Canon Canon G1X III: Finally, a huge step forward for the G X series ...

[MoonMind]

Any other suggestions?

I'd like to get my hands on some images (RAWs) from the 45mm end, preferrably real life with lots of detail all over in order to check lens performance across the frame; if they are ISO 3200, all the better, but ISO 100 is also very interesting. I have looked at RAW shots at ISO 6400 - too much chrominance noise for my taste, but then, I'd never expect an APS-C sensor to perform sufficiently well there. But I'd like to get an impression about ISO 3200 - that's what my Nikon D5500 handles pretty graciously. DXOMark results suggest that the 24MP Canon sensor should deliver in more or less the same way (minus a bit of DR and colour accuracy, but I can work around that - I have done so with the LX100 pretty successfully, and that sensor sure is no match ...). But then, DXOMark doesn't tell the whole story - if you look at the difference in ISO performance between the "mighty" D750 and the "frugal" D5500, you'd expect a huge difference. In truth, it's about one solid stop and a bit of colour conservation you get from the FF sensor. Sure, the D750 is a fantastic workhorse camera that beats its little sibling in almost every way - but truth be told, most of the time, the APS-C results are so good that the difference is negligible (if you're not into printing large). And actually, the D5500's files are a lot less fiddly to process, too ...

That said, I stumbled upon a couple of sample shots today (not from the infamous low light promotion event you find pictures from all over the web - hardly any good shots in that set!) which indicate that the lens is up to my needs and expectations at the tele end; I'll keep looking around for another couple of days (maybe only hours ) before deciding.

EDIT (in the morning, after sleeping over it once more): I've ordered. I don't think I'll be disappointed - and I'll finally have the compact no-worries camera I've been looking around for for years.

M.

 

[tonyturley]

I just picked up the rental package a short while ago. The Pelican Storm Case in which it came packed would probably probably let the camera survive being dropped off a cliff . . . or run over by a truck. Don't think I'll try those tests, though.

Here's the G1X MIII next to an old SX110 IS for comparison. Battery was already charged, so let the games begin. Second pic is a quick test inside my office, showing my X-T1 next to the SX110 IS. You can see the X-T1, which isn't a large camera, is somewhat larger than the G1X MIII. This was just a JPEG; RAW testing will come later. Both of these were at ISO 1600. The Fuji was 1/40 + f4, and the Canon 1/40 + f5.6.

I tried to set the lens as close to the FL of the X-T1 as I could; however, using the same exposure settings on the G1X as I used on the X-T1 for the previous image resulted in an image overexposed by almost 2 stops. My initial impression - without pulling out the manual - is that it's going to take some getting used to a power zoom and no dials for ISO or Aperture. All exposure settings are done either by spinning the front or rear dial and watching the display, or via a quick menu, not completely unlike the Fuji's "Q" menu.

Join to see EXIF info for this image (if available)

Join to see EXIF info for this image (if available)

 

[tonyturley]

Let’s play with some RAW files. I tried all sorts of combinations to test diffraction and/or force slow shutter speeds. All converted with RawTherapee’s default preset. No sharpening on any images. First image has no noise reduction, second used RT’s denoise algorithm. The first two were in a room with the lights off, lit only by a large window.

F5.6, 1/125, ISO 3200, 15mm:

Join to see EXIF info for this image (if available)

F5.6, 1/100, ISO 3200, 45mm:

Join to see EXIF info for this image (if available)

Playing with the ND filter. I’m wondering if this camera is going to be a dust pump like the LX100. Note the dust spot in one of my images. Easily fixed in post, but I’ve only had it in my possession a few hours. There were other images like that, as well.

F13, 1/100, ISO 200, 15mm:

Join to see EXIF info for this image (if available)

F13, 1/50, ISO 200, 45mm:

Join to see EXIF info for this image (if available)

F5.6, 1/5, ISO 200, 43.4mm:

Join to see EXIF info for this image (if available)

Trying to catch my dog and the neighbor’s cat playing in the leaves. Way too slow shutter speed for most images, but even with the ND off, f5.6 didn’t help much at the longer end in the failing light. Could have bumped up the ISO, but wanted to see the results.

F5.6, 1/6, ISO 200, 40.4mm:

Join to see EXIF info for this image (if available)

F5.6, 1/6, ISO 200, 40.4mm:

Join to see EXIF info for this image (if available)

F5.6, 1/30, ISO 200, 39.8mm:

Join to see EXIF info for this image (if available)

One thing I noticed is the lens is at f2.8 only at its widest; it closes down quickly as one zooms, and the MFD (which is shown on the screen), goes from 3.9” at 15mm to 11.9” at 45mm. This lens is s.l.o.w.

