Horshack

Horshack

Lives in United States United States
Has a website at http://horshack.smugmug.com
Joined on Jun 7, 2008

Comments

Total: 585, showing: 1 – 20
« First‹ Previous12345Next ›Last »

Some of the best test images I've seen posted on the DPR front page.

Direct link | Posted on Feb 12, 2016 at 04:33 UTC as 27th comment
In reply to:

Horshack: Roger dutifully sheds light on a surprisingly complex issue. But let's not throw the baby out with the bathwater when it comes to forgiving the QC sins of these companies on the altar of complexity and perceived manufacturing/testing cost. The fact is we don't have enough information to quantify and project how QC could improve relative to the additional marginal costs of those improvements. But in the absence of such information (which could be used to either forgive or indict their QC sins), I think it's useful to look at one essential clue - how is it that these companies have such meager manufacturing+service testing capabilities for what are precision, high-value products? Is it really due to the cost of such equipment and the personnel needed to operate them? Or is it instead a 'see no evil do no evil' scenario, where plausible deniability is utilized to mask a simple unwillingness to deliver a consistent quality product irrespective of how small the impact on margins may be.

I agree. My contention is that we shouldn't necessarily presume there aren't other reasons behind QC lapses besides cost.

Direct link | Posted on Feb 12, 2016 at 01:31 UTC
In reply to:

c45: On different but related topic, according to many experts, the optical performance of modern lenses had deteriorated in many areas. Lens manufactures react to overwhelming pressure coming from digital-photography consumers (many of them just pixel peepers) to eliminate easily detectable chromatic aberration and deliver corner-to-corner sharpness. Both goals can accomplished at design level by adding optical lens-elements to the design, it's not uncommon now days to see 20+ glass elements in a lens. Unfortunately, adding glass element causes deterioration of micro-contrast, perceptual depth of field, color saturation etc.
And, unfortunately not all of it can be corrected in post processing.
Many photographers are buying older low glass element-count lenses (often manual focus ZEISSs, Nikons or Voighlanders) to get maximum IQ.

@Airy, don't additional elements also have the potential to reduce light tranmission (t/stop)? Or is that not necessarily the case?

Direct link | Posted on Feb 12, 2016 at 01:26 UTC
In reply to:

Mike FL: So, basically the message of this article tried to deliver by LensRentals is very clear:

If you want picture perfect, rent lens from my LensRentals b/c:
- I have $ XXX USD test bench for the lens
- Lens makers do not know what are they doing
- I can do it right, but too hard for Lens makers do it right.

Again, please rent the lens from me. Please.

@neodp, I guess when one has a hammer everything and everybody looks like a nail. Against that unfortunate reality I suppose I will in fact have to just agree to disagree with you and Mike on this matter.

Direct link | Posted on Feb 11, 2016 at 18:37 UTC
In reply to:

Mike FL: So, basically the message of this article tried to deliver by LensRentals is very clear:

If you want picture perfect, rent lens from my LensRentals b/c:
- I have $ XXX USD test bench for the lens
- Lens makers do not know what are they doing
- I can do it right, but too hard for Lens makers do it right.

Again, please rent the lens from me. Please.

I don't understand how you got that message from this article. Anybody who has read Roger's articles knows he does what he does out of genuine curiosity and a desire to help others understand complex topics. On a more personal level Roger offered and perform a favor for me a few years back that required a lot of time and expense on his part without any prompting and refused all offers for compensation.

Direct link | Posted on Feb 11, 2016 at 17:42 UTC
In reply to:

J A C S: There is another side of the story. A design that would be more forgiving of small manufacturing tolerances.

Roger addresses this in the article. He doesn't call it by its name but in design/manufacturing circles it's referred to as "designing for manufacturability".

https://en.wikipedia.org/wiki/Design_for_manufacturability

Direct link | Posted on Feb 11, 2016 at 17:02 UTC

Roger dutifully sheds light on a surprisingly complex issue. But let's not throw the baby out with the bathwater when it comes to forgiving the QC sins of these companies on the altar of complexity and perceived manufacturing/testing cost. The fact is we don't have enough information to quantify and project how QC could improve relative to the additional marginal costs of those improvements. But in the absence of such information (which could be used to either forgive or indict their QC sins), I think it's useful to look at one essential clue - how is it that these companies have such meager manufacturing+service testing capabilities for what are precision, high-value products? Is it really due to the cost of such equipment and the personnel needed to operate them? Or is it instead a 'see no evil do no evil' scenario, where plausible deniability is utilized to mask a simple unwillingness to deliver a consistent quality product irrespective of how small the impact on margins may be.

Direct link | Posted on Feb 11, 2016 at 16:59 UTC as 26th comment | 3 replies
On Connect post Instagram now supports multi-account use (12 comments in total)

Will be funny to see when people accidentally post a personal image to their "business" account.

Direct link | Posted on Feb 10, 2016 at 23:34 UTC as 2nd comment

For small bodies I go for the thinnest and lightest strap I can find. Comfort isn't an issue since the body doesn't put much weight around your neck. These straps look nice but a little unwieldy and heavy.

