Personally, I see value in using automated software to analyze the resolution chart images (no personal judgment involved) and I am not surprised to see that measured resolution doesn't start dropping until after ISO 3200. That matches my own experience pretty closely.
The results aren't particularly surprising either. A 100MP camera with a good lens should outresolve a 61MP camera with a good lens. A 100MP medium format digital camera should outresolve a 100MP 4/3" digital camera. And that's what the Imatest results show.
I agree it will be good to see some comparisons made with real live test targets like the DPReview target, assuming the 28mm equivalent lens and minimum focus distance for the M4P doesn't prevent that.
When I get my M4P I will do some same scene / light comparisons and post the RAW files online. ...Show more →
Thanks for the link about the testing. When you see what they did, you can soon realize how little information it is providing:
Here is their description of the resolution test:
We use a resolution chart based on ISO-12233 from Applied Image inc to indicate the limit of the camera’s vertical resolution at the centre of the frame. The higher the value, the better the detail resolution.
What is the problem here? it doesn't tell us what aperture they used to do these tests and it doesn't tell us anything about resolution except in the center of the frame. Why is that a problem?
Would we say a lens that is sharp in the center at let's say f/5.6, but crappy across the Fram at f/2.8 and shows lots of deterioration of performance at f/8 is as good as a lens that is sharp across the frame at f/2.8 and has that same sharpness at f/11. I find it bad enough when test sites just provide center, edge, and corner resolution at a range of apertures (that still leave out some pretty important information like potential mid zone dips in performance, but with these tests we literally only know about center performance and then only at presumably the best aperture. That is really a paucity of information.
Steve Spencer wrote:
What is the problem here? it doesn't tell us what aperture they used to do these tests and it doesn't tell us anything about resolution except in the center of the frame. Why is that a problem?
Would we say a lens that is sharp in the center at let's say f/5.6, but crappy across the Fram at f/2.8 and shows lots of deterioration of performance at f/8 is as good as a lens that is sharp across the frame at f/2.8 and has that same sharpness at f/11. I find it bad enough when test sites just provide center, edge, and corner resolution at a range of apertures (that still leave out some pretty important information like potential mid zone dips in performance, but with these tests we literally only know about center performance and then only at presumably the best aperture. That is really a paucity of information....Show more →
I agree with you generally speaking. We definitely don't have good measurements across the entire image field yet. One thing to keep in mind is many serious landscape photographers using drones will stitch images, so they are using the sharpest, center section of the lens for many of their images.
Also, the author did discuss apertures in the review, and if I remember correctly he mentioned f/2 and smaller than f/8 wasn't up to 100MP, he recommended f/2.8-f/5.6 with f/4 at peak sharpness.
flash wrote:
I’ll have to be more careful with DR and noise. Likely more need for HDR blends. OTOH it’s a lot less conspicuous than the Inspire. Certainly there are many occasions where the images might be indistinguishable from each other.
Gordon
My workflow with the M3 is to ALWAYS shoot with exposure bracketing and HDR the RAW images in post.
I am kind of crushed that I didn't figure the huge IQ improvement from that until after a year and a half of using the drone.
Jun 24, 2025 at 08:52 PM
Steve Spencer Online • • • • • • • Upload & Sell: On
lightskyland wrote:
I agree with you generally speaking. We definitely don't have good measurements across the entire image field yet. One thing to keep in mind is many serious landscape photographers using drones will stitch images, so they are using the sharpest, center section of the lens for many of their images.
Also, the author did discuss apertures in the review, and if I remember correctly he mentioned f/2 and smaller than f/8 wasn't up to 100MP, he recommended f/2.8-f/5.6 with f/4 at peak sharpness.
Yeah, he mentioned different apertures, but we don't even know what apertures are being used in that chart and we don't know if f/2.8 is a lot worse that f/5.6 or basically the same. Just a dearth of information and far far less than is typical in such reviews, which I still think tells us less than ideal.
The point that it can hang with the Inspire and the a7r3 is fantastic. Makes me want to upgrade my Mavic 2 pro.
flash wrote:
I have both and no, it doesn’t.
It is super impressive for something so small. The best I’ve seen. I rarely flew my Mavic 3 Pro as my Inspire 3 stomped on it. This is not the case with the 4. While both the Inspire and A7R5 produce files with far more DR and less noise and that means more *usable* resolution, the M4 is actually very very usable for larger prints. I’m still in the testing stage and the weather has been horrid here but so far it’s looking more A7R3 than A7R5, which is still incredibly impressive.
I’ll have to be more careful with DR and noise. Likely more need for HDR blends. OTOH it’s a lot less conspicuous than the Inspire. Certainly there are many occasions where the images might be indistinguishable from each other.
The tester doesn’t say anything about his testing methodology. were they tested side by side on the same day? Is the A7R5 historical data. What lens on the Sony? jpegs or rawls? Which raw processor? Whatever he did his results don’t match mine, so far. But as I said, it’s early day for me.
