CSC SYS 50 as a replacement for a cabinet saw?

[tjbnwi]

Please explain how read from both ends of a shared ruler works using left edges. And why there’s no mark at the very ends.

My Starrett rules are marked all the way to both ends.

I don’t understand the first question.

Tom

 

[smorgasbord]

Take a close look at the Starrett C635-E metal ruler:

It's marked from both ends. Note the tick marks for 20mm from the left and 130mm from the right. If what you claim is the correct way to read rulers, then the left edge of the 20mm tick would line up with the right edge of the 130mm mark (as viewed in this orientation). But, that's not the case - the 20mm and 130mm tick marks have their centers lined up, as do all the tick marks on opposite sides of the ruler.

Now take two metal rules and use one to measure the other's tick marks, but offset them by at least one tick mark. Line up the tick marks and see where the edge of the ruler intersects the tick mark on the other. It'll be at the center.

That the NIST engineer used the center of the lines, not the left edges, is also indicative that the correct way to read rulers is to use the center.

Try measuring your 4" block with a metal hook ruler as well.

 

[mino]

Take a close look at the Starrett C635-E metal ruler:

It's marked from both ends. Note the tick marks for 20mm from the left and 130mm from the right. If what you claim is the correct way to read rulers, then the left edge of the 20mm tick would line up with the right edge of the 130mm mark (as viewed in this orientation). But, that's not the case - the 20mm and 130mm tick marks have their centers lined up, as do all the tick marks on opposite sides of the ruler.

Now take two metal rules and use one to measure the other's tick marks, but offset them by at least one tick mark. Line up the tick marks and see where the edge of the ruler intersects the tick mark on the other. It'll be at the center.

That the NIST engineer used the center of the lines, not the left edges, is also indicative that the correct way to read rulers is to use the center.

Try measuring your 4" block with a metal hook ruler as well.

Hate to merge in, but you are unintentionally conflating two things here.

When you check/calibrate a tape measure for precision, you check for TWO things:

A) Is the tape exact and uniform along its length - I am not sure of NIST procedure here, but both DIN(German) and ČSN (Czechoslovak) standards would require that calibrating a tape measure checks every 200 mm of its length for deviations. This is because you can have a tape that is exact end-end by accident but have big variations through the scale with "short" and "long" sections.

The standard calls for maximum accuracy here, so the most precise checking method is used - reference of a mark edge as is the rule in all metrology. If you ever did metrology, or even analytical chemistry, when you take visual readings of a scale, and you are after maximum precision, you never go to the centers of the mark. The way a human eye works is it is reference-based. Meaning you will be more accurate if you take two "left" readings or two "right" readings of a scale and subtract.

When proper science classes are*) done, this is hammered into kids as the first thing as it is useful in daily life. Pretty sure it was at our 6th year labs (K7 US terms).

B) Is the end stop zeroed? To do this, you check the end position vis-a-vis a given mark close to it, say at 200 mm. You do it using magnification kit ideally.

When you did both these readings, you then combine them into a single calibration value.

----
In short, there are two rules to read a scale for best precision:
1) Whenever possible, calibrate (and read) either the left or right sides of the marks. Do not read by the center, excepting point 2).
2) Always keep note of how a given instrument was calibrated if you you cannot follow rule 1).

Scales starting at zero cannot use the left/right approach thus are calibrated to the center of the mark. While this is the less accurate method, a scale starting at zero cannot be made other way.*)

---
What this all means for practice:
1) When you can, always take a relative (subtractive) reading using the left or right edges of the marks on both ends of your measurement.
/With tape measures this is even more important as the end stops are the things the most abused/damaged. A tape measure will usually keep its relative accuracy even after its end stop gets bent inevitably./

2) Be aware that the ends of zero-starting measuring instruments are calibrated to the center of the mark, so when using the zero start/stop, you must use that.

