There are some weird people out there. Great video, by the way. I learned a lot about a topic I had very little knowledge of. The big shame though is that the methodological differences make it harder to compare lenses from different manufacturers. That's why I like DxOMark, despite its flaws. What's your take on their lens benchmarks? I remember you had some issues with their sensor tests, but I'm not sure about the lens tests. Any idea if we can trust their sharpness measures?
Never mind. I see someone had already asked the same question and you already answered. thanks Edit: At first glance, Cicala's blog looks high quality, but it's no DxOMark replacement. The number of lenses tested is relatively puny, and it only seems to cover sharpness. And maybe I missed it, but there's no easy way to compare two lenses, or rank lenses. I get your point about DxOMark's lack of transparency about their methodology, but I think there's nothing out there that's quite like what they offer. They're the only thing we have.
Ill have some thoughts on this coming soon. In the end, any camera tester that publishes data and scores without methodologies should raise red flags all over the place, especially if they are claiming a scientific position. DXO's marks very well could be legitimate, but until they publish how they are coming up with these numbers, it they cannot be taken seriously. The strange thing is, even their Perceived Megapixel scores rely on MTF charts, which makes it more strange, because it appears they are applying a penalty to the entire sensor, but even that cannot be confirmed or denied.
I would encourage you to look deeper there. 107 pages, with 5 articles per page, all high quality stuff and when he does lens tests, you know where all the data is coming from. best wishes
It does look good. However, what I am missing is something like this: www.dxomark.com/best-lenses-under-13000-dollars That's a list of well over 500 lenses, all easily compared and ranked in one place by sharpness, distortion, vignetting, transmission, and chromatic aberration. And then you can click on each one to get a ton more detail. That's pretty neat, and I don't think you can find that anywhere else.
Michael, this was really helpful. I've been taking pictures for 50 years now and never knew how to read an MTF chart. Most of the explanations I've read have just made my eyes glaze over. Thanks!
6:03 The explanation of contrast and resolution are totally misconception. Read the subscription, The red lines (Solid and dash) are S10 and M10 (NOT contrast), the blue lines are S30 and M30 (NOT resolution). Some of the MTF charts go further, can have S50/M50, and S70/M70 etc. You can also put in S40/M40 if you want to.
Mike - I always thought it was MODULATION transfer function - not modular. I used to work in a camera facility (on the electrical & microwave side) - the "LENS GURU" wouldn't stop talking about the MTF for a given lens. "modulation" has a different meaning to microwave engineers than optical engineers. Your video is still great - I just wish more manufacturers would publish their MTF charts online, especially for some older glass. I've really enjoyed your videos! Keep up the good work!
Despite it is 3 years old, I have to correct your explanation at time 5:54 The lines does not represent contrast and resolution. The lines only represent contrast. The resolution is given in number of lines per mm. The upper line in the example show the resolution at 10 lines per mm. The lower line show the resolution at 30 lines per mm. Maybe that is why some people give you a thumbs down.
@@eksund1900 oh great news! When will you be publishing your video? PS- I should probably warn you I carefully researched this topic before making it, but I won’t spoil it for you.
Very clear explanation. But what is the practical utility of keeping MTF charts in mind if you can’t compare them between different lens companies. I understand you could compare lens from that same manufacturer but is that it?
sir kindly can you share me your facebook id? i work in a mobile assembly factory.there are lot of camera testing tools with MTF. i have more query which i want to share with you personally. pls sir .....
Well presented. I was really sure those vertical and horizontal window blinds in the background were going to become part of a demonstration. Guess not.
What is your opinion on arguments like this posted at www.kenrockwell.com/tech/mtf.htm quote: Reality Check: 1.) Most MTF curves are merely plotted from calculations, meaning they are only true in the manufacturer's wildest dreams. No real lens from the assembly line would ever be able to equal this, since tolerances are never perfect. Real MTF curves are measured from real samples of lens. 2.) MTF curves look best when plotted for only one color, and worse if plotted for several colors integrated at once. They look worst for white light, which is how we actually use them photographically. A monochromatic (single-color) MTF curve conveniently ignores chromatic aberrations! An MTF curve is meaningless unless the curve specifies if the results are measured or calculated (dreamed), for what color, colors or bands of light and with what mixing percentages, and at what focused distances (reproduction ratios). Lens makers rarely specify which conditions they are using, meaning the graphs are meaningless, and are especially dangerous to use to compare between different brands. Of course they also need to specify aperture and focal length settings, which is all they usually do. Thank you!
