Pages: [1] 2 3   Go Down

Author Topic: LenScore / SenScore (Lens and Sensor Reviews)  (Read 13336 times)

John Koerner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 866
  • "Fortune favors the bold." Virgil
    • John Koerner Photography
LenScore / SenScore (Lens and Sensor Reviews)
« on: December 18, 2015, 10:55:07 pm »

Does anyone know who runs this site or what criteria they use?

These twin sites seem to be akin to the DxO Mark, but different.

Bernard first brought these sites to my attention, about a year ago, and I find them a lot more user-friendly than DxO Mark. (They also seem to be more up-to-date too.)

LenScore / SenScore

Their rating system seems to make a whole lot more sense too.

For example, according to the DxO Mark, in evaluating the DSLR cameras themselves, the Nikon 810 has the #2 camera sensor, with the new Sony A7 RII being #1, while the Nikon D4s occupies a lowly #17 spot. This doesn't seem to make any sense to me, as the D4s is so much more expensive than the D810. (And, surely, Nikon knows the value of its own cameras.)

By contrast, SenScore rates the Nikon D4s the #1 camera across the board, boasting more dynamic range than the D810 (1321 to 928), more color range (1369 to 1073), more tonal range (1139 to 10035), it just does not have the resolving power of the D810 (563 to 1051), as the D4s only has half the megapixels of the D810. I can follow this logic, so the SenScore website makes more sense to me.

The SenScore database compiles its sensor ratings based on Noise, Dynamic Range, Color Range, Tonal Range, and Resolving Power, as well as the combined aggregate (the Final SenScore), all of which are more relevant to me than what the DxO Mark tries to put out. (You can also click on each of these criterion and get a hierarchical order of magnitude based on which of each individual attribute is most important to you.)

For example, if you click on Resolving Power, up top, you see the results in this criteria alone, and it is a landslide for the new Canon 5DS R (1357 compared to the D810's 1051).

By contrast, if you click on Dynamic Range, up top, the new Canon 5DS R has a paltry 850 score compared to the D810's 1028.
(Interestingly, however, the D810 is not at the top of their list for DR, the D4s is.

Anyway, I am just curious as to the credentials of who puts the information out and what criteria they use ... does anyone know?

Jack
Logged

NancyP

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2513
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #1 on: December 18, 2015, 11:11:14 pm »

Actually it makes a lot of sense that a large-pixel sensor could out-perform a small-pixel sensor, other things being equal. More photons per pixel = more head-room in your computations to deal with noise, color accuracy, etc, OTBE.
Logged

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #2 on: December 18, 2015, 11:11:21 pm »

Does anyone know who runs this site or what criteria they use?
These twin sites seem to be akin to the DxO Mark, but different.

you might also what to see http://photonstophotos.net, for sensors only... and unlike sensorscore you can talk with the author how things are measured (if site is not enough) = http://www.dpreview.com/members/9263714680/forums/posts and see what is the deal per ISO, not some god knows how calculated "scores"...
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13880
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #3 on: December 19, 2015, 12:28:36 am »

If you just look at the single numbers of DxO they will give the best value among ISO, which is why the D810 has by far the best DR of any camera. If you bother to look at the curves you have much more useful data.

On the other hand, senscore computes an average accross ISO, which is why they rate the D4s as having the best DR.

There is little reason to only consider the summary rating, but if you do then DxO can be considered to be a landscape rating why senscore is a more generic one.

Cheers,
Bernard

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11310
    • Echophoto
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #4 on: December 19, 2015, 06:17:42 am »

Hi,

There is a great difference between the two sites in that DxO-mark discloses all (or almost all) measured data. Both site also gives figures of merit but with some information how those values are calculated from DxO and none at all from SenScore/LenScore.

DxO has quite a lot of credibility on the sensor side, as their measurements show good agreement with other tests at least regarding DR.

With sensors, it matters a lot if you look at base ISO or at high ISO numbers. For a tripod shooter, like me, high ISO is simply irrelevant. For one shooting sports low ISO probably quite irrelevant.

With lenses it may be a bit more complex, as DxO does not exactly say what their evaluations are based on. Still they present a lot of data across the field, so lenses can be compared.

