Reading mode: Light Dark

MTF maybe 2x better than reported?

Go to last post
L

Larry Lueck

Senior Member
Messages
1,258
Reaction score
0
Location
US
In another thread, Joe ('the lens guy') mentioned that the MTF for
the new lenses wasn't very good. But I stumbled across the
following quote at the address listed below. There's diagrams
and such there as well. I wonder how this affects Joe's comments?

http://www.olympusamerica.com/e1/feat_quality_elements.asp

Begin Quote:

In the Four Thirds MTF chart, 20 lines per millimeter is used vs. the 10 lines per millimeter for 35mm film lenses and 60 lines per millimeter instead of 30 lines per millimeter for the 35mm film format.

A measurement of contrast at 60% or higher in a lens is considered acceptable. If the contrast measures at 80% or higher it is considered to be a very fine lens.

End Quote.
 
Not really. They have to use twice as much of lines because the 4/3 system is twice as small as the 35 mm format, in order for the test to be meaningful.

(i.e. if it resolves 10 lines per mm in 35 mm format, it has to resolve 20 lines per mm in 4/3 format in order for the photos they produce to look equivalent, since the magnification to say, 8x10 print is twice as much for the 4/3 system)
Larry Lueck said:
In another thread, Joe ('the lens guy') mentioned that the MTF for
the new lenses wasn't very good. But I stumbled across the
following quote at the address listed below. There's diagrams
and such there as well. I wonder how this affects Joe's comments?

http://www.olympusamerica.com/e1/feat_quality_elements.asp

Begin Quote:

In the Four Thirds MTF chart, 20 lines per millimeter is used vs.
the 10 lines per millimeter for 35mm film lenses and 60 lines per
millimeter instead of 30 lines per millimeter for the 35mm film
format.

A measurement of contrast at 60% or higher in a lens is considered
acceptable. If the contrast measures at 80% or higher it is
considered to be a very fine lens.

End Quote.
Click to expand...
--
Gallery: http://violin.deviantart.com/gallery
 
Tony Ku said:
Not really. They have to use twice as much of lines because the
4/3 system is twice as small as the 35 mm format, in order for the
test to be meaningful.
Click to expand...
Which would prove that running Canon's 35mm lenses on the 10d
is a HORRIBLE mistake...

I'm surprised you couldn't come up with a better troll/bash, Tony...
 
LOL
Hi Larry - our Tony really did walk straight into that one!

I'd guess you don't get a reply there!

kind regards
jono slack
Larry Lueck said:
Tony Ku said:
Not really. They have to use twice as much of lines because the
4/3 system is twice as small as the 35 mm format, in order for the
test to be meaningful.
Click to expand...
Which would prove that running Canon's 35mm lenses on the 10d
is a HORRIBLE mistake...

I'm surprised you couldn't come up with a better troll/bash, Tony...
Click to expand...
--
Jono Slack
http://www.slack.co.uk
 
Who's trolling? A nyone who doesn't agree with you is trolling?

Geez Larry -- my Canon glass resolves FAR MORE than my 10D sensor can take. that's my point... and hence, the E-1 system does not give much advantage at 5 megapixel level.

If you're talking about 20 megapixel sensor in a 10D format, then you have a point, but we're not comparing that, are we?

Remember: SLIDE FILMS have about the equivalent of 25-30 megapixel resolution in full 35 mm format. a 10D sensor when resized to the full frame is only about 15 megapixels, far far less than what slide film can record.

Also, it'll be interesting to see if Olympus CAN make glass that can go twice as much resolution as normal 35 mm lenses. :) That should be what you're worried about.....
Larry Lueck said:
Tony Ku said:
Not really. They have to use twice as much of lines because the
4/3 system is twice as small as the 35 mm format, in order for the
test to be meaningful.
Click to expand...
Which would prove that running Canon's 35mm lenses on the 10d
is a HORRIBLE mistake...

I'm surprised you couldn't come up with a better troll/bash, Tony...
Click to expand...
--
Gallery: http://violin.deviantart.com/gallery
 
Just to clarify where the 15 megapixel came from:

if you take a 10D sensor's resolution, 6 megapixels, and use the same desnity to FIT a 35 mm full frame, you will get approximately 15 megapixels.

Slide films goes to about 25-30 megapixels or so, at least the finest grain, and the lowest ISO films such as Velvia is acapble of resolving. And they work fine with current 35 mm lenses.

