FF vs APS-C camera - worth trying a 90D? (1 Viewer)

[Nick Leech]

Many people seem to own a FF camera for general use, and an APS-C camera for wildlife shooting, where the greater "pixels on the bird" should give some IQ advantage.

I have a 5D4 (FF) and used to also have a 7D2 (APS-C). I once did some detailed comparison shots using a 600mm lens on both bodies, cropping heavily into both images and comparing IQ on crops at the same field of view. I expected to see the 7D2 give an advantage, but was surprised to find the more heavily cropped 5D4 images gave slightly better IQ. I decided that this was due to the 5D4 having a 30MP sensor, compared to the 20MP sensor in the 7D2. Maybe also, the newer design of sensor on the 5D4 was giving additional IQ advantages?

Consequently I sold my 7D2 as I could not see any "reach advantage" from the crop sensor camera.

Now a year or so on the 90D is available with a 32.5MP sensor.

I am wondering whether, if I bought a 90D and repeated my test against my 5D4, I may now find the 90D giving slightly better IQ than the 5D4 on heavily cropped bird photos?

What do people think?

I'd especially like to hear from people who have a 5D4 and a 90D and have done such comparisons on cropped bird photos!

Thanks.

 

[Chosun Juan]

Interesting question. I don't own those Canon's but know someone with a 7D II and they are pretty much near suicidal when comparing that sensor to the one in the Nikon D7200.

Apart from the actual resolutions, sensor technology plays a big part in it, and I think this is what is accounting for the bulk of the differences you are seeing between the 5D IV and the 7D II.

The resolution of the 5D IV is 30.1MP with a 5.4um pixel pitch.
When cropping to the same size equivalent focal length multiplier as the APS-C cameras (1.6x) this becomes 11.8MP. Compare that to:
7D II 20MP 4.1um pixel pitch,
90D 32.3MP 3.2um pixel pitch.

The sensor in the 90D seems far better than the one in the 7D II. It is pretty comparable with other brands on the market (It seems that the Olympus OM-D EM-1 II offers the best performance considering crop factor of all the sensors on the market as the photonstophotos.net photographic dynamic range charts show).

This 90D sensor will form the basis of Canon's next 83MP high resolution full frame offering in a year or two - simply directly scaled to full frame sensor size. As such the 90D would need high resolution technique and quality lenses to match (L glass).

Definitely throwing the kitchen sink at it - sturdy tripod, mirror up, remote or timed release, cooperative subjects etc would help. I have seen some handheld stuff with a Sony 61MP a7R IV (which would be ~24MP Canon APS-C equivalent crop) which was pretty good, so likely there is some latitude for walk around fun with the Canon 90D and say the 100-400 II, the DO 400 II + 1.4×TC, or the 300 f2.8 II + 2×TC, etc.

Given those things I would expect that you would see better IQ with the 90D when cropping the subject to the same size (ie. focal length equivalent - not pixel size).

It would come down to what lenses you will put on it, whether the focal length is suitable, shooting techniques, and light levels suitable for the smaller pixel pitch (say ISO 200 to 3200 .... don't expect miracles outside of that).

NB: I have read that the 90D's through the viewfinder AF tracking isn't as good as the 7D II's or likely the 5D IV's for that matter). Using just centre point (or expanded) rather than iTTR is likely to close that gap, and shooting in Live View definitely does - that could even be comparable to the 5D IV - you are down to 7fps in that mode though. You would also want to capture in lossless RAW at the highest bit rate and process fully from there, as the brief look at samples I had exhibited aggressive jpeg noise reduction which sacrificed detail.

A few things to weigh up. :cat:

Certainly worth a go I reckon if you are mainly doing reach limited photography. Should tide you over a few years at minimal expense until the hi-res R model lands ...... then that would be an all-in-one camera ......





Chosun :gh:

 

[Chosun Juan]

Canon's 32MP Sensor

Interesting assessment of Canon's 32.3MP Sensor (as found in the 90D) .....
https://m.dpreview.com/articles/8673278060/canon-s-32mp-chip-marks-the-end-of-the-24mp-aps-c-era





Chosun :gh:

 

[Chosun Juan]

Here's something unexpected - a DSLR 5D V is in the works ! :eek!:
It seems the 'R' mount may be more of an each way bet than a full steam ahead transition ..... :cat:
https://www.canonrumors.com/forum/i...anon-eos-5d-mark-v-is-in-the-works-cr2.37724/




Chosun :gh:

 

[marcsantacurz]

Nick,

There have been a lot of sensor improvements since 2015, when the 7dII came out. Things like back side illumination (i'm not sure if canon has this in production yet), readout speed, noise performance, etc.

There's really nothing different between a crop sensor and a full frame sensor except the size, so if one uses the same technology and just a smaller area, one would expect similar IQ. Usually, however, crop sensors use different technology to get the MP rating up.

I think it would be worth giving the 90d a try, especially if you can borrow or rent one.

As Chosun pointed out, the main thing you fight in crop is a smaller pixel pitch to get a similar MP resolution as a full frame. Smaller pitch will usually result in less DoF, more sensitivity to diffraction, more sensitivity to shake, and worse noise performance. usually, not always, as some technologies like backside illumination actually increase the available surface area by 30% or so (if I remember correctly).

I used a Nikon d500 for a while, then switched to the full frame d850. If I crop down 1.5x, I still liked the d850 output better. But it's close. I wouldn't be surprised if the 90d is close in quality to a cropped 5dIV.

The 5DIV is a DIGIC 6+ and the 90d is a DIGIC 8. The 8 is used in the R and RP and the M6. I'm sure they have different AF systems too.

