[Coiseam]

FF getting cheaper is nice but I think it's being hobbled by two factors - the lenses are generally very expensive (especially in the case of Canon) and Canon's crappy sensors that they insist on churning out and putting in cameras which otherwise might be fairly good.

I'll take an APS-C camera with a good sensor in it over a FF with a crappy sensor any day. FF with crappy sensor means it's only real benefit is lower DoF (if you like that), no extra dynamic range.

I don't have any personal experience with Canon sensors, but from what I've read it seems that their crappiness is rather overestimated. I mean that can be a point if you need to do +4EV adjustment, but if you ever need to, that means that something's gone terribly wrong before. And - at least in my case - that happens not that often And as a free bonus to Canon sensors you get pleasing Canon colours.

What for the price - Canon's new RF full frame 35 1.8 is not much more expensive than Fuji's APS-C analogs. Fuji's APS-C 50-140 2.8 is slightly cheaper than Pentax full frame 70-200 2.8... May be it makes sense to compare exect lense types from different manufacturers.

P. S. About sensors - I own an X-T20 and I've noticed that there highlights get blown much easier than on Pentax K-5 ii s. I use 35 and 23 2 mainly. You too has both Pentax and Fuji gear, have you noticed the same? I just don't know what to blame - lenses or sensor.

---

Last edited by Coiseam; 07-03-2019 at 05:39 AM.

[Jonathan Mac]

I don't have any personal experience with Canon sensors, but from what I've read it seems that their crappiness is rather overestimated. I mean that can be a point if you need to do +4EV adjustment, but if you ever need to, that means that something's gone terribly wrong before. And - at least in my case - that happens not that often And as a free bonus to Canon sensors you get pleasing Canon colours.

What for the price - Canon's new RF full frame 35 1.8 is not much more expensive than Fuji's APS-C analogs. Fuji's APS-C 50-140 2.8 is slightly cheaper than Pentax full frame 70-200 2.8... May be it makes sense to compare exect lense types from different manufacturers.

P. S. About sensors - I own an X-T20 and I've noticed that there highlights get blown much easier than on Pentax K-5 ii s. I use 35 and 23 2 mainly. You too has both Pentax and Fuji gear, have you noticed the same? I just don't know what to blame - lenses or sensor.

I haven't found blown highlights to be more common on the X-T20, but what I have found is that it tends to under-expose by 1/3 to 2/3 of a stop, sometimes more (especially with adapted lenses), so this was maybe done to protect highlights. The K200D did the same. My Fujifilm X-M1 exposed perfectly almost always, without any exposure comp being needed when using either native or adapted lenses, so I find it very strange that the X-T20, which is two generations newer than the X-M1, should suffer from this problem.

I suspect that the X-T20 and K-3 sensors are the same other than the X-Trans array on the Fujifilm one. Likewise for 16MP sensor in the K-5 and X-M1. So in either case, a similarity in responses to highlights and shadow would not surprise me.

[Winder]

I think Canon's crappy sensors are overstated. They have a bit less dynamic range at base iso, but still have less shadow noise throughout the iso range and they match any of Sony's sensors once you get to iso 400 and above with regard to dynamic range too.

To me, the benefits of full frame sensors are that you get better dynamic range in high iso settings (along with less noise) compared to smaller sensor size. If you are only shooting iso 80 or 100 you will do as well with a K5 IIs (except for resolution, if you need that).

One of the consistent comments you hear from people who switch from Canon to Sony is that the image have more detail which doesn't make sense when you consider the Canon 5DIV (what most of them were using) has 30MP while the A7III has only 24. Julia (first video)is a long time Canon shooter who just recently switched over to Sony.

You Tube

You Tube

You Tube

You Tube

I won't link all the different videos, but it seem pretty consistent. As Sean Tucker points out while comparing portraits, Canon is smoothing skin (NR) even in the RAW files. This is why their 30MP images have less detail than the 24MP images. Maybe Canon is using AI to only apply this to faces in RAW files, or maybe they are applying it across the entire image, I don't know, but it's pretty obvious when you look at Canon 5DIV images next to the Sony A7III images.

[Rondec]

One of the consistent comments you hear from people who switch from Canon to Sony is that the image have more detail which doesn't make sense when you consider the Canon 5DIV (what most of them were using) has 30MP while the A7III has only 24. Julia (first video)is a long time Canon shooter who just recently switched over to Sony.

Zeiss FE 35mm f1.4 + Sony A7III Natural Light Photoshoot BTS - YouTube

Skin Tones in Portraits: Canon vs Sony and How to Correct - YouTube

I won't link all the different videos, but it seem pretty consistent. As Sean Tucker points out while comparing portraits, Canon is smoothing skin (NR) even in the RAW files. This is why their 30MP images have less detail than the 24MP images. Maybe Canon is using AI to only apply this to faces in RAW files, or maybe they are applying it across the entire image, I don't know, but it's pretty obvious when you look at Canon 5DIV images next to the Sony A7III images.

Less sharpening out of camera? Different AA filters? Tough to say, isn't it?

