Reviewers have negative bias LX2 vs F30

I've noticed a definite anti-LX2/pro-Fuji F30 bias in several reviews. And it hits the greatest differential at ISO100-400. Because I'm lazy (and as a warning not to take anyone too seriously without doing your own follow-up research), I'll demonstrate this using just one reviewer's sample photos. But that reviewer's results are no different than the other reviews I have read, so he is largely in accord with his profession's standards of evaluation, and the photos I'm about to point out actually appear to support his case, at first. So I feel this is just part of the human condition.

Now for it:

Go to DCRP Reviews and look at their Fuji F30 night shots of San Francisco( http://www.dcresource.com/reviews/fuji/finepix_f30-review/ . The 400 ISO photo is here:

http://www.dcresource.com/reviews/fuji/finepix_f30-review/nightshot400.jpg

Now look at their same night shot for the LX2 ( http://www.dcresource.com/reviews/panasonic/dmc_lx2-review/ . Here's the 400 ISO photo:

http://www.dcresource.com/reviews/panasonic/dmc_lx2-review/nightshot400.jpg

Regarding the F30 night shots, the reviewer comments:

"The night shot looks very good... The camera took in plenty of light... The buildings are all nice and sharp, and noise is minimal. There's a fair amount of purple fringing in the photo... Things look very clean though ISO 400."

And regarding the LX2, the same reviewer comments:

"The LX2 turned in a fairly good night shot... The camera took in plenty of light, everything is very sharp, and purple fringing is at a minimum. The issue here is ... the loss of detail from too much noise reduction. Look at the edges of the buildings -- they look smudged. Also, the sky is quite mottled, with some banding in places... Since the LX2 already has problems with fuzzy details at ISO 100, you know it's only going to get worse..."

The reviewer continues:

"You don't have to be a rocket scientist to describe what's going on here. At ISO 200 details start to really get fuzzy, looking more like a watercolor painting than a photograph. Things only get worse at ISO 400... Needless to say, I would never take the LX2 above ISO 200 when shooting in low light, unless you're absolutely desperate."

So, when you look at the 100% crops, things at first do appear to be just as the reviewer says. That's part of the problem. You need to look at equivalent sizes. I opened both ISO 400 shots in Microsoft Photo Editor (came with Windows). Then I resized as follows: F30=100%, LX2=70%. What did I see? Try it yourself. See if you can spot which camera smears detail more. In many places, the F30 looks less detailed, but I'll call the two nearly equivalent because the F30 took in less light and the smearing is more noticeable in those darker patches. I'd give the edge to the LX2, though, but that could just be my bias. Look for those fuzzy building outlines the reviewer mentions. It's not so clearly just an LX2 phenomena, is it? In fact, the noise reduction applied to both cameras seems to have much in common.

Nevertheless, the reviewer says he "would never take the LX2 above ISO 200", but the F30 looks "very clean though ISO 400".

Now compare the ISO 100 shots on both cameras:

http://www.dcresource.com/reviews/fuji/finepix_f30-review/nightshot.jpg

http://www.dcresource.com/reviews/panasonic/dmc_lx2-review/nightshot.jpg

Again, use 70% for LX2, 100% for F30.

The reviewer says "the LX2 already has problems with fuzzy details at ISO 100". Take a good look at those two shots and consider just how much more details and sharpness are in the LX2 shot vs. the F30.

Makes it sound like you wouldn't want to use the LX2 at ISO100, huh?

You'll see similar language used throughout this forum as well as on several other reviewers' sites, but admittedly this was the most extreme case I could find. Remember, the pictures support this reviewer's case if you fail to take size into consideration.

It was because of this language and lack of precision in doing comparisons that I seriously considered the F30 and several Canon cameras. I was even hopeful about the A710 until I resized it to match the photo size of the LX2... and saw that it HAS MORE NOISE! Then I did the same for the F30, and saw that IT SMEARS DETAIL--BADLY! And I reevaluated from scratch.

Now I realize the LX2 will work better for me than I thought. Yes, I still may get the F30 for my wife to have a camera that can freeze action in low light. But this whole experience has been quite a revelation.

What on earth are you trying to say here? That a reviewer is being biased because he views and rates the images based on looking at them at 100% magnification? Or that the LX2 should only be judged as a 6.x megapixel camera?

--
http://psygnotist.smugmug.com

MrXpress said:

What on earth are you trying to say here? That a reviewer is being
biased because he views and rates the images based on looking at
them at 100% magnification? Or that the LX2 should only be judged
as a 6.x megapixel camera?

--
http://psygnotist.smugmug.com

Click to expand...

Correct on both accounts.

Obviously, I'm saying that if you expected the LX2 to perform as well as a 10MP SLR, you would be sadly disappointed.

Equally obvious is that you must therefore make adjustments in your expectations when you evaluate compact cameras.

Not so equally obvious: when a reviewer says F30 da' kine, LX2 puke, they aren't using an even playing field--far from it. I caught one such reviewer with an obvious contradiction to my own conclusions at ISOs 100 and 400. And I called it a "bias" because I disagreed and felt others would too.

So I claim an