Medium format look – a topic debated at nauseam but now with a visual twist

[JimKasson]

Thanks for dropping by, Jack. Always good to hear from you.

 

[JimKasson]

Three images, three sensor sizes. All the required settings for equivalence.

Camera jpgs, only changes are the crop to 16x16 and the stitch to 64x64. Apologies for the stitching artifact, it was windy.

Images downsized to the lowest common resolution

Hi, have you stitched photos from the smaller sensor to replicate the larger sensor size? Then downsampled the higher resolution files.

Isn't that like limiting the speed of 2 cars to 70mph, the larger engine car gets to 70mph first but once they get there they are equal?

And in actual use you'd take the larger sensor photo at lower iso.

If you have plenty of light.

 

[BackToNature1]

Sensors of different sizes come with different quality. Same goes for sensors of the same size. Lens, ditto. So to come to any real definitive conclusions on the Look per se would prove to be quite difficult. However that being stated, I truly feel there are big differences between the different sensor sizes, just on personal experience alone.

 

[Stan Disbrow]

Hi,

And a bad amplifier sounds like utter crap.

Stan

 

[DavidMillier]

There is no debate about the facts that pixel size, pixel count make a difference. A 100MP MF sensor looks different to a 20MP m4/3 because of resolution difference alone.

However, let's imagine we had gigapixel sensor tech of excellent pixel level quality.

Would shooting a m4/3, APS-C, full frame or MF possess an intrinsic look then? Obviously lens quality, depth of field and angles of view would all need to be equalised, but if somehow this could be done, would there be a specific look down to sensor size alone?

No is the answer.

 

[DavidMillier]

Hi,

And a bad amplifier sounds like utter crap.

Stan

By definition, a bad amplifier is any amplifier that does anything other than make the sound louder.

 

[BackToNature1]

I shoot with a vary of sensor sizes so I am pretty much certain about what I see. I can't speak for others. Your mileage may vary. At the end of the day, everybody gets to shoot with the Gear they feel best about. Too bad prices are going up though. Mostly.

 

[tex]

• I know of no "optical bench testing" that exists that measures tonality. Jim can correct me I guess.

You can measure SNR. But unless you're going to put more photons on the MF sensor (abandoning equivalence), you won't see a big SNR difference with same sensor tech.

Yes, of course, but does that measure tonality transitions? Or can we extrapolate from SNR tests anything about transitions?

 

[JimKasson]

• I know of no "optical bench testing" that exists that measures tonality. Jim can correct me I guess.

You can measure SNR. But unless you're going to put more photons on the MF sensor (abandoning equivalence), you won't see a big SNR difference with same sensor tech.

Yes, of course, but does that measure tonality transitions? Or can we extrapolate from SNR tests anything about transitions?

Noisy transitions don't look smooth. Inadequate dithering, however, can look unnatural. It's the Goldilocks effect, but the acceptable range is very large.

 

[JimKasson]

There is no debate about the facts that pixel size, pixel count make a difference. A 100MP MF sensor looks different to a 20MP m4/3 because of resolution difference alone.

However, let's imagine we had gigapixel sensor tech of excellent pixel level quality.

Would shooting a m4/3, APS-C, full frame or MF possess an intrinsic look then? Obviously lens quality, depth of field and angles of view would all need to be equalised, but if somehow this could be done, would there be a specific look down to sensor size alone?

No is the answer.

To a first approximation.

But there are higher-order effects:

Format size and image quality - the last word

The format rhetoric seems to be heating up again. Maybe it never cooled off. In the last 7 or 8 years, I’ve owned and made many images with APS-C (with crop factors between 1.3 and 1.5), full frame (FF), and 33x44 (MF, or crop MF, depending on your level of precision – and maybe your

blog.kasson.com

 

[JimKasson]

I shoot with a vary of sensor sizes so I am pretty much certain about what I see. I can't speak for others. Your mileage may vary. At the end of the day, everybody gets to shoot with the Gear they feel best about. Too bad prices are going up though. Mostly.

In the MF world, over the last 10 years, prices have come down. A lot.

 

[Doppler9000]

I shoot with a vary of sensor sizes so I am pretty much certain about what I see. I can't speak for others. Your mileage may vary. At the end of the day, everybody gets to shoot with the Gear they feel best about. Too bad prices are going up though. Mostly.

The 50 MP FSI Hasselblad XID came out over four years ago at $9,000. The 100 MP BSI X2D, which added a very effective IBIS system, and various other improvements, is $8,200.

 

[Erik Kaffehr]

• I know of no "optical bench testing" that exists that measures tonality. Jim can correct me I guess.

You can measure SNR. But unless you're going to put more photons on the MF sensor (abandoning equivalence), you won't see a big SNR difference with same sensor tech.

Yes, of course, but does that measure tonality transitions? Or can we extrapolate from SNR tests anything about transitions?

Hi,

SNR defines the resolution of tonal differences, at least as I understand it. So a system with higher SNR will be able to detect more tonal differences.

Now, talking about tonality transitions, what is the definition definition of it?