Very helpful, thank you. It's surprising that Nikon and others' MTF test charts show lens results only when wide open. Their marketing departments can't be too keen on that I'm betting! A couple of typos/mispronunciations: MTF = "Modulation" (not "modular") Transfer Function; "tangential" (not "tangental").
Nice video that explains what MTF is. Thanks a lot. I hope you could show a real test setup that's used for MTF measurement. do you test the lens only with external camera, or you need to put the lens on an imaging sensor to actually take a image?
Mike, Thanks for great explanation. I did not hear details about the chart for a prime lease where we see four sets of lines, two with darker colors (e.g. blue and black) and two with lighter shades of those colors. I believe the darker color pairs belong to S and M lines for focal length of Prime lens since it is fixed, but what does lighter color pairs represent !
Thanks. An excellent down to Earth explanation and with caveats on why there will be variations from different manufacturers and even between same models of the same lens. Well done.
Great explanation. Took me a bit to catch up on the nomenclature but well worth the effort. Thank you ! BTW, the thumbs down guy probably suffers from Dunning-Kruger effect and wears a red hat.
Lens designer here. I’m three minutes in, and found several mistakes. 1) It’s modulation not modular. 2) The x axis is on the sensor not the lens - 0 is the center of the sensor and 21 is the corner of the sensor. 3) MTF is NOT relative to the center of the lens, its relative to a perfect lens. 4) MTF has nothing to do with signal to noise. Sensor noise is a COMPLETELY different subject. 5) The fact that prime lenses are sharper has nothing to do with the number of elements. Zoom lenses are fuzzier because the focal length is adjustable, which is very hard - it takes more elements I could design a zoom that was sharper but it would need so many elements that it would be so expensive nobody would buy it.
Thank you. 2- if zero is the center of the sensor then would it not also be the center of the lens? (Seems so) 3- It’s not “relevant” it’s “relative” (isn’t it fun to be nit picky? See your point 1) 4- where did I say that 5- contradictory - either primes are sharper than zooms or they are not.
@@MMaven Fixed my typo. 😁 Yes primes are sharper (in general, but not always) but you said it’s because they have fewer elements…. That’s not right. It’s a common misconception…. An expensive zoom might be sharper than a cheap prime. And the zoom would have way more elements.
@@MMaven You can compare MTF curves from different manufacturers if F/# and cycles/mm are the same. And if they’re both diffraction MTF or both geometric (but those are only really different at very slow F/#). You said the sensor matters - no, the MTF is just the lens by itself. As far as I know, no manufacturers publish MTF of the whole camera plus lens system. That would be interesting but I’ve never seen it.
@@luomoalto Your argument is: “Zooms aren’t as sharp as Primes (generally speaking) because it would be too expensive to do it right.” While theoretically, yes, you could make a certain zoom more sharp than a poorly/carelessly made prime, that’s generally not the case in actual practice. Even with your argument, my position remains the same “Primes are (generally) sharper than zooms, because they are easier to make sharp, because there are fewer lens elements.” It’s literally the same point. Also, introductions new glass creates diffraction issues with every element, which would have to be corrected, and makes it increasingly complex, so while in theory/extreme cases yes, some zooms can be made to be sharper than primes, again that’s generally not the case, again because more elements are involved. It seems like it’s literally talking around in circles. However even in these extreme cases we would still have the problem of differences in T stops. You made no mention of this. But what do I know I am not a lens designer
Another great video in your series... your teaching style (and how you break things down) is so perfect for subjects like this. I am still not sure exactly how I would use these charts in real world decision making about lenses- there are so many other ways to collect information and reviews about overall lens usefulness - but I did always kind of hate that there were charts out there that I knew little about reading and had always just ignored. You definitely solved that! Still, focus speed and accuracy and build quality (weather) are probably going to always rule for me over the minor variances shown in these charts when it comes to actually picking a lens for capturing fleeting moments.
Wilson Newman all excellent points and I agree that your lens characteristics and strengths depend on what your needs are. There are a number of questions and issues (and misperceptions) that MTF charts answer if one is familiar with them, something I'll be addressing in a future video.
Work Pics I do not endorse, recommend or support DXO marks tests because they do not publish their full methodology. I'll be addressing this specifically in an upcoming video. Just need to edit it.
I guess the question though is where is the most reliable data that does publish not only their full process but also provide a good enough cross section (including multiple copies of the same lens). Tough call... the process is very time consuming and possibly of yet unreliable benefit. Looking forward to your edits... Thanks.
Its a very important question and as of this moment, I believe it is Roger Cicala's (lensrentals.com) blog: www.lensrentals.com/blog/ He publishes his full methodology and usually tests 10 lenses, with 4 rotations each and then averages. Its outstanding work.