More tests tomorrow, only with using more realistic settings. Awfully early to say for sure, but I’m not overly impressed with either the haptics of this camera or the IQ of the lens.

 

[bartjeej]

I'm following your findings with great interest, Tony! Much appreciated.

 

[rayvonn]

Very informative and appreciated, thanks.

 

[tonyturley]

I'm following your findings with great interest, Tony! Much appreciated.

Very informative and appreciated, thanks.

You're welcome, guys. I spent some time in the wee hours of the morning today testing JPEG quality in low light. Resized and a small amount of sharpening applied in GIMP. Default JPEG settings seem to be a bit soft. IS works fairly well - I let the SS get pretty slow. Lens also doesn't seem to flare badly in strong direct light. Definitely better than yesterday's results.

f/5.6, 1/50, ISO 200, 16mm:

Join to see EXIF info for this image (if available)

f/5.6, 1/80, ISO 200, 16mm:

Join to see EXIF info for this image (if available)

f/5.6, 1/25, ISO 200, 30mm:

Join to see EXIF info for this image (if available)

f/5.6, 1/15, ISO 200, 45mm:

Join to see EXIF info for this image (if available)

f/5.6, 1/15, ISO 200, 45mm:

Join to see EXIF info for this image (if available)

f/8, 1/125, ISO 200, 15mm:

Join to see EXIF info for this image (if available)

f/8, 1/320, ISO 200, 30mm:

Join to see EXIF info for this image (if available)

 

[MoonMind]

I should get mine this week; will join in the fun shortly, hopefully.

@tonyturley Just out of curiosity: The speck of dust seems to be gone - did you clone it out or did it dissolve somehow? The most irritating thing would be debris from manufacture in a dust-proof camera ...

M.

 

[tonyturley]

Just out of curiosity: The speck of dust seems to be gone - did you clone it out or did it dissolve somehow? The most irritating thing would be debris from manufacture in a dust-proof camera ...

M.

No Matt, I didn't do anything. It just went away.

Here are a few testing cropping on JPEGs, all shot with Aperture Priority. First was f8, 1/100, ISO 200, 45mm:

Join to see EXIF info for this image (if available)

The squirrel was about 100 feet away. f8, 1/200, ISO 200, 45mm:

Join to see EXIF info for this image (if available)

f8, 1/320, ISO 200, 26.1mm:

Join to see EXIF info for this image (if available)

 

[rayvonn]

So Tony, how do you compare it against your Fuji?

 

[tonyturley]

So Tony, how do you compare it against your Fuji?

Ray, the Fuji seems to be slightly sharper, even though I have the sharpening on the X-T1 at its minimum setting. I also prefer the way the Fuji handles.

Here are a few JPEG comparisons between the G1X Mk III and the X-T1 with 23mm f2 WR lens. No sharpening applied to any images. First the Canon.

f8, 1/3, ISO 1600, 15mm:

Join to see EXIF info for this image (if available)

f8, 1/4, ISO 1600, 20mm:

Join to see EXIF info for this image (if available)

f8, 1/2000, ISO 1600, 15mm:

Join to see EXIF info for this image (if available)

Next, the Fuji. By the time I got the Canon off the Ultrapod and the Fuji mounted and secured to the same fence post, the sky had brightened considerably. The way Fuji reports ISO values also affects exposure values. Many sources claim Fuji over-reports ISO by at least a full stop.

f8, 1/2, ISO 1600, 23mm:

Join to see EXIF info for this image (if available)

f8, 1/2, ISO 1600, 23mm:

f9, 1/500, ISO 1600, 23mm:

Join to see EXIF info for this image (if available)

 

[drd1135]

Just to note, the compression algorithm or whatever the forum uses makes these look somewhat unsharp on my monitor. They look much better once i click on the image and it expands a bit.

 

[tonyturley]

Just to note, the compression algorithm or whatever the forum uses makes these look somewhat unsharp on my monitor. They look much better once i click on the image and it expands a bit.

You're right, Steve. That makes a big difference.

