ANSI Standards and digital lenses

[MIOPE]

We recently started with freeform equipment and digital designed lenses. The freeform equipment is supposed to work with a precision of 0.01 D.
Some of this lens designs are compensated. The original prescription is altered with calculations that include, vertex distance, panto, wrap, etc.
This compensations are measured with a precision of 0.01 too.
My question is. Are there different ANSI standards for freeform lenses?

For example. I calculated a prescription of -3.00 with the calculator on Opticampus http://www.opticampus.com/tools/ansi.php
Used Panto=9 and Facial Wrap=5 on a Poly lens. The result is -2.88 DS -0.09 DC × 61. On the other hand. I see on the Ansi Standards Opticampus http://www.opticampus.com/tools/ansi.php that the tolerance on sphere from 0.00 to +-6.50- is 0.13D. and
cylinder up to -2.00 it is also 0.13D. On progressive lenses the tolerance is a little higher.

Why worry to compensate on lower prescriptions if when processing the lens, a error up to 0.12D is accepted?

Also. Consider how difficult is to find two lensmeter that measure exact the same on precision 0.01D. Most of the prescriptions are written on steps of 0.25D. Most nearsighted patients will notice easier a little weak lens than a little strong one.

[Barry Santini]

Biggest issue here is not the compensated Rx, nor ANSI tolerance.

1. It is NOT knowing how fully corrected for far DV vision the Rx presented to you is, and...

2. The unspoken or habitual vision needs of the wearer.

Bottom line; NEVER bounce a well-compensated Digital FF lens UNTIL you have tried it on them FIRST.

Period. That is all.

B

[lensgrinder]

Are there different ANSI standards for freeform lenses?

No, there is no need to have separate standards. ANSI if for the fabrication of an Rx, in this case you are manufacturing a compensated Rx.

For example. I calculated a prescription of -3.00 with the calculator on Opticampus http://www.opticampus.com/tools/ansi.php
Used Panto=9 and Facial Wrap=5 on a Poly lens. The result is -2.88 DS -0.09 DC × 61. On the other hand. I see on the Ansi Standards Opticampus http://www.opticampus.com/tools/ansi.php that the tolerance on sphere from 0.00 to +-6.50- is 0.13D. and
cylinder up to -2.00 it is also 0.13D. On progressive lenses the tolerance is a little higher.

You would apply the standard to the -2.88 -0.09 X 61
Keep in mind the reason for compensation is to make up for the refraction. When we are refracted there is no tilt in wrap with the phoropter.
If this Rx was made without compensation the patient would look through -3.03 -0.10 X 61, so we compensate the Rx to account for the wrap and tilt.

[rbaker]

Whose compensation? Isn't this somewhat like measuring a turd?

[RT]

Section 5.1 of ANSI Z80.1-2010 states:

For lenses produced with compensations to account for as worn correction, the tolerances in the tables in clause 5 apply to those values specified by the manufacturer and not to the prescribed RX.

Thus, your tolerances would apply to the target compensated Rx.

Last year The Vision Council's Lens Division published a paper on the repeatability and accuracy of lensometer readings across a variety of lensometers. The conclusion was that there was scatter of +-0.06 in readings under the best of conditions. So while somebody's process may be accurate to +-0.01, you probably can't measure it.

[Barry Santini]

Section 5.1 of ANSI Z80.1-2010 states:

For lenses produced with compensations to account for as worn correction, the tolerances in the tables in clause 5 apply to those values specified by the manufacturer and not to the prescribed RX.

Thus, your tolerances would apply to the target compensated Rx.

Last year The Vision Council's Lens Division published a paper on the repeatability and accuracy of lensometer readings across a variety of lensometers. The conclusion was that there was scatter of +-0.06 in readings under the best of conditions. So while somebody's process may be accurate to +-0.01, you probably can't measure it.

dispense the eyewear and troubleshoot.

b

[Lab Insight]

Biggest issue here is not the compensated Rx, nor ANSI tolerance.

