Inversed Ronchi testing

[Chriske]

Some amateurs use Ronchi testing technique, I never did in the past.
Ronchi is a good testing methode but it lacks accuracy.
So most of amaterur mirror grinders end up with Foucault testing because it gives the necessary figures to know whether a mirror is close to perfect or not.
Foucault is not the only, there are a few other, and also 'null'testing like Dall or Ross.
Null testing is also very good to because you end up checking on a spherical surface. Needless to say the eye is very sensitive when it comes to checking on spherical mirrors. Differences of 1/50 Lambda can be seen, that's about the limit.
Straight lines are also good to be judged because the eye is very sensitive to it. A minor bump in a straight line will be seen no matter how small the bump or curve.

There's the Ronchi test that uses a straight line grating to project curved lines on the mirror to judge the accuracy of your parabola. If you see a bump in the curves of in your test result, you immediately will  see it. There too the eye is very sensitive.
Problem with Ronchi is how far should the curves under test bend. How does a Ronchi user deals with it. Because every other mirror need a completely other set of projected curves.

But there's a solution using a (different) Ronchi test, the 'Inverse Ronchi test'. Its goal is exactly the same as the original Ronchi test, except it uses curved lines in it's grating instead of straight lines. There is that little program(1.9Mb) developed in 1998 by Mark D. Holm. InvRon he called it. The software calculates the curves for a specific mirror, the only thing you need to do is print the pattern onto transparent and there you have your grating. Well the thing is you need to start parabolizing until you end up with straight lines. And again the human eye is very sensitive to straight lines.
I'd like to try it and compare the result with regular Foucaulttesting,
Years ago I've downloaded InvRon but never used it.
A few days ago I've tried installing InvRon but I get error messages, Even in Linux it does not work.

My question : has someone ever tried to install InvRon or used it...?

Caution : maybe the file needs to be scanned for viruses, in my case there were no problems with 2 OS'ses. I have no scanner on board.
There's a readme file with lots of explanation about InvRon. Opened a few URL's mentioned in the readme, but got a few times 404...:-(

InvRon.zip

[Astrobits]

1 hour ago, Chriske said:

 

There's the Ronchi test that uses a straight line grating to project curved lines on the mirror to judge the accuracy of your parabola. If you see a bump in the curves of in your test result, you immediately will  see it. There too the eye is very sensitive.
Problem with Ronchi is how far should the curves under test bend. How does a Ronchi user deals with it. Because every other mirror need a completely other set of projected curves.

 

InvRon.zip 1.7 MB · 1 download

That has always been my objection to the Ronchi test, trying to compare curves. The InvRonchi sounds like a useful  solution.

Nigel

[Chriske]

I see there are a few downloads of InvRon. Somebody tried it..?

[Chriske]

Found this :

https://web.archive.org/web/20130420122726/http://www.atmsite.org/contrib/Holm/invron/invron.html

[Chriske]

It will only work on a 32bit Win system.

[Astrobits]

Yes, I have downloaded it and now have this programme running in Ubuntu 20.04 thanks to my computer guru son. Still have to read how to use it but other tasks await my attention at the moment.

Nigel

[Astrobits]

Didn't take long to get the hang of this prog. The only drawback is that it was intended to be printed large onto paper and then reduced photographically using high definition transparency film. Now I don't have a working film camera any more, having gone digital, and therefore I cannot use this unless I purchase some very clear transparency film and print it direct with an inkjet printer. The question is: is printer transparency film of sufficient optical quality for this purpose and can my printer print accurately enough? I don't know the answers.

Here's a screengrab of the prog in action:

Nigel

[benzomobile]

On 27/05/2022 at 19:06, Chriske said:

It will only work on a 32bit Win system.

It works in Windows 10  64bit too.

Beppe

[Chriske]

You tried it..???

[Chriske]

Just tried it myself, works perfectly...

So the story about only running in 32bit is pure nonsense ...
What a waste of time...🙄

[skybadger]

That's cos win10 supports 64 and 32 bit.

[Chriske]

Ok now this little thing works I need to figure out how it works.
It is not my goal to use and finish a mirror with it to the very end. Parabolizing is a delicate task.
I need figures to see how close or near I am from a perfect mirror. And that is done, as you all know, by using Foucault.

[Chriske]

I've put in 'some' telescope to see what the result would be.
On the right you see the printer settings.