 

[tonyturley]

Playing with shadows. Slight sharpening added in post.

f8, 1/1250, ISO 1600, 24mm:

Join to see EXIF info for this image (if available)

f8, 1/500, ISO 1600, 15mm:

Join to see EXIF info for this image (if available)

f8, 1/500, ISO 1600, 18mm:

Join to see EXIF info for this image (if available)

f8, 1/1250, ISO 1600, 15mm:

Join to see EXIF info for this image (if available)

f8, 1/500, ISO 1600, 25mm:

Join to see EXIF info for this image (if available)

 

[MoonMind]

Playing with shadows. Slight sharpening added in post.

f8, 1/1250, ISO 1600, 24mm:
View attachment 130654

f8, 1/500, ISO 1600, 15mm:
View attachment 130655

f8, 1/500, ISO 1600, 18mm:
View attachment 130656

f8, 1/1250, ISO 1600, 15mm:
View attachment 130657

f8, 1/500, ISO 1600, 25mm:
View attachment 130658

Works pretty well, I would think. I'm charging up batteries at the moment (got a second one - and I think that was a wise move); yes, the camera has arrived and already been mostly customised. I need to figure out a few things still (like if there's any way to get Zebra to judge blown out highlights reliably - love that on the LX100 ...). But that'll have to wait until I can power up the camera again.

M.

 

[MoonMind]

Okay, it's getting interesting. I like a lot of things about the new camera; for one thing, in spite of what I had read beforehand, it's a speedy camera in operation, but also, crucially, when it comes to start-up - it's just a little slower(!) than the LX100 (that can be a bit fickle at times - mainly, it's pretty decent) and the GR which I find mostly fantastic in that regard (though if I force the shutter - which is actually configured that way - before the camera is fully ready - I get corrupt files). Focus action is good, the EVF is nice (a little smaller, but crisper than the LX100's). Sadly, no Zebra, but it's simple to call up a histogram. This is one configurable camera, too - I've figured out a couple of neat things, but I have to test this stuff further before writing about it - for now, I'm more or less feeling my way along ...

Now, for some shots:

I know it may sound hard to believe, but except for the EVF hump, the G1X III is actually smaller and slimmer than the LX100 - even when I remove the automatic lens cap from the latter. The lens' diameter is also smaller - a clear sign that compactness trumps mass (i.e. amount of glass) in this case. Handling is surprisingly nice - everything falls nicely under the fingers; I have to learn the paradigm (this is my first serious Canon camera - apart from the older and more frugal S95). Image taken with the D5500 and Sigma 18-35mm f/1.8 Art zoom (probably the best zoom for APS-C ever built).

This is just a bit of fun - but it clearly shows just how small the G1X III is - the camera in the background is the Nikon D5500, itself one of the smallest APS-C DSLRs on the market; same resolution. Of course, the huge Sigma 18-35mm f/1.8 I on there overexaggerates the effect, but I did this on purpose - even the smallest lens I own for the D5500 would appear massive compared to the Canon. Oh, and the image was taken with the LX100 - and I have to say that it did a fine job.

Finally, this shot was taken with the G1X III; I didn't correct anything on capture to see how it would do. This is about as bad as it can get under reasonably bright artificial lighting if you do most things wrong - this was shot wide open at 15mm, very close to the subject (to exacerbate any distortion) at ISO 1250; I did zero perspective correction. I think it's quite okay for what I did. btw. The two cameras in the image came back from CLA/repairs today, both of them are battered beauties, now fully working again. I'd say the sharpness at focus point is very nice, overall rendition is solid, but the APS-C sensor played a slight trick on me (that I could have avoided had I wanted to): DoF is a bit shallow

This time, I did what I could (in short order) to get the best result without changing conditions, only settings, so this is at medium zoom (I have to figure out how to check this on capture; there is a step zoom feature I have to activate ...). I know it's in the EXIF somewhere, but since the editor doesn't make it available, I'll skip that for now. The important stuff: 1/50s, f/5.6, ISO 3200. Pretty crisp and even, lots of fine detail, little noise, good colour recovery possible - I'm really satisfied with the performance. Actually, the image is so detailed that you can observe some compression artifacts (not for the first time in this thread) - I have to keep this in mind since I could have avoided this by resizing before uploading. What looks like image noise in the left front corner isn't - I've looked at 400%. Likewise, the writing around the lens of the Retina looks a bit overprocessed in the uploaded image - it's not, I checked in the original. There *is* some noise visible in the original - but it's mild at best. Good show!

All images were processed with Polarr - so it's basically an enhanced JPEG. Polarr doesn't offer a lot of advanced features - like sophisticated WB correction. So this is what you get if you don't do a lot (or can't). While I have to wrap my head around the camera (the same process Tony's going through), I think I'll enjoy it a lot.

M.

