Quote:"MTF software says nothing about the lens"
That is a silly bit of paraphrasing gone wrong right there.
Quote:Oh, good. It's more for nice and decorative charts, right?
Once again I can't follow your reasoning or would have to disagree if I tried to. But that's alright, I leave it to others to discuss that since I know close to nothing about Sony stuff these days. -_-
As MTF is always looking at the combination of a given sensor and lens, you just can't say the size of the vertical scale says nothing about the lens because this scale is not only sensor resolution but a recorded resolution of lens and sensor which were placed in front of a pattern. And the Sony combination gives more details about that pattern than the (in that case) Canon combination. At lensrentals, that Sony A7R and 35/2.8 had a bigger resolution than a D800E with 35/1.4 or Zeiss 50/2. They redid the test because at first they didn't believe it was max aperture and thought they made a mistake. Source to be find here
I do not mind if a given resolution of lens and sensor is 80% of a chart and another combination is 95% of another, smaller chart, if at the end of the day that means the 80 % will give much more resolution in absolute terms. We're not talking about 3 or 5% - in total there is a difference of 20% more LW/PH (4252 against 3512). Sensor sizes are about the same.
But as I said, I'm no Sony expert. And also, I don't care about "resolution over everything else". Matter of fact, since I saw some prints of those Sigma Merrill cameras with the Foveon sensor, I want to rent one and make my own comparison. Also, what good is the superresolution of one lens if the rest of the system is questionable in terms of reliabilty or (in the Sony case) big shutter vibration. Again, I only read about.
I can only state what I stated before, in the hope that it sinks in. The numbers when comparing are meaningless, the bar heights in relation to the graph height are not all that meaningless. The numbers are higher for the A7r tests because of a higher resolution sensor, not because of a significantly sharper lens.
The aim for PZ is to test lenses, not camera/lens combinations. That is why the higher numbers do not reflect in higher ratings.
Lens rentals often makes odd mistakes in logic. If they what to show how sharp that 35mm f2.8 is, why do they not do the obvious (which is: to put that Nikkor 35mm f1.4 and/or 50mm f2 Zeiss also on the A7r)? THEN you can say something about the lens. The AA-filterless D800E is not really AA-filterless (Nikon says it has two which cancel eachother out). It shows in images. It is not like there is no way to mount those lenses on that A7r, after all.
Anyhow, I hope you get the point. PZ scores lenses, not camera/lens combinations. The bar heights (with disregard of the particular numbers) more or less show how good a lens is, sharpness wise. Look at the "excellent / very good / good / fair / poor" markings instead of pure numbers.
Well I thinik this is fallacy in your statement in that it takes a good lens to produce high resolution on a dense sensor; but having said that a good lens cannot produce a high resolution on a low density sensor.
-
This is one thing that always bugged me about lens testing. They test with the camera so it is system testing; kind of wish there was an easy way to test without the camera to determine the quality of the lens (though then we run into the crap with interaction between the angle of the light and the micro lenses on the sensor as well as other issues). Hum. Oh well.
Next time you'd better check PZ's own statements before speaking in their name, Brightcolours:
Quote: Q: Why are the quality ratings different from system to system ?
As mentioned above the lens quality is affected by the sensor "system". Every additional step in the pipeline decreases the output quality, specifically the low-pass filter in front of the sensor. Assuming you mount the same lens on different system its maximum resolution will vary according to the max. quality of the sensor system. There're also evolutions regarding the RAW converter quality so more recent system tests starts can benefit from this - e.g. Canon/Olympus RAWs are/were converted using ACR 3.2 whereas Pentax/Sony RAWs are/were converted via ACR 3.7 and there was an increase in converter quality with ACR 3.4). This must all be taken into account regarding the rating system.
To be read here
I try to avoid becoming personal, BC, but I put it in this way: I do trust lensrentals a couple of times more than your reasoning. Their, and especially Roger's, comments do make sense to me. While your reasoning appears to me from of a different world with some special optic rules. So, no, nothing of your last post will sink in.
Do you really and seriously believe, the lack of AA filters leads to an increase of 20% MTF values? And if you think, putting an adaptor in between lens and body is a good way to compare lenses - why don't you do it yourself? Lenstesting is time consuming and lensrentals runs a main business and a very informative blog - with enough experience, Imatest and optical benches. Trying to bring them into discredit leads only to discredit your own posts in my eyes.
02-12-2014, 07:24 PM
(This post was last modified: 02-12-2014, 07:28 PM by Brightcolours.)
Quote:Next time you'd better check PZ's own statements before speaking in their name, Brightcolours:
You have to read and try to understand what it all means. YOU were the one trying to compare the numbers, while PZ explains exactly there why that is not a good idea. So... Yeah. You read it again.
Quote:To be read here
I try to avoid becoming personal, BC,
Coming from you, that is rich. Now I wonder where in my post i got "personal".
Quote:but I put it in this way: I do trust lensrentals a couple of times more than your reasoning.
Oh well. If you think it is faulty reasoning to suggest to test the SAME lens on the two different cameras.. what can I say. Only that way you know which part of the results is due to the different camera.
Quote:Their, and especially Roger's, comments do make sense to me. While your reasoning appears to me from of a different world with some special optic rules. So, no, nothing of your last post will sink in.
Do you really and seriously believe, the lack of AA filters leads to an increase of 20% MTF values? And if you think, putting an adaptor in between lens and body is a good way to compare lenses - why don't you do it yourself? Lenstesting is time consuming and lensrentals runs a main business and a very informative blog - with enough experience, Imatest and optical benches. Trying to bring them into discredit leads only to discredit your own posts in my eyes.
Oh well.
Quote:
Thanks for the test!
Well - there was not much more to expect looking at the MTF Sony published.
The 35 2.0 on the RX1 shows superior MTF curves when compared to the FE 35 2.8.
The FE 55 1.8 should play in the same league with the RX1 35 2.0 although at 55mm it should be easier anyway to have high resolution at the border.
Seeing the test at dxomark showing the flat sharpness distribution I can't see how this goes together with the resolution figures in your test.
Is it sample variation or are we comparing contrast with resolution e.g. mtf10 and mtf40 ?
I can never place the DXO "colours". So I end up disregarding their "findings" due to the weird nature of them.
|