Best regards

Erik

 

[BackToNature1]

I shoot with a vary of sensor sizes so I am pretty much certain about what I see. I can't speak for others. Your mileage may vary. At the end of the day, everybody gets to shoot with the Gear they feel best about. Too bad prices are going up though. Mostly.

In the MF world, over the last 10 years, prices have come down. A lot.

I have noticed that at least for Fuji, that has certainly been the case. I don't know however it that's the same for other MF companies.

 

[BackToNature1]

I shoot with a vary of sensor sizes so I am pretty much certain about what I see. I can't speak for others. Your mileage may vary. At the end of the day, everybody gets to shoot with the Gear they feel best about. Too bad prices are going up though. Mostly.

The 50 MP FSI Hasselblad XID came out over four years ago at $9,000. The 100 MP BSI X2D, which added a very effective IBIS system, and various other improvements, is $8,200.

I should have mentioned concerning other Sensor sizes, not MF in my first post. I have closely followed the prices of the Fuji latest models for MF an have seen those prices decreases. I have not followed MF prices for other brands.

 

[just Tony]

So why do so many get this wrong? Because many photographer's idea of equivalence is based on maintaining equivalent composition and therefore changing the camera position or focal lengths between shots.

I don’t know of a way to maintain a composition if I am changing the camera position, in the general case of subjects that are not planar like charts and documents.

 

[JimKasson]

So why do so many get this wrong? Because many photographer's idea of equivalence is based on maintaining equivalent composition and therefore changing the camera position or focal lengths between shots.

I don’t know of a way to maintain a composition if I am changing the camera position, in the general case of subjects that are not planar like charts and documents.

Don't change the camera position when you're testing equivalence.

 

[JimKasson]

I shoot with a vary of sensor sizes so I am pretty much certain about what I see. I can't speak for others. Your mileage may vary. At the end of the day, everybody gets to shoot with the Gear they feel best about. Too bad prices are going up though. Mostly.

In the MF world, over the last 10 years, prices have come down. A lot.

I have noticed that at least for Fuji, that has certainly been the case. I don't know however it that's the same for other MF companies.

Hasselblad, too. And Pentax broke price barriers a while back.

 

[geo444]

.
Some testing shots with the GFx 100s + GF 32-64, very close to OoC Jpegs,
- just downsized plus a bit of USM - both with IrfanView.
Downsized to that point because Unreliable WiFi in the hotel where I was,
- it took me 80 minutes to upload the 6 first pictures...
those from the Ferris Wheel down to the Opera :
https://www.dpreview.com/forums/post/65838551
- Please, enlarge them up to 200% !

With a FF ILC, many of them would show an aspect ratio closer to 24x32,
so, should I have shot them with a rare ~22~23 mm wide angle lens ?
https://www.dpreview.com/galleries/7467909648/albums/marseille-by-night
- a 100% crop from the Original Jpeg is inserted in each picture.



--
http://mecastronics.free.fr/
https://en.astronomike.net/m/6809.html
https://en.astronomike.net/m/134251.html
https://www.dpreview.com/galleries/7467909648
https://www.dpreview.com/members/7467909648/challenges

 

[DavidMillier]

There is no debate about the facts that pixel size, pixel count make a difference. A 100MP MF sensor looks different to a 20MP m4/3 because of resolution difference alone.

However, let's imagine we had gigapixel sensor tech of excellent pixel level quality.

Would shooting a m4/3, APS-C, full frame or MF possess an intrinsic look then? Obviously lens quality, depth of field and angles of view would all need to be equalised, but if somehow this could be done, would there be a specific look down to sensor size alone?

No is the answer.

To a first approximation.

But there are higher-order effects:

https://blog.kasson.com/the-last-word/format-size-and-image-quality/

The flavour of the "medium format look" claim is pretty bold, though. We are not debating some infinitesimally small esoteric difference that exists only under extreme test conditions. At least I'm not. My umbrage is with the basic argument is that there is something uniquely special about medium format systems that makes it obviously look superior from other formats. And this difference isn't just half a stop of dynamic range or extra resolution. It's something caused by the size of the sensor, not the tech of the sensor.

I understand this argument when applied to 6x7 film portraiture where the superior sharpness, detail, fine grain and tonal smoothness compared to 35mm film when enlarged to the same print size is usually obvious. Or even the mid distance DoF fall off of the 105mm f/2.4, an effect people weren't used to from 35mm equipment, presumably because equivalent lenses weren't commonly in use.

But the argument applied to GFX doesn't make a lot of sense. The sensor is barely larger than full frame and it's the same tech. As I said to Greg, shoot in FF crop mode and if you still think the 4:3 image looks better, the magic clearly resides in the pixels of the 8mm strip around the periphery of the sensor.

I have started testing my GFX50s; so far the files look similar to my smaller cameras, just bigger. Which is hardly a surprise given that m4/3 files look similar to 1" files, just bigger and Aps-c files look similar to m4/3 files, just bigger etc etc. It's a continuum, there is no discontinuity when it gets to 44x33.

--
Website: http://www.whisperingcat.co.uk/ (2022 - website rebuilt, updated and back in action)
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)