Go Back   Steve's Digicams Forums > Digicam Help > What Camera Should I Buy?

Reply
 
Thread Tools Search this Thread
Old Apr 27, 2007, 4:04 PM   #11
Senior Member
 
TCav's Avatar
 
Join Date: Sep 2005
Location: Washington, DC, Metro Area, Maryland
Posts: 13,544
Default

kenbalbari wrote:
Quote:
One thing you could look at that generally has an impact is sensor size or pixel pitch. Like any other one number though, it's not always that meaningful.

But in general, a 1/2.5" sensor is better than a 1/3", and a 1/1.8" is better than a 1/2.5". But you have to be comparing contemporary models, and image processing plays a large role as well and varies between manufacturers. It used to be some of the highest quality digicams had a 2/3" (same as 1/1.5") sensor, but nowadasy the largest seem to be in the 1/1.8-1/1.6" range. And there aren't many.

There are also many smaller sensors which produce excellent image quality, especially at lower ISO. But this is at least one specification that is available and probably more meaningful than megapixels.
"More meaningful than megapixels"? I don't think I could go along with that. Perhaps you misspoke?

kenbalbari wrote:
Quote:
But best bet for image quality is still to pay more attention to reviews, reputation, and samples than specs. And the lens often plays a bigger role than the sensor, but there's no simple way to measure lens quality either.
Without a doubt.

kenbalbari wrote:
Quote:
There's alot more to a good lens than resolution.
To be sure, but if a lens has bad resolution, then nothing else matters.
TCav is offline   Reply With Quote
Old Apr 27, 2007, 5:32 PM   #12
Senior Member
 
bernabeu's Avatar
 
Join Date: Jan 2005
Posts: 990
Default

RIGHT ON -> LENS + SENSOR QUALITY (pixel pitch)



fuji 3800 6x zoom 3.1 megapix - printable photo quality @ 10x8

Last edited by bernabeu; Jun 27, 2015 at 4:25 PM.
bernabeu is offline   Reply With Quote
Old Apr 27, 2007, 5:57 PM   #13
Senior Member
 
Join Date: Feb 2007
Location: Wisconsin
Posts: 1,241
Default

peripatetic wrote:
Quote:
As you say Canon and Sigma publish their MTF figures. And Nikon do too on their Japanese website. But their figures aren't really comparable between the other manufacturers though they do a decent job of allowing you to compare the different lenses in their own lineups.
Every Canon MTF chart I've ever seen fails to label anything. This renders the thing meaningless. If someone can tell me what the axis and linesshould be labeled, then it has value. The way it is now, they don't even have to be uniform from lens to lens.

From what I understand Nikon publishes MTF charts based on what the lens SHOULD do in theory, and not in actual testing. Once again, not as valuable.

OLympus, Leica, and Sigma do tests, but use different criteria from each other. With no standard, there can be no comparison.

The world seems MTF nuts. Until there is a published standard of testing it is mumbo-jumbo.
fldspringer is offline   Reply With Quote
Old Apr 29, 2007, 1:46 PM   #14
Super Moderator
 
peripatetic's Avatar
 
Join Date: Nov 2004
Posts: 3,599
Default

That's rather over the top.

MTF can be understood by reading any book on optics, and has been in use for many (50+) years. Every manufacturer also puts up an idiots guide next to the charts by following a link. The one under the Canon MTF charts says "How to Read MTF Chart?" if you have managed to miss that on every occasion it's hard to see how Canon is to blame.

Try this link and click on "How to Read MTF Chart?" below the chart.

Sigma likewise has a link. Nikon I presume has links too, but I can't read Japanese, so I am just guessing really.

Leica and Zeiss have a wealth of detailed information about MTF on their websites.

The reason that Canon & Nikon don't put up actual measured MTF, but rather calculated MTF is that there is sufficient sample variance off their manufacturing lines that it would be problematic.

Photozone.de and SlrGear.com use ImaTest and DXO Analyzer respectively and test actual samples of lenses. Both programs are readily available to buy and have a wealth of information about MTF.

I don't understand the General Theory of Relativity, but the books are there and the mathematics available if I am interested enough. The fact that I don't understand it doesn't make it mumbo jumbo.