3) When calibrating scales on machines, for maximum precision calibrate to the left or right edges of the marks and make sure every user of the equipment is aware which side it is calibrated to on that specific machine.
If you cannot ensure all users are aware which side a scale a machine was/is calibrated to, a common scenario, then it is preferable to eat the precision loss and calibrate to the center than having people using the "wrong" side. In machine shops it was common to have shop-wide policies of "left of mark" or "right-of-mark" readings when manually controlled machines were in use by multiple people. In such settings, many workers had no clue why and just followed the policy, so lots of anecdotal "this is the correct way" stuff is out there. These days of CNCs, it is mostly irrelevant as anything needing to be precise is not made by hand/eye.


---
Now to put this all in context.
We are talking about precision differences in the 0.1 mm (0.005") range and smaller. Talking about this stuff in the context of studs positioning would be just ridiculous. But above is how it goes if you need the best precision from a given scale reading.

If you do not need it, or multiple users are involved, just stick to using mark centers which are unambiguous. No point wasting time/energy on precision that is not required.

----
*) were done, notice the "proper" term too. To align us to US/Western "quality education", Physics labs at middle school level were abolished 20 yrs ago and the Chemistry labs were castrated to a teacher show by EU about 10 yrs ago to "protect the kids from chemicals knowledge". These days one will probably hit this only at a trades school or a natural sciences higher edu course. It still is the fundament of visual instrument reading, nonetheless.

 

[smorgasbord]

A) Is the tape exact and uniform

B) Is the end stop zeroed?

I guess you didn't watch the NIST video.

First, are you suggesting that tape measures are designed to be read differently than metal rules? Sure, you can modify the hook end of your tape measure such that the reading is at the left edge of the tick marks, but that's not how they're designed to be read. And, on a metal rule, you can't add material to the zero end to make it read that way.

Here's the Starrett metal rule and a modified version when I thickened two tick marks, keeping the left edge intact, to emphasize what is continually being ignored here:

As you can see, the thickened ticks don't line up because Starrett didn't design them to read from the left edge, but from the middle of the tick. The middles of the ticks on opposing sides does line up.

Go grab yourself a metal hook rule, such as the nice ones Woodpeckers makes (here and here, for instance) and see for yourself where your accurate 123 blocks line up.

Now, sure, you can burn-an-inch for outside measurements to use the scale any way you want, BUT, for measuring from the end of the tape (most common), or using a hook rule or taking an inside measurement, you're pretty much stuck with how the ruler was designed. And that is to use the center of the tick marks.

 

[mino]

Um, this is indeed tricky to convey. Not being a native speaker is not helping either.

When you use edges of marks, you always compare two marks on the same scale and their respective edges. I.e when you need to measure 2", you look at the left edge of the 1" mark and the left edge of the 3" mark. Alternatively (which side is more suitable is chosen) you look at the right edge of the 2" mark and the right edge of the 4" mark, etc.

As there is no "edge" to the 0" mark on that particular rule, the start at zero cannot be read that way, so all marks are calibrated as "centered" with respect to the 0-starting edge of the measuring instrument. That is the "special case" I refer to when zero-starting commodity measuring instruments are at play. It is that way because that is the ONLY way how to make them practical to use. That is a limitation of sorts, forced by them starting at zero, not a "feature". There is no reason to abide by that when one is not using the zero-start capability, outside the additional precision is not being needed and approximate measurement being more than adequate for the use in question, which does happen to be most woodworking scenarios.

This is the exact same as when marking stock. One always marks such that the cut is to be located on the edge of the mark. Not the center.
---
With a ruler that is not starting at zero, thus has a proper 0 mark, you can use left, right (more precise) or center (less precise reading). This holds as long as the marks are same-width. Which they are on any serious measuring instrument. For exactly this reason so edge-reading is possible.


Add:
When you compare two scales, you then leverage both edges, center is ignored as that is how a human eye works when comparing the relative placement of two objects of a similar width. You cannot see a center or a different-color line, you /your brain/ can only approximate it. But you do see its edge(s).

 

[smorgasbord]

all marks are calibrated as "centered" with respect to the 0-starting edge of the measuring instrument. That is the "special case"...

Glad you agree with me on how rulers are designed, but using tick centers is not a "special case," it's the standard.