1. It is NOT knowing how fully corrected for far DV vision the Rx presented to you is, and...

2. The unspoken or habitual vision needs of the wearer.

Bottom line; NEVER bounce a well-compensated Digital FF lens UNTIL you have tried it on them FIRST.

Period. That is all.

B

How can you try it on them first? Most ECP's don't have immediate access to refract compensated Rx's generated by the FF software while the patient is still sitting in the chair. It's not a feasible solution IMHO.

And troubleshooting also isn't a viable option - very costly to your remakes and lab relationship, and a huge inconvenience for the patient's time, not to mention would seriously tarnish your reputation if you had to remake every Rx twice to achieve the best V/A results.

[MIOPE]

I agree with "Lab Insight". I do not expect a final precision of 0.01 but if the calculation lower the minus sphere, the sphere of that lens should be very close to the calculated one or between the calculated and the original one.

[lensgrinder]

How can you try it on them first? Most ECP's don't have immediate access to refract compensated Rx's generated by the FF software while the patient is still sitting in the chair. It's not a feasible solution IMHO.

And troubleshooting also isn't a viable option - very costly to your remakes and lab relationship, and a huge inconvenience for the patient's time, not to mention would seriously tarnish your reputation if you had to remake every Rx twice to achieve the best V/A results.

You could not refract the compensated Rx anyway. The Rx is compensated because of the wrap, tilt, and vertex of the frame, not the phoropter. When you refract a CL wearer you would not put them on the phoropter at the compensated CL Rx.
Barry is saying(not to put words in his mouth) to trust the labs compensation. If you properly fit the frame and taken PoW measurements then the Rx should work out. The success rate is not going to decrease because it is a compensated Rx.

I agree with "Lab Insight". I do not expect a final precision of 0.01 but if the calculation lower the minus sphere, the sphere of that lens should be very close to the calculated one or between the calculated and the original one.

Not necessarily, depends on the tilt, wrap and vertex.

[MIOPE]

I speak spanish and is not easy to me to express in english.
Maybe I should say it in another way.
If freeform equipment is very precise (0.01). Calculations are made to 0.01 too. Why should I accept from the lab a lens with a prescrition 0.08 to 0.12 away of the Rx. Calculated?
If this lens is a progressive, this means to me that the map of the progressive could also be away from what it is suppose tp be.
But this lens is acceptable according to ANSI standards.

[Robert Martellaro]

If freeform equipment is very precise (0.01). Calculations are made to 0.01 too. Why should I accept from the lab a lens with a prescrition 0.08 to 0.12 away of the Rx. Calculated?

Because that degree of error is inconsequential visually.

If this lens is a progressive, this means to me that the map of the progressive could also be away from what it is suppose tp be.
But this lens is acceptable according to ANSI standards.

I suspect that that level of deviation, even globally, is still inconsequential, especially when compared to centration errors made by the optician when measuring IPDs and tilts, and from edging the lenses.

[MakeOptics]

Unless using a legacy design with an atoric back surface, you are getting what the lab finds acceptable. This could means anything, deviations from intended design are not covered by any standard I know of yet. A few years back I sent the same exact design, same Rx, same frame parameters, from the same lab out to be mapped and they came back with two significantly different maps. The cream on top, is that the distance reference point, prism reference point, and the near reference points were the same and passed ANSI, yet the deviation across the lens surface varied as much as 0.50D, this to me indicates a lab that has an inconsistent process and no way to measure the surface.

Legacy designs fair a bit better because the front surface molded lens design has to conform to ISO standards or 0.08D (don't quote me on this), of course the back surface is still digitally produced as an atoric which is still difficult for office equipment to verify, but the lenses are at least for the most part rotationally symmetrical allowing for a little bit of flexibility in the process tolerance. Process tolerances are the ugly little secret that no one wants to talk about and difficult to verify, eventually office based verifiers will become more prevalent and labs will have to clean up their act, until then make sure your lab has always had a knack for accuracy.

