This will be a very long post, taking an in-depth look at two of Canon's best 35mm EOS cameras and how they performed for me personally, under test conditions that matterd most to ME. Keep that in mind. I have specific requirements that you may not have so if you find my review to be way off base, that is probably the reason. If you find yourself having similar demands from your equipment, you MIGHT fnd this lengthy review worth the read. If I was shooting landscapes or other static subjects under ideal light conditions, any EF mount 35mm body would probably do what I need done. For that matter, for static subjects under ideal lighting, I would be content to just stick with my Pentax manual focus bodies or my Mamiya RB67 Pro S.

I've been wanting the EOS 3 for probably better than 5 yeas but I took a very long and convoluted path to get here. It arrived today and I ran a short test roll through it. In my quest to find the perfect 35mm EOS, I also recently purchased an Elan 7S. Since I have both now, I figured that I should share my thoughts on the comparison to maybe help someone else who might be in the same place I was in trying to decide between the two. I'm not too demanding on features and things like that so my primary concern with these cameras has been the AF capability, particularly in low light.

First of all, the Elan 7s arrived about two weeks ago so I've had a bit of time to play with it. My first impression was that although is is much lighter than my 5D kark IV and my 7D mark II, the Elan 7s doesn't feel "cheap" or "plasticky" like some people have said. I did notice that the battery door hinge is just a thin strip of plastic and will certainly eventually break off after years of opening and closing, along with the degradation of the plastic. My copy also had the dreaded "sticky grip," in spite of the camera being in otherwise mint condition. I searched the web and decided to try the fix for this that made the most sense to me. I aggressively scrubbed the grip with a wash cloth soaked with HEET in the yellow bottle. This is nothing more than 99% methyl alcohol and cleaned the grip up perfectly to like new condition.

Performance wise, the Elan 7s reminds me of my old D30, 10D, or 5D mark I. It has just what it needs to be a decent general use camera with nothing that stands out to make the camera spectacular in any way by today's standards. I discovered, as many others have as well, that the ECF feature can be somewhat unreliable. I have probably run the calibration procedure over a dozen times with various lenses, and under various lighting conditions and the camera still chooses the adjacent sensor to the one I'm looking at, at least 25% of the time. I could have gotten the Elan 7N for less that half of what I paid to have the ECF gimmick.

The main reason that I decided to get an EOS film camera is because I was having difficulty focusing indoors at night under normal room lighting with my Pentax MF 35mm cameras. I thought that one of the latest, highest end EOS cameras might get the job done. The Elan 7s didn't impress me under these low light conditions. The lighting that I tested it under was typical to a worst-case scenario that I had phoped the camera could handle. the exposure was 4 seconds at f4 with 100 iso film (EV 2). I turned off the ECF and selected the center focus point and the camera could not find focus anywhere on my daughter's face. I attempted several targets around the room and found only three that it could focus on at all, and then only about 30-40% of the time. The common factor between the three targets is that the were primarily bright areas (white, very light tan, very light grey) with at least a reasonable amount of contrast.

Because the Elan 7s performed so marginally, I decided to step up to the EOS 3, supposedly a magical camera when it comes to its focusing capabilities. In my testing, the ECF was no better than that of the Elan 7s, frequently selecting an adjacent focus point, rather than the one I was looking at, regardless of the fact that I calibrated the camera over a dozen times. The EOS 3 has very similar CF features to the Elan 7s and I didn't find many of them useful to my photography (YMMV). The ultimate test would be the low light focus ability.

Under the exact same lighting conditions, with exactly the same lenses, the EOS 3 BARELY outperformed the Elan 7s. Performance was identical with the exception of only one target, the light tan one. Where the Elan 7s took only one attempt to find focus coming from a close focus lens setting, the EOS 3 took two attempts. Also, the EOS 3 took one attempt to focus coming from infinity and the Elan 7s was unable to achieve focus coming from infinity. I gave the VERY SLIGHT edge to the EOS 3 because it was able to achieve focus on this target probably 80% of the time, while the Elan 7s only got it about 40-50%.

Both cameras focused quickly or racked the lens quickly while looking for focus with the 24-105 f4 L version 1 and the 70-200 f2.8 L IS version 1. Bothe cameras also focused, or racked the lens PAINFULLY slowly withe the 100-400 f 4.5-5.6 L IS version 2. In other words, in low light, f4 lenses will respond quickly enough to make me happy, while the f4.5 lens literally had me wondering at times if the camera was driving the focus at all.

