proxy

proxy

Joined on Jan 11, 2012

Comments

Total: 75, showing: 1 – 20
« First‹ Previous1234Next ›Last »
In reply to:

Sdaniella: most small sensor systems lenses have considerable lens aberration correction, especially if going very wide fov, so it starts to get harder to tell how much of the distortion is from variation in optics, or from the standardized correction of those expected distortions as designed.

the Canon G7X RAW (seen shown elsewhere, shared on DPR, from a French test comparison webdite, forgot which) at 24mm eq, looks extremely distorted, its a wonder how it can possibly be corrected fully 'rectilinear' in JPEGs without plenty of evident softness in the corners (and at the edges of the frame).

most of it, would not be of great concern, knowing such massive correction is going on, on corner details where most are likely to ignore, and not notice unless specifically looked for, or typically removed during 'creative cropping' for final work presentation.

anyone choosing from compacts Canon S120 to Canon G7X will cheer 'yay'!

ditto: LX7 to LX100 (a bigger sensor size jump; yields bigger IQ gains)

thanks for posting, quite revealing and at the same time disappointing, wide... the lens is too wide and distorts heavily at 24mm equiv., mid and tele: the performance at tele is really excellent sharpness-wise. Then there's CA's, bit disappointing as well, overall interesting camera for jpeg shooting due to auto-correction but a bit scary when using raw, let's wait and see what's Canon going to do about that. More pictures and more info in coming weeks will shed more light on these issues. I have not seen a full review published yet. It's a mixed bag, in some aspects it's better then rivals in others is slightly behind, not a bad camera as it's capable of producing stunning images but...

Direct link | Posted on Oct 15, 2014 at 23:11 UTC
In reply to:

Sdaniella: most small sensor systems lenses have considerable lens aberration correction, especially if going very wide fov, so it starts to get harder to tell how much of the distortion is from variation in optics, or from the standardized correction of those expected distortions as designed.

the Canon G7X RAW (seen shown elsewhere, shared on DPR, from a French test comparison webdite, forgot which) at 24mm eq, looks extremely distorted, its a wonder how it can possibly be corrected fully 'rectilinear' in JPEGs without plenty of evident softness in the corners (and at the edges of the frame).

most of it, would not be of great concern, knowing such massive correction is going on, on corner details where most are likely to ignore, and not notice unless specifically looked for, or typically removed during 'creative cropping' for final work presentation.

anyone choosing from compacts Canon S120 to Canon G7X will cheer 'yay'!

ditto: LX7 to LX100 (a bigger sensor size jump; yields bigger IQ gains)

I checked a lot more incl. IR, Flicker, some examples here if you're interested:
http://www.dpreview.com/forums/thread/3732635?page=2#forum-post-54569340

I never intended to do a full evaluation, I'm posting to put some opinions expressed here in perspective as many (not me) judge a camera after a superficial examination of one picture. You can find more data points if you you look for them. Obviously you prefer to pick on me instead of spending your time getting more acquainted with the camera itself and test results. They are not complete, its a new camera but worth exploring, this is what I'm trying to say. Definitely not as bad as some may think, time will show. You'll get all your data points in time, different FL's ans stop values, just not yet. In the meantime, if you're curious you can find out more then just from DPR test, judging by what you wrote I can see you missed most of my posts or thread.
Camera used here had visible lens issue so how could I possibly accept it.

Direct link | Posted on Oct 15, 2014 at 13:56 UTC
In reply to:

intruder61: Too many guys believe the BS hype the manufacturers are throwing at you about how good these compacts are, then when they don't perform to what you want it to, out come the guns....compacts are compacts are compacts, small sensors, nothing special, it is what it is....and here is the special, you get shafted by paying premium price....entry level DSLRs with APSC sensors for less $$ are a whole lot better value...oh yeah i forgot, the DSLRs dont fit in your back pocket while you're having coffee down at the cafe.

Although still too expensive for what it offers you have to admit it produces some really decent images by any measure.
Plus it fits in your pants pockets, a huge plus anyway you look at it, don't you think?

