08-29-2012, 08:17 AM
(This post was last modified: 08-29-2012, 08:25 AM by Brightcolours.)
[quote name='Rover' timestamp='1346218500' post='19868']
Most of the times the results across these sites are comparable - of course, they should be if the lens is the same and the methodology is valid at all sites - but sometimes there are discrepancies. Lenstip is the one that sticks out more often.
Wonder what you mean by saying SLRGear gets weird results very often? Which lenses do you have in mind? I don't remember any of them sticking out too much from what the other 3 sites find. You tell me.
[/quote]
The list of weird or even nonsensical results from SLRgear is too long to just list them all!
http://slrgear.com/reviews/showproduct.p...353/cat/11
Canon EF-S 17-55 IS USM:
"Chromatic Aberration
CA is a bit of a weak point for this lens, at least at its widest focal lengths. At 17mm, maximum CA is quite high, although average CA is much lower, indicating that the worst CA is limited to the edges and corners of the frame. CA decreases as you zoom to longer focal lengths, reaching a reasonable level (but still higher than we'd like) at 28mm. CA from 35-55mm is acceptably low."
"The Tamron has better CA (sometimes much better) across the focal length range"
In reality, this lens has very low CA.
Photozone (correctly) finds:
"Lateral chromatic aberrations (color shadows at harsh contrast transitions) tend to be a weak spot of most zoom lenses in this range but the EF-S 17-55mm f/2.8 USM IS produces an unusually moderate degree of around 1px at the image borders."
![[Image: ca.gif]](http://www.opticallimits.com/images/8Reviews/lenses/canon_1755_28is_50d/ca.gif)
About that Tamron that has "much better" CA across its focal range, photozone (correctly) finds:
"Chromatic aberrations (color shadows at harsh contrast transitions) are high towards the wide-end of the zoom range especially at large aperture settings with a peak of almost 2.2px on the average at the image borders at 17mm. Stopping down reduces the issue but you will need to stop down to f/8 in order to suppress the problem to an uncritical level at 17mm. CAs are a lesser problem towards the long end of the zoom range."
![[Image: ca.png]](http://www.opticallimits.com/images/8Reviews/lenses/tamron_1750_28_d7000/ca.png)
Checking any non-corrected images from each lens confirms this.
According to SLRgear, the Tamron 70-300 VC and Canon 70-300 IS are sharper than the Canon 70-300mm L IS USM:
"It also offered slightly better results for sharpness" (canon)
"Despite the dramatically lower price tag, the Tamron actually produced as-sharp or sharper results than the Canon."
70-300 L IS USM:
![[Image: mtfx.png]](http://www.opticallimits.com/images/8Reviews/lenses/canon_70300_456lis_ff/mtfx.png)
70-300 IS USM:
![[Image: mtf.png]](http://www.opticallimits.com/images/8Reviews/lenses/canon_70300_456is_ff/mtf.png)
70-300 VC USD:
![[Image: mtf.png]](http://www.opticallimits.com/images/8Reviews/lenses/tamron_70300_456vc_eos_ff/mtf.png)
Just reading the 24-85mm review on SLRgear I see a number of sentences that are odd:
"Diffraction limiting sets in by ƒ/11"
They tested on a D800 and D7000. For both these cameras diffraction sets in at f5.6.
"Mounted on the D800, the worst barrel distortion shows at +1% in the corners at 24mm, and -0.1% pincushion over 50mm. "
I know that the "percentage" numbers mean little from website to website. But, Lenstip correctly shows the pincushion distortion to be very strong, and gives a number that is about half that of the barrel distortion at 24mm. We can see this is correct from the images Markus has posted. So... -0.1% pincushion? Really? When the barrel distortion is said to be +1%?
"CA presents as purple fringing in areas of high contrast, in the corners." Uhmm... what??
So now SLRgear does not even know what CA means? And they are measuring it? Ok....
Lenstip shows this lens to have blue/yellow CA.
![[Image: 3677_nik24-85_abISO.jpg]](http://www.lenstip.com/upload3/3677_nik24-85_abISO.jpg)
SLRgear says:
"The use of ED glass helps a bit here, as chromatic aberration is kept quite low on both the D7000 and D800 bodies"
Remember that they said the CA on the Canon 17-55 f2.8 IS USM is a weak point? This is what it looks like:
![[Image: 1610_can17-55_abISO.jpg]](http://www.lenstip.com/tmp/1610_can17-55_abISO.jpg)
So... on the Canon 17-55mm quite high according to SLRgear. And this 24-85mm the CA is kept quite low. Right. Forgive me to trust the Lenstip results a lot more. Want to bet on what Markus will measure?