What you and @tjbnwi are talking about is a special way to use a ruler that can only be applied, as I said, when you do something like "burn an inch." But for tape measures, which is how this tangent got started, I'll argue that tick mark bleed variation makes trying to use the edges problematic, if not even more inaccurate. that the NIST engineer chose to use the center for his tick mark to tick mark accuracy measurement was not by accident or lack of knowledge.

 

[ReneS]

I might be missing something here, but if you were to read the ruler (above) from centers, shouldn't you have a half a tick mark at the beginning edge?

It seems to me that the units on your ruler are one white space and one tick mark. So in this case, I think the measurement point of reference would be just to the right of the tick mark.

It looks like the first unit of measurement is one white space plus one tick mark. So the second unit would also be one white space and one tick mark.

Additionally, half a tick mark at the beginning and end of measurement is the same as one tick mark (plus any full tick marks that fall between the two). Since there is no tick mark at the beginning (or half a tick mark) seems center to center won't work.

I might be wrong, but it is fun to think about this stuff.

Edit: The picture of the ruler is fuzzy, if there is a tick mark or half a tick mark at the very beginning, then I would rethink this.

 

[smorgasbord]

I might be missing something here, but if you were to read the ruler (above) from centers, shouldn't you have a half a tick mark at the beginning edge?

I've not seen a 0-end ruler that has a half-thickness tick mark at the end.

BTW, these are images from Starrett's web site.

 

[ReneS]

I've not seen a 0-end ruler that has a half-thickness tick mark at the end.

BTW, these are images from Starrett's web site.

So logically, if starting from the end, wouldn't you go to the far edge of the tick mark of the measurement you are making?

 

[smorgasbord]

So logically, if starting from the end, wouldn't you go to the far edge of the tick mark of the measurement you are making?

No.

To figure it out for yourself, take two of the same rulers and hold them edge to edge, line up one of the tick marks, and see where the end of one of the rulers intersects the corresponding tick mark.

 

[tjbnwi]

Take a close look at the Starrett C635-E metal ruler:
View attachment 375988
It's marked from both ends. Note the tick marks for 20mm from the left and 130mm from the right. If what you claim is the correct way to read rulers, then the left edge of the 20mm tick would line up with the right edge of the 130mm mark (as viewed in this orientation). But, that's not the case - the 20mm and 130mm tick marks have their centers lined up, as do all the tick marks on opposite sides of the ruler.

Now take two metal rules and use one to measure the other's tick marks, but offset them by at least one tick mark. Line up the tick marks and see where the edge of the ruler intersects the tick mark on the other. It'll be at the center.

That the NIST engineer used the center of the lines, not the left edges, is also indicative that the correct way to read rulers is to use the center.

Try measuring your 4" block with a metal hook ruler as well.

Per your request.

Hooked left to right.







Flipped this rule around to read right to left.





Looking back at an earlier post it was mentioned he used the 2” mark on the old tape because there is no zero, he states he used the 2” mark because he was set up for something else.

Tom

 

[smorgasbord]

Flipped this rule around to read right to left.

The only way your photos are accurate is if your 6" ruler isn't actually 6" long, but is 6" long plus the width of single tick mark. Seems unlikely, no?

If measuring from the left is accurate to the left of the tick mark AND measuring from the right is also accurate to the right of the tick mark, which is what your photos purportedly show, then this is what's going on:

In the diagram above, the red box is exactly 2" long and the blue box is exactly 4" long. Both are placed to align with the appropriate cut end of the ruler. Yet, somehow, each reads to the appropriate side of the tick mark. But, since we know the red and blue add up to exactly 6", then the ruler must be longer than 6" : by exactly the width of one tick mark.

I'll bet real money that isn't the case.

If the ruler design intent was to read from the left end of the ruler to the left of the tick mark, then the 4" block from the right should have also lined up with the left edge of the 2" tick mark, not the right edge of the tick mark as photographed.

What we're likely seeing here is the result of parallax. Maybe try using the good technique below:


Although we're talking not doing the "burn an inch" technique here. With that technique you can choose edge or middle of the tick marks as you feel convenient.
This "good technique" is harder to do with a tape measure, which is how this rabbit hole got started. But, your Starrett hook rule should be fine, except it has a hook at only one end.