Other impressions of the EOS 3 are that although it is noticeably louder that my digital EOS bodies, and considerably louder than the Elan 7s, I didn't find it offensively loud like others have complained it to be. I would compare it to the likes of my Pentax Super A with the motor drive A installed or my Pentax MX with the Winder MX installed. The EOS 3 also has a strange feel to it. As heavy as it is, I would expect it to feel more solid, like my 5D4 or 7D2, but the plastic shell leaves me wanting. If I have to tolerate a plastic exterior, I want the camera to feel like the Elan 7s, light and compact. If I have to tolerate a lot of weight, I expect a rock solid magnesium exterior like my 5D4 or 7D2. The EOS 3 just doesn't fall on the right side of either extreme for me.

One other complaint about both cameras is that their aggressive winders have both ripped the film from the canister, causing me to fog the final frames of several rolls when I opened the backs. The EOS 3 is the bigger offender, pulling the film from the grip of a piece of Gorilla brand duct tape that I used to secure the film to the spool. I guess that I need to overlap the tape onto the film by a bit more than the half inch that I thought was sufficient.

My final thoughts are that they are both acceptable, and in my experience, nearly identical performers. Each has pros and cons and if I didn't have a hoarder's personality, I'd get the 1V and decide which of the three of them makes me the happiest and get rid of the other two. As it is, I'll probably get the 1V to see if it performs any better in low light and I'll probably end up keeping all three. I paid too much for the Elan 7s at $225 shipped, in mint condition. I did ok on the EOS 3 at $200 shipped, in mint condition. I'd be really hard pressed to choose between the 2 but if I could only keep 1, it would probably be the EOS 3 because of the prestige, the theoretically greater longevity, and the slightly better low light performance.

The EOS 3 is not a professional level camera, despite popular commentary to that effect. It was pitched at advanced amateurs (whatever they are). It certainly was not a camera with magical autofocus.

Lenses with IS will, as a characteristic, retard AF whether they are from the L-series or non-L series lineup. If you are using the 2CR5 battery, performance will also not be at or near established benchmarks for AF performance, especially with L-series lenses.

Like the EOS 5 and a couple of others, weak spots with the EOS 3 were the lens release button and the rear cover latch, both of which were serviceable by Canon (not users). Annoyingly, various cameras had these weaknesses: the Elans, EOS 5, the EOS 50/50E and EOS 30/33 and EOS 3. Professional-level EOS bodies did not use the same materials for these known failure points, and just as well: a rear latch breaking mid-assignment is catastrophic. There were a large number of reports of this with the EOS 5 bodies, additional to stripped command dial indents and display driver failure (also affects the EOS 30, 33 and 3).

All EOS bodies "read" the film line tension. Self-loaded films have always caused problems with premature rewinding or forced removal from the canister. I have no historical evidence that the cameras treat normal 35mm loaded film aggressively. And certainly in my experience it has never happened. What are you doing with the loading of film, specifically the first wind on the spool?

ECF never took off, much as people wax enthusiastically about it (it sold cameras, for a while, and got people talking...then what?).
It is known that it works most effectively in brightly lit situations. And you are correct in assessing it as occasionally unreliable, and that's why it was dropped by Canon. If eye-controlled focus was as effective, efficient and error-free and a time-saver going forward, Canon would have carried it over to is digital cameras (this was considered almost 20 years ago). It did not because the amount of variability of users (eye size, pupil, the presence of eye defects like cataracts and other defects) is so great that ECF simple could not be tailored precisely. That was established way back in the day of the EOS 50E, which was descended from the EOS 5 with ECF re-jigged for horizontal, vertical and diagonal operation, but still prone to failure.

Custom functions are very useful but only with a vast amount of experience where they can be applied on-the-fly, specifically steps of exposure, leader in/out on rewind, button reassignment (etc., etc.). They should be seen as an extension of the visible controls of the camera and how they can change both the way the camera behaves and how exposures are dealt with, particularly switching from slide to negative film (exposure steps). I think for a skilled user it would be remiss to dismiss custom functions as "not very useful".