Regarding lens tests, it matters a lot if we are shooting at large apertures, medium apertures or small apertures. Building a highly corrected lens for f/1.4 is no easy task. But, a relatively simple lens can be very good at medium apertures. So, if you don't know how performance at different apertures is weighted in the figures of merit are actually of very little use.

In addition, there is some sample variation. Some would argue that a lens should not be judged based on a single sample.

I would suggest that both sites do analysis based on adequate measurements, but the way DxO presents their data it is much more useful.

This site offers a lot of good info: http://www.the-digital-picture.com

The site also show comparisons shots: http://www.the-digital-picture.com and also MTF data: http://www.the-digital-picture.com/Reviews/MTF.aspx

The MTF data here is coming from the optical lab at Lensrentals. (*)

This site is testing lenses at the Hasselblad factory working with the engineer in charge of lens developmemt and testing at Hasselblad. Their data is comparabe with data coming from Zeiss or Leica: http://www.lensfreaks.com

Best regards
Erik

(*) For completeness, I posted some questions regarding MTF plots on some lenses to Roger Cicala, the owner of Lensrentals/Olaf Testing. From the answer I got they are looking into some issues. Another point may be that it is quite possible that cover glass effects need to be taken into account. Conventional wisdom used to be that cover glass thickness doesn't really matter for DSLR lenses, but it may play a greater role than expected.


Does anyone know who runs this site or what criteria they use?

These twin sites seem to be akin to the DxO Mark, but different.

Bernard first brought these sites to my attention, about a year ago, and I find them a lot more user-friendly than DxO Mark. (They also seem to be more up-to-date too.)

LenScore / SenScore

Their rating system seems to make a whole lot more sense too.

For example, according to the DxO Mark, in evaluating the DSLR cameras themselves, the Nikon 810 has the #2 camera sensor, with the new Sony A7 RII being #1, while the Nikon D4s occupies a lowly #17 spot. This doesn't seem to make any sense to me, as the D4s is so much more expensive than the D810. (And, surely, Nikon knows the value of its own cameras.)

By contrast, SenScore rates the Nikon D4s the #1 camera across the board, boasting more dynamic range than the D810 (1321 to 928), more color range (1369 to 1073), more tonal range (1139 to 10035), it just does not have the resolving power of the D810 (563 to 1051), as the D4s only has half the megapixels of the D810. I can follow this logic, so the SenScore website makes more sense to me.

The SenScore database compiles its sensor ratings based on Noise, Dynamic Range, Color Range, Tonal Range, and Resolving Power, as well as the combined aggregate (the Final SenScore), all of which are more relevant to me than what the DxO Mark tries to put out. (You can also click on each of these criterion and get a hierarchical order of magnitude based on which of each individual attribute is most important to you.)

For example, if you click on Resolving Power, up top, you see the results in this criteria alone, and it is a landslide for the new Canon 5DS R (1357 compared to the D810's 1051).

By contrast, if you click on Dynamic Range, up top, the new Canon 5DS R has a paltry 850 score compared to the D810's 1028.
(Interestingly, however, the D810 is not at the top of their list for DR, the D4s is.

Anyway, I am just curious as to the credentials of who puts the information out and what criteria they use ... does anyone know?

Jack
Logged
Erik Kaffehr
 

kers

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3876
    • Pieter Kers
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #5 on: December 19, 2015, 06:55:26 am »

If you buy a high pixel camera like the D810- 7RII or the Canon 5dsr you are surely into low iso value photography.
So the low iso quality counts. But also the lens quality will count.
I have not seen too many wideangle lenses that do 36MP in the corners, but a lot do 50 MP in the centre.
The Canon has less Dynamic range than the other two.
So it depends on your type of photography what camera is best for you.

To average out the sensor as Senscore does is not very interesting.
for instance they say:
"Why is the score of the D3x so low?
While being able to produce outstanding images at base ISO, the D3x's image quality rapidly diminishes with higher ISO settings."

True!  I had one and used it beneath 400 asa. Above 400 it became noisy and lost dynamic range.
The 100 asa setting however was best in class for some years and for my photography that was important.

I just have worked with the Nikon 4ds and i was not impressed. I liked the shots with my D810 better below 3200 asa.
Above 3200 asa there was a clear advantage. It is a speed king for getting the shot in every circumstance.

