least partly open.

Unfortunately, I think that corporate voices carry too much power with government for change to ever be realized in this matter.

It is a tug of war.  Open-source options are attractive for highway agencies because management of legacy material is a huge issue for them, and the openness of the source code helps guarantee against obsolescence.  On the other hand, the vendors of closed-source software have a robust business model and a steady funding stream and that gives them the resources to maintain sales pressure, e.g. by persuading their customers that they "need" features which they were quite happy to do without in the past.  A case in point is the various attempts that have been made to sell 3-D functionality to highway design professionals (it is of very doubtful value except as a way of checking coordination of horizontal and vertical curves in perspective--a friend tells me that adding 3-D attributes to an existing design is an expensive and time-consuming proposition, which is normally resorted to only when the honchos demand it specifically).

[Scott5114]

...I do not know of a single open-source CAD program.

There is one called QCad, which I have installed on my computer once before through the Linux software repositories, but I have never used it for anything, and since I don't have any CAD experience I don't really know how it compares to professional software and really don't have much use for it.

[J N Winkler]

http://en.wikipedia.org/wiki/QCad

Apparently it is an AutoCAD clone (uses DXF format both internally and externally) and development began in 1999.  Ports to Windows are unofficial, and only the "community" edition is made available through the GNUPL--the functionality it offers is restricted compared to the full version.  3-D is not supported in any version (though highway design professionals like my friend would say, "You don't need it").

In contradistinction, the first AutoCAD release came out in 1982, while the first version of MicroStation that was not a viewer (2.0) came out in 1987.  This means a minimum 10 years when state DOTs would have had to wait for open-source CAD alternatives if they didn't want to buy into closed-source programs.

There are a couple of programs called FreeCAD or some variants thereof, but both were designed single-handed and are oriented toward design of 3-D solids.

http://en.wikipedia.org/wiki/Comparison_of_CAD_software

MicroStation, it turns out, has an interesting history:  it was originally a viewer (PseudoStation) for design files produced using Intergraph's Interactive Computer Graphics Design System, which required expensive, high-end VAX workstations.  (Who uses VAX anymore?)  PseudoStation allowed users to use a PC instead, the savings being so high that the Bentley brothers were able to sell a decent number of copies at their asking price of $7,943.  This prompted them to develop their own PC-based CAD package which would incorporate functionality from the ICGDS.

http://en.wikipedia.org/wiki/MicroStation

Intergraph has been around since 1969 and is of military origin.  The original motivation for developing computer graphics capability was to allow visualization of missile trajectories.

http://en.wikipedia.org/wiki/Intergraph

Wikipedia says that the development of PC-based CAD allowed companies to replace four or five draftsmen with a single CAD operator:

http://en.wikipedia.org/wiki/Computer-aided_design

I am pretty sure that computerization of highway design (though not of plans production) began in the US in the late 1950's/early 1960's with the use of punch-card machines for optimization of earth-moving.  Pen plotters were also available by the late 1960's at latest, and around 1970 MnDOT started plotting designs for freeway guide signs (for insertion in construction plans sets) which were pattern-accurate except for the route shields.  I think these designs must have been generated through key entry at a dumb screen terminal (then a new technology), or possibly even with punch cards.  I understand that the remote ancestor of the current SignCAD program is the collection of routines which MnDOT used to confect freeway guide sign designs, and that the current SignCAD company is a spinoff/privatization which retains close links to MnDOT.

It was actually not all that common for state DOTs to prepare pattern-accurate sign designs before about 2000, when the required functionality became easily accessible within signing CAD packages.  Before then, what designers typically expected from a signing CAD program was a set of drawings showing correct positioning of shields and legend on each sign panel.  In the age of demountable copy, it was not strictly necessary to have the correct outline of a given letter:  all that was needed was the correct location of a consistent reference point such as the top left of the letter.  Quite a few designers and state DOTs discounted the ability of pattern-accurate sign designs to serve as a quick visual check on the fabricated sign.  Instead, they prepared dimensioned sign sketches, and relied on secondary guidance (such as sign drawings books and spacing tables for the various traffic sign alphabets) to ensure the signs were correctly fabricated.

A few states did bother with pattern-accurate sign designs and a variety of techniques were used to produce them.  One common approach was to form the letters on little pieces of paper, which were attached in the correct positions using rubber cement; the finished result was sent out for reproduction.  I think this technique was used by Arizona DOT and Nevada DOT in the late 1950's/early 1960's, by MnDOT through the 1960's, and by PennDOT from the late 1950's to the mid-1980's.  It was also possible to buy lettering stencils with the correct outlines of the individual letters in the traffic sign alphabets, and these were used by Arizona DOT from about 1965 to about 1985, when CAD took over.  The I-10/I-17 stack in Phoenix was probably the last major Arizona DOT project to have hand-drawn sign design sheets.

The list of states which I know didn't bother with pattern-accurate sign design sheets in the pre-2000 "dark ages" is rather long--CA, OR, WA, KS, OK, TX, MO, MI, OH, KY (Kentucky Turnpike was the only one which did have pattern-accurate sign design sheets--using pre-1948 unrounded typefaces, no less), NC, GA, FL, and MS.  Many of these states did have pattern-accurate standard plan sheets, which often (as in the case of MI) used idiosyncratic typefaces with glyphs for some letters and digits which failed to match the FHWA standards.

[LeftyJR]

I don't mind the Clearview letters so much - it's the numbers that bug me.

IF states are going to switch to Clearview, they need to keep the FHWA Series numbers (at least on the route markers).

I have found that PA has been keeping the FHWA numbers, while replacing the letters with Clearview. 

The newer Clearview signs in PA look much better than the old ones do!  The new signage on I-376 looks good.

[Alps]

http://en.wikipedia.org/wiki/QCad

Apparently it is an AutoCAD clone (uses DXF format both internally and externally) and development began in 1999.  Ports to Windows are unofficial, and only the "community" edition is made available through the GNUPL--the functionality it offers is restricted compared to the full version.  3-D is not supported in any version (though highway design professionals like my friend would say, "You don't need it").

Not to stray off topic, but 3-D in AutoCAD can be very useful.  Or Microstation, which also supports it.  Right now you have vertical information stored in project files that are then imported for graphs and elevations.  Why not simply have a 3-D file with the elevations built in?  I think now that CAD programs have been 3D for a number of years, standards will start to come around to incorporate that functionality.

[wytout]

Hadn't driven northbound on 9 in a while in CT lately.  When I saw this I almost threw up on my shoes.  I had to go back around for a shot.  It must be very new.  It's the last NB exit at Corbin's Corner just before the northern CT-9 Terminus at I-84E/W.

[joseph1723]

Here in Ontario I've only seen around 3 BGSs done by MTO in clearview probably as a test. The rest that I've seen installed this year are all still done in FHWA. The City of Toronto loves clearview though, they use clearview on all their BGSs and street name signs.  

Here's a pic of a clearview sign from onthighways.com:


Look like MTO still uses FHWA for the numbers like everyone else.

However it looks like that MTO might switch to clearview in the future, according to this article about clearview from their website. Their halation sample looks like it was done in arial though.

[agentsteel53]

However it looks like that MTO might switch to clearview in the future, according to this article about clearview from their website. Their halation sample looks like it was done in arial though.

I think that website just crashed my firefox

sounds like appropriate behavior for a Clearview page!