[photoptimist]

One of the consistent comments you hear from people who switch from Canon to Sony is that the image have more detail which doesn't make sense when you consider the Canon 5DIV (what most of them were using) has 30MP while the A7III has only 24. Julia (first video)is a long time Canon shooter who just recently switched over to Sony.

Zeiss FE 35mm f1.4 + Sony A7III Natural Light Photoshoot BTS - YouTube

Skin Tones in Portraits: Canon vs Sony and How to Correct - YouTube

I won't link all the different videos, but it seem pretty consistent. As Sean Tucker points out while comparing portraits, Canon is smoothing skin (NR) even in the RAW files. This is why their 30MP images have less detail than the 24MP images. Maybe Canon is using AI to only apply this to faces in RAW files, or maybe they are applying it across the entire image, I don't know, but it's pretty obvious when you look at Canon 5DIV images next to the Sony A7III images.

Canon's "smoothing" is caused by the optical low pass filter (OLPF) in the camera that is required to avoid artifacts with high detail scenes shot with high-quality lenses. It seems like the Canon 5DIV has the filter and the Sony A7III does not.

There's no right or wrong design decision for OLPFs because both choices "suck" under different circumstances of subject matter and lens choice. Canon's use of OLPF is the safer choice but it results in 100% of the images looking "softer" even in RAW (correctable in post). Sony's lack of OLPF creates sharper images at first glance but can create nasty artifacts under some conditions. In this regard, Pentax anti-aliasing simulator is a superior feature which is why Pentax is so much better than Canon or Sony.

[Coiseam]

I won't link all the different videos, but it seem pretty consistent. As Sean Tucker points out while comparing portraits, Canon is smoothing skin (NR) even in the RAW files. This is why their 30MP images have less detail than the 24MP images. Maybe Canon is using AI to only apply this to faces in RAW files, or maybe they are applying it across the entire image, I don't know, but it's pretty obvious when you look at Canon 5DIV images next to the Sony A7III images.

This link leads to one Russian guy's blog, in his post he compares pics taken by Canon 5D mark IV and Sony A7RIII, and his point is that Sony's RAWs somehow lack tonal range in middle tones and highlights and because of this Sony's pics look "cheap". Canon has smooth transition of colours, Sony doesn't. He writes also that it's a problem on newer Sonies, 1st gen A7's didn't have that. It seems to me that can be seen on Sean Tucker's video, too.

[Winder]

Canon's "smoothing" is caused by the optical low pass filter (OLPF) in the camera that is required to avoid artifacts with high detail scenes shot with high-quality lenses. It seems like the Canon 5DIV has the filter and the Sony A7III does not.

They both have AA filters. The 30MP Canon should have a weaker on than the 24MP Sony. I'm not sure you really need one on the 30MP sensor. Nikon and Pentax took it off for the D810/D850 and the K-1. The Sony A9 doesn't have an AA filter and its 24MP. Regardless, the AA filter doesn't explain it.

---------- Post added 07-04-19 at 03:28 PM ----------

This link leads to one Russian guy's blog, in his post he compares pics taken by Canon 5D mark IV and Sony A7RIII, and his point is that Sony's RAWs somehow lack tonal range in middle tones and highlights and because of this Sony's pics look "cheap". Canon has smooth transition of colours, Sony doesn't. He writes also that it's a problem on newer Sonies, 1st gen A7's didn't have that. It seems to me that can be seen on Sean Tucker's video, too.

I can't read Russian, so I can't really comment on it.

---------- Post added 07-04-19 at 03:30 PM ----------

Less sharpening out of camera? Different AA filters? Tough to say, isn't it?

We are talking about RAW files so the sharpening shouldn't be an issue. The Canon should have a weaker AA filter than the Sony. As you go up in MPs the AA filter isn't needed. The K-1 doesn't have one and its not much more MPs than the Canon's 30MP.

[Rondec]

They both have AA filters. The 30MP Canon should have a weaker on than the 24MP Sony. I'm not sure you really need one on the 30MP sensor. Nikon and Pentax took it off for the D810/D850 and the K-1. The Sony A9 doesn't have an AA filter and its 24MP. Regardless, the AA filter doesn't explain it.

---------- Post added 07-04-19 at 03:28 PM ----------



I can't read Russian, so I can't really comment on it.

---------- Post added 07-04-19 at 03:30 PM ----------



We are talking about RAW files so the sharpening shouldn't be an issue. The Canon should have a weaker AA filter than the Sony. As you go up in MPs the AA filter isn't needed. The K-1 doesn't have one and its not much more MPs than the Canon's 30MP.

I understand that you are wholly in the Sony camp at this point, but I have seen enough to know that Sony does massage their RAW files as much (or more) than anyone out there -- both in terms of fixing lens issues and in terms of adjusting for sensor deficiencies. In the case of Sony, RAW isn't really RAW.

(And no, I don't own a tin foil hat)...

[Winder]

I understand that you are wholly in the Sony camp at this point, but I have seen enough to know that Sony does massage their RAW files as much (or more) than anyone out there -- both in terms of fixing lens issues and in terms of adjusting for sensor deficiencies. In the case of Sony, RAW isn't really RAW.