The $220 you paid for a used EOS 3 is cheap, but they can go lower. Dealers here think is will be around 2-3 years when an EOS 3 will go available used for $150 (body only).
Maybe go for the EOS 1V, which has been seen around here for $700 with PDBE-1 (high speed drive/battery booster). Better still, put an EOS 1V (or an EOS 1N) in parallel with the tests of your other two cameras! It is rare for anybody of skilled substance to complain about the EOS 1N and its later brother, the 1V.

Thanks for your input. Many of my observations about these cameras are not necessarily meant to be complaints, but just items to note if someone is considering a similar purchase. These are both great daily use cameras in my opinion. The only area that I would even consider a disappointment is the low light AF. Of course I don't expect the blazing speed and precision of my 5D4 from a camera that was built nearly 20 years prior, but I did have hopes (not necessarily expectations) that the EOS 3 would be a little better in this area than it is. When I purchase the 1V, I have every intention of adding a comparison of it here in this thread. With the 1V though, I'll be far more critical of it's ability to nail focus in low light better than its slightly lower end siblings. As the flagship, I do expect it to be noticeably better at the things that are truly important to most users. As for my seemingly dismissive comment about the custom functions, I didn't mean to say that they are useless, just that many of them aren't very useful to ME for MY style of shooting.

OK. I was not able to edit my post yesterday because there is an error being returned when such an action is initiated.

What I was going to add is that the EV2 you mentioned sounded to me to be quite low for any AF to reliably work, possibly with the exception of the beam emitter to help lock AF.
I would be very surprised however if even the EOS 3 could not lock focus onto a portrait subject in the frame in that condition, with or without the beam emitter. By the sound of things you are not aiming for accessory flash exposures where the pre-flash effectively directs the AF of the camera in low light -- effectively extending low light AF performance.

The clustered contrast-detection matrix of AF points in the EOS 3 and EOS 1V (also the 5 points in the 1N) are said to improve accuracy, and in a fast moving sports situation -- motor sports for instance, that would undoubtedly hold true. Your testing situation is unusual because it is taking the cameras down the other end of performance, or in the camera's own mind, "not interested!" -- it's wants speed and lots of it to play with! I'm reminded that the use of an f4 lens would also further hobble the results: try something like an f2.8 -- it need not be an L-series lens, but the Ls have been optimised for AF focus for the professional bodies. Load a very fast film e.g. Delta P3200 and boost it up-range.

The "most users" of the EOS 3 and EOS 1V are sports, wildlife, science and documentary, and the camera excels at speed. I can admit to never having seen an EOS 3, EOS 1V being used in a low light situation; it might even be found to be completely unsuitable, but that is no great shame considering poking around in near-darkness is not the camera's major cut and thrust. But a good experiment anyhow.

This is all very interesting. I might take my inveterate EOS1N out for a tat-tat over the weekend to rekindle its performance (or lack of) in low light. I have a wide variety of L-series lenses from f1.2 up to f4 and a solitary EF 20mm f2.8 (my first EF lens in 1994 after outplaying $4,000 for the 1N body only.

.::GRH

The specs for the EOS 3 say that it will focus from EV 0 to EV 18 so it should do well in EV 2 light. The reason that it's important that the camera I use be a good performer in low light is because I use it indoors for family type pictures, whether they be in the living room at night when the light levels are EV 3-4 or when my kids are on the stage when the lights are giving EV 4-5 levels. I typically shoot in black and white so there is usually a filter out front that drops the value by one stop, bringing much of my shooting into the EV 2-4 range. I wouldn't say that these are unusual circumstances for a high end consumer (prosumer) camera to find itself in. I don't believe that I will use the camera in levels below EV 2, so I decided to do my testing there.

I did quite a bit more testing tonight with only the 70-200 f2.8 L IS tonight and found that it does seem to be a bit better at grabbing focus at EV 2 and is more consistent than the 24-105 f4 L. The problem is that about half or more of my indoor shots are made with that lens. The remainder would be shot with eithe the 70-200 f2.8 L IS or the 100-400 f4.5-5.6 L IS II when using my 5D4. Unfortunately, the 100-400 is probably not an option with either the EOS 3 or the Elan 7s. Additionally, I might be better served for the bulk of my shots by getting a 24-70 f2.8 L II. The only way to effectively replace my 100-400 would be to get the 200-400 f4 L for $11,000. Ain't gonna happen.