If you were interested in understanding MTF there are thousands of web pages and dozens of books that you can read.


peripatetic is offline   Reply With Quote
Old Apr 29, 2007, 3:12 PM   #15
Senior Member
 
Join Date: Feb 2007
Location: Wisconsin
Posts: 1,241
Default

The response was intended to be over the top. The fact that manufacturers publish MTF charts and their own guides as to how to read their own charts is the exact point I'm trying to make.

The fact that a chart is compiled with anticipated data vs actual is additional insult. If the manufacturing tolerances are that wide, perhaps they need to publish a MMTF (Mean Modulation transfer function) where the first "mean" is the statistical mean of the varience of the manufacturer tolerance.

Its one thing to understand the MTF and how it applies to optics. Its another to make testing standards uniform so comparisons can be made across the manufacturer spectrum. Its that lack of a system I'm calling mumbo-jumbo. Just as I don't trust a drug manufacturer to choose how to test its new medications, I don't trust a lens manufacturer to cook up its own formula to publish these charts. The tendancy is to come up with respectable results by lowering testing standards. If these manufacturers aren't even testing to verify, the charts lack what little value they would have.

I graduated with honors and I had my share of time spent in the optics lab. If I switched between four or five sets of measuring criteria, and finally decided to settle on theoretical values, I'd spend a whole lot more time in that lab because I'd be told to start from scratch. That's science.

By the way, here's the link to a Canon lens and I fail to see the "How to read" part. I have seen countless lenses and haven't seen that link. Pleae help.

http://www.usa.canon.com/consumer/co...p;modelid=7443

fldspringer is offline   Reply With Quote
Old Apr 29, 2007, 3:47 PM   #16
Senior Member
 
Join Date: Feb 2007
Location: Wisconsin
Posts: 1,241
Default

A link that points to what I'm trying to say:

http://www.zeiss.com/de/photo/home_e...4?OpenDocument
fldspringer is offline   Reply With Quote
Old Apr 30, 2007, 2:00 AM   #17
Super Moderator
 
peripatetic's Avatar
 
Join Date: Nov 2004
Posts: 3,599
Default

Hmm, sorry about the link - forgot to paste one in.

And yes it does seem that the Canon US site doesn't put the MTF explanation in. The Canon UK doesn't put the MTF chart in at all.

I tend to use the Canon HK site as the best source of info.

It doesn't appear to be available as I'm writing this, but the main site address is http://www.canon.com.hk


[Edit: here's one - the site is back up...

http://www.canon.com.hk/en/Consumer/...p;tag_id=11106

]


I have some sympathy for the calculated v measured. I suppose it's obvious why they don't put up their measured results. It would depend on the sample size, the manufacturing line used, the year produced, etc, etc. Sure it would be better for consumers if they did that, but hardly in their own interest. And the incentive to cheat if it was done by all manufacturers would be strong enough to make the results presented somewhat doubtful anyway.

But Zeiss and Leica have very precise manufacturing lines, and in fact as far as I know every lens is measured and their premium lenses are hand made and calibrated by craftsmen. This is the reason that they are so incredibly expensive and so incredibly good and have so little variance between copies.

At any rate Sigma do publish their "What is MTF" guide on all their websites I believe.

http://www.sigma-imaging-uk.com/lens...hart_guide.htm

But I guess the real point is that the independent review sites do compare across manufacturers, but the results there are biased because they do not test enough samples to give a figure on the mean measurements. They tend to just test one copy and unless it's obviously flawed they don't bother with another.

I guess the only solution is to buy yourself a copy of Imatest, make sure you get your lenses from a shop that will let you return them and test them before you keep them.





peripatetic is offline   Reply With Quote
Old Apr 30, 2007, 6:04 AM   #18
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

peripatetic wrote:
Quote:
I guess the only solution is to buy yourself a copy of Imatest...
Instead of doing this, I tend to agree with peripatetic and trust the independent reviewers like Photozone and Photodo who already run Imatest and spot the trend of their outcomes:

http://www.photodo.com/category_2.html
http://www.photozone.de/8Reviews/
NHL is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 3:42 PM.