Logged
Pieter Kers
www.beeld.nu/la

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11310
    • Echophoto
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #6 on: December 19, 2015, 11:01:09 am »

Hi,

Some interesting reading from the SenScore FAQ: http://www.senscore.org


"The D810 offers far superior resolving power and is able to provide much better image quality than the D4s under favorable conditions. However, dynamic range, tonal range and color range drop off significantly with higher ISO settings while the D4s holds up astonishingly well. The break-even point is below 400 ISO, at 400 ISO the D4s is already marginally ahead. The D810 retains superior resolving power throughout the measured ISO range, but resolving power is only 20 percent of the overall score, which is why the D4s comes out atop. Thus, if a resolution of significantly more than 16MP is required, the D810 may very well be the better camera for the job even at high ISO, but if 16MP is enough, then the D4s will provide higher image quality. Many people erroneously believe that downsampling will make up for a lack of dynamic range, tonal range and color range, but that's not true. However, downsampling done right can hugely improve the noise characteristics of an image, and in case of the D4s vs the D810, this pretty much eliminates the D4s' advantage in noise performance."

Best regards
Erik

Does anyone know who runs this site or what criteria they use?

These twin sites seem to be akin to the DxO Mark, but different.

Bernard first brought these sites to my attention, about a year ago, and I find them a lot more user-friendly than DxO Mark. (They also seem to be more up-to-date too.)

LenScore / SenScore

Their rating system seems to make a whole lot more sense too.

For example, according to the DxO Mark, in evaluating the DSLR cameras themselves, the Nikon 810 has the #2 camera sensor, with the new Sony A7 RII being #1, while the Nikon D4s occupies a lowly #17 spot. This doesn't seem to make any sense to me, as the D4s is so much more expensive than the D810. (And, surely, Nikon knows the value of its own cameras.)

By contrast, SenScore rates the Nikon D4s the #1 camera across the board, boasting more dynamic range than the D810 (1321 to 928), more color range (1369 to 1073), more tonal range (1139 to 10035), it just does not have the resolving power of the D810 (563 to 1051), as the D4s only has half the megapixels of the D810. I can follow this logic, so the SenScore website makes more sense to me.

The SenScore database compiles its sensor ratings based on Noise, Dynamic Range, Color Range, Tonal Range, and Resolving Power, as well as the combined aggregate (the Final SenScore), all of which are more relevant to me than what the DxO Mark tries to put out. (You can also click on each of these criterion and get a hierarchical order of magnitude based on which of each individual attribute is most important to you.)

For example, if you click on Resolving Power, up top, you see the results in this criteria alone, and it is a landslide for the new Canon 5DS R (1357 compared to the D810's 1051).

By contrast, if you click on Dynamic Range, up top, the new Canon 5DS R has a paltry 850 score compared to the D810's 1028.
(Interestingly, however, the D810 is not at the top of their list for DR, the D4s is.

Anyway, I am just curious as to the credentials of who puts the information out and what criteria they use ... does anyone know?

Jack
Logged
Erik Kaffehr
 

John Koerner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 866
  • "Fortune favors the bold." Virgil
    • John Koerner Photography
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #7 on: December 19, 2015, 12:51:59 pm »

Thanks to all for the responses (and thanks especially to Eric and Bernard for providing some additional insight as to how Len/SenScore work).

I agree the "sum total" scores of Sen/LenScore are not the most valuable scores to read, but their easily-switched Individual Criteria Columns are, which become especially useful because it is a cinch switch to to the categories that matter (to me/you).

As far as I am concerned, the DxO mark setup is less intuitive and less complete.
It seems to be setup by people who measure a very limited range statistics, and who may not actually be using the cameras in the real world.

For example, Base ISO measurements have almost zero relevance to a nature photographer (bird photographer/macro shooter).
I honestly don't know of any bird photographer or nature photographer who uses ISO 80 or 100 in their image-making. It simply almost never happens.

I personally enjoy macro photography, and I try my best to stick to natural light, so it is difficult to shoot below ISO 320.
(When you get extremely close to a tiny object, even if it is completely motionless, your available light is drastically-reduced).
Typically, there is always the possibility for movement, so I can only drop my shutter speed so low before I run the risk of the subject (usually an insect or spider) moving, changing positions, or even being affected by wind.

As a macro shooter, my primary hot buttons are resolution and bokeh, trying to get as much detail as possible of the subject, while blurring the background, with color and contrast being important elements also.
(Most subjects don't have a super-wide color gamut, but some do.)