I am looking forward to your video on DXO mark, partly because someone criticised me on my own channel for not conforming my lens test scores with ones that agree with those of DXO mark, something that i deny is necessary, but at the same time i don't have much of a ready made answer for the devout DXO mark believers. Do you have any quick advice before your video comes out?
And yet if they did care about MTF performance, they’d more make better informed lens purchasing decisions. That doesn’t mean they’d always buy a lens with superior MTF charts, they’d at least better understand the limitations of the “vintage look” lenses they’re interested in.
I see a lot of ignorant people down below, who are unfortunately, badly informed. Before you buy a lens - new or use - learn how to read the MTF chart. To me, it's like buying a car without checking Consumer Reports, MTF charts are a guide on how well your lens performs, buy wisely. Thank you for this video.
LOL there is always that one guy out there who always gives me a thumbs down as soon as I post. :)
There are some weird people out there. Great video, by the way. I learned a lot about a topic I had very little knowledge of.
The big shame though is that the methodological differences make it harder to compare lenses from different manufacturers.
That's why I like DxOMark, despite its flaws. What's your take on their lens benchmarks? I remember you had some issues with their sensor tests, but I'm not sure about the lens tests. Any idea if we can trust their sharpness measures?
Never mind. I see someone had already asked the same question and you already answered. thanks
Edit: At first glance, Cicala's blog looks high quality, but it's no DxOMark replacement. The number of lenses tested is relatively puny, and it only seems to cover sharpness. And maybe I missed it, but there's no easy way to compare two lenses, or rank lenses.
I get your point about DxOMark's lack of transparency about their methodology, but I think there's nothing out there that's quite like what they offer. They're the only thing we have.
Ill have some thoughts on this coming soon. In the end, any camera tester that publishes data and scores without methodologies should raise red flags all over the place, especially if they are claiming a scientific position.
DXO's marks very well could be legitimate, but until they publish how they are coming up with these numbers, it they cannot be taken seriously. The strange thing is, even their Perceived Megapixel scores rely on MTF charts, which makes it more strange, because it appears they are applying a penalty to the entire sensor, but even that cannot be confirmed or denied.
I would encourage you to look deeper there. 107 pages, with 5 articles per page, all high quality stuff and when he does lens tests, you know where all the data is coming from. best wishes
It does look good. However, what I am missing is something like this:
www.dxomark.com/best-lenses-under-13000-dollars
That's a list of well over 500 lenses, all easily compared and ranked in one place by sharpness, distortion, vignetting, transmission, and chromatic aberration. And then you can click on each one to get a ton more detail. That's pretty neat, and I don't think you can find that anywhere else.
Michael, this was really helpful. I've been taking pictures for 50 years now and never knew how to read an MTF chart. Most of the explanations I've read have just made my eyes glaze over. Thanks!
I have never heard about MTF charts, but you explained it really well!
6:03 The explanation of contrast and resolution are totally misconception. Read the subscription, The red lines (Solid and dash) are S10 and M10 (NOT contrast), the blue lines are S30 and M30 (NOT resolution). Some of the MTF charts go further, can have S50/M50, and S70/M70 etc. You can also put in S40/M40 if you want to.
Mike - I always thought it was MODULATION transfer function - not modular. I used to work in a camera facility (on the electrical & microwave side) - the "LENS GURU" wouldn't stop talking about the MTF for a given lens. "modulation" has a different meaning to microwave engineers than optical engineers. Your video is still great - I just wish more manufacturers would publish their MTF charts online, especially for some older glass. I've really enjoyed your videos! Keep up the good work!
ef is correct
Absolutely a clear, concise and informative video! Thank you so much!
Despite it is 3 years old, I have to correct your explanation at time 5:54 The lines does not represent contrast and resolution. The lines only represent contrast. The resolution is given in number of lines per mm. The upper line in the example show the resolution at 10 lines per mm. The lower line show the resolution at 30 lines per mm. Maybe that is why some people give you a thumbs down.
Are you certain? Please make a video about it if you are.
@@MMaven Yes Michael I am certain. Even your picture in your video have the two lines resolution noted. Red line 10 mm, blue line 30mm
@@eksund1900 oh great news! When will you be publishing your video?
PS- I should probably warn you I carefully researched this topic before making it, but I won’t spoil it for you.
Very clear explanation. But what is the practical utility of keeping MTF charts in mind if you can’t compare them between different lens companies. I understand you could compare lens from that same manufacturer but is that it?
sir kindly can you share me your facebook id?
i work in a mobile assembly factory.there are lot of camera testing tools with MTF.
i have more query which i want to share with you personally.
pls sir .....