So my point is, the 10D sensor can resolve just fine at 6 megapixels. It'll have a problem if the same sensor is at 20 megapixels (i.e. it won't be able to get to the 20 meagpixel resolution). But we all knowwhere Canon's direction is going: to full frame 35 mm format.

Good luck to Olympus convincing anyone else but Olympus to make glass for the 4/3 systems, except for the two film companies that are in trouble due to their dependency of earnings on film sales....... neither of them makes good glass. If they get Nikon to sign up, i'll be impressed and real competition will begun, if not, those 80% DSLR market ownership of Canon, will probably approach 90% in a year!
Tony Ku said:
Geez Larry -- my Canon glass resolves FAR MORE than my 10D sensor
can take. that's my point... and hence, the E-1 system does not
give much advantage at 5 megapixel level.

If you're talking about 20 megapixel sensor in a 10D format, then
you have a point, but we're not comparing that, are we?

Remember: SLIDE FILMS have about the equivalent of 25-30 megapixel
resolution in full 35 mm format. a 10D sensor when resized to the
full frame is only about 15 megapixels, far far less than what
slide film can record.

Also, it'll be interesting to see if Olympus CAN make glass that
can go twice as much resolution as normal 35 mm lenses. :) That
should be what you're worried about.....
Larry Lueck said:
Tony Ku said:
Not really. They have to use twice as much of lines because the
4/3 system is twice as small as the 35 mm format, in order for the
test to be meaningful.
Click to expand...
Which would prove that running Canon's 35mm lenses on the 10d
is a HORRIBLE mistake...

I'm surprised you couldn't come up with a better troll/bash, Tony...
Click to expand...
--
Gallery: http://violin.deviantart.com/gallery
Click to expand...
--
Gallery: http://violin.deviantart.com/gallery
 
Jono Slack said:
I'd guess you don't get a reply there!

kind regards
jono slack
Larry Lueck said:
Tony Ku said:
Not really. They have to use twice as much of lines because the
4/3 system is twice as small as the 35 mm format, in order for the
test to be meaningful.
Click to expand...
Which would prove that running Canon's 35mm lenses on the 10d
is a HORRIBLE mistake...

I'm surprised you couldn't come up with a better troll/bash, Tony...
Click to expand...
--
Jono Slack
http://www.slack.co.uk
Click to expand...
Below is part of a thread posted by Tony in the Sigma Forum. He was literally made to eat his words when confronted by the response to his remarks about the quality of the SD9. I like the bit where he says...'but Canon can still produce nice images'....what? only nice.

I have no doubt that he will be made to eat his words again when there is an aboundance of supior than D10 images posted in this Forum and I for one intend to demonstate this.

Richard

Extract of Tony's post.....

4. It's true Foveon can produce incredible images RIGHT off of the camera, but as I have pointed it out, it's all a matter of adjustments post camera that matters. Canon will probably require more post-camera work. The images you posted are SATURATED colors. The images I posted are have not really been saturated that much. Canon 10D has very conservative scaling.

5. Foveon will be sharper than Canon, and canon images will require more sharpening. This is where Foveon has the real edge, IMHO. Especially when it renders leaves or similar colors.

That being said, you can still produce nice images from Canon 10D.
 
Tony Ku wrote:
a load of old dingo's kidneys.

Tony - there are two problems with your posts (well, at least two)

1. you quote as 'established facts' observations which are, at least, arguable, and often simply untrue.

2. you have some idea that using Because and Therefore in your arguments confers some sort of logical validity.

I could think of some other, more personal reasons, but will restrain myself.

You are causing a great deal of irritation to a number of people here - I'm sure that isn't your intention. Why don't you bugger off, wait until there are a lot of samples on the net, then you can come back and say:

"I told you so", and we'll all say "That Tony, he's such a clever guy"

I, for one, have done with you, arguing with you is like arguing with a bowl of jelly, and, quite frankly, you are about as well informed.

I hope you have a happy time with your 10D, but why don't you have a happy time in the Canon forum

kind regards
jono slack
Tony Ku said:
Larry Lueck said:
Tony Ku said:
Not really. They have to use twice as much of lines because the
4/3 system is twice as small as the 35 mm format, in order for the
test to be meaningful.
Click to expand...
Which would prove that running Canon's 35mm lenses on the 10d
is a HORRIBLE mistake...

I'm surprised you couldn't come up with a better troll/bash, Tony...
Click to expand...
--
Gallery: http://violin.deviantart.com/gallery
Click to expand...
--
Jono Slack
http://www.slack.co.uk
 
Tony Ku said:
Who's trolling? A nyone who doesn't agree with you is trolling?