About vignetting.... I have not been able to make sense of the graph! But the text seems silly again:
"On the D800 however, there is considerable corner shading at several focal lengths and apertures. The worst offender is 24mm and ƒ/3.5, where we note extreme corners that are over a stop darker than the center of the frame."
Over a stop. It of course CAN mean over 2 stops, as lenstip finds. But if they mean just over 1 EV in the corners wide open, that is sensational for a 24mm on FF. Want to bet about what Markus will find?
Most of the times the results across these sites are comparable - of course, they should be if the lens is the same and the methodology is valid at all sites - but sometimes there are discrepancies. Lenstip is the one that sticks out more often.
Wonder what you mean by saying SLRGear gets weird results very often? Which lenses do you have in mind? I don't remember any of them sticking out too much from what the other 3 sites find. You tell me.
[/quote]
The list of weird or even nonsensical results from SLRgear is too long to just list them all!
http://slrgear.com/reviews/showproduct.p...353/cat/11
Canon EF-S 17-55 IS USM:
"Chromatic Aberration
CA is a bit of a weak point for this lens, at least at its widest focal lengths. At 17mm, maximum CA is quite high, although average CA is much lower, indicating that the worst CA is limited to the edges and corners of the frame. CA decreases as you zoom to longer focal lengths, reaching a reasonable level (but still higher than we'd like) at 28mm. CA from 35-55mm is acceptably low."
"The Tamron has better CA (sometimes much better) across the focal length range"
In reality, this lens has very low CA.
Photozone (correctly) finds:
"Lateral chromatic aberrations (color shadows at harsh contrast transitions) tend to be a weak spot of most zoom lenses in this range but the EF-S 17-55mm f/2.8 USM IS produces an unusually moderate degree of around 1px at the image borders."
About that Tamron that has "much better" CA across its focal range, photozone (correctly) finds:
"Chromatic aberrations (color shadows at harsh contrast transitions) are high towards the wide-end of the zoom range especially at large aperture settings with a peak of almost 2.2px on the average at the image borders at 17mm. Stopping down reduces the issue but you will need to stop down to f/8 in order to suppress the problem to an uncritical level at 17mm. CAs are a lesser problem towards the long end of the zoom range."
Checking any non-corrected images from each lens confirms this.
According to SLRgear, the Tamron 70-300 VC and Canon 70-300 IS are sharper than the Canon 70-300mm L IS USM:
"It also offered slightly better results for sharpness" (canon)
"Despite the dramatically lower price tag, the Tamron actually produced as-sharp or sharper results than the Canon."
70-300 L IS USM:
70-300 IS USM:
70-300 VC USD:
Just reading the 24-85mm review on SLRgear I see a number of sentences that are odd:
"Diffraction limiting sets in by ƒ/11"
They tested on a D800 and D7000. For both these cameras diffraction sets in at f5.6.
"Mounted on the D800, the worst barrel distortion shows at +1% in the corners at 24mm, and -0.1% pincushion over 50mm. "
I know that the "percentage" numbers mean little from website to website. But, Lenstip correctly shows the pincushion distortion to be very strong, and gives a number that is about half that of the barrel distortion at 24mm. We can see this is correct from the images Markus has posted. So... -0.1% pincushion? Really? When the barrel distortion is said to be +1%?
"CA presents as purple fringing in areas of high contrast, in the corners." Uhmm... what??
So now SLRgear does not even know what CA means? And they are measuring it? Ok....
Lenstip shows this lens to have blue/yellow CA.
SLRgear says:
"The use of ED glass helps a bit here, as chromatic aberration is kept quite low on both the D7000 and D800 bodies"
Remember that they said the CA on the Canon 17-55 f2.8 IS USM is a weak point? This is what it looks like:
So... on the Canon 17-55mm quite high according to SLRgear. And this 24-85mm the CA is kept quite low. Right. Forgive me to trust the Lenstip results a lot more. Want to bet on what Markus will measure?
About vignetting.... I have not been able to make sense of the graph! But the text seems silly again:
"On the D800 however, there is considerable corner shading at several focal lengths and apertures. The worst offender is 24mm and ƒ/3.5, where we note extreme corners that are over a stop darker than the center of the frame."
Over a stop. It of course CAN mean over 2 stops, as lenstip finds. But if they mean just over 1 EV in the corners wide open, that is sensational for a 24mm on FF. Want to bet about what Markus will find?