 

[tjbnwi]

Never measured the rule before.





Starrett lists the tick line width at 3/64 (0.046875)

CH604R-6 Spring-Tempered Steel Rule with Inch Graduations

The Starrett 6" Spring-Tempered Steel Rule with Inch Graduation features a satin chrome finish. Graduations at 8ths, 16ths, Quick-Reading 32nds, 64ths 6", 4R Grad, Reversible Hook

www.starrett.com

Tom

 

[luvmytoolz]

@tjbnwi It's not an exhaustive test unless you now measure your digital gauge! ;-)

 

[smorgasbord]

Starrett lists the tick line width at 3/64 (0.046875)

No, they don't.

And, just think about it: How can a ruler with gradations at every 1/64" have tick marks that are three times as wide?

Here's a close-up of a good quality ruler with 1/64" gradations, and lines drawn in Adobe Illustrator to show tick width and spacing:


This is rough, of course, but it appears the tick marks are about 0.006" thick. That's less than 1/100", which is the finest gradations of a rule that Starrett makes. This would make the ticks slightly wider than the spacing between them.

But, even if the ticks are half that, say 1/200" (making the 1/100 gradations as thick as the spacings between them), that's 0.005", which is still more than the length overage (0.0035") you measured. And that's assuming you cleaned the rule carefully and measured properly.

So, no, metal rules aren't longer than their stated measurement by the width of tick mark. See my red/blue diagram above for the flaws in edge gradation reasoning.

Practically, if you care about the center/edge of a tick mark on a metal rule, you're using the wrong measurement tool. But, this tangent started with tape measures, which have much wider tick marks.

 

[tjbnwi]

You’re correct on the thickness, I read the thickness spec incorrectly.

You’re saying Starrett does not make a good quality ruler? Inat would be shocking to many.

Cleaned the rule and jaws of the caliper.

What brand and model is your “good quality ruler”?

Tom

 

[smorgasbord]

You’re saying Starrett does not make a good quality ruler

Never said nor implied that.

 

[fritter63]

I'm certainly guilty of obsessing about accuracy of cuts, but it's important to remember that there are only some operations where it actually matters.

As you progress through a project, there comes a point where the actual measurement isn't as important as CUTTING IT TO FIT WHERE IT WILL GO BETWEEN OTHER parts.

As an example, attached is a picture of my current project. I'm making two of these (king and queen). The legs are 2 3/4" square. Does it matter if they're PRECISELY that? Actually no, once they're built, as long as I register off consistent faces for the mortise and tenons (or in my case for the domino joinery), things will work just fine.

For the footboard, the length of the vertical slats is not as important as them all being the same length. I'm not going to mark each of them for final dimensioning, I"m going to mark one, use it to set up a stop on my miter gauge or whatever I'm using, then cut them all to that stop. Similarly, for where the rails join into the legs, I'm only going to position ONE joint accurately to get it flush the top of the legs, then I'm going to make the other domino cuts on the "medium" setting so that the length of those vertical slats is also not critical.

etc etc etc. If you follow the right order, you aren't as concerned about actual lengths. And when you are, you should probably be marking with a pencil or Japanese knife where it intersects rather than marking to a measurement.

making a door with raised (or flat) panel, there is slop in the rails and stiles so it doesn't have to be exact.

And... what was the OP's question?

 

[mino]

...
Practically, if you care about the center/edge of a tick mark on a metal rule, you're using the wrong measurement tool. But, this tangent started with tape measures, which have much wider tick marks.

Sorry, I have been busy to disprove some stuff written here, but above is just as false statement as it gets.

For anything longer than an foot or so, quality metal rulers are the ONLY feasible option. The only thing better are calibers which is purely theoretical in a woodshop setting.

As such you absolutely do care on how you read out stuff.

Also, as per earlier, the width of a mark tells you NOTHING about the accuracy of its placement. Anyone familiar/trained to use precision instruments will read the edges of the marks so no one cares a iota on how wide the marks are as long as their width is identical and they are placed precisely.

When I have some spare time to dig around or make illustrations, I will try to explain this in less technical matter.

