How accurate is accurate enough?

[Mike B]

Much is made these days of the accuracy of powertools, saws, planes etc and indeed the greater the accuracy of the component parts the greater the accuracy and quality of the finished article.Additionally, this is largely up to each individuals own standards/pride but as a relative newbie "how accurate is accurate enough?"

By this I mean in reality how accurately does a piece need to be made to look professional when placed in your living room (and I do not mean anything to do with design, colour, type of finish...).

For example, if I were to make a 1.2m tall bookcase and the sides are 1mm bowed over the length will anyone (except me) notice?? Obviously, if the dovetails joining the sides to the top have 1mm gaps then it will look terrible!!

So, I was wondering what "rules of thumb" / years of experience other users employ to know when to stop planing and paring joints and get the project finished??

As a relative newbie find it all too easy to spend hours and hours trying to eradicate every chink of light under the straightedge or paring every joint, only to get frustrated by the wood "settling" a little overnight or seemingly never getting anything finished!!

Cheers
Mike

 

[mr]

As another newbie I take the view that you should be as accurate as you are able. But whats more important that how accurate a piece has to be to look right to me is how accurate should it be for you to feel right about it. If you're happy to think something is good enough at whatever point then thats all the accuracy you need. If you're not that sort of person then good enough never will be. Chances are that as your skills increase so will your accuracy and your requirement for it. Obviously the finer (in terms of tolerances) the piece you're making then the greater the need for accuracy or the thing wont go together and a level of accuracy is required for structural integrity in any case. But to me it's more about how it feels. If you take care and spend time, don't rush etc the accuracy should come out of that naturally. Of course taking time and care also involves the "back office processes". The amount of care you put into the tools and so on. Its not just about the wood on the bench.
To my mind youre answering your own question a bit when you mention never getting the project finished. That suggest that you know how accurate you want the piece to be and if thats the case nothing else will ever satisfy you so dont worry about it and unless someones waiting with a cheque for the piece just enjoy it.
Cheers Mike

 

[Steve Maskery]

ROT - If it looks right it is right!


Cheers
Steve

 

[mr]

Rot? Isnt that the same thing by a different route?

 

[Steve Maskery]

ROT - Rule OF Thumb!

It wasn't a dimissal!



Cheers
Steve

 

[mr]

AHHH I never remember what these TLA s stand for,though as I say I think you're actually kind of saying what I tried to but more succinctly.
Mike

 

[Mike B]

Guys, thanks for the responses.

I totally agree with what you are saying - "if it looks OK to you then it is OK", but I know I for one get bogged down when preparing each component of a project trying to reach a level of perfection that is a) above my skill level (obviously it's useful practice) and b) probably not necessary anyway.

What I was trying to achieve here was some kind of suggested minimum tolerances that the pile of parts on the bench should reach in order to yield a professional looking result.

Obviously everyones own personal standards will reflect their own set of tolerances, but I thought this may be interesting and useful to those like me with little experience to get a better idea of when enough is enough.

After all wood moves and I'm pretty sure that even the pros do not manage to achieve mega tolerances on ALL of their components (especially so when time is money), but do have the experience to know what will/wont show on the final piece.

Cheers
Mike

 

[mr]

Just a thought but it seems that a lot of "pros" will use stable secondary materials which they veneer - plywood, mdf etc. Nothing wrong with that it takes away a lot of the issues of seasonal movement and possibly drops the cost of production. Others however will tend to use solid / real wood for want of a better word and take movement into account in the joinery, some / most will use a mixture of both. As far as I can see if you have measurements marked out that you want to work to theres no reason for not achieving them with time and care, the secret is probably to evaluate what you need to achieve in terms of your materials (mdf and veneer or solid). A large part of success is probably to be found in the design which allows for seasonal movement of materials if required etc rather than purely in the "engineering" or the actual cutting. As for achieving mega tolerances for me the thing is not to make something that never moves due to having achieved mega tolerances but making something which moves gracefully and is allowed to have a life of its own. You can't stop that so rather than try it's probably better to incorporate it. Decades of experience (that I haven't got) are probably invaluable here.
(Not sure that I'm helping really - reaching for coat).
Mike

 

[Scrit]

I'm with Steve. If it looks right, then it is right. Sometimes customers have told me about little mistakes they've found - and I've several times offered to put them right, but I find the response is generally (fortunately) along the lines of "oh no, that's all right, it shows it was hand made......."

I'm mindful to repeat the old saw that the only difference between an amateur and a professional is that the professional knows how to hide his mistakes better :roll: Then again, maybe not...............

Scrit

 

[woodbloke]

I agree with lots of the posts here but would add and additional guide which is checking diagonals after glueing. Equal diagonals is ideal but somewhere in my 'bible' (Techniques of Furniture Making) I recollect that a discrepancy of about 1 or 1.5mm over about 1m each way is acceptable - Rob

 

[Scrit]

... I recollect that a discrepancy of about 1 or 1.5mm over about 1m each way is acceptable

Precisely (or not ). If you look at the legs on an original 19th century Windsor chair you'll like as not find all four are slightly different, after all they were hand bodged on a pole lathe in the middle of the woods. But the chair as a whole looks OK, has a sense of cohesion, partly because our brain adjusts the image...... I think

Scrit

 

[Colin C]

... I recollect that a discrepancy of about 1 or 1.5mm over about 1m each way is acceptable

Precisely (or not ). If you look at the legs on an original 19th century Windsor chair you'll like as not find all four are slightly different, after all they were hand bodged on a pole lathe in the middle of the woods. But the chair as a whole looks OK, has a sense of cohesion, partly because our brain adjusts the image...... I think

