Advice please on the 24-70....
- Thread starter newtonikon
- Start date
Yeah, DxO must think they are testing camera sensors. The sensor sampling obviously does limit what the lens can do (insufficient sensor resolution fails to deliver what the lens did, which is what causes moire - using less than minimum sampling for the detail that exists). That's always been true, we've always needed anti-aliasing filters, but today, we are beginning to dare to remove a few anti-aliasing filters, meaning that our sensor resolutions are finally just barely approaching the MINIMUM sampling of the detail that the lens can deliver (except the filter decision did not know which lens.
). And the Minimum is just Minimum, oversampling obviously has much to be said for it.So to rate lens resolution in megapixels is just plain stupid, They just made that up, and either they don't know, or they are just hoping to standout somehow by impressing beginners that don't know either. Reporting lens results based on insufficient sampling (less than the known minimum requirement) is stupid, or fraudulent, or something very bad. Certainly it only depends on the sensor used, so the plan is just dumb. They may imagine if this is the sensor used, more lens doesn't matter, but this ignores oversampling, which is always a very good known thing. DxO is nonsensical meaningless numbers, we should look for more orthodox standard lens testing methods that actually test lenses.
Digital sampling sort of has its own rules, regardless if DxO knows them or not.
See obvious evidence at Have we hit a megapixel resolution limit?
Last edited:
You obviously know your stuff about lens testing.
What do you make of the sharpness testing that photography life do using IMATEST.
I tend to trust what they say more than DXO.
https://photographylife.com/reviews/tamron-24-70mm-f2-8/4
What do you make of the sharpness testing that photography life do using IMATEST.
I tend to trust what they say more than DXO.
https://photographylife.com/reviews/tamron-24-70mm-f2-8/4
No, I don't know much about lens testing, except a few basics how it is done. It has been done for more than 100 years, so a lot of experience is known, but it is still a very difficult problem to concoct meaningful numbers.
I may know a little more about digital sampling of that lens resolution. The big mistake we make is to imagine there is some one to one correspondence between pixels per mm and line pair per mm. There is Nyquists Minimum 2x sampling rule about aliasing (false detail due to insufficient sampling), but no maximum. More sampling is always better, at least within reason.
Measuring line pair per mm is difficult, determining when we can just make out that two fuzzy lines are distinguishable (not really the same criteria by which we judge our photos). Contrast of those fuzzy lines is a very big deal, so Modulation Transfer Function (MTF) is a big deal thought useful by lens manufacturers - but hard for many to understand, and again, not what we see in our photos. What we like to see is sharp photos, but measuring that in meaningful numbers is very difficult. We just know what we like.
Imatest manufactures printed test charts for others to use. Which seem very popular, and they do observe ISO standards (DxO makes up their own crap about megapixels). The Imatest chart is an ISO standard used by many tests - www.photozone.de too, except they do not duplicate your links report about the Tamron at 70mm peaking at f/11. I trust Photozone more, they are not amateurs. The job sort of needs experienced people too.
Tamron AF 24-70mm f/2.8 SP Di USD VC (FX) - Review / Test Report - Analysis
Isn't that heresy?? i don't understand those that hang on their every word, and wait for a DxO score before they decide on their gearDxO?
I do use, trust, and like Welcome to Photozone!.
"The Tamron AF 24-70mm f/2.8 SP Di USD VC is not a perfect lens - none really are - but regarding the sum of its qualities it is a worthy competitor for Nikon's AF-S 24-70mm f/2.8 lens. In any case it is very hard to beat from a price perspective!"
I'd agree, it obviously seems a worthy competitor and hard to beat on price. I suspect nobody doesn't like it. I suspect either will do any job well.
Price definitely is a major factor, some are not into it enough to ever have either lens. And extra price normally does suffer diminishing returns, but some are always interested in greatest returns.
Just numbers, but almost all of the Photozone numbers are in fact better for the Nikon. I doubt either is actually accurate. Certainly it does not mean the Tamron is not a worthwhile bargain.
Even your own link (Lens Comparison page) says:
The Tamron 24-70mm shows impressive performance at the shortest focal length of 24mm. It has excellent sharpness across the frame and it clearly beats the Nikon 24-70mm at all apertures. However, as the Nikon 24-70mm is zoomed in beyond 24mm, it produces better resolution across the frame. While Tamron starts out strong at the shortest focal length and weakens towards the telephoto range, the Nikon shows a completely reverse picture. Except for 24mm, the Nikon is clearly better, especially when stopped down to f/5.6-f/8. It also produces smoother bokeh and has a little less vignetting problems to deal with. At the same time, the Nikon 24-70mm suffers from heavier distortion issues (stronger barrel distortion at 24mm and pincushion distortion at all other focal lengths)
However, their wording is strangely strong, even for their own data...
According to Photozone, the Nikon is better, and even at the problem 24mm, they compare this way:
Probably more equal than not there, but I see an edge for Nikon there too. In the numbers.... which are just numbers, always suspect, probably cannot be duplicated in either case. What we need is both lenses to go take actual pictures with. We need to see OUR pictures, we'll know what we like.
Likewise, their statement about distortion, instead Photozone says Nikon is better (by good margin) at 24 and 70mm, but not at 40mm.
Who're you gonna believe?
I think we cannot be too concerned with numbers on the internet.salukfan111
Senior Member
If you want a "trial" lens you could resell with no depreciation, try a 28-85 or a 24-85 VR. You can pick up the first for 60 bucks on ebay and the second maybe 200. There is a 35-70 f/2.8 that should be a available on ebay for about 200 as well but I don't have that (I really like the previous two). You may find you're not too into that range at all and it's better to find that out before plunking down a bizillion dollars for that "trinity" lens.
I have a 20-35 nikon lens that is really sweet for walking around shooting with and is reasonably priced as well and is a tiny fraction of the weight of Tamron 15-30 but the pictures aren't quite as good.
I have a 20-35 nikon lens that is really sweet for walking around shooting with and is reasonably priced as well and is a tiny fraction of the weight of Tamron 15-30 but the pictures aren't quite as good.
Ruidoso Bill
Senior Member
One other thing to consider is build quality, I know nothing of the tamron 24-70, in fact the only tamron lens I have ever owned is the 150-600 and I bought it for a specific purpose, the Nikon Trilogy (hate it called the Trinity) I use on a daily basis, I shoot about 1000 clicks per day and use the 14-24 and the 24-70 heavily. I have never in several years of ownership had a failure, they continue to remain tight and just like new, that is a great statement considering build quality.