 

[MoonMind]

I have to make a little correction (already did it in my last post, but it's important): Start-up of the G1X III is fluid, but not that quick; just tried it again - the LX100 can be a bit unpredictable, but if it works well, it's faster than the Canon; the GR is faster than both, but sometimes, the image engine needs a sort of second start which can lead to corrupted files. Sorry for any confusion - the LX100 was at its worst when I did the first comparison. Thus, the G1X III is the slowest (but not slow), but most reliable, the LX100 brings up the middle ground (but strangely, not always), the GR is fastest, but the file corruption issue is real (in five consecutive test rounds, it produced three unreadable files - it happens if I press the shutter a fraction of a second too soon; in real life, this means I tend to hesitate a bit - which eats away any speed advantage anyway ... ).

Some more observations: Zoom speed is pretty quick when using the rocker, control via the lens ring, on the other hand, is slow, but very precise. Wake up from sleep is quick. The shutter button is a bit spongy - tribute to the weather-sealing, I guess. Images are very(!) detailed and stand up well to scrutiny up to ISO 3200, and that's all I possibly need (I think I could even get away with JPEGs at ISO 6400 if need be - but RAW gets rough at that speed). Something I've experienced before: AWB can be totally off due to adjusting to some part of the image that's definitely not the main subject - the same thing can happen on the S95. This may lead to me using Custom WB or WB presets more often than with other cameras. If AWB works, it's fine, though.

Two more things I really like: The mode selector is locked by default - no knocking possible; the dedicated exposure compensation dial is quite stiff - no knocking probable. Very nice, very welcome - the same dial on the LX100, while practical in use, is a constant cause for irritaion because it doesn't stay put. The same goes for the aperture ring, btw. - it's not quite as bad, but sometimes I either end up in Auto mode or the aperture is way smaller than what I remember setting.

The G1X III is a very well rounded camera, that's already clear. When considered on its own, no real weaknesses stand out. We'll see how things develop, but I'm really optimistic now. It'll have to prove itself in marginal weather this weekend (in conditions I wouldn't use the LX100 - that's what's in store for us around here during the next couple of days) - then I'll know all I need to know.

M.

 

[rayvonn]

Playing with shadows. Slight sharpening added in post.

f8, 1/1250, ISO 1600, 24mm:
View attachment 130654

f8, 1/500, ISO 1600, 15mm:
View attachment 130655

f8, 1/500, ISO 1600, 18mm:
View attachment 130656

f8, 1/1250, ISO 1600, 15mm:
View attachment 130657

f8, 1/500, ISO 1600, 25mm:
View attachment 130658

What a wonderful time of today to have a camera!

 

[tonyturley]

Two more things I really like: The mode selector is locked by default - no knocking possible; the dedicated exposure compensation dial is quite stiff - no knocking probable. Very nice, very welcome - the same dial on the LX100, while practical in use, is a constant cause for irritaion because it doesn't stay put. The same goes for the aperture ring, btw. - it's not quite as bad, but sometimes I either end up in Auto mode or the aperture is way smaller than what I remember setting.

The G1X III is a very well rounded camera, that's already clear. When considered on its own, no real weaknesses stand out. We'll see how things develop, but I'm really optimistic now. It'll have to prove itself in marginal weather this weekend (in conditions I wouldn't use the LX100 - that's what's in store for us around here during the next couple of days) - then I'll know all I need to know.

M.

Agreed on the knobs . . . . those are nice. However, being somewhat old school, I've always preferred labeled control knobs and aperture rings. I've been using G1X Mk III on manual most of the week, and the exposure values are easily observed on the display screen or in the EVF, so it hasn't been any sort of inconvenience. Just different.

The big question for this lens is low light. I'm going out on an excursion tomorrow where I'll be bike riding along along a steep-walled ravine that drops off a mountain into a deep canyon. Lots of opportunities for waterfall pics in very low light. We'll see how we do with the Mk III.

 

[drd1135]

The big question for this lens is low light. I'm going out on an excursion tomorrow where I'll be bike riding along along a steep-walled ravine that drops off a mountain into a deep canyon. Lots of opportunities for waterfall pics in very low light. We'll see how we do with the Mk III.

That sounds dangerous. Maybe your should upload your pictures in case you fall in. :eek-54: OTOH, this is radio free West Virginia we're talking about.

 

[tonyturley]

That sounds dangerous. Maybe your should upload your pictures in case you fall in. :eek-54: OTOH, this is radio free West Virginia we're talking about.

LOL . . . I can actually get a cell phone signal on the lake shore at the bottom of the canyon. I've ridden my bike down the gravel road that follows the ravine a few times. It is a steep, windy road, and I always ride on the side away from the ravine, which has a sheer drop as much as 25 feet from the road. There was actually a death there a few years ago. Mill Creek is a popular spot for both fishing and photography. The rail trail on the other side of the ravine is a lot easier climb out of the canyon.