July 25, 2014, 03:01:15 PM

Author Topic: Canon may be expensive but...  (Read 19585 times)

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 13543
    • View Profile
Re: Canon may be expensive but...
« Reply #45 on: February 01, 2013, 06:31:00 AM »
Wondering why would Canon, Nikon would NOT measure actual lenses?
It's a numbers game. If they produce ten thousand lenses of a particular lens (I picked that number out of the air - I could be waaaaaaay out on that one...) and each MTF chart has at least eight squiggles, and each squiggle needs at least fifteen points (with zoom, double that - two charts...) that is a lot of points! So which point do you publish? Best? Worst? Average? Median? Mode? I guess simpler not to publish and statistically QA lenses assuring quality is as expected...

I wouldn't think so, I don't thing anyone is suggesting that they QC every lens that rolls off the line with a full MTF curve test.  However, Zeiss does publish real, empirical MTF curves for their production lenses, whereas Canon and Nikon choose to publish idealized MTF charts.  That choice is likely a marketing decision - it makes their lenses look better (but does Zeiss test a whole bunch of lenses, then publish the best curve?  Most likely).  In the end, it probably does not matter...but I'd prefer to see real data from a real lens, rather than (not-real) ideal data which a real lens may never match.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

canon rumors FORUM

Re: Canon may be expensive but...
« Reply #45 on: February 01, 2013, 06:31:00 AM »

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 13543
    • View Profile
Re: Canon may be expensive but...
« Reply #46 on: February 01, 2013, 06:53:22 AM »
To all of you , then read the discussion again, and read Neuros answer.

By all means, people should read that thread and observe Mikael 'Half The Photons' Residal's staunch defense of his explanation that the general mechanism of HTP is that the camera reduces by half the amount of light hitting the sensor, his consistent refusal to admit that his 'explanation' applies only at ISO 100, his avoidance of specific questions from several posters to describe how HTP works at any other ISO setting than 100, etc.  Note how here in this thread, he reposts TheSuede's correct description of how HTP works, which matches what others in the thread were saying, but not his own flawed and incorrect explanation.

The answer is , I have been at Photokina since 30years back  as a member of the press every second year and discussed MTF, Lenses,Cameras, scanners with technical chiefs from Nikon, Olympus, Canon , Leica, you name it and I know they have own MTF equipments like Hasselblad

That's all you had to say, Mikael.  Why did you feel in necessary instead to lead off by insulting my knowledge and intelligence?

I can tell you people that I have got stringent regulations about my use of   language from CR,

Clearly, there's a reason for that.  Those regulations seem to have moderated your overt behavior somewhat (and even that seems to be backsliding), but not your general attitude.  Frankly, I find many of your posts to be rude and condescending.  I respect your 30 years experience as a photographer, but rather than using your experience and knowledge to help others here on these forums, you choose mainly to comment on Canon's poor sensor performance with regard to DR (you're correct about that, but really, most of us understand it already, so why do you keep beating that dead horse?), argue with people, insult others, and generally make these forums a more hostile place. 
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 13543
    • View Profile
Re: Canon may be expensive but...
« Reply #47 on: February 01, 2013, 09:53:41 AM »
Neuro:  Do not try to talk away your mistakes and that you  dont know its needs  a head room to create HTP


Do we really need to go over this again?  I have already pointed out your mistake, the incorrect extrapolation of the case of setting the camera to ISO 100 then enabling HTP while in an autoexposure mode (Av/Tv/P), where the camera will then change the exposure by adjusting the shutter speed or aperture one stop, to be a general explanation that the mechanism of HTP is for the camera to adjust the shutter speed or aperture by one stop, when if fact, that only occurs at ISO 100.  That's something that TheSuede clearly understood and explained very well, but you still seem unable to grasp.

Of course, headroom is needed to preserve the highlights. We agreed on the 'what' - your error concerned the 'how'.  My point was that HTP achieves the additional headroom by exposing at a one-stop lower ISO than is actually selected, not by reducing the exposure so that 'half the photons' hit the sensor as you repeatedly stated.

All of that was hashed out to the nth degree in the other thread, there's no point in restating those arguments - if you want to continue to argue your points, please go back to that thread.

As for MTF - read carefully what you wrote:

Page2 I wrote:
we dont know if this is estimated MTF, Canon and Nikon has a predilection to exhibit estimated MTF results  to impress

and then I wrote:
MTF tests from Nikon, Canon, Leitz, Zeiss , Hasselblad are real MTF tests and of the lens only


First, you state that Canon shows estimated MTF results.  Then you state that Canon's MTF curves are "real MTF tests".  Do you understand that your statements conflict with each other?  Canon shows MTF plots for their lenses (one for primes, two for zooms) - those data are either calculated/theoretical or they are real/empirical.  First you state it's the former (correct) then you state it's the latter (wrong).  Your statement about Canon having the equipment to actually measure MTF is irrelevant - fine, they can measure real MTF, but they do not show those data for their lenses, they show only theoretical data. 