Scrit

I do agree with you Srit but there is one small thing use missed about the chair is that the wood would have been green wood and it all moves differently( you will find that they are no longer round ), but this will not be seen unless you put your hand to them

 

[woodbloke]

I can see this post running for longer than the London Marathon. IMHO it all depends on what you're looking at; a fantastic piece by Robert Ingham which is engineered down to the 'n' th degree or a medieval hammer beam roof as in Westminster Hall. Its in the eye of the beholder and generally if it looks right, it probably is right. After all, this is how, I believe, shipwrights of old built ships. Even after a thousand years, a Viking longship still looks a serious piece of boat - only my view - Rob

 

[engineer one]

for my two pennorth.
as an engineer it is a trial to work with a material which moves on an
irregular basis, but then that is why i make things in wood.

if you are making things from sheet materials, then the most accuracy
that you can produce is a must i would suggest since basically sheet
materials "do not move" ( i know that is not true, but we have to believe)

however when using "real" wood, then there are many tricks that you can use to hide the "errors" which are in fact only allowances for movement.

may i suggest you go to look at furniture in shops, and then work to somewhat closer tolerances. i am constantly amazed by the gaps in
pine dressers. there is so much slop you wonder whether the drawers
ever fit properly.

you are looking for good tight joints, but where the movement is going to
happen, you can put "shadow lines" or cover the gaps by cosmetic
items.

the first imperative is to finish the b***** piece, then the next one, and the next one, as you get more experienced, both in the manufacture, but
also in the handling of wood, you will find that your techniques allow
you to tighten up your tolerances.

maybe the most important thing to remember is that ALL the pillars on
the greek temples are not straight, they bow slightly in the middle, but
look straight.

so what am i saying, wood moves, accept it, work round it, and aim for
a piece that looks good to other people. you will always find fault, but
most of the time it is so small that the user/customer /swmbo
will not notice until you tell them.

also if the way in which most people hang paintings, posters and mirrors is
anything to go by they cannot see the odd angles anyway.

every job you finish will be better, and the tolerances will be tighter,
so FINISH, then improve.

paul :wink:

 

[woodbloke]

amazed by the gaps in
pine dressers. there is so much slop you wonder whether the drawers
ever fit properly

I had a look at a very nice top endy Ercol dresser a few days ago; the drawers fitted where they touched which wasn't all that frequently, not very impressed. Having said that, when I worked in the trade some years ago we were advised to make drawers on the slack side as some pieces were shipped to the caribbean where its moderately warm and humid :lol: ,well fitted drawers would swell and stick, not very clever in a piece costing several grand - Rob

 

[richie]

This is an excellent question for any newcomer as it can save you a lot of hassle.

Basically, don't bother trying to get engineering tolerances. Yes, you can spend an hour sharpening and honing your plane blade 'til it's like a mirror, then take off 1 thou' thick shavings to get two pieces of wood together without any air or light between them, but that's not woodwork, it's a technical stunt. And as you have seen, wood never stays exactly the way you made it!

If you haven't been doing it for very long, start simple. A well-made simple project is infitely nicer than a botched over-ambitious one. Even practicing joining scraps of wood can help.

And after a while you will get to know wood a bit better, and how you and your tools perform, and you will find that things start to come together a bit easier.



r

 

[DomValente]

My eldest son is an engineer and regularly visits my works.
His favoured dig" dad that cabinet is out about a half mil."
I apprenticed for a year in Oxford with some great restorers none of whom seemed to be much under 80. One wag told me that a cabinetmaker works to the nearest sixteenth, a carpenter to the nearest inch and a chippy to the nearest house.

With apologies to all chippies( not my words).

 

[builderchad]

Interesting what Paul said about other people not noticing. I read a similar thread in another forum and someone chipped in and said "The best way to hide you mistakes is to NOT TELL ANYONE ABOUT THEM! They won't notice them that way."

When I started WW last year my perfectionism really showed through and SWMBO noticed this trait in me for the first time after being together over 10 years. Previously my perfectionism only showed in my programming code and that only other developers saw. My first project was a pine toolchest and the wood movement and inaccuracy was so great that I was kind of shocked into the aforementioned "if it looks right, it is right" mode of thinking. Since then I have had a much better time of working wood because I am not spending endless, fruitless hours trying to achieve some unachievable perfection.

Ofcourse using the basic rules of thumb about measuring and marking out go a hell of a long way to keeping things accurate. Like marking two adjacent faces with the cursive f once they are square to eachother and squaring the other two and measuring/marking only from those. And about not going all the way around the piece in one direction marking with the try square as that just compounds any error. And using a knife or propelling pencil with its 0.5mm lead to keep marks as accurate as possible. And using a marking gauge whereever you can instead of a ruler and marking knife...etc etc.

 

[engineer one]

just a quick thought. was in john lewis yesterday, looked at a set of bedroom furniture made by "frank hudson", very nice, BUT
every drawer had a moulding glued on, and the glue lines were
an absolute joke. now i know this is a production piece, but
600 quid, i think not.

so make each piece, then make the next one better, then again, and
then again. every time we make something we learn more, then
as we learn we can experiment more. so as has been said so many times
in this thread, first finish, then move to the next stage of accuracy.

paul :wink:

 

[Scrit]

And using a marking gauge where ever you can instead of a ruler and marking knife...etc etc.

And using a rod to hold dimensions rather than measuring at all, and usin dividers to transfer measurements rather than measuring at all....... That sort of approach can really upset some people :wink:

Scrit