---
For now, let me just say that I would not have won a regional Chemistry Olympiad lab part back in the day if I was reading mark centers ... the accuracy difference between me and the second placed fella was less than 1% .. about half the .4 mm wide mark at the concentrations that were in play. Yes, how to read marks was a major part of prepping for those labs.

ADD: For fun, yeah, I screwed up the written part prior to the labs, so barely made it into the top ten overall. But those evenings spent in the school lab did pay off.


---
Example:
Notice the "suitable for calibration" comment, that is directly related to the use of laser markings for the ruler which makes the marks consistent.

https://shop.k-met.com/en/steel-thin-ruler-1000mm-0-5-mm-laser-marking-suitable-for-calibration-0-5-mm-p9204993c6c242

 

[smorgasbord]

Sorry, I have been busy to disprove some stuff written here, but above is just as false statement as it gets.

You should be sorry: You're flat-out wrong about the falseness of my statement: Caring about the edge/middle of a metal rule's tick mark means you care about a difference of 0.0025". I double-down that if you care about that kind of tolerance, a metal rule is the wrong measuring tool.

For anything longer than an foot or so, quality metal rulers are the ONLY feasible option. The only thing better are calibers which is purely theoretical in a woodshop setting.
As such you absolutely do care on how you read out stuff.

I didn't say one doesn't "care on how you read out stuff" I said that if you care about differences in center/edge of a metal rule you shouldn't use a metal rule. As that's 0.0025" of difference, you should be using gauge blocks and/or calipers. As for the caliper length restriction, do you care about 0.0025" in anything over a foot long?

Please get your attributions straight before you go accusing me of the worst kind of falsehoods.

Also, as per earlier, the width of a mark tells you NOTHING about the accuracy of its placement. Anyone familiar/trained to use precision instruments will read the edges of the marks so no one cares a iota on how wide the marks are as long as their width is identical and they are placed precisely.

Again, we all agree that measurements where you can measure tick mark to tick mark, using the edges (with magnifying glasses) is as good as any.

But, for the original tangent discussion - tape measures with hooks - and by my extension to hook rulers, we have to work within how those end-registering tools are designed and manufactured. And that is by the center of the tick marks.

Now, for tape measures, you can bend the hook inward (by half a width of a tick mark) such that the correct outboard reading will have moved from the tick mark center to the edge of the tick mark nearest the hook. That modification has some drawbacks:
1) The inside measurement is probably still more accurate on the tick centers since the hook's pivot (the bend) remains in the same spot.
2) A bent hook may be accurate only at the tip, and so outside measurements of thin boards won't have been compensated the same amount.

For metal hook rulers, I know of no modification that would make reading the near edge of the tick marks more accurate. One theoretically could add half a tick mark's width of material to the inside of the hook and then read off the far edge of tick marks, but that would seem hard to do accurately and probably hard to remember in practice.

Notice the "suitable for calibration" comment, that is directly related to the use of laser markings for the ruler which makes the marks consistent.

https://shop.k-met.com/en/steel-thin-ruler-1000mm-0-5-mm-laser-marking-suitable-for-calibration-0-5-mm-p9204993c6c242

Typically, photo etching is the preferred method for creating tick marks on metal rules, over laser engraving.

The DIN standards for rules are specified on page 93 of this document:

https://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CONSLEG:2004L0022:20130101:en:PDF

For rules a meter or less, EC1 for gradations is +/-0.2mm, and for EC2: +/-0.5mm (at 20ºC).
For measurements from an end, which is what we're talking about here, an additional 0.1mm is allowed, so +/-0.3 for EC1 and +/-0.6mm for EC2. I don't know if that decreases for measurements under a meter, as I've seen some manufacturers claim, but maybe I'm not reading the DIN spec right.

0.2mm is just under 0.008", so already we see that even an EC I metal rule may be off by an amount larger than the width of a tick mark. My initial statement, which you erroneously claimed was "as false statement as it gets," is thus actually extremely truthful, accurate, and something to keep in mind. If you need accuracy of half a width of a tick mark (0.06mm), then a metal rule is not the right tool.