Higher ISO capability is something I need to upgrade to, because I am not enjoying this right now with my 7D (I can't go above 640 without wanting to junk the file, so I am pretty much "stuck" at shooting between ISO 320 and 640 right now).

The scores across the board for my antiquated 7D are accelerating my desire to get rid of it, and finally make an upgrade, but the thought of "switching systems" has not been much of a motivator, as I really like the Canon lenses. (However, the excellence of Sigma has made me abandon 2 Canon lenses lately ...)

When it first came out, the Sony A 7R II originally seemed like the perfect solution (specs-wise), because it has the Exmor sensor and can be compatible with Canon lenses ... however for people who use cameras for nature photography, most of the reports in actual field use have been photographers "loving the images" (in the shots they "do get" with the Sony A7R II) ... but they can't stand to use it (poor ergonomics, etc.).

Right now, I am able to get plenty of decent shots with my 7D, but I am wanting to elevate from "decent" to spectacular, and I am wanting my next camera purchase to last me for a few years (not just one season).
I am not necessarily "brand loyal" (as my favorite macro lens is a Sigma 180), but I neither do I want to jump ship without careful consideration.

I was excited about the Canon 7D II, but (ultimately) while offering great ergonomic features, the 7D II simply ranks too low (across the board) in image quality for me to be satisfied with the purchase.
To me, all the bells and whistles in the world fall short if the image-quality is only so-so.

When I first got the original Canon 7D, it was a class-leading APS-C camera, but it has fallen so far behind now, that I need to pull the trigger and upgrade.
I have been tolerating the so-so image quality of my 7D for a couple of years, and now there is the recent Canon 5DS R to consider.
The DxO specs on the 5DS R (and several reviews from landscape photographers) didn't help my feeling of letdown on the new Canon FF ...
However, after hearing what many macro shooters have had to say about the 5DS R, and in going over the SenScore specs, I am 99% positive I will be buying the 5DS R by springtime.

If the Nikon D810 has been the benchmark (for landscape shooters), for a number of years, then the 5DS R has to be the macro shooter's benchmark.
While the 5DS R may still fall short of the D810 in some important respects, as a landscape camera, to me it surpasses the D810 as a macro photography camera.

Let's face it, resolution is everything in macro photography (that and bokeh).
And, in Resolution, the 5DS R beats the D810 by a wide margin (1357 to 1051) ... not to mention beating my 7D by a country mile (1357 to 475).

I am not worried about the D810 beating the 5DS R in Dynamic Range (1028 to 850), because I will not even bother taking a natural light macro shot unless I have even light.
Therefore, "High Dynamic Range" means nothing to me; it has no relevance to the way I shoot.

However, one thing that does bother me is the fact the D810 is listed as having the better Color Range (1073 in the Nikon to a measly 751 in the Canon). That really does bother me :o

Still, the Color Range rating for the 5DS R (751) is a vast improvement over the rating for my current 7D (105 :-\). Therefore, although the 5DS R rating is low, I am encouraged by the fact I have taken some pretty colorful photos with my 7D ... so I should be exhilarated at what the 5DS R can do, with more than 5x the color range of the 7D. (Also makes me wonder how much of the "theoretical" color potential is actually see-able by the human eye ...)

More than this, as far as Noise Levels go, the 5DS R beats the D810 (1041 to 1010) and in Tonal Range the 5DS R again edges the D810 (1041 to 1035).
Realistically, they're both nearly identical in these respects.

Therefore, for a macro shooter, the Canon 5DS R is actually the better camera than the D810.
While it is lacking in respect to color, I am not sure what the real-world difference is in the color disparity.
The DR makes zero difference, while the Canon surpasses the D810 in every other category, most importantly in Resolution (and widely-so).

My sticking point is the color issue. Color is important, and some insect/spider species have scintillating, metallic colorations.
That said, despite the fact that my own Canon 7D has a lowly Color Rating of 105, I have gotten a gazillion compliments on the "colors" of my images, and have been well-pleased with the colors myself.
So I believe the color range of the 5DS R will be more than enough.

The bottom line is, I have *always* been chagrined by the LACK of total resolution in my images, as there is no way around the fact the 7D is a "soft" camera.
Even while using a tripod, and a remote switch, highly-critical precision-focus of tiny subjects remains a challenge for me.