So my compromise kit for film would be the two bodies, add the 1V, use the 70-200 f2.8 L IS for long shots, and purchase the 24-70 for $1500 for use around the house. The only big downside to this is the loss of reach from 200-400 at school functions. I guess I could always just flex on all of the aggressive moms that try to intimidate other parents out of the prime photo spots in the gym. Either that or I could just use my position in the district as a teacher to get those prime spots without having to give any moms the beat down. My only other option is to wait until I can get the 1V and see if it will outperform the other 2. I'm not holding my breat, as most people seem to state that the 1V has the same focus system as the EOS 3, minus the ECF. I don't mind the prospect of having three of Canon's most legendary film bodies though. After all, I AM somewhat of a collector, having currently over 20 cameras that I use. I'm not giving up on the ones I have unless the 1V just completely blows them away with 7D2 or 5D4 like performance. I would bet that the more I use them, the more I'll get to know and like them. In the end, I'm sure that any of the threee will suit me just fine within their dated technological limitations.

P.S. my 50mm f1.4 is out on loan to a friend at the moment. I'll be looking forward to seeing how it performs around the house with these two bodies.

They sure are pretty thought, aren't they?

Looks expensive, those lenses with the AF selection switches and maybe vibration reduction.

Have you tried the eye-control focus on the 7s?

I'm addicted to feature - the geek in me wants it to work all the time, although in practice that's not the case.

I've tried both. The 7s seems to pick up my eye better but only because the points are bigger and farther apart so I think that makes it easier for the sensor to register where I'm looking. It's much less frustrating than using the ECF in the EOS 3. The trouble with the EOS 3 is that the sensor isn't precise enough to accurately and consistently able to pick up the correct AF point for me. The trouble with the 7s is that since the points are larger and farther apart, there is not often a point exactly where I want to focus so I usually have to focus and recompose anyway. The feature sounds good in theory but isn't quite good enough to meet my needs. I'd like for it to work but I've used focus/recompose for years, from the original digital EOS, the D30, to the 10D, 20D 40D, 5D, 5D3, and now for still subjects, the 5D4 and 7D2 that I'm currently using.

Here are my final thoughts on the ECF feature in the EOS 3 that I had posted on another site. My thoughts on the 7s are pretty much the same although the 7s is more consistent on getting the right AF point:

Like many who have tried it, I'm not impressed. If I could look at the eyeball of the subject that I'm photographing and have it nail focus there 90% of the time, I'd be happy with it. With the EOS 3, I'm lucky if it picks the AF point next to the one I want better than half the time. Way too often it picks a point two or more away. I want focus on the eye and I get a chin or an ear instead. I know what people are thinking already too, "You have to run the calibration several times to improve the accuracy." I cleared the cal channels when I first got the camera and calibrated it to my eye both horizontally and vertically at least two dozen times each, under varying lighting conditions. I'd almost say that the more times I run the cal, the more erratic the performance becomes.

Having said that, when I treat it like the 20 year old technology that it is and use center point focus/recompose like I did with my old D30 or my 10 or 20D, it performs exceptionally well, even in low light, far better than the D30 ever did and probably better than the 10D and 20D. My 5D4 and 7D2 don't have eye control and I'm able to make them work for me. The EOS 3 is no different when I use it within its design limitations. There may be situations where the ECF feature is useful, but I just haven't found one yet.

They sure do look pretty, but these are working tools.
"You have to run the calibration several times to improve the accuracy."
Yes, that is correct. It is a factor. No more than 4x should be required. It will be less accurate overall if you wear glasses.
Calibration is done with your eyes wide open (e.g. like the look of shock), and in standard daylight (for one or more calibrations) and then lower light for other calibrations.
If ECF was useful, it would have been carried over to the flagship 1V. You know, professionals can work these beasts masterfully without gimmicks like ECF. I have never missed it.

Long story short, I did pick up my 1V a few weeks ago and after spending some time with all three cameras I discovered that there is nearly NO PERCEIVABLE DIFFERENCE in the low light performance of the AF systems in any of the three of these cameras (here's your caveat) when set up and used the way that I use them. As a result, I decided to keep the 7s for its ultra quiet shutter and light weight, and the 1V for its amazing features and ruggedness. As awesome as the EOS 3 is, I just can't justify keeping all three bodies.