Direct link | Posted on Oct 15, 2014 at 13:02 UTC
In reply to:

Tonkotsu Ramen: There must be some crazy sample variation. On the bottom left cards, the G7 X is so soft it looks like motion blur. This isn't the case from any of the other comparisons i've seen though, where I need to go back n forth to look for differences.

my point exactly, I'd ask for my money back, ...yet it was elevated to be a test sample here

Direct link | Posted on Oct 15, 2014 at 12:51 UTC
In reply to:

Discovery Of Light: Something is up with the G7X. They seem very blurry especially around the playing cards. Did somebody run into the tripod?

'Designed to compete with the above?' Designed to compete with what?

DSLR's... if you missed that. Most DSLR's are still APS-C sensor based (or alike) not FF. Not comparing with FF cameras (DSLR's). You seem to be shooting in all possible directions, I don't get it.
First you say "it's just a compact" then you compare to FF, to say at the end "smaller sensor puts far more demands on the resolving power of the lens" so... the lens has to be good or better with small sensors with high Mpix count. In summary, first you say it's just a compact so lower your expectations, then you say the opposite considering its high Mpix count.
Well, you tried every trick, I give you that, none of which applies to issue I raised.

Direct link | Posted on Oct 15, 2014 at 12:40 UTC
In reply to:

Sdaniella: most small sensor systems lenses have considerable lens aberration correction, especially if going very wide fov, so it starts to get harder to tell how much of the distortion is from variation in optics, or from the standardized correction of those expected distortions as designed.

the Canon G7X RAW (seen shown elsewhere, shared on DPR, from a French test comparison webdite, forgot which) at 24mm eq, looks extremely distorted, its a wonder how it can possibly be corrected fully 'rectilinear' in JPEGs without plenty of evident softness in the corners (and at the edges of the frame).

most of it, would not be of great concern, knowing such massive correction is going on, on corner details where most are likely to ignore, and not notice unless specifically looked for, or typically removed during 'creative cropping' for final work presentation.

anyone choosing from compacts Canon S120 to Canon G7X will cheer 'yay'!

ditto: LX7 to LX100 (a bigger sensor size jump; yields bigger IQ gains)

Our opinions differ based on pictures you and I have seen. Let's leave it at that. I cannot take your hand and walk you through all steps I went through so I'm inclined to drop it here.

"And using DPR samples for evaluating lens performance is quite silly."
Regarding the DPR samples... well, now I know that from you and Rishi. I was under the impression that test pictures were done using lenses as well not sensors only what you seem to be suggesting.
So IQ that I see in tests is a function of sensor performance but not lens performance even though the light had to go through a lens be4 it reaches the sensor. And all it produces is just "some idea" of an image but not a full image, "not much more" as you put it. I happen to believe that it is more.
Isn't it a purpose of testing to produce an image? And this image is to be used only to check for noise? and sensor performance?? If DPR uses FE55mm you want me to believe that this is to test noise and sensor only? That is quite bizarre.

Direct link | Posted on Oct 15, 2014 at 12:18 UTC
In reply to:

proxy: Something's wrong with Canon G7X and Sony RX100 III namely Canon gets blurry towards lower left corner and Sony get blurry towards upper right. Looking at cards it becomes very obvious.... corners confirm that. Seems that both cameras have issues with lens folding mechanism or lenses are not centered properly. IMO such camera should not be tested unless it's confirmed that several other cameras display the same behavior which is then a good reason for a recall (never happens) and have to be considered a serious fault. So, has DPR tested faulty cameras or... all or most RX100 III and G7X are faulty? Which one is true?
Btw. had THIS issue with Canon 850IS years ago, gave it as a gift brand new but year later closely looked at pics I'm getting as obviously something was wrong, called Canon & went to Canon repair facility and 1 mth after warranty expiration Canon fixed it (lens module replacement), no charge. Excellent service and no more issues later. Bravo Canon. Has DPR overlooked it ?