Geez Larry -- my Canon glass resolves FAR MORE than my 10D sensor
can take. that's my point... and hence, the E-1 system does not
give much advantage at 5 megapixel level.
Click to expand...
Tony, before you embarass yourself further, I'd suggest
you spend some time (actually, a LOT of time) here:
http://clarkvision.com/imagedetail/index.html

Your Canon glass does NOT resolve more, not while retaining
high contrast. It is suspected that the E-1 lenses will do
even better than the E-10/20 did. And the combination of
E-10/20 CCD and that lens has been shown to produce more
contrast in small details than designed-for-last-milleniums'
35mm lenses do.
 
Tony Ku said:
Good luck to Olympus convincing anyone else but Olympus to make
glass for the 4/3 systems, except for the two film companies that
are in trouble due to their dependency of earnings on film
sales....... neither of them makes good glass. If they get Nikon
to sign up, i'll be impressed and real competition will begun, if
not, those 80% DSLR market ownership of Canon, will probably
approach 90% in a year!
Click to expand...
If you are referring to Fuji, you should do your homework before you talk.

Have you heard of EBC Fujinon lenses? Of course not. Otherwise you wouldn't make such a stupid remark.

I guess Hasselblad made a mistake by having Fuji manufacture their latest professional camera, including all their lenses.
I own a couple of Fuji MF cameras and their lenses are top notch.

Jeff
 
Actually this is probably one of the sources for the info, but his conclusion is wrong. he's assuming that 15 megapixels equate Velvia. Yes, MTF tests wise it does, but not the current Bayes sensors.

Velvia grain senses a COLOR (combination of R/G/B). Bayes sensor only detects ONE color, red green or blue. Unless it's 15 megapixels on a FOVEON sensor, his conclusion is not correct. It's correct for resolution but not color/contrast. You'll need 45 megapixels of Bayes type (any CCD or CMOS sensor) to match this.

In terms of apparent resolution, that is, what you see on a photograph, contrast and color matters more than the newspaper test MTF graphs.

Even if you take his conclusion to be valid, and 15 megapixels is the magic number. Well, the equivalent 10D sensor when scaled to 35 mm would be 15 megapixels, which means that the 6 megapixels within the 10d size sensors would be more than adequate.

Another thing to consider: bigger glass = MORE light onto the film plane. More light = more contrast. This is an area where no one is consdering and where MTF tests will not show. Unless you plan on shooting newspapers, i'd suggest that you look into it more.

And your last statement: show me the tests that prove this :-)

Believe me, guys, I'd love to be proven wrong when the E system comes out. But what I find really funny is that everyone here is hyping on it and thinking it'll be much better than anything out there when it's all just VAPORware at this moment.
Larry Lueck said:
Tony Ku said:
Who's trolling? A nyone who doesn't agree with you is trolling?

Geez Larry -- my Canon glass resolves FAR MORE than my 10D sensor
can take. that's my point... and hence, the E-1 system does not
give much advantage at 5 megapixel level.
Click to expand...
Tony, before you embarass yourself further, I'd suggest
you spend some time (actually, a LOT of time) here:
http://clarkvision.com/imagedetail/index.html

Your Canon glass does NOT resolve more, not while retaining
high contrast. It is suspected that the E-1 lenses will do
even better than the E-10/20 did. And the combination of
E-10/20 CCD and that lens has been shown to produce more
contrast in small details than designed-for-last-milleniums'
35mm lenses do.
Click to expand...
--
Gallery: http://violin.deviantart.com/gallery
 
There is one thing different Richard:

These people actually might bge using a system that have a chance of surviving :-) At least they're not stuck with Sigma glass.
Richard V Lamprecht said:
Jono Slack said:
I'd guess you don't get a reply there!

kind regards
jono slack
Larry Lueck said:
Tony Ku said:
Not really. They have to use twice as much of lines because the
4/3 system is twice as small as the 35 mm format, in order for the
test to be meaningful.
Click to expand...
Which would prove that running Canon's 35mm lenses on the 10d
is a HORRIBLE mistake...

I'm surprised you couldn't come up with a better troll/bash, Tony...
Click to expand...
--
Jono Slack
http://www.slack.co.uk
Click to expand...
Below is part of a thread posted by Tony in the Sigma Forum. He was
literally made to eat his words when confronted by the response to
his remarks about the quality of the SD9. I like the bit where he
says...'but Canon can still produce nice images'....what? only nice.

