Sony a77 and Minolta lenses?

I've been lens testing since the early 1980's. I usually used a test chart at around 2 metres from the camera and measured the lens to each corner to ensure flatness. I used that as an initial guide for flat field sharpness across the frame and so that I could determine corner sharpness and 'best' aperture.
However, I also went out and shot actual images using all apertures and on a tripod with a cable release (often mirror up). Clearly, some lenses have a curved field and can be sharp at the edges wide open with an unsharp centre.
Although I wasn't always looking for the sharpest lens, I recall one Nikkor 20mm lens being so bad I returned it to the shop and exchanged for another, different lens.
Nowadays, with objective lens tests easily available online for pretty much every lens available, I'm more interested in 'the look' and I don't usually bother repeating what I can see already done. It's more about what I might do with that particular lens.
I note the discussion on formats. I WOULD recommend full frame over APS-C or smaller formats. Once a lens's best resolution is reached, a format (whether film or digital), is not adding any more detail; it's just removing sensor/film noise but that DOES make an image 'look' better.
The small formats do struggle at larger sizes and I print to 24x36" on photo paper and A2+ with my large format printer. APC-=C just doesn't cut it at those sizes. If a really good lens achieves 250ppm, then, at best, APS-C is maxxed out at about 24MP. HOWEVER, most lenses don't get very close to that resolution and even when they do, the shooting method (hand held, non optimal apertures etc) usually means it can't achieve its best. I believe most people will be getting only about 120-150ppm. Therefore, the same lens on FF can reach over 50MP (although, for the same reasons; they usually don't). Even in a small print of say 8x10 / A4; the difference is quite obvious. I've experimented with this using my Sony A7R as the FF and my Sony QX-1 as the APS-C using the same lens at different distances to achieve the same image.
So any 'lens' test should also be a system test because format and nowadays, sensor design etc all add up to create an over all IQ. However, the biggest variable is always how the photographer uses the equipment.

Ditto for film, film format combined with lppm (line pairs per mm) affects noise and high res films look far more 'digital' than std films. I have a collection of 16mm cameras and use microfilm to reduce the noise massively. That does make them look better than shots with consumer film. Some microfilms can resolve up to 3-500lppm when processed for a wide tonal range in Pyro or special low contrast developers. Although they can reach over 1200lppm at 1000:1 contrast; that's not useful in photography so is irrelevant. This process gives a nice 'clean' look but the lens is still resolving a paltry 80-150lppm.

For me, it's a combo of things but the 'look' is the main one. I do want corner sharpness at wide apertures but, as I'm not a macro shooter, I don't need that as a flat field. It will be the contrast, colour, OOF, and any distortions I can utilise that will attract me.

So, in the end, lens tests mean different things to different people. Some want flat field sharpness corner to corner, others want creamy bokeh, others want a particular kind of distortion etc. I reckon the best lens tests include all parameters so that different photographers can take what they need for their photography from such testing.

The shots below are from some of those film tests. Although they are relatively grainless and therefore low noise; they don't have any extra detail; just a smoother 'look' and it's the 'look' I am chasing. The films used all resolve better than 350lppm (700ppm).

A. Reference image. 16mm (10x14mm neg) using consumer 100 iso film (very grainy, not great tonal separation)


1. Viscawide 16mm panorama camera 10x46mm neg. Fuji microfilm

Maratai Beach by Michael J Breen, on Flickr

2. Minolta 16 camera with Rokkor 23mm f3.5 lens for a 10x14mm neg. Fuji microfilm

Selfie: all 61 years of me by Michael J Breen, on Flickr

3. Minolta Alpha 7, 16mm f2.8 fisheye lens. Fuji Eterna RD-1 Copy Film (1.5 iso)
Playskool by Michael J Breen, on Flickr

So, for me, format matters but a specific focus on lens sharpness doesn't so much: It's about my mental perception of IQ