Well presented. I was really sure those vertical and horizontal window blinds in the background were going to become part of a demonstration. Guess not.
Thank you very much, very well explained.
What is your opinion on arguments like this posted at www.kenrockwell.com/tech/mtf.htm
quote: Reality Check:
1.) Most MTF curves are merely plotted from calculations, meaning they are only true in the manufacturer's wildest dreams. No real lens from the assembly line would ever be able to equal this, since tolerances are never perfect. Real MTF curves are measured from real samples of lens.
2.) MTF curves look best when plotted for only one color, and worse if plotted for several colors integrated at once. They look worst for white light, which is how we actually use them photographically. A monochromatic (single-color) MTF curve conveniently ignores chromatic aberrations!
An MTF curve is meaningless unless the curve specifies if the results are measured or calculated (dreamed), for what color, colors or bands of light and with what mixing percentages, and at what focused distances (reproduction ratios).
Lens makers rarely specify which conditions they are using, meaning the graphs are meaningless, and are especially dangerous to use to compare between different brands. Of course they also need to specify aperture and focal length settings, which is all they usually do.
Thank you!
Very helpful, thank you. It's surprising that Nikon and others' MTF test charts show lens results only when wide open. Their marketing departments can't be too keen on that I'm betting! A couple of typos/mispronunciations: MTF = "Modulation" (not "modular") Transfer Function; "tangential" (not "tangental").
Great explanation. Maybe they should call the YMMV charts.
Modulation Transfer Function
I just turned the numbers to 5-6-7-8!
Really appreciate the video
Thank you very very much, really helpful. Probably the best explanation of MTF charts I've seen so far.
Enjoying your lessons. MTF is Modulation Transfer Function, no Modular Transfer Function
Perfect! Thank you! Excellent presentation!!!!!
Nice video that explains what MTF is. Thanks a lot. I hope you could show a real test setup that's used for MTF measurement. do you test the lens only with external camera, or you need to put the lens on an imaging sensor to actually take a image?
Mike, Thanks for great explanation. I did not hear details about the chart for a prime lease where we see four sets of lines, two with darker colors (e.g. blue and black) and two with lighter shades of those colors. I believe the darker color pairs belong to S and M lines for focal length of Prime lens since it is fixed, but what does lighter color pairs represent !
Thanks Michael, I just wonder why in most of those MTF charts they do not show results at different f-stops, rather showing it in general.
You sir just earned a subscriber
Thanks! I didn't know how the hell to read these until this video. Thank you for this valuable information.
We’ll done ! Thank you.
Best explanation.
Very good detail
As a compensation I just gave you a "Thumb up" for this nice explanation ! - Thanks !
Thanks. An excellent down to Earth explanation and with caveats on why there will be variations from different manufacturers and even between same models of the same lens. Well done.
Thank you .
many thanks for this explanation. do you have any recommendations or suggestions for how i can do MTF tests on my own lenses?
Excellent explanation, I definately learnt something new here as MTF charts have always baffled me 😃
Mike, thank you for the brief and clear summary on how to read MTF charts
Great explanation. Took me a bit to catch up on the nomenclature but well worth the effort. Thank you ! BTW, the thumbs down guy probably suffers from Dunning-Kruger effect and wears a red hat.
Thank you Ian. Its ok, I get this a lot on youtube LOL
This helps a lot. You explained it very well and I appreciate your explanation. Well done and thank you.
Really want get you review on Canon 6D mark2, is it worth to buy it? Because I see so many people complain about no 4K video....
+Jie Zeng I'm really not that interested in it
Fantastic video. You communicate your subject matter really well. Thank you
Gavin Steiner thank you Gavin
Wow excelent , very clear , thanks a Lot
Really useful and easy to follow - thank you!
very clear explanation,,,thx
Lens designer here. I’m three minutes in, and found several mistakes. 1) It’s modulation not modular. 2) The x axis is on the sensor not the lens - 0 is the center of the sensor and 21 is the corner of the sensor. 3) MTF is NOT relative to the center of the lens, its relative to a perfect lens. 4) MTF has nothing to do with signal to noise. Sensor noise is a COMPLETELY different subject. 5) The fact that prime lenses are sharper has nothing to do with the number of elements. Zoom lenses are fuzzier because the focal length is adjustable, which is very hard - it takes more elements
I could design a zoom that was sharper but it would need so many elements that it would be so expensive nobody would buy it.