Everything within reason... I hope we agree. And if at first sight I see that there's something wrong with the test image I would discard it and investigate why. Bad cameras are sold daily despite QC. But that bad camera should be spotted right away in the hands of an expert. That is all. I'm not a pixel peeper, I happen to be one when looking at a new product that I'd like to have an opinion of. I may adopt yours or develop my own that depends on quality of produced material. I don't think we're that different. Most members are pixel-peeping here, that's the nature of the beast. You don't trust blindly. Trust but verify. I did my DD and expect you to do the same considering website's mission and responsibility of serious journalism. Are we in disagreement here? I spot a problem and post here, it's that simple. Problem that is easily identifiable without specialized knowledge. The rest is up to you, whatever you please but don't tell me my eyes lie to me. Deal with it I'd say.

Direct link | Posted on Oct 15, 2014 at 11:57 UTC
In reply to:

Zeisschen: - the APSC-sensor in the a5000 is still quite a bit ahead of all of them
- the lens on the a5000 (which one was it) is much better (raw)
- the raw performance on the G7X and the RX100 looks very similar, which is expected due to the same sensor
- the Panasonics raw performance is between the two 1 inch sensors and the APS-C sensor like expected
- the Canon lens show massive CA compared to the RX100 (check raw)
- the Panasonic lens is between the Sony and the Canon (at this focal length)

Overall picture quality order for me based on those samples as well as the earlier posted picuter samples:

1. Sony a5000 > best sensor and best lens
2. Panasonic LX100 due to better ISO performance than The Canon and Sony RX100
2. Sony RX100 due to better lens compared to LX100 based on picture samples, exqualizing the slightly better ISO performance of the LX100 for me
3. Canon G7X > ISO performance like RX100 but lens show some weekness in CA, no suprise for a Canon lens.

"Unfortunately, that can't be said about some of our readers here apparently, even after being told repeatedly that the studio scene is not a measure or absolute word on lens performance." I agree fully and precisely that is why checking at other FL's and F values gives a more complete picture.
Coincidentally this is exactly what Imaging Resource is doing. Good for them. In that spirit DPR tests are merely an indicator of possible camera performance but not necessarily so due to testing scope and method and associated limitations of such testing if I read you correctly. We are in full agreement so far. Difference is that based on such limited testing DPR awards camera ratings expressed in pretty precise numbers followed by overall score.

You have just discredited tests at DPR by proving how unreliable they may be, how far reaching caution should be while assessing them, yet you firmly stand by the results???!!
Re. 4 cams... well it's 50/50. It'll be fair to say it loud and clear.

Direct link | Posted on Oct 15, 2014 at 11:29 UTC
In reply to:

Sdaniella: most small sensor systems lenses have considerable lens aberration correction, especially if going very wide fov, so it starts to get harder to tell how much of the distortion is from variation in optics, or from the standardized correction of those expected distortions as designed.

the Canon G7X RAW (seen shown elsewhere, shared on DPR, from a French test comparison webdite, forgot which) at 24mm eq, looks extremely distorted, its a wonder how it can possibly be corrected fully 'rectilinear' in JPEGs without plenty of evident softness in the corners (and at the edges of the frame).

most of it, would not be of great concern, knowing such massive correction is going on, on corner details where most are likely to ignore, and not notice unless specifically looked for, or typically removed during 'creative cropping' for final work presentation.

anyone choosing from compacts Canon S120 to Canon G7X will cheer 'yay'!

ditto: LX7 to LX100 (a bigger sensor size jump; yields bigger IQ gains)

Well I'm not sure if she's thinking what you're thinking.
Re. IR... I already posted several hrs ago a crop from resolution chart that you refer to that proves the opposite, you'll need to check your facts again, you can find mine in this thread: http://www.dpreview.com/forums/thread/3732635?page=2#forum-post-54567793
if you expand the resolution chart in comparometer view you'll be able to directly compare corners on IR website, that may lead you to opposite conclusion to the one you expressed here. (resol. chart and multi)

Direct link | Posted on Oct 15, 2014 at 10:36 UTC
In reply to:

Peng Bian: Anyone know what lens is on the 5100? I'm on my mobile can't see exif.

Well if this is the only way you can think of... by all means, for instagram's sake.
I posted crops to support my point, more persuasive then your dancing naked. Didn't think about that?
I refer you to IR comparometer with both cams resolution chart, & please check chart corners if you do that. That may cure your blindness. Btw. why you're tormenting me instead of checking facts?

