rscheffler wrote:
Tariq - It will be interesting to see your 50 Cron-R images in comparison.
As for your point about the softer center - I would suspect it's field curvature. IIRC, someone in one of the XP1 threads here said it's pretty obvious with the 35 - was it you?
The 50 cron looks nice on the X-Pro1. I also tested a Pentax SMC 35 3.5 which looked sharper than the Fuji 35 1.4 in the center (this was the landscape test though where I focused the Fuji on the corners) but terribly soft in the corners - much softer than the lens looks on any other camera, even the NEX-7, in fact. I guess I should perform one more test to settle the matter but I suspect that the X-Pro1 is very picky even with some wider SLR lenses.
rscheffler wrote:
Regarding sharpness:
FWIW, I posted the images below in https://www.fredmiranda.com/forum/topic/1100128/0#10497815 as a reference of an image pulled from my M9, though shot with the 28 Summicron. Please forgive the crooked horizon. Exif says f/9.5... but with the M9 the aperture is an estimation. From what I recall, it was at or near f/8, though at ISO 320... peak sharpness might have been slightly better with this lens at f/5.6..
The overall image and 100% crops, unsharpened, from the unsharpened LR export (though -10 red/cyan CA correction was applied):
Indeed, that example looks much softer than I would have expected from the M9 unsharpened with a good lens (though, admittedly, I have never shot the M9 myself). Did you use a tripod? I saw the shutter speed but in my experience, sometimes even a higher speed can be required for critical sharpness (even with wider lenses).
This is similar to comparing lowlight performance at the pixel level vs. at the same output size. Comparing camera's without any sharpening may be academically interesting, but you can't really get a sense of which performs better in use without optimizing sharpening for each, which means very different sharpening routines for cameras with and without AA filters.
Yes, I will show at least one example of two different lenses of the same focal length on the X-Pro1 versus one another as well as how these compare to the NEX-7 just to get give an idea as to the "relative to other cameras" and the "relative to other lenses" question. All examples with no sharpening applied in the raw converters to hopefully help level the playing field (this will better show raw sensor and raw lens performance hopeully). There is still the variable of the raw converter (at least with the X-Pro1 vs NEX) as well as the question of how much, if any, Fuji manipulates the raw files which are made with Fuji X lenses behind the scenes (No way to answer that question as of yet). Yes, if there is time and will I will show both sharpened and non-sharpened.
I think an unsharpened RAW file does tell a little tale on how well a lens performs when compared to another similar lens on the same camera.
I also think that comparing cameras with or without AA filters is interesting but not more interesting or important than the actual image output of a finalized image.
The X Pro 1 RAW output seems a little softer than expected, but it seems to take to sharpening well, so I don't see that as a major problem even though I would have liked to see a little more bite.
My GXR M mount RAW files are quite sharp and crisp straight out of the converter with no sharpening applied, it makes it very easy to compare lenses. However when I compare a finalized image from the NEX5N of a similar subject I can often not tell them apart, they are both equally good with the more recent NEX5N sensor having low light/high ISO advantage.
When RAW converters are supporting the X Pro 1 we will see how well it will perform, I am sure it will hold it's own against the other APS-C cameras and it does seem to hold a moiré advantage over the M9 for sure.
I think the M9 takes to sharpening beautifully and I am not sure how well the X Pro 1 will stack up against it.
I think high ISO images from the X Pro 1 will be better or at least easier to process judging from what I've seen so far.
PS
Thanks Freaklikeme, Carsten and Tariq for the B-day comments, l celebrate my 40th with a sprained back and cheese cake. Luckily I get professional in home treatment by the missus who is schooled in cracking bones and bending limbs.
kosmoskatten wrote:
Thanks Freaklikeme, Carsten and Tariq for the B-day comments, l celebrate my 40th with a sprained back and cheese cake. Luckily I get professional in home treatment by the missus who is schooled in cracking bones and bending limbs.
Well, at least there is cheese cake and the missus! Just be sure and keep her happy so she doesn't go overboard on the "cracking bones and bending limbs" thing!
Re the Ricoh A12 M module. That's sort of the pixel level bite I was expecting out of the X-Pro1 with no sharpening. It's also important to note that the sharpening that is performed on the x-Pro1 files may be somewhat proprietary. It's certainly different and seems more encompassing in what it does versus sharpening in LIghtroom/ ACR. Thus, you can't really compare sharpened images without dealing with that variable which can't be equalized with other files (ie. using the same sharpening techniques). As a total system - lens, sensor, software - Fuji has done a great job with the files (as long as they are not pushed too much).
Yes, I think the Fuji X Pro 1 seems like a well balanced concoction of lens, sensor and software. I think it might be a little bit crippled with the sensor design for wide angle lenses as I have yet to see that they will pull of the even wider glass that they have on the road map. Also I am hoping they will not have to resort to software corrections too much as it might hamper use of other non proprietary RAW converters.
Haven't got a clue who Zack is, but have to confess that I'm not into these breathless Steve Huff-type of reviews.