I have no doubt that he will be made to eat his words again when
there is an aboundance of supior than D10 images posted in this
Forum and I for one intend to demonstate this.

Richard

Extract of Tony's post.....

4. It's true Foveon can produce incredible images RIGHT off of the
camera, but as I have pointed it out, it's all a matter of
adjustments post camera that matters. Canon will probably require
more post-camera work. The images you posted are SATURATED colors.
The images I posted are have not really been saturated that much.
Canon 10D has very conservative scaling.

5. Foveon will be sharper than Canon, and canon images will require
more sharpening. This is where Foveon has the real edge, IMHO.
Especially when it renders leaves or similar colors.

That being said, you can still produce nice images from Canon 10D.
Click to expand...
--
Gallery: http://violin.deviantart.com/gallery
 
Tony Ku said:
Actually this is probably one of the sources for the info, but his
conclusion is wrong. he's assuming that 15 megapixels equate
Velvia. Yes, MTF tests wise it does, but not the current Bayes
sensors.
Click to expand...
Bayer, Tony, bee aye why eee arr
Tony Ku said:
Velvia grain senses a COLOR (combination of R/G/B). Bayes sensor
Click to expand...
There's three layers, Tony. Each layer responds to ONE color. And
I don't even shoot FILM!!!!
Tony Ku said:
only detects ONE color, red green or blue. Unless it's 15
megapixels on a FOVEON sensor, his conclusion is not correct. It's
Click to expand...
'his conclusion is not correct'... Sure, Tony.
Tony Ku said:
correct for resolution but not color/contrast. You'll need 45
megapixels of Bayes type (any CCD or CMOS sensor) to match this.

In terms of apparent resolution, that is, what you see on a
photograph, contrast and color matters more than the newspaper test
MTF graphs.
Click to expand...
EXACTLY! Which is why digicams look SO good compared to film!
Tony Ku said:
Even if you take his conclusion to be valid, and 15 megapixels is
the magic number. Well, the equivalent 10D sensor when scaled to
35 mm would be 15 megapixels, which means that the 6 megapixels
within the 10d size sensors would be more than adequate.
Click to expand...
Except you forget that the BAYER pattern provides 1 blue pixel, 1
red pixel and TWO green pixels, so they AREN'T strictly comparable.
Tony Ku said:
Another thing to consider: bigger glass = MORE light onto the film
plane. More light = more contrast. This is an area where no one
is consdering and where MTF tests will not show. Unless you plan
on shooting newspapers, i'd suggest that you look into it more.
Click to expand...
I don't need to look into it, Tony! I understand F-stops. (Hope the
rest of you get the joke, too!)
Tony Ku said:
Believe me, guys, I'd love to be proven wrong when the E system
comes out. But what I find really funny is that everyone here is
hyping on it and thinking it'll be much better than anything out
there when it's all just VAPORware at this moment.
Click to expand...
You also don't understand the term 'vaporware'. That means an
announced product that has NOT BEEN PRODUCED. What, exactly
do you think all the reviewers were holding, cumquats?

This was funny for awhile, Tony...but now you're to the point
of being absurd.
 
Larry Lueck said:
You also don't understand the term 'vaporware'. That means an
announced product that has NOT BEEN PRODUCED. What, exactly
do you think all the reviewers were holding, cumquats?
Click to expand...
Oh my, such a lovely mental image!

(But, but... The already on the market Canon 10K kumquat is able to use existing peelers, and because of the size of its pulp's grain it contains more actual fruit than the Olymous Eat-1 kumquat!)

-Dylan
 
Larry Lueck said:
In another thread, Joe ('the lens guy') mentioned that the MTF for
the new lenses wasn't very good. But I stumbled across the
following quote at the address listed below. There's diagrams
and such there as well. I wonder how this affects Joe's comments?

http://www.olympusamerica.com/e1/feat_quality_elements.asp

Begin Quote:

In the Four Thirds MTF chart, 20 lines per millimeter is used vs.
the 10 lines per millimeter for 35mm film lenses and 60 lines per
millimeter instead of 30 lines per millimeter for the 35mm film
format.

A measurement of contrast at 60% or higher in a lens is considered
acceptable. If the contrast measures at 80% or higher it is
considered to be a very fine lens.

End Quote.
Click to expand...
--
Ciao!

Joe
 
Joseph S. Wisniewski wrote:
--
Ciao!

Joe
 
The resulting Velvia film is ONE color (components of R/G/B as i have said)!