I am therefore quite sure that elevating from my Camera Resolution Score from 475 (in the 7D) to a class-leading score of 1357 (in the 5DS R) will make my macro photography take a quantum leap forward.

In closing, my apologies for "thinking out loud" here, but I hope it helps some others with their own "chin-rubbing" as to weighing the different criteria as to their next purchase.
In the end, I believe the DxO Mark system of scoring sensors is limited in value, and ultimately very misleading in their conclusions.

I think the LenScore/SenScore site is much more user-friendly, and much more helpful and realistic, in assessing the Lenses and Sensors in a more "real world"-applicable way, so I am glad Bernard shared this resource with me because it is definitely bookmarked.

Jack
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13880
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #8 on: December 19, 2015, 05:57:12 pm »

Indeed, the 5DS is a great camera and should serve you well across the board and especially so for macro work.

Cheers,
Bernard

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #9 on: December 19, 2015, 08:39:22 pm »

For example, Base ISO measurements have almost zero relevance to a nature photographer (bird photographer/macro shooter).
I honestly don't know of any bird photographer or nature photographer who uses ISO 80 or 100 in their image-making. It simply almost never happens.

did you miss the fact that sites like Photonstophotos (or DxoMark) do give you the details for various ISOs (nominal ISO in case of Photonstophotos or Ssat ISO in case of DxOMark) ? so if you are interested in DR or S/N @ high nominal (or Ssat equalized) ISOs then you can simply look at the relevant data (ignoring scores).

Logged

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #10 on: December 19, 2015, 08:49:51 pm »

Hi,

Some interesting reading from the SenScore FAQ: http://www.senscore.org


"The D810 offers far superior resolving power and is able to provide much better image quality than the D4s under favorable conditions. However, dynamic range, tonal range and color range drop off significantly with higher ISO settings while the D4s holds up astonishingly well. The break-even point is below 400 ISO, at 400 ISO the D4s is already marginally ahead. The D810 retains superior resolving power throughout the measured ISO range, but resolving power is only 20 percent of the overall score, which is why the D4s comes out atop. Thus, if a resolution of significantly more than 16MP is required, the D810 may very well be the better camera for the job even at high ISO, but if 16MP is enough, then the D4s will provide higher image quality. Many people erroneously believe that downsampling will make up for a lack of dynamic range, tonal range and color range, but that's not true. However, downsampling done right can hugely improve the noise characteristics of an image, and in case of the D4s vs the D810, this pretty much eliminates the D4s' advantage in noise performance."

Best regards
Erik

CFA characteristics do not change based on "ISO"... so for as long as a specific method of downsampling helps to increase S/N it will help with "dynamic range, tonal range and color range"... because CFA will be the same, if they are good @ base ISO in terms of color separation then they are as well good @ high ISO _if_ we manage to increase S/N... what is the math behind their statement that changing (if increasing) S/N will not affect for example DR per unit of area (not for a fixed one sensel) is not very clear.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11310
    • Echophoto
Re: LenScore / SenScore (Lens and Sensor Reviews) An example.
« Reply #11 on: December 20, 2015, 04:55:09 am »

Hi,

A practical example: according to LenScore the 24-70/4L is better than the 16-35/2.8LII which is better than the 17-40/4L and the worst is the 24-70/2.8LII.

According to other tests, like Photozone the 24-70/4L suffers from spherical aberration and is not really up to the 24-70/2.8LII.

The DxO data indicates that the 24-70/4L trails the 24-70/2.7LII quite a bit at large apertures but catches up nicely at f/8.

In the lens scores DxO-mark gives 35 for the 24-70/2.8LII and 25 for the 24-70/4L.

Best regards
Erik

Logged
Erik Kaffehr
 

John Koerner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 866
  • "Fortune favors the bold." Virgil
    • John Koerner Photography
Re: LenScore / SenScore (Lens and Sensor Reviews) An example.
« Reply #12 on: December 20, 2015, 09:32:32 am »

Hi,

A practical example: according to LenScore the 24-70/4L is better than the 16-35/2.8LII which is better than the 17-40/4L and the worst is the 24-70/2.8LII.

According to other tests, like Photozone the 24-70/4L suffers from spherical aberration and is not really up to the 24-70/2.8LII.

The DxO data indicates that the 24-70/4L trails the 24-70/2.7LII quite a bit at large apertures but catches up nicely at f/8.