Thank you. 2- if zero is the center of the sensor then would it not also be the center of the lens? (Seems so) 3- It’s not “relevant” it’s “relative” (isn’t it fun to be nit picky? See your point 1) 4- where did I say that 5- contradictory - either primes are sharper than zooms or they are not.
@@MMaven Fixed my typo. 😁
Yes primes are sharper (in general, but not always) but you said it’s because they have fewer elements…. That’s not right. It’s a common misconception….
An expensive zoom might be sharper than a cheap prime. And the zoom would have way more elements.
@@MMaven You got radial and tangential lines right, which a lot of people screw up. 😁
@@MMaven You can compare MTF curves from different manufacturers if F/# and cycles/mm are the same. And if they’re both diffraction MTF or both geometric (but those are only really different at very slow F/#). You said the sensor matters - no, the MTF is just the lens by itself. As far as I know, no manufacturers publish MTF of the whole camera plus lens system. That would be interesting but I’ve never seen it.
@@luomoalto Your argument is: “Zooms aren’t as sharp as Primes (generally speaking) because it would be too expensive to do it right.” While theoretically, yes, you could make a certain zoom more sharp than a poorly/carelessly made prime, that’s generally not the case in actual practice.
Even with your argument, my position remains the same “Primes are (generally) sharper than zooms, because they are easier to make sharp, because there are fewer lens elements.” It’s literally the same point.
Also, introductions new glass creates diffraction issues with every element, which would have to be corrected, and makes it increasingly complex, so while in theory/extreme cases yes, some zooms can be made to be sharper than primes, again that’s generally not the case, again because more elements are involved. It seems like it’s literally talking around in circles.
However even in these extreme cases we would still have the problem of differences in T stops. You made no mention of this.
But what do I know I am not a lens designer
Great Video!!!!!!!!!!
Great explanation. Well done.
Another great video in your series... your teaching style (and how you break things down) is so perfect for subjects like this.
I am still not sure exactly how I would use these charts in real world decision making about lenses- there are so many other ways to collect information and reviews about overall lens usefulness - but I did always kind of hate that there were charts out there that I knew little about reading and had always just ignored. You definitely solved that!
Still, focus speed and accuracy and build quality (weather) are probably going to always rule for me over the minor variances shown in these charts when it comes to actually picking a lens for capturing fleeting moments.
Wilson Newman all excellent points and I agree that your lens characteristics and strengths depend on what your needs are. There are a number of questions and issues (and misperceptions) that MTF charts answer if one is familiar with them, something I'll be addressing in a future video.
Nice... I look forward to it. This is a way for me to fill in on a topic that I have known little about.
Thank you for this video
Very useful to have a clear explanation and finally I understand this, thank you
Ken McDougall thank you Ken!
good work done! wanna more pls
The best explaination i have ever seen
Thanks for the knowledge transfer. I always wondered how to read those things.
+Cactus Tweeter lol...I see what you did there - u are welcome
What are your thoughts on how DXO do this to ensure consistent measure processes?
Work Pics I do not endorse, recommend or support DXO marks tests because they do not publish their full methodology. I'll be addressing this specifically in an upcoming video. Just need to edit it.
I guess the question though is where is the most reliable data that does publish not only their full process but also provide a good enough cross section (including multiple copies of the same lens). Tough call... the process is very time consuming and possibly of yet unreliable benefit. Looking forward to your edits... Thanks.
Its a very important question and as of this moment, I believe it is Roger Cicala's (lensrentals.com) blog: www.lensrentals.com/blog/
He publishes his full methodology and usually tests 10 lenses, with 4 rotations each and then averages. Its outstanding work.
I am looking forward to your video on DXO mark, partly because someone criticised me on my own channel for not conforming my lens test scores with ones that agree with those of DXO mark, something that i deny is necessary, but at the same time i don't have much of a ready made answer for the devout DXO mark believers. Do you have any quick advice before your video comes out?
99.9 % don't care about this, sorry🙈
lol unfortunately this is also true! I hope to change that a little
I would go farther and say this is irrelevant for 99.9% of photographers.
You are reaching for the stars Alan! lol
And yet if they did care about MTF performance, they’d more make better informed lens purchasing decisions. That doesn’t mean they’d always buy a lens with superior MTF charts, they’d at least better understand the limitations of the “vintage look” lenses they’re interested in.
I see a lot of ignorant people down below, who are unfortunately, badly informed. Before you buy a lens - new or use - learn how to read the MTF chart. To me, it's like buying a car without checking Consumer Reports, MTF charts are a guide on how well your lens performs, buy wisely. Thank you for this video.