Can Pentax make A Decent Lens? Ricoh have not fixed QC issues...

Started 4 months ago | Discussions thread
Trevor G
Veteran MemberPosts: 6,466Gear list
Like?
Re: Please use better experiment protocol - wait till you see some results!
In reply to rgibbons, 4 months ago

rgibbons wrote:

you have some people that are trying to help you (and a few that aren't), and you seem to be arguing too much with those that are trying to help.

Maybe because I know something that a lot of people don't?

Are you brave enough to name those on both sides, because I certainly don't know who is actually giving me informed help...

Trevor G wrote:

Except that you are wrong, very wrong.

Fo the first time ever I used a DOF Calculator.
Depth of field Near limit 42.8 m

No Trevor, you misunderstand the DOF calculators, The old formulas were meant to show what is acceptably in focus, for a print on the wall, viewed from a "normal" viewing distance. As the print size, and viewing distance change, the acceptable DOF changes. The calculator you used doesn't seem to have an input for these variables, and the result of 42.8m is NOT meant to be used for pixel peeping.

We;ll, thanks for trying to help.

Here's Cambridge in Colour's DOF Calculator - is this a little more accurate?

Even if I set the print size to 48" (twice my monitor's size) and keep the viewing distance at 25cm (half my viewing distance) I still get a DOF of 120m starting at (nearest acceptable sharpness) 90m, which is where the red lamp post is.

If this is right, there is no way that I could be off to one side enough to affect the focus on the right hand side. I could move around at a 45 degree angle, focus on the centre of the building and, with a good lens, both near and far extremities should be in acceptable focus.  NB The building is less than 80m wide at the extremities shown.

The issue is that the lens is very sharp on the left and progressively gets worse on the right.  Even the centre is softer than the left.

There was no temperature effect  affecting the image on the right hand side - we know that, because the matching shots from the Nikon D200 and 50mm f1.8D lens on it show soft focus on both sides and stronger, sharper focus in the centre.  That's how most lenses turn out.

Trevor G wrote:

The subject is far enough that hot air turbulence can fuzz out minute details.

But it wasn't happening in this case. It's Autumn here.

Optics has been my main hobby for 40 years, I spend a lot of time with my telescope, trying to get maximum magnification. Air turbulence ("seeing" distortions) exist all year round, Hot or cold, night and day, It is almost always a problem and a the limiting factor, the building will not have equal turbulence on both sides. You might reduce this variable in your experiment by rotating the camera 180 degrees (take a picture upside down), and see if the left side of the building stays sharper.

It depends on the lens.

My DA16-45 was very fuzzy on the left while my Sigma 17-70 was very fuzzy on the right.

Seriously, I know what temperature effects look like.  I have been shooting for a little longer than 3 weeks.

The softer side of your photo of the building also has trees causing additional diffraction around some of the more blurry sections. The bricks far from the trees are better test targets (and yes, the bricks on the left do appear to be sharper).

Thank you.

Not just the bricks, check out the lattice work and follow it across.

For your Mixer board photo (I like the model, I spend time on one every Sunday), when I look at the left and right sides at maximum magnification for many minutes, I can't say one side is sharper than the other; rather it appears that the right side comes into focus a few rows of knobs higher than the left side, (either the board is not perfectly square with the camera, or the lens doesn't focus at the same distance for right and left side).

You've got it!

I have some much better examples of skew coming up shortly.  You might not believe what a bad lens looks like when you map out its DOF characteristics like this.

I truly am amazed that so few people understand the principle behind what I believe is the most common lens abnormaility, and even fewer seem to have seen it.  But don't leave me, I'll either post here or start a new thread

But for my purposes, I'd say the lens is good enough that I'd use it, rather than return it and get someone else's problems when the Pentax service center sends back a rebuilt lens that was only tested for center focus.

Use it with that amount of defocus on the right?  Wow, not me.

There's no way a lens like that could have scored 18 at DXO Mark

I'm not sure why you can accept that 19 of 20 Pentax lenses are bad, but not accept that 4 of 4 bodies might have trouble, (they are statistically the same results as far as margin of error for the sample sizes).

The bodies went back to Pentax to see what they could do.  They were no different when they returned.

I spoke to their chief optical customer relations guy today and they simply will not entertain any thought of viewing a 3D map such as I have produced.  That's because if they did, they would have to fail a lot more lenses.

Their method is to focus on a newspaper at whatever distance and view corner sharpness.  Of course, if they  set focus on a central target and get it right you can have softer sides because one side is forward focused (via lens irregularities) while the other side is back focused when compared with the centre target.

I'll show you sonme great examples shortly.

I like optics problems, and just wanted to offer a possibility to the question of your original post. I've read the entire thread thus far, I'm not a Pentax Fanboy. I applaud the effort and testing you have done thus far to prove your idea; however, even good science benefits from peer reviews.

I appreciate your thoughtfulness. 

-- hide signature --

Cheers
Trevor G
Silkypix tutorials at: http://photo.computerwyse.com

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow