Hello David,
For what it's worth, we had a good presentation on the 90D at the Fremantle conference by the Canon WA rep who firmly stated that this is not a replacement for the 7DII but a replacement for the 80D. Some members had loan cameras in the field and those I talked with were delighted with the the overall feel and performance of the 90D. I suspect you will not know the answers to your questions until you get your hands on a 90D and compare it with the AF performance of the 80D on birds in flight. I hope one of our experienced members will soon do just that and give us their impressions.
Regards, Ian

Thanks for that, Ian. I wish I'd been there - the first question I would have asked the Canon rep is that if that's the case, what upcoming camera model do Canon have in mind to replace the popular 7DII? I find the present advice from Canon (both here and in the US) about what the 90D represents a little odd, given that up till recently it seems to have been represented as a model which was anticipated to cover the expectations of both current 80D and 7DII owners - and I distinctly remember reading that this included previous indications from within Canon itself.

My personal interest in the 90D has been within this context, i.e. would it be a better camera for bird photography (including BIF) than the 7DII, particularly when used in combination with the EF 100-400 L II zoom sans TC? Further reviews and impressions from users may throw some light on this, but the DPReview assessment was not encouraging in regard to BIF use. So it still leaves that question about the mystical 7DIII (or its mirrorless equivalent).

Cheers, David

I'm not so worried about whether this is technically a replacement for camera X or Y. For me it is a question of whether this is a worthwhile upgrade (for about $1900). At this stage I am holding fire on this as I suspect that there will be downwards adjustment in prices in the not too distant future (although web sites seem to be saying that there is limited stock of the 90D so perhaps Canon is managing supply at this stage to keep prices up). Also my current pp software of choice does not yet handle the files.

The 80D is a fine camera and has done me very well for bird photography over the last 3 years or so. It has proven reliable and robust, getting lugged around every where including international trips in relatively rough environments. I have never had a problem with the overall image quality for my purposes. I have never even had a significant dust on sensor problem as essentially I use this as a fixed lens camera with the 100-400 lens pretty much permanently attached. It has been dropped onto a hard surface (though only with a light lens attached at that time!). It has plenty of life left in it. The main problem I have with focus is not so much tracking BIF in servo mode, but that the focus points become 'confused' relatively easily where backgrounds are relatively messy or the light is not strong, resulting in focus hunting. This is to some extent inherent in the ways DSLRs focus, but it would be nice to have a camera which was a bit more sensitive in this regard particularly in the centre focus point. I am not sure if Live View would be any better in this respect as I don't use it, although Live View is much better at things like face tracking or subject tracking based on colour (and this aspect seems to be given a lot of weight in the DP Review assessment). Equally I have never used the 7D Mk 2 so don't know if its viewfinder focus is better in relation to its sensitivity and 'lock on' ability in situations where there is a lack of contrast.

I have no doubt that the 90D would be a good overall camera for bird photography, given my experience with the 80D. It seems to be improved in worthwhile ways. The extra megapixels would be worthwhile although the linear (x and y axis) resolution increase is only about 15% over 24 megapixels. The new DIGIC processor seems to cope well with the increased data sampling, and the buffer levels are not really a problem for me as I tend to shoot small groups of fast shots and then refocus and recompose. The focusing through the viewfinder is at least no worse than the 80D, as far as reports go.

If Canon produces a genuine 7D Mk 2 replacement at some stage, what advantages would it offer? One might anticipate more robust chassis (incorporating alloy), more robust shutter with perhaps slightly improved burst speed, more focus points in the viewfinder, possibly a larger buffer, possibly slightly higher magnification viewfinder etc. But equally, I imagine it will be $1000 more than 90D. Perhaps also Canon might wind back the megapixels to increase shooting rates, as per the Nikon equivalent. Will any of these make it a better general purpose bird photography camera? The jury is out, for me.

I would not be surprised if Canon were to produce a 'pro' level APSC camera as a mirrorless rather than DSLR, perhaps with the R mount. They seem to be investing a lot of resources in the Live View experience, including around focusing, and the 90D seems to be in some ways a transition camera designed more around its Live View experience than DSLR experience. Canon faces the classic market leader problem, how do you stay at the front while keeping your vast numbers of satisfied customers happy, some of whom rely on your cameras to make a living. Sony chose to just dump their DSLR customers when they went mirrorless, but they could do this as there weren't many of them. Clearly Canon cannot do this, but it is difficult to see how much innovation they can (and would want to) bring to the traditional DSLR which is already well refined.

Cheers
Simon

Thanks Simon for posting your detailed thoughts. I guess my contributions to this thread could now be considered somewhat off-topic, since Canon started confirming worldwide that they do not consider the 90D to be a replacement for the 7DII. FWIW, In reponse to your thoughts on a genuine 7DII replacement, I think highest-priority improvements in a 7DIII might include:

Sensor similar to the 90D, same pixel count, perhaps lower noise characteristics;
Dual Digic 8 processors;
Perhaps more focus points in the viewfinder;
Better and more stable viewfinder AF in servo mode than the 7DII, capable of reliably tracking rapidly approaching (or approaching and cross-tracking) birds in flight.

I agree that such a camera might cost $1000 more than the 90D, but that may be about what you would expect if the 7D series line was both kept going and improved. Looking at international forums, there would certainly be no shortage of buyers. If a 7DIII does not appear, then apart from the upcoming 1Dx III the only other future high-performance camera on Canon's present roadmap is the R series one predicted to appear in 2021 - time will tell if this proves to be a mirrorless 7DIII equivalent, or a migration of the 1Dx series into mirrorless, but it's a fair way off.

In regard to DSLR focusing systems, I think Nikon has proven that there is more scope for development there than Canon has managed to get into most of its DSLR models. Given Canon's resources I've wondered about this, but it is possible that Nikon's patents may have proven a constraint on Canon in that area (it can sometimes be the case in engineering that the first with a new design idea tends to retain the front running).

Incidentally, just one thought on the hunting during focus tracking you mention with the 80D - I don't know whether the 80D gives you access to adjusting the AF parameters for scenario-specific situations, but I think the 90D has some such capability. The 7DII in my experience certainly needs parameter adjustment, particularly the Tracking Sensitivity parameter, which needs to be pulled back to -1 or -2 to suppress that tendency to lock on to unwanted detail in the background while tracking. In this case you are actually reducing the sensitivity of the AF to cut the system's tendency to twitchiness (an appropriate term for bird photography?).

Cheers, David

PS I'd better promise not to harp on further about the 7D series in this thread!

Some interesting throughts from Gordon Laing at Cameralabs.com in relation to the autofocus (cut and paste below):

"When shooting action, the 90D is a tale of two autofocus systems with quite different capabilities..... When it came to photographing Brighton’s seagulls using the EF 70-200mm f2.8L mostly at 200mm f2.8, I found the 90D’s viewfinder AF system proved quite hit and miss. Sometimes it’d deliver crisp results, but at others just seemed overwhelmed by the speed of the birds. I repeated the tests with different AF areas and settings across several days, but found the 90D’s viewfinder AF system just wasn’t well-suited to birds in flight. To be fair, it’s no worse than the EOS 80D, but no improvement on it either in this regard. Switching to live view on the 90D delivered far more confident performance for autofocus, but it’s hard to aim a DSLR with a long lens when using the screen alone – in fact it’s almost impossible to photograph birds in flight without a viewfinder, and that’s the ironic thing about the 90D. It does have a viewfinder, but its best autofocus is during live view when you can only use the screen....At this point I need to mention the M6 Mark II which shares the same sensor, but as a mirrorless camera is always in live view allowing you to use the screen or viewfinder with the same Dual Pixel AF system. The M6 II also shoots faster than the 90D at 14fps, while maintaining confident autofocus, and offers a RAW burst mode which captures 30fps bursts with a half second pre-buffer, a feature that’s strangely absent on the 90D’s live view menu. So in a surprise move, the M6 II is actually more compelling than the 90D for sports, action and birds in flight, so long as you fit its optional viewfinder accessory. But then in an even stranger twist of fate, the 90D becomes preferable to the M6 II for video and even vlogging...."

His conclusion: "....crucially, the 90D’s 45-point viewfinder AF system is not a patch on the 63-point system on the 7D II, in my tests delivering fairly lacklustre performance when photographing birds in flight with the viewfinder. Switch to live view and the 90D’s autofocus becomes much better, but the speed falls to 7fps and it’s almost impossible to aim a long lens for fast action using the screen alone. This is something the M6 II does much better, allowing you to enjoy its even faster 14fps bursts and best autofocus system with the screen or optional electronic viewfinder. So Canon bird photographers should go for the M6 II over the 90D...."

Make of this what you will!

Simon

Thanks for that Simon. It's interesting to see some real-world testing of the 90D on birds in flight, and it does appear to confirm the results from the DPReview standard moving-subject tests. I think that despite the improvements in other areas that it incorporates, anyone thinking of migrating to the 90D for bird photography may need to take into account this AF limitation.

It does leave a question about where current 80D users in bird photography should look for an upgrade. The 7D II is well due for an upgrade itself, and as I mentioned previously the various Canon full-frame body options leave you feeling a bit short of focal length if your main birding lens is the EF 100-400 mm L II zoom - which tends to compel you to add the 1.4x TC, dropping maximum aperture to f/8 and noticeably slowing the autofocus when servo-tracking birds in flight. A bit of a Catch-22, at least within the Canon world currently.

One thing the 90D appears to demonstrate is that Canon's main advances in AF are being made in Live View AF, which should have application in through-viewfinder AF in the mirrorless R series once Canon optimises that system. If the R series remains full-frame only, there will still be that issue with the 100-400 zoom for birds, although the need for teleconvertors may be avoided once sensors above 30 megapixels are added to the cameras.

Cheers, David