Wētāpunga wrote: addy landzaat wrote: Miranda F wrote: I wouldn't disagree with Addy, except that many of the reviews (though interesting and useful) are essentially worthless as comparative tools. The strength of the lensdatabase is the number of different experiences. That is why you should read ALL the reviews of a lens you're considering... Agreed, I also find the (more detailed) written reviews to be more useful than the actual scoring. One of the issues is the ranking system (1-5) is very coarse and does not allow for a lot of separation of lenses by the traits given. Another (assuming the ranks haven't been updated) is that zoom lenses and prime lenses have a different value for 5. A prime that scores 4.5 for sharpness say, meets the same qualification for a zoom at 5. It's not widely adhered to of course, but it reflects the problem of the grading. With a large enough sample, the average scores across a lens do say something I think, but for many lenses we don't have big samples. What I like to see is a good detailed review that outlines the real life strengths and weaknesses of the lens. How does it work at the apertures you're likely to use. What applications will it excel in etc? I like to read them all, for overall impression.

In a truly objective test; there should be no difference between the score of a zoom vs a prime: if that's not how Dyxums' lens reviews are scored, they are rather useless.
For the most part, I use them as a guide in general only: if a lens scores over 4.5; I'll probably look at it but will still consider scores down to 4 if the contrast, OOF, tilt/shit/macro/aperture etc attract my attention.
For properly objective tests, IMO it's better to stay with one testing org who uses the same equipment for every test and completely standardizes them. As a DxO software user, I naturally tend to their tests because many of my old lenses were listed as are many I'd choose to use with DxO Photo Lab 4.
However, as you say, these are not necessarily useful for evaluating an overall 'look' or personal IQ. For those requirements, I look at about 5 reviewers who comment on the things I look for.
One of my favourite lenses in E-Mount is a Russian 50mm f2.8 mounted on a rubber tube. It has about 12 blades (I'll count them one day) and allows me to tilt/shift and control focus and background / foreground blur. In the end, it probably only resolves 100-150ppm but it's final product is very interesting.

Wētāpunga wrote: Your Russian lens sounds...interesting.

Ha! ha! I wonder what it would score on the Dyxum scale? Maybe a 2.1? That would miss the point completely!



Here's a couple of pics from it:


Toi Whakairo by Michael J Breen, on Flickr


DSC00470DxOa by Michael J Breen, on Flickr



Edited by Idleidolidyll - 12 October 2021 at 02:20

addy landzaat wrote: Miranda F wrote: I wouldn't disagree with Addy, except that many of the reviews (though interesting and useful) are essentially worthless as comparative tools. The strength of the lensdatabase is the number of different experiences. That is why you should read ALL the reviews of a lens you're considering. IMO your reviews are also useless as comparative tools as your use differs from mine. And in several cases my experience in my use differ massively from your and QuietOC's reviews. I sometimes wonder whether you ever use the lenses in the real world or just test them. You never share any of your pictures. At least QuietOC participates in some challenges.

I take exception to your remarks and feel no need to justify myself, but I will point out that despite passing nominal retirement age I am still working full-time and have many other interests besides photography, including maintaining a number of vehicles and an old house. I take a lot of pictures but don't feel many are worthy of sharing even if I had the time and energy to do so. You can make whatever inference from that you like.

Frozenpixels wrote: This forum needs a like button

NNOOO!!!!!! The 'like' button is the death knell for discussion.

Frozenpixels wrote: To everybody who has helped me out here, thank you very much. As per a few people’s suggestions, I just picked up a 16-50mm 2.8 to play with. So far it looks pretty cool. I haven’t uploaded any shots yet. I must say for a size the darn thing is very heavy! But it makes my old Sony 16-55 kit lens look like a dollar store toy. I’ll probably pick up a wider APS-C wide lens to mess around with also. If I feel that what I’ve got is making me happy I might just stick with the a77 and keep the Minolta lenses for whatever is useful or maybe even try film until I feel an a99 is warranted/or in my budget.

Good choice. I like the a77. A very capable bit of kit. The SAL1650 is an excellent lens. Please remember to post some images.