[Tallboy]

That is a fantastic point MO, and that is why I'm hesitant to try just any labs FBS processing, I have to implicitly trust my lab in ways that I used to only have to trust the lens design companies.

[Robert Martellaro]

Unless using a legacy design with an atoric back surface, you are getting what the lab finds acceptable. This could means anything, deviations from intended design are not covered by any standard I know of yet. A few years back I sent the same exact design, same Rx, same frame parameters, from the same lab out to be mapped and they came back with two significantly different maps. The cream on top, is that the distance reference point, prism reference point, and the near reference points were the same and passed ANSI, yet the deviation across the lens surface varied as much as 0.50D, this to me indicates a lab that has an inconsistent process and no way to measure the surface.

Legacy designs fair a bit better because the front surface molded lens design has to conform to ISO standards or 0.08D (don't quote me on this), of course the back surface is still digitally produced as an atoric which is still difficult for office equipment to verify, but the lenses are at least for the most part rotationally symmetrical allowing for a little bit of flexibility in the process tolerance. Process tolerances are the ugly little secret that no one wants to talk about and difficult to verify, eventually office based verifiers will become more prevalent and labs will have to clean up their act, until then make sure your lab has always had a knack for accuracy.

.50 D is a lot of error, especially if is it is asymmetrical! I might lose some sleep after reading this.

I wonder if legacy (semifinished) designs surfaced on free-form generators can still have significant defects in the optics, maybe even more so than full backside designs. For example, I'm aware of one PAL manufacturer that changes the corridor length from standard to short on some of their semifinished PALs on a free-form generator; I could see where the alignment of these two surfaces is extremely critical, something that is generally not a concern with most FBS designs.

I'm aware of one coating company that requires an unnamed lab to send a sample lens/coating from each daily production run to inspect for quality. I wonder if any of the PAL companies require the same daily quality audit to watch for mechanical and software errors, almost guaranteeing a very high level of quality control, compared to less frequent spot inspections.

[Barry Santini]

... eventually office based verifiers will become more prevalent and labs will have to clean up their act, until then make sure your lab has always had a knack for accuracy.

Disagree here, MO. It'll be no different than when Dr.'s say "the Rx is off", but I find it compliant.

B

[drk]

Section 5.1 of ANSI Z80.1-2010 states:

For lenses produced with compensations to account for as worn correction, the tolerances in the tables in clause 5 apply to those values specified by the manufacturer and not to the prescribed RX.

Thus, your tolerances would apply to the target compensated Rx.

Last year The Vision Council's Lens Division published a paper on the repeatability and accuracy of lensometer readings across a variety of lensometers. The conclusion was that there was scatter of +-0.06 in readings under the best of conditions. So while somebody's process may be accurate to +-0.01, you probably can't measure it.

And, I will add (cautiously, because I don't have all the solid facts) that lensometry is index-dependent.

So chill out, Edwin Hubbles.

Here's some weaksauce that may be interesting: a decente refractionista will tilt their phoroptorista about 5-10 degrees for some panto.

[Robert Martellaro]

And, I will add (cautiously, because I don't have all the solid facts) that lensometry is index-dependent.

Not with manual lensometers. True with most auto lensometers, although we can select the correct index electronically for the proper reading.

Here's some weaksauce that may be interesting: a decente refractionista will tilt their phoroptorista about 5-10 degrees for some panto.

I would worry about power variance due to a too long vertex distance, if their brow isn't snug against the back of the phoropter. It probably doesn't matter because almost nobody measures it anymore.

[optical24/7]

Here's some weaksauce that may be interesting: a decente refractionista will tilt their phoroptorista about 5-10 degrees for some panto.