Direct link | Posted on Feb 10, 2016 at 03:31 UTC as 101st comment
On article Heavy hitter: Sony FE 24-70mm F2.8 GM first shots (337 comments in total)
In reply to:

Rishi Sanyal: Have any of our readers spotted any samples elsewhere showing F22 sunbursts from this lens? The Canon 24-70 F2.8L II sets an impossibly high bar for sunstar rendition, which'll be important to landscape shooters considering this lens.

@role_of_72, the 14-24mm has pretty good sunstars. Here's a photo with mine @ 14mm f/13:

https://horshack.smugmug.com/Nature/Lake-Tahoe/i-jtmnctv/0/X2/tahoe_465-X2.jpg

Direct link | Posted on Feb 6, 2016 at 15:45 UTC
On article Heavy hitter: Sony FE 24-70mm F2.8 GM first shots (337 comments in total)

Thanks for the images. I immediately went looking for any that are suitable for gauging edge sharpness and centering quality. Image #1 (wide shot of woman on phone) @ 67mm f/3.5 has the edges near within DOF and the edges look promising. Image #3 (wide shot of fighters) @ 24mm f/2.8 looks even more so - the right extreme edge with the white brick looks excellent - the left edge looks worse (couldn't find anything sharp except for sliver of floor) but the shot appears taken with a slight right-to-left orientation so it's hard to say for sure.

Direct link | Posted on Feb 6, 2016 at 15:34 UTC as 50th comment

If every camera maker focused on software things would get a lot better for users.

Direct link | Posted on Feb 5, 2016 at 20:09 UTC as 37th comment | 1 reply
On article Delayed: Nikon Japan pushes D500 to end of April (138 comments in total)

Better to keep the camera's actual performance a mystery (ie no reports from actual owners vs Nikon-selected enthusiastic pre-owners) than to have the camera start entering the mainstream and possibly hinder the pre-release buzz (and sales), esp in regards to expectations for High ISO performance.

Direct link | Posted on Feb 4, 2016 at 22:55 UTC as 19th comment
On article Hands on with Sony's a6300 and G Master lenses (276 comments in total)

If the 24-70 performs as well as the Canon but I can find a well-centered copy then the A7rII will become my main landscape camera.

Direct link | Posted on Feb 4, 2016 at 02:21 UTC as 50th comment
In reply to:

Horshack: DOF looks pretty deep in this video, equal to about 3-4 seconds of motion of the car. This is based upon how much of the road appears to be in focus at any point in time.

@JACS, that looks more like motion blur than being OOF.

Direct link | Posted on Feb 3, 2016 at 22:44 UTC
In reply to:

Horshack: DOF looks pretty deep in this video, equal to about 3-4 seconds of motion of the car. This is based upon how much of the road appears to be in focus at any point in time.

Thanks @sibuzaru, looks impressive. The pigeon stays about the same distance from the camera so it's hard to evaluate AF engagement but the tracking looks impressive.

Direct link | Posted on Feb 3, 2016 at 22:38 UTC
In reply to:

Horshack: DOF looks pretty deep in this video, equal to about 3-4 seconds of motion of the car. This is based upon how much of the road appears to be in focus at any point in time.

Classifying it as a promo video doesn't affect its applicability unless you believe Sony rigged the demonstration. I'm taking the video at face value, no different from any other source, which again includes comparing the DOF vs competing models in order to discount and compare its performance to other models.

Direct link | Posted on Feb 3, 2016 at 21:47 UTC
In reply to:

Horshack: DOF looks pretty deep in this video, equal to about 3-4 seconds of motion of the car. This is based upon how much of the road appears to be in focus at any point in time.

It's a demonstration of video tracking from the actual camera - what else would we use to quantitatively evaluate it? We have tracking examples from other cameras so they can be compared. Part of that evaluation includes comparing the DOF used to see how wide the AF tracking envelope is.

Direct link | Posted on Feb 3, 2016 at 21:20 UTC
In reply to:

Horshack: DOF looks pretty deep in this video, equal to about 3-4 seconds of motion of the car. This is based upon how much of the road appears to be in focus at any point in time.

That's fair but consider Sony is the source of the video, presumably so that the camera's tracking performance can be discussed as we're doing here.

Direct link | Posted on Feb 3, 2016 at 21:14 UTC
In reply to:

Horshack: DOF looks pretty deep in this video, equal to about 3-4 seconds of motion of the car. This is based upon how much of the road appears to be in focus at any point in time.

lol, how so? If it can track as well as the NX1's video AF then it will be very impressive. Here are some NX1 samples:

https://www.youtube.com/watch?v=0DigAgYD-QY
https://www.youtube.com/watch?v=znoDaqWLt9E

Direct link | Posted on Feb 3, 2016 at 20:54 UTC
Total: 585, showing: 1 – 20
« First‹ Previous12345Next ›Last »