The guy shot only JPGs (for reasons he explains in the comments), but I would have preferred a more analytical approach to the results. Hey, I'm into a "non-exquisite aesthetic" and like a fluid and loose style — and I blow highlights with abandon on a daily basis — but I don't care for his pictures of Gabby in the cab and the one of "Gabby and the Cabby" in this respect, about which he states, "and the sun was just right". But, then, these two pictures show more about the photographer than about the dynamic range of the X-1 Pro.
What disturbs me more is that he says nothing of the rendition of Gabby's red dress in the second picture after "Gabby and the Cabby" — a rendition similar to that of the red car in the comparison with the M9 earlier in this thread. Look also at the reds of the picture further up in the series of the Indian woman in the sari.
I'm all for "real world reviews", but they need real world analysis. For example. he could have written about the different JPG film simulations.
So, all I'm left with from this review is wondering what sort of eye the guy has and confirmation that the reds produced by the JPGs of this camera are problematic, as well as a concern about shutter lag.
To look at the red channel issue, I shot ColorChecker (x-rite) panels along with some bright red thread, using a Sony A77 and the Fuji XP1. I exposed both cameras about -0.3 to -0.7 under. The out of camera JPGs looked similar; the Sony appeared slightly warmer. All settings were default as I typically only use RAW files.
RAW files were processed with the latest version of Silkypix Studio Pro (v 4.1.55.0). Neither camera had issues with over exposing the red channel. While the histograms slightly differ, the resulting color charts look remarkably similar with good red, pinks, blues and greens. It would seem the over-exposed red channels are simply over-processed JPGs. The RAW data is just fine.
I get the impression that most of the images shown on the web with over-exposed red channels have little or no adjustments being done to default camera settings. Instead, default settings are being compared. Defaults are just average estimates by the manufacturer, and not all encompassing ideal settings. By using the JPG adjustments (that's why they're there), the red channel issue is not an issue. Planning---considering the subject matter and lighting---should dictate the JPG settings (Astia instead of Provia; AWB with less red in the next option box, and so forth), if one wants to avoid working with RAW files and not have red channel exposure problems.
As an aside, if one were to search on UniWB, a method of properly exposing RAW files (I cannot see how to do so on the Fuji), they'll see that red is almost impossible to over-expose (unless shooting a red subject on a red background!) since there are twice as many green pixels. A UniWB JPG looks green and will optimally expose the given sensor while providing the lowest noise. Unfortunately, no manufacturer seems willing to give us this tool.
Mitch Alland wrote:
Haven't got a clue who Zack is, but have to confess that I'm not into these breathless Steve Huff-type of reviews.
+1. This guy goes through a wild range of opinions within a single review with a single camera. Clearly if you are not fully harmonized with his mood swings you can't learn much from him.
I felt the review was pretty clear about that. He initially expected an SLR replacement and was disappointed. Once he reset his expectations, his opinion changed.
Zack doesn't write analytical, dpreview style reviews. I wouldn't lump him in with Steve Huff, because Steve is writing from a different perspective, but the style of their reviews is somewhat similar.
Tariq Gibran wrote:
Unfortunately, the "watercolor" effect is in both raws (as processed by Silkypix at least) and OOC jpegs. The examples I posted are from raws, not OOC jpegs.
I don't really should landscapes, mostly people and products but I was wondering, is taking panoramic shots an option for you? Having few photos stitched together would take care of the resolution (no more "watercolor effect") and would allow you to use longer lenses (like the 35/1.4) so wouldn't have to worry about soft corners of the 18mm.
nandemofoto wrote:
I don't really should landscapes, mostly people and products but I was wondering, is taking panoramic shots an option for you? Having few photos stitched together would take care of the resolution (no more "watercolor effect") and would allow you to use longer lenses (like the 35/1.4) so wouldn't have to worry about soft corners of the 18mm.
It's not a viable option for me personally but it certainly might make sense for others.
I just returned the X-Pro1 today. I actually felt a little loss handing it over to the UPS guy and I suspect I'll miss it a bit. Reviewing some of the shots I took with it, it's hard not to be impressed with the color and huge dynamic range. In both respects, it's easily as good as any FF camera I have used and a bit above other APS cameras imo. I still need to get around to posting some of the tests I did with various alt lenses. Anyway, once better raw support shows up and a few firmware refinements occur, I may very well jump back in. It is a very pleasurable camera to shoot.
Tariq Gibran wrote:
It's not a viable option for me personally but it certainly might make sense for others.
I just returned the X-Pro1 today. I actually felt a little loss handing it over to the UPS guy and I suspect I'll miss it a bit. Reviewing some of the shots I took with it, it's hard not to be impressed with the color and huge dynamic range. In both respects, it's easily as good as any FF camera I have used and a bit above other APS cameras imo. I still need to get around to posting some of the tests I did with various alt lenses. Anyway, once better raw support shows up and a few firmware refinements occur, I may very well jump back in. It is a very pleasurable camera to shoot. ...Show more →
I'm really curious if this effect can be minimized by using lenses with better resolution, like the Zeiss 18mm f3.5 AIS.
I found the XF 18mm to be rather average (put it on sale in a different form) instead I'm using VC Color Scopar 20mm f3.5 AIS which has much better corners and overall better resolution.
I will show both sharpened and non-sharpened.