Direct link | Posted on Oct 15, 2014 at 10:24 UTC
In reply to:

Tonkotsu Ramen: There must be some crazy sample variation. On the bottom left cards, the G7 X is so soft it looks like motion blur. This isn't the case from any of the other comparisons i've seen though, where I need to go back n forth to look for differences.

@Rishi,
worst thing is when instead of identifying the issue and saying "we'll verify that" or "we'll look into this" you come up with lengthy explanations to justify your position and then later when confronted about specifics you blame the messenger for pointing it out to you and DPR.
My impression is that you'd rather quickly come up with some line of defense on the forum and put out the fire right away then diligently verifying stuff on your end to make sure you're not misinforming the public. If, say, out of 3 cameras all 3 displayed this type of lens distortion it's a flag for users about some serious inconsistency in design or manuf. process. That is why public checks reviews to avoid disappointments. This is also your journalistic duty to spell it out clearly. It seems that you'd rather engage in exchange here instead of doing your DD. Quite strange and disappointing.

Direct link | Posted on Oct 15, 2014 at 09:43 UTC
In reply to:

Sdaniella: most small sensor systems lenses have considerable lens aberration correction, especially if going very wide fov, so it starts to get harder to tell how much of the distortion is from variation in optics, or from the standardized correction of those expected distortions as designed.

the Canon G7X RAW (seen shown elsewhere, shared on DPR, from a French test comparison webdite, forgot which) at 24mm eq, looks extremely distorted, its a wonder how it can possibly be corrected fully 'rectilinear' in JPEGs without plenty of evident softness in the corners (and at the edges of the frame).

most of it, would not be of great concern, knowing such massive correction is going on, on corner details where most are likely to ignore, and not notice unless specifically looked for, or typically removed during 'creative cropping' for final work presentation.

anyone choosing from compacts Canon S120 to Canon G7X will cheer 'yay'!

ditto: LX7 to LX100 (a bigger sensor size jump; yields bigger IQ gains)

well, I do & sorry if that tone offended anyone, it's space limitation that's kind of putting me in a different mode, can't help but I prefer specifics, proofs and examples...

Direct link | Posted on Oct 15, 2014 at 09:12 UTC
In reply to:

Tonkotsu Ramen: There must be some crazy sample variation. On the bottom left cards, the G7 X is so soft it looks like motion blur. This isn't the case from any of the other comparisons i've seen though, where I need to go back n forth to look for differences.

I'm pretty sure IR and other sites will do what you have missed as they post more test pics with different FL/F values as well as multiple test scenes as opposed to just one here. No in-length explanation or statement of intentions will supersede test pictures so if viewers cannot find what they came for here they will go elsewhere, it's that simple. I'm just surprised by myopic point of view you presented here.

Direct link | Posted on Oct 15, 2014 at 09:00 UTC
In reply to:

Sdaniella: most small sensor systems lenses have considerable lens aberration correction, especially if going very wide fov, so it starts to get harder to tell how much of the distortion is from variation in optics, or from the standardized correction of those expected distortions as designed.

the Canon G7X RAW (seen shown elsewhere, shared on DPR, from a French test comparison webdite, forgot which) at 24mm eq, looks extremely distorted, its a wonder how it can possibly be corrected fully 'rectilinear' in JPEGs without plenty of evident softness in the corners (and at the edges of the frame).

most of it, would not be of great concern, knowing such massive correction is going on, on corner details where most are likely to ignore, and not notice unless specifically looked for, or typically removed during 'creative cropping' for final work presentation.

anyone choosing from compacts Canon S120 to Canon G7X will cheer 'yay'!

ditto: LX7 to LX100 (a bigger sensor size jump; yields bigger IQ gains)

I'd personally prefer if you named the site, the picture and refrain from general thoughts which you cannot support by any serious argument either scientific or statistical. So put in the time please, find the site and picture and let us judge for ourselves instead of you spreading the "word of truth" here. You don't know to what degree cameras correct "in general" or "in this specific case" so why these wild sweeping statements unless you can attach some specifics to them. This all of course if you want to be taken seriously.