I would worry about power variance due to a too long vertex distance, if their brow isn't snug against the back of the phoropter. It probably doesn't matter because almost nobody measures it anymore.

Personally, I don't want a comped refraction. I'm the guy designing the lenses. There are too many factors on lens design, fit and POW's.

[MIOPE]

I really appreciate all this comments. I am new to freeform and i our case, we own the lab and the stores. So, If something go wrong, we are the only responsible.
Your comments are teaching me what could be wrong so I can pay attention to it. Thanks all.

[drk]

I was taught to tilt the phoroptor, but I should re-examine it, now.

If you don't, then you get a power with no panto comp. Do we routinely panto comp? Not when I was taught in the mid-80s.

Do we, now?

I know we drop the OC so when the eye rotates through the lens the powers are close on downgaze, etc.

[DanLiv]

I was taught to tilt the phoroptor, but I should re-examine it, now.

If you don't, then you get a power with no panto comp. Do we routinely panto comp? Not when I was taught in the mid-80s.

Do we, now?

I know we drop the OC so when the eye rotates through the lens the powers are close on downgaze, etc.

The problem is the prescriber never knows if there is going to be compensation of not, or even simple OC measurement or compensation at all. Are they going to Haute-Couture-Optik for a $1000 pair, or getting 8 pair for $9.95 with free shipping? In cases when we do compensate we are assuming a base zero panto, zero wrap, vertex 12.0 (refracted vertex is of course just guesswork) refraction, because we have no information otherwise. If the prescriber is already compensating, then obviously final compensation will off.

I think it would be cool if a doc did issue scripts with *actual* phoropter panto, wrap, and vertex values on the Rx. Those values would usually be closer to final frame fit than 0,0,12 so even if patients went with simple conventional lenses they should get a better final result, and then if we did compensate from there we need only measure and calculate deviation from the phoropter values.

[drk]

I think you're right.

So you know (if you don't) there's no panto indicator on the phoroptor. It is not wrapped in distance refracting, ever, and it has a negative face form (to converge the optics) for near testing.

Vertex distance will vary wildly, but it can never be less than 14 mm on my phoroptor.

Panto compensation on a "compensated-when-refracted" Rx will cause overcompensation. (Like when we discuss how "there was shrinkage!")

[MakeOptics]

Disagree here, MO. It'll be no different than when Dr.'s say "the Rx is off", but I find it compliant.

B

That might be the case on occasion. If you personally had a verifier and access to a version of the design file and you saw that consistently the lab was off by a good measure I am sure it wouldn't take long before trying another lab. Most offices don't "lab hop", so losing an account is a big deal, once lost it maybe some time before the office will even consider coming back and reputations tarnished are hard to polish back up again.

[MakeOptics]

The problem is the prescriber never knows if there is going to be compensation of not, or even simple OC measurement or compensation at all. Are they going to Haute-Couture-Optik for a $1000 pair, or getting 8 pair for $9.95 with free shipping? In cases when we do compensate we are assuming a base zero panto, zero wrap, vertex 12.0 (refracted vertex is of course just guesswork) refraction, because we have no information otherwise. If the prescriber is already compensating, then obviously final compensation will off.

I think it would be cool if a doc did issue scripts with *actual* phoropter panto, wrap, and vertex values on the Rx. Those values would usually be closer to final frame fit than 0,0,12 so even if patients went with simple conventional lenses they should get a better final result, and then if we did compensate from there we need only measure and calculate deviation from the phoropter values.

Great suggestions, in the past I have as well advocated for inscriptions of the panto, wrap, and vertex on the lens design to indicate to opticians that the lens has been compensated should an optician downstream try to duplicate a Rx from a pair of glasses. A simple reversal in compensation can help to arrive at the original compensated values -or- a difference in measures can be used to compute additional compensation. The game is changing as players it's nice to hash these things out even if they will not e implemented until years down the line.

[drk]

How about no duplication without a prescription? I like that better.