In the lens scores DxO-mark gives 35 for the 24-70/2.8LII and 25 for the 24-70/4L.

Best regards
Erik

Two things: I once owned the original 24-70 and it was an exceptional lens, which easily outranked the 24-105 and 16-35.

Another thing is, LenScore only tests 1 lens of each type, and so copy variance is (well) copy variance.

I looked under their FAQ and here is what they say about their testing system:

  • Why create another lens rating system?

    Existing lens rating systems are flawed by two severe problems. The first and somewhat less significant problem is that not all lenses are tested on the same camera, making it virtually impossible to compare the results. All lenses should be tested on the same camera. The second problem is that many lenses outperform even the best full frame sensors currently available. It is impossible to determine the performance of a lens if the lens outresolves the sensor used for testing. The sensor used for testing lenses should have more resolving power than the best lens. Based on these findings, we've created a lens testing system based on a custom-built camera using a 200MP super high resolution CCD sensor, an apochromatic high-precision lens group custom-made by one of the world's best industrial optics manufacturers and 5 exchangeable lens mount adapters. The sensor of our measuring camera is very slow and not designed for low light applications, but these restrictions are of no significance when measuring the optical performance of a lens. We test every lens on one and the same camera with a sensor that easily outperforms every full frame lens in existence.
  • Why are there so few lenses in the database?

    The almost fully automated measuring process takes about 16 hours for primes and 48 hours for zoom lenses, data analysis takes another 2 to 6 hours of number crunching. The goal is having all current full frame Nikon, Canon, Sony, Zeiss, Leica and Sigma lenses in the database plus selected Tamron, Tokina and Schneider lenses, but it will take some time to get there."
  • What can we do as a manufacturer if we suspect the lens tested was defective?

    Get in touch. We will provide you with the serial number of the measured lens and schedule testing with a different copy. If the results of the second test are significantly better, we will update the database accordingly. We do not test lenses provided by manufacturers. Re-testing will be done with a lens acquired independently. We will not do more than one re-test for any given lens model.

So that might be your answer right there.

Jack
Logged

John Koerner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 866
  • "Fortune favors the bold." Virgil
    • John Koerner Photography
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #13 on: December 20, 2015, 09:47:29 am »

Under their About page:

  • About LenScore™
     
    LenScore™ is a rating system for full frame digital camera lenses. It is based on the optical performance of a lens at 8 different apertures from f/1.4 to f/16 at 5 different focusing distances. Zoom lenses are measured at 3 different focal lengths. The score does not take into account a lens' price, weight, size, build quality, AF speed and image stabilization system. It is purely a rating of optical performance. The score is calibrated based on the Nikon AF-S Nikkor 85mm f/1.4G measurements, i.e. the Nikon AF-S Nikkor 85mm f/1.4G is the reference with a score of 1000 in each category.

    The overall score ( = LenScore™ ) is a weighted average of the following 10 scores:


    Resolving Power
    Amount of detail a lens is able to project onto the 24x36mm square within the image circle used by full frame cameras. weight = 7

    Contrast
    Microcontrast performance. weight = 5

    Color
    Color fastness. weight = 3

    Bokeh
    Rendering of out-of-focus areas in front of and behind the focal plane. weight = 5
    How to rate bokeh?

    Star
    Rendering of direct light sources at small apertures. weight = 1

    Distortion
    Geometrical distortions. weight = 5

    Falloff
    Light falloff from center to the edges (vignetting). weight = 3
    How to rate falloff?

    Flare
    Amount of contrast reduction and ghosting caused by direct light sources within and ouside the field of view. weight = 5

    LaCA
    Lateral chromatic aberrations. weight = 3

    LoCA
    Longitudinal chromatic aberrations. weight = 5
  • Declaration
    We have never received any equipment from Nikon, Canon, Sony, Sigma, Zeiss, Leica or any other digital camera lens manufacturer. We only test production lenses purchased and paid for in full from regular camera stores.

    Our sister site SenScore™ started out as a small engineering project associated with a PhD, then more people became interested, ideas for testing lenses started to fly and here we are. Both SenScore™ and LenScore™ are strictly non-commercial. We have never received any form of payment from anybody. There is no advertising and no tracking or profiling of any kind on the websites. We are completely independent and can and will say so when a sensor's or lens' performance is disappointing without worrying about hurting affiliate sales, and when we praise a sensor's or lens' performance, then we do it because we are genuinely impressed, not because we will profit from doing so. All of us are passionate about both photography and engineering, and we believe that, for any task, chosing the right tools is important, knowing fully well that photography is so much more than numbers. Good light!