Direct link | Posted on Oct 15, 2014 at 08:46 UTC
In reply to:

Discovery Of Light: Something is up with the G7X. They seem very blurry especially around the playing cards. Did somebody run into the tripod?

don't forget these are DSLR priced compacts (!) designed to compete with the above, do not throw them in the same basket with $150 compacts or if you believe otherwise test them against cheap compacts, downplaying issues with sample cameras is not helpful

Direct link | Posted on Oct 15, 2014 at 07:05 UTC
In reply to:

Zeisschen: - the APSC-sensor in the a5000 is still quite a bit ahead of all of them
- the lens on the a5000 (which one was it) is much better (raw)
- the raw performance on the G7X and the RX100 looks very similar, which is expected due to the same sensor
- the Panasonics raw performance is between the two 1 inch sensors and the APS-C sensor like expected
- the Canon lens show massive CA compared to the RX100 (check raw)
- the Panasonic lens is between the Sony and the Canon (at this focal length)

Overall picture quality order for me based on those samples as well as the earlier posted picuter samples:

1. Sony a5000 > best sensor and best lens
2. Panasonic LX100 due to better ISO performance than The Canon and Sony RX100
2. Sony RX100 due to better lens compared to LX100 based on picture samples, exqualizing the slightly better ISO performance of the LX100 for me
3. Canon G7X > ISO performance like RX100 but lens show some weekness in CA, no suprise for a Canon lens.

you just said it that "your copy seems better then the test one", loose "quote" sorry but this is what you wrote in essence, use your copy then, it's that simple
if 1/3 of the picture is affected it is very much out of the ordinary especially for a test website, may not be out of the ordinary with a casual user buying his/her first camera if this is what you meant, but keep in mind you are giving ratings based on your test, so testing HAS to be done with due diligence and extreme caution to be believable, you're not an average user, YOU ARE THE AUTHORITY and that means responsibility and accountability equally to the users and camera makers

Direct link | Posted on Oct 15, 2014 at 06:59 UTC
In reply to:

Peng Bian: Anyone know what lens is on the 5100? I'm on my mobile can't see exif.

prove it

Direct link | Posted on Oct 15, 2014 at 06:50 UTC
In reply to:

proxy: I checked at IR and none of the camera lenses (G7X and RX100III) show misalignment there. It's DPR that overlooked it. Overall G7X and RX100III show similar performance with Canon showing minimally higher in-camera sharpening. The result is razor sharp, better edges on G7X then Sony. Higher Canon CA's too in both DPR and IR tests (software correctable with lens profile). Very good lens indeed on G7X.

that's what I did, not only I did that but also wrote about CA and confirmed that. Yet it is still far from "horrible" as you put it.
Issue is that CA can be corrected with lens profile (1 click in post even if you don't use it) and if you DO use lens profile it's done automatically without your input whereas lack of sharpness as displayed in the crops can not. Big difference.

Direct link | Posted on Oct 15, 2014 at 06:45 UTC
In reply to:

Tonkotsu Ramen: There must be some crazy sample variation. On the bottom left cards, the G7 X is so soft it looks like motion blur. This isn't the case from any of the other comparisons i've seen though, where I need to go back n forth to look for differences.

agree, this is to confirm bad camera sample used by DPR, yet G7X corner performance is still better then RX100III, check other corners as well, G7X is consistently better/sharper

Direct link | Posted on Oct 15, 2014 at 06:33 UTC

Since no pics can be posted here I updated the crops with more samples.
Since my first crops post all the critics/bashers of G7X disappeared together those who proclaimed RX100III early superiority especially regarding the lens.
Now, here's some more for you to consider. Check this link again.
http://www.dpreview.com/forums/thread/3732635?page=2#forum-post-54568080

Again, these are DPR test crops, no changes. Enjoy.

Direct link | Posted on Oct 15, 2014 at 06:03 UTC as 55th comment
Total: 75, showing: 1 – 20
« First‹ Previous1234Next ›Last »