It's obvious that we could go around and around about this just like we did with your 'half the photons' argument, and it's equally obvious that there's no point in doing so.  You know as well as I that Canon's published MTF curves are theoretical/calculated, and not 'real MTF tests' resulting from actual measurement of real lenses.  There is really no point in discussing it further, and therefore, I will not.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

sanj

  • 1D X
  • *******
  • Posts: 1425
    • View Profile
Re: Canon may be expensive but...
« Reply #48 on: February 01, 2013, 10:37:32 AM »
I felt cheated on learning that Canon publishes hypothetical MFT charts and does not actually test a sample lens.

Zv

  • 1D X
  • *******
  • Posts: 1151
    • View Profile
    • Zee-bytes
Re: Canon may be expensive but...
« Reply #49 on: February 01, 2013, 11:03:02 AM »
Theoretical is fine by me, I mean theoretically I should be working right now and not reading about MTF charts!  :P
5D II | 17-40L | 24-105L | 70-200 f4L IS | 135L | SY 14mm f/2.8 | Sigma 50 f/1.4

EOS M | 22 f/2 | 11-22 IS

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 13543
    • View Profile
Re: Canon may be expensive but...
« Reply #50 on: February 01, 2013, 11:03:28 AM »
I felt cheated on learning that Canon publishes hypothetical MFT charts and does not actually test a sample lens.

Maybe.  But practically, it probably doesn't matter.  How are you going to use those MTF charts? 

  • If it's to compare one Canon lens to another, e.g. is the 135L sharper than the 28-135?, then the theoretical charts are fine.
  • If it's to compare a Canon lens to another manufacturer's lens, e.g. is the Canon 24-105L sharper than the Nikon 24-120?, they aren't valid, but that's ok because you can't easily Canon/Nikon lenses on the other manufacturer's bodies, so directly comparing those lenses via MTF curves isn't terribly useful.
  • If it's to compare your purchased copy of the lens to those MTF curves, unless you have the >$100K equipment to test bare lens resolution, neither theoretical nor empirical MTF data would be useful to you.

In fact, Canon quite likely does test some lenses empirically, as part of the QC process for setting up the manufacturing lines - they just don't publish those data.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 13543
    • View Profile
Re: Canon may be expensive but...
« Reply #51 on: February 01, 2013, 12:16:30 PM »
Therefore Photodos MTF tests where so valuable (world largest MTF tests collection) (today the magazine "Photo" in Sweden are alone  using Hasselblad MTF test equipment [/b]) and  to see if  a  lens have for example been improved or not regarding resolution. The Swedish magazine Foto are also testing new lenses from all manufactures and provide the readers with lens  MTF tests


Unfortunately, at least for Photodo, they have MTF measurements for only ~60 of their ~140 listed Canon lenses, and mostly for old ones, just a few of the current lenses have MTF data - e.g., none of the three 24-70mm lenses (but they do have MTF data for the long-discontinued 28-70L), the non-IS 70-200 lenses have data, but none of the three IS versions, etc.  But for the lenses that they happen to have tested, it's great that they have published those data!

As for Foto magazine, I wish I read Swedish.  :)

Roger's data are definitely useful. 

Reikan FoCal is also now publishing some aggregated sharpness and AF consistency data, but I'm still a bit skeptical about that, since the actual testing is in the hands of many users, not one, which has implicatations for standardization and consistency of results.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

canon rumors FORUM

Re: Canon may be expensive but...
« Reply #51 on: February 01, 2013, 12:16:30 PM »

rpt

  • Canon EF 300mm f/2.8L IS II
  • *******
  • Posts: 2118
  • Could not wait for 7D2 so I got the 5D3
    • View Profile
Re: Canon may be expensive but...
« Reply #52 on: February 01, 2013, 12:42:20 PM »
Good thing tomorrow is Saturday ;)

We can type all night...

BTW, does anyone need to go to work?

Rienzphotoz

  • Canon EF 300mm f/2.8L IS II
  • *******
  • Posts: 3320
  • Peace unto all ye Canon, Nikon & Sony shooters
    • View Profile
Re: Canon may be expensive but...
« Reply #53 on: February 01, 2013, 12:53:01 PM »
Their 800mm lens isn't $18000.


http://www.thedigitalpicture.com/News/News-Post.aspx?News=4073

$18000 is not an issue for me ... coz I can't afford that much ;D ... and I also don't need/want an 800 mm lens.
Canon 5DMK3 70D | Nikon D610 | Sony a7 a6000 | RX100M3 | 16-35/2.8LII | 70-200/2.8LISII | 100/2.8LIS | 100-400LIS | 40/2.8 | 50/1.4 | 85/1.8 | 600EX-RTx2 | ST-E3-RT | 24/3.5 T-S | 10-18/4 OSS 16-50 | 24-70/4OSS | 55/1.8 | 55-210 OSS | 70-200/4 OSS | 28-300VR | HVL-F43M | GoPro Black 3+ & DJI Phantom

rpt

  • Canon EF 300mm f/2.8L IS II
  • *******
  • Posts: 2118
  • Could not wait for 7D2 so I got the 5D3
    • View Profile
Re: Canon may be expensive but...
« Reply #54 on: February 01, 2013, 01:10:24 PM »
The new Photodo do not show any longer the old MTF tests manage by Lars Klellberg the founder of Photodo, since years back Photodo owns by a UK  company
Please start your own threads if you want to argue a point that does not directly apply to the threads you usually write into. Please do not hijack threads of others. Please stay on topic on other's threads. It is simple politeness. We are a community. Please help build it. We ALL share this great forum. Contribute. Life is about sharing. Not about who won an argument! >:(

J.R.