Jack
Logged

John Koerner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 866
  • "Fortune favors the bold." Virgil
    • John Koerner Photography
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #14 on: December 20, 2015, 10:29:16 am »

SenScore:

  • About SenScore™
     
    SenScore™ is a rating system for digital camera sensors. It is based on the performance of a sensor at all available ISO settings within the 100 to 25600 range as well as the sensor's resolving power across this range. The score is indicative of a camera's image quality across a wide range of real-world conditions. However, the score does not take into account a camera's speed, buffer size, AF, metering, build quality etc. and is therefore not a camera rating. The score is calibrated based on the Nikon D800 measurements, i.e. the D800 is the reference with a score of 1000 in each category. Measurements are based on raw files. JPEG engine performance is not part of SenScore™. Video performance is not part of SenScore™.

    Nikon F-mount cameras are measured with: Nikon AF-S Nikkor 85mm f/1.4G
    Canon EF-mount cameras are measured with: Canon EF 85mm f/1.2L II USM
    Sony A-mount cameras are measured with: Sony Planar T* 85mm f/1.4 ZA

    The overall score ( = SenScore™ ) is the average of the following 5 scores:

    Noise
    The amount of noise added by the sensor and AD conversion process.

    Dynamic Range
    The ratio between the lowest amount of light a sensor is able to detect and the amount of light that pushes the sensor into saturation. Not to be confused with tonal range. A sensor with a 4-bit readout able to distinguish only 16 levels of lightness could theoretically have a very high dynamic range. We measure the dynamic range of a camera sensor with a custom built device for which we are in the process of seeking patent protection.

    Color Range
    The number of colors a sensor is able to distinguish.

    Tonal Range
    The number of lightness levels a sensor is able to distinguish. The imaginary 4-bit sensor mentioned in the Dynamic Range section offering an extremely high dynamic range would score very low on tonal range.

    Resolving Power
    Amount of detail a sensor is able to record. The lens plays an important part in this. All F/EF/A/E mount cameras are measured with the same F/EF/A mount lenses. The 3 lenses used for SenScore™ measurements never leave the lab and are used strictly for this purpose. There is variation in lenses. The F/EF/A mount lenses used for SenScore™ measurements have been carefully selected for their performance.
Logged

John Koerner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 866
  • "Fortune favors the bold." Virgil
    • John Koerner Photography
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #15 on: December 20, 2015, 10:41:32 am »

  • How does SenScore™ differ from other sensor ratings?
    Most sensor ratings represent the image quality of a sensor under ideal conditions, i.e. base ISO, which is very important and useful for selecting a camera for studio use. However, most cameras do not stay in a studio or on a tripod all the time, they are taken to every conceivable dark and challenging corner of the world, and most of the time, they are used to take images under less than ideal conditions. So for selecting a camera for street photography or shooting concerts, for sports and family and the great outdoors, or maybe for a year of backpacking around the globe, SenScore™ is more useful, because it is indicative of a camera's image quality under a wide range of real-world conditions. Also, we think that resolving power is very important and must be taken into account when rating a digital camera sensor.
  • Why is the score of the D3x so low?
    While being able to produce outstanding images at base ISO, the D3x's image quality rapidly diminishes with higher ISO settings.
  • How can the D4s beat the D810?
    The D810 offers far superior resolving power and is able to provide much better image quality than the D4s under favorable conditions. However, dynamic range, tonal range and color range drop off significantly with higher ISO settings while the D4s holds up astonishingly well. The break-even point is below 400 ISO, at 400 ISO the D4s is already marginally ahead. The D810 retains superior resolving power throughout the measured ISO range, but resolving power is only 20 percent of the overall score, which is why the D4s comes out atop. Thus, if a resolution of significantly more than 16MP is required, the D810 may very well be the better camera for the job even at high ISO, but if 16MP is enough, then the D4s will provide higher image quality. Many people erroneously believe that downsampling will make up for a lack of dynamic range, tonal range and color range, but that's not true. However, downsampling done right can hugely improve the noise characteristics of an image, and in case of the D4s vs the D810, this pretty much eliminates the D4s' advantage in noise performance.
    (Eric quoted this earlier.)
  • Upcoming tests:
    Sony A7 II
Logged

kers

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3876
    • Pieter Kers
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #16 on: December 20, 2015, 01:36:03 pm »