(And no, I don't own a tin foil hat)...

I'm not in any camp. I would be just as happy shooting Panasonic, Olympus, Fuji, Pentax, or whom ever if they delivered the results that I want. Currently, I shoot Sony. I'm not a fan of the bodies, but I can't argue with the results I get. I've played with the Panasonic S1R twice now and I really want to like it, but Panasonic isn't really there yet. Give them 3 years and who knows. I really like a lot about the S1R body and controls. Panasonic just needs the glass and the AF. Panasonic IQ is really good.

But what I'm currently shooting really doesn't have anything to do with Canon sensor performance or the softness the files have.

[Coiseam]

I can't read Russian, so I can't really comment on it.

Sorry, he doesn't write in English... I thought that pictures in that post speak for themselves and gave a rather brief summary of what he was writing about.

but I have seen enough to know that Sony does massage their RAW files as much (or more) than anyone out there -- both in terms of fixing lens issues and in terms of adjusting for sensor deficiencies.

At the end of the post that guy voiced the same suspicions - Sony somehow compute colours in their RAWs in a way that a pixel is a kind of average of nearby ones. And it's not about colour science. And, for example, A7R (1st gen) didn't have that feature. Perhaps that helps with noise reduction. Canon pics look more noisy, Sony's are more clean but luck tonal range. I would prefer to keep tonal range, it is always possible to kill it and denoise an image at post processing, whereas to get tonal range is impossible.

[Winder]

Sorry, he doesn't write in English... I thought that pictures in that post speak for themselves and gave a rather brief summary of what he was writing about.

I have not seen any other source indicating that Sony lacks tonal range compared to Canon. Are there some other sources showing this same issue? Was he shooting in compressed RAW? There are several variable and without being able to read Russian its had for me to comment one way or the other. It may very well be true. I have read quite a few comparisons and this has not come up.

[Coiseam]

I have not seen any other source indicating that Sony lacks tonal range compared to Canon. Are there some other sources showing this same issue? Was he shooting in compressed RAW? There are several variable and without being able to read Russian its had for me to comment one way or the other. It may very well be true. I have read quite a few comparisons and this has not come up.

His point was that whereas Canon has smooth tonal transition, Sony has a kind of a border between 2 different colours. Kind of that. Here is another his post where he compares Sony A7ii and Sigma Quattro H (look at the lips - there is that smoothness in Sigma case, though lenses used are different - Sigma Art vs Samyang). Pictures here are his own.

About that guy - as far as I understand he has owned and / or tested Canons 5D something and 5DSR (his current camera), Pentaxes 645D and K1 (liked both, used to own 645D), Sony A7ii (sold because pics looked "cheap" for him) and Sigma Quattro H (tested, mixed feelings).

In the first quoted post pics are from Internet. In comments some Sony owner wrote the same - that perhaps they were shot with compressed RAW and that was the point.

Really sorry for Russian - just it's the only source where I've read about that issue in any language.

[reh321]

Sure.

Canon will be fine. If nothing else, they are the biggest camera brand and have a lot of other businesses to help support cameras if photography gear takes a down turn for a bit. I'm just pointing out that there is a general fallacy that it is enough just to drop a MILC on the market and your brand will gain market share.

Canon doesn't think the same as other companies - they are much more interested in retaining market share.

[MarkJerling]

His point was that whereas Canon has smooth tonal transition, Sony has a kind of a border between 2 different colours. Kind of that. Here is another his post where he compares Sony A7ii and Sigma Quattro H (look at the lips - there is that smoothness in Sigma case, though lenses used are different - Sigma Art vs Samyang). Pictures here are his own.

About that guy - as far as I understand he has owned and / or tested Canons 5D something and 5DSR (his current camera), Pentaxes 645D and K1 (liked both, used to own 645D), Sony A7ii (sold because pics looked "cheap" for him) and Sigma Quattro H (tested, mixed feelings).

In the first quoted post pics are from Internet. In comments some Sony owner wrote the same - that perhaps they were shot with compressed RAW and that was the point.

Really sorry for Russian - just it's the only source where I've read about that issue in any language.

Google translate seems to do an OK job with translating Russian. There's a huge difference in quality between those two shots.

[Coiseam]

I haven't found blown highlights to be more common on the X-T20, but what I have found is that it tends to under-expose by 1/3 to 2/3 of a stop, sometimes more (especially with adapted lenses), so this was maybe done to protect highlights. The K200D did the same. My Fujifilm X-M1 exposed perfectly almost always, without any exposure comp being needed when using either native or adapted lenses, so I find it very strange that the X-T20, which is two generations newer than the X-M1, should suffer from this problem.

I suspect that the X-T20 and K-3 sensors are the same other than the X-Trans array on the Fujifilm one. Likewise for 16MP sensor in the K-5 and X-M1. So in either case, a similarity in responses to highlights and shadow would not surprise me.

Thanks for feedback! I gave a Fuji camera to a friend till the end of the summer, but when it's back I can make a little test with my Pentax and Fuji gear using similar lenses (K-mount via adapter) and camera settings.