  • 1D X
  • *******
  • Posts: 1461
  • A Speedlight Junkie!
    • View Profile
Re: Canon may be expensive but...
« Reply #55 on: February 01, 2013, 01:30:51 PM »
The new Photodo do not show any longer the old MTF tests manage by Lars Klellberg the founder of Photodo, since years back Photodo owns by a UK  company
Please start your own threads if you want to argue a point that does not directly apply to the threads you usually write into. Please do not hijack threads of others. Please stay on topic on other's threads. It is simple politeness. We are a community. Please help build it. We ALL share this great forum. Contribute. Life is about sharing. Not about who won an argument! >:(

There was a time when you could log into CR looking for friendly banter but this >:( >:( >:( ... . It's almost 1:30 am where I am so I beg you ... I've run outta popcorn ... Goodnight!
Light is language!

J.R.

  • 1D X
  • *******
  • Posts: 1461
  • A Speedlight Junkie!
    • View Profile
Re: Canon may be expensive but...
« Reply #56 on: February 01, 2013, 01:32:31 PM »
Good thing tomorrow is Saturday ;)

We can type all night...

BTW, does anyone need to go to work?


My weekend is screwed ... Gotta catch a flight at 10AM and gotta work almost 6 hours straight when I land! Maybe I can get my hands around the camera on Sunday ... OTW I'll be prowling CR over the weekend ...  Nite!
Light is language!

neuroanatomist

  • CR GEEK
  • ********
  • Posts: 13543
    • View Profile
Re: Canon may be expensive but...
« Reply #57 on: February 01, 2013, 01:40:05 PM »
The new Photodo do not show any longer the old MTF tests manage by Lars Klellberg the founder of Photodo, since years back Photodo owns by a UK  company
Please start your own threads if you want to argue a point that does not directly apply to the threads you usually write into. Please do not hijack threads of others. Please stay on topic on other's threads. It is simple politeness. We are a community. Please help build it. We ALL share this great forum. Contribute. Life is about sharing. Not about who won an argument! >:(

I don't see the discussion of MTF curves as being off topic because:

1) The very impressive theoretical MTF curve of the new Nikon 800/5.6 was brought up on the first page, and

2) Once you've said, "Nikon just announced an 800mm f/5.6 VR lens that costs $18,000," really, how much more can you say that's actually 'on topic' besides, "Wow" or "Damn, that's expensive!"?!?
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

canon rumors FORUM

Re: Canon may be expensive but...
« Reply #57 on: February 01, 2013, 01:40:05 PM »

J.R.

  • 1D X
  • *******
  • Posts: 1461
  • A Speedlight Junkie!
    • View Profile
Re: Canon may be expensive but...
« Reply #58 on: February 01, 2013, 01:58:00 PM »
The new Photodo do not show any longer the old MTF tests manage by Lars Klellberg the founder of Photodo, since years back Photodo owns by a UK  company
Please start your own threads if you want to argue a point that does not directly apply to the threads you usually write into. Please do not hijack threads of others. Please stay on topic on other's threads. It is simple politeness. We are a community. Please help build it. We ALL share this great forum. Contribute. Life is about sharing. Not about who won an argument! >:(

I don't see the discussion of MTF curves as being off topic because:

1) The very impressive theoretical MTF curve of the new Nikon 800/5.6 was brought up on the first page, and

2) Once you've said, "Nikon just announced an 800mm f/5.6 VR lens that costs $18,000," really, how much more can you say that's actually 'on topic' besides, "Wow" or "Damn, that's expensive!"?!?

I understand. But this doesn't have to go so downhill at such a fast pace again...

As usual, you appear to be right whatever Mikael would like us to believe so ... What's the point?
Light is language!

yogi

  • Guest
Re: Canon may be expensive but...
« Reply #59 on: February 01, 2013, 08:50:44 PM »
J.R, you made me get a strong craving for popcorn last night and i didnt have any! Darn. Will have to get some so the next time i will be prepared.
Has anybody mentioned IQ yet?  I would bet that Neuro's IQ is higher than Risedal's. Start a vote thread anyone? Maybe we should ask Marylyn.

canon rumors FORUM

Re: Canon may be expensive but...
« Reply #59 on: February 01, 2013, 08:50:44 PM »