Well - Jack- or - John,
you copied about all the text of their site!
maybe it is quicker to use an URL...  :)

Then might as well copy this too:
"LenScore™ is a rating system for full frame digital camera lenses. It is based on the optical performance of a lens at 8 different apertures from f/1.4 to f/16 at 5 different focusing distances. Zoom lenses are measured at 3 different focal lengths"

My idea is that if you calculate a medium value of sharpness of a wide angle lens- you still now nothing.
If you do the same for a telelens it has a lot more value...
etc..
How to measure a medium value for "star"?

« Last Edit: December 20, 2015, 01:51:58 pm by kers »
Logged
Pieter Kers
www.beeld.nu/la

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995
Re: LenScore / SenScore (Lens and Sensor Reviews) An example.
« Reply #17 on: December 20, 2015, 02:37:16 pm »

according to LenScore the 24-70/4L is better than the 16-35/2.8LII which is better than the 17-40/4L and the worst is the 24-70/2.8LII.

Two things: I once owned the original 24-70 and it was an exceptional lens, which easily outranked the 24-105 and 16-35.

your gear list does not show that you ever owned any FF Canons... so your experience with the original 24-70/2.8 (Erik's text was about 24-70/4 ) on APS-C body is relevant how ?


Quote
DSLRs

    Canon EOS 7D (x2)
    Canon EOS 50D - retired
    Canon Digital Rebel T2i - retired


Quote
>Zooms

    Sigma 150-600mm f/5-6.3 DG OS HSM S
    Canon EF 24-70mm f/2.8L II USM
    Canon EF-S 10-22mm f/3.5-4.5 USM
    Canon EF-S 15-85mm f/3.5-5.6 IS USM - retired
    Canon EF 24-70mm f/2.8L USM - retired
Logged

John Koerner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 866
  • "Fortune favors the bold." Virgil
    • John Koerner Photography
Re: LenScore / SenScore (Lens and Sensor Reviews)
« Reply #18 on: December 20, 2015, 02:40:14 pm »

Well - Jack- or - John,
you copied about all the text of their site!

I sign my name Jack, so please use that.

And, um, no, I didn't copy all their text; just the relevant parts to what Erik had posted.



maybe it is quicker to use an URL...  :)

Unfortunately, their URL is static and doesn't go to specific pages; hence the copying/pasting required.



Then might as well copy this too:
"LenScore™ is a rating system for full frame digital camera lenses. It is based on the optical performance of a lens at 8 different apertures from f/1.4 to f/16 at 5 different focusing distances. Zoom lenses are measured at 3 different focal lengths"

I already did post that (see Reply #13).

Maybe you should slow down, and actually take the time to read what's posted, before offering a knee-jerk response :P



My idea is that if you calculate a medium value of sharpness of a wide angle lens- you still now nothing.
If you do the same for a telelens it has a lot more value...
etc..

Then maybe you should create your own site.

I am interested in knowing more about LenScore's testing methods, as I happen to like their site and their setup (as well as the fact they don't take $$ and refuse to test more than 1, at most 2, copies.)



How to measure a medium value for "star"?

That is not a value that concerns me.

But you should ask them, not me.

Jack
Logged

John Koerner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 866
  • "Fortune favors the bold." Virgil
    • John Koerner Photography
Re: LenScore / SenScore (Lens and Sensor Reviews) An example.
« Reply #19 on: December 20, 2015, 02:47:53 pm »

your gear list does not show that you ever owned any FF Canons... so your experience with the original 24-70/2.8 (Erik's text was about 24-70/4 ) on APS-C body is relevant how ?

Good catch.

What's relevant is that the original 24-70/2.8 was a really good lens, that was better than either the elder 16-35 or the 24-105, and so it is conceivable that a particular copy of the original 24-70 lens might outscore one of the newer versions 16-35 or 24-70 on their (LenScore's) setup.

But I made a mistake in not seeing the subject was the new f/4.

Still doesn't change the point though ...
Logged
Pages: [1] 2 3   Go Up