Bayesian sensors only have one COMPONENT color PER pixel, the results are interpolated. Thus it's not "true" pixel. The comparison of 15 megapixels versus Velvia is not valid, as the pixel it generated is an approximation. In order to get true information, it will need 3 times that to compare to Velvia film.

This won't matter for MTF tests, because it does have luminosity values, but it will matter for real color/contrast, which is much more important than MTF. And thus, it will require digital cameras a lot more pixels to equate to film! (unless they're using something like Foveon).

You actually think digicam photos look better than film???? You gotta be kidding, right?
Larry Lueck said:
Tony Ku said:
Actually this is probably one of the sources for the info, but his
conclusion is wrong. he's assuming that 15 megapixels equate
Velvia. Yes, MTF tests wise it does, but not the current Bayes
sensors.
Click to expand...
Bayer, Tony, bee aye why eee arr
Larry Lueck said:
Velvia grain senses a COLOR (combination of R/G/B). Bayes sensor
Click to expand...
There's three layers, Tony. Each layer responds to ONE color. And
I don't even shoot FILM!!!!
Larry Lueck said:
only detects ONE color, red green or blue. Unless it's 15
megapixels on a FOVEON sensor, his conclusion is not correct. It's
Click to expand...
'his conclusion is not correct'... Sure, Tony.
Larry Lueck said:
correct for resolution but not color/contrast. You'll need 45
megapixels of Bayes type (any CCD or CMOS sensor) to match this.

In terms of apparent resolution, that is, what you see on a
photograph, contrast and color matters more than the newspaper test
MTF graphs.
Click to expand...
EXACTLY! Which is why digicams look SO good compared to film!
Larry Lueck said:
Even if you take his conclusion to be valid, and 15 megapixels is
the magic number. Well, the equivalent 10D sensor when scaled to
35 mm would be 15 megapixels, which means that the 6 megapixels
within the 10d size sensors would be more than adequate.
Click to expand...
Except you forget that the BAYER pattern provides 1 blue pixel, 1
red pixel and TWO green pixels, so they AREN'T strictly comparable.
Larry Lueck said:
Another thing to consider: bigger glass = MORE light onto the film
plane. More light = more contrast. This is an area where no one
is consdering and where MTF tests will not show. Unless you plan
on shooting newspapers, i'd suggest that you look into it more.
Click to expand...
I don't need to look into it, Tony! I understand F-stops. (Hope the
rest of you get the joke, too!)
Larry Lueck said:
Believe me, guys, I'd love to be proven wrong when the E system
comes out. But what I find really funny is that everyone here is
hyping on it and thinking it'll be much better than anything out
there when it's all just VAPORware at this moment.
Click to expand...
You also don't understand the term 'vaporware'. That means an
announced product that has NOT BEEN PRODUCED. What, exactly
do you think all the reviewers were holding, cumquats?

This was funny for awhile, Tony...but now you're to the point
of being absurd.
Click to expand...
--
Gallery: http://violin.deviantart.com/gallery
 
Unless the ccd can use it it is useless. I recall the hype Olympus pulled over ED glass in the Exx: it was functionally useless according a lens designer who studied their patent...

I can say that going from an E10 to a D60 with 'mtf challenged' lenses was a revelation. I MUCH prefer the D60 to the E10 - even with cheap Sigma glass. It kills it dead! There is much more to this than 'mtf graphs' and the like. Watch out for the hype. Use your eyes. In which case, we will have to wait for REAL results.

I hope Olympus makes it happen this time - a new standard with upgrades to follow. I hope this is not just more hype and build quality to hide mask noise and going-nowhereness. Genuine compeition and choice are good for all of us.

Dave (Plooph!)
Larry Lueck said:
In another thread, Joe ('the lens guy') mentioned that the MTF for
the new lenses wasn't very good. But I stumbled across the
following quote at the address listed below. There's diagrams
and such there as well. I wonder how this affects Joe's comments?

http://www.olympusamerica.com/e1/feat_quality_elements.asp

Begin Quote:

In the Four Thirds MTF chart, 20 lines per millimeter is used vs.
the 10 lines per millimeter for 35mm film lenses and 60 lines per
millimeter instead of 30 lines per millimeter for the 35mm film
format.

A measurement of contrast at 60% or higher in a lens is considered
acceptable. If the contrast measures at 80% or higher it is
considered to be a very fine lens.

End Quote.
Click to expand...
--
http://www.photo.net/photodb/user?user_id=570619
http://www.usefilm.com/browse.php?mode=port&data=15884
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top