Marc

 

[Pauhana]

Think i would rent before I would buy. I don't know why but for some reason the 90D just has no appeal to me. But that's just me.



Randy

 

[neocacher]

I went with the 90D (32 megapixels) + 100-400mm f5.6 is II USM. I still need to get a 1.4x MIII extender. If money were no object I would get a 5DS R (50 megapixels) + 300mm f/2.8 + 1.4x + 2x= 840mm and crop when necessary or if used on a 90D that would give you 1344mm. I think megapixels are king if you have to crop some anyways.

 

[MJN]

I went with the 90D (32 megapixels) + 100-400mm f5.6 is II USM. I still need to get a 1.4x MIII extender. If money were no object I would get a 5DS R (50 megapixels) + 300mm f/2.8 + 1.4x + 2x= 840mm and crop when necessary or if used on a 90D that would give you 1344mm. I think megapixels are king if you have to crop some anyways.

Well you can't physically piggyback Canon convertors.
I'm not too sure about third party products but even if you could manage it, the hits to both IQ and af would completely rule it out for me.

 

[Chosun Juan]

I went with the 90D (32 megapixels) + 100-400mm f5.6 is II USM. I still need to get a 1.4x MIII extender. If money were no object I would get a 5DS R (50 megapixels) + 300mm f/2.8 + 1.4x + 2x= 840mm and crop when necessary or if used on a 90D that would give you 1344mm. I think megapixels are king if you have to crop some anyways.

The IQ of the 300 f2.8L IS II holds up quite well with the 2XTC III on it. On the 90D this would give you 960mm equivalent @f5.6 with 32.3MP ...... crop in post from there. Why would you want to further degrade the image by putting another TC on it? (also bumping up the weight of the lenses to 2.947kg ~ 6&1/2lb in the process).

If you want more reach on FF you would be better off starting with the 400 f2.8 + 2xTC for 800mm @f5.6 (and 3.165 kg ~ 7lb), or the 600 f4 + 1.4xTC for 800mm @f5.6 (and 3.272kg ~ 7&1/5lb). These would only be 218grams and 325grams heavier respectively and you would still get full AF with both. On the 50.3MP 5DS R if you cropped in post to 960mm equivalent you would still have 34.9MP and 38.5MP respectively. You would also only have 5fps (though at full AF) and a significantly lighter wallet !

Of course you could also rock either of those supertelephoto lens setups on the 90D for 1280mm equivalent and 1344mm equivalent respectively with 32.3 APS-C MP and full 10fps AF. (If you were to crop the 50.3MP 5DS R to those equivalent focal lengths you would be at 19.6MP)

I suppose the ultimate question is how does the quality of those MP's compare? which gets back to what the OP was asking - does the sheer number of MP reveal more detail?




Chosun :gh:

 

[MJN]

I have all the equipment mentioned by Neocacher apart from the 90D and I can tell you that it's just not a sensible thing to do.
The af on the 5DSR is pretty basic and for BIF (this is bird forum so presumably that would be Neocachers intention)For BIF you would not use that set up with one convertor never mind two.

If you want extreme mm's you might consider the 500mm f4 which will take a 2X to give 1000mm and can produce acceptable results in controlled circumstances.
I used the 500 with a 5DSR +1.4 X almost exclusively on a trip last year with results that I was happy with but I could only use that combo on static subjects.

It's also worth noting that the 5DSR is severely restricted with regard to ISO capability, I don't like to go over iso 800 or 1250 max because excessive noise creeps in. Adding convertors to the mix may create more problems than it solves because you won't have much wriggle room.

Quote "The IQ of the 300 f2.8L IS II holds up quite well with the 2XTC III on it. On the 90D this would give you 960mm equivalent @f5.6 with 32.3MP"

It holds up very well indeed, particularly when stopped to f8.

 

[stevo]

Well you can't physically piggyback Canon convertors.
I'm not too sure about third party products but even if you could manage it, the hits to both IQ and af would completely rule it out for me.

mk2 canon extenders you can piggyback them,mk3's you can't.


Steve.

 

[MJN]

:t:

 

[Nick Leech]

As per my original post, have any 5D4 owners bought a 90D and compared IQ on equivalent images with the same lens?

 

[Nick Leech]

I have now bought a 90D and have done many tests vs my 5D4 using the same subject, lighting, lens and settings.

Also, I have repeated the tests using 4 different lenses (3 "L" Canon lenses plus a Tamron lens).

In comparing 90D vs 5D4 I cropped both images to give the same effective field of view (ie greater cropping on the 5D4 images).

I am finding that the 90D is consistently giving me slightly more detail in the image than the 5D4, due to having more "pixels per duck".

The 90D is not perfect - the AF system is not as good at tracking Birds In Flight as the 5D4 (or 7D2). However, I am finding it quite usable. People who specialise in BIF (especially small birds), may be less happy with the 90D. Personally, I shoot static or slow moving birds more often than BIF, so I will be mainly using the 90D for wildlife from now on (whilst keeping the 5D4 for other types of photography).

Mind you - if and when a 7D3 appears - I will probably switch from the 90D to that.........

Still interested to hear from others who have been able to make a direct comparison between 90D and 5D4. Also, any comparions between 90D and 7D2.

 

[h14nha]

On other forums people ( birders ) who I respect greatly, are reporting the 90D's AF system through the view finder to be erratic and unreliable. The live view they say is superb, and, because of this they are using Lazer Red Dot Finders with great results. But, that means shooting the great white super teles at arms length via live view, not for me, but great IQ can be achieved.

https://www.amazon.com/Omegon-Finder-DSLR-Flash-Shoe/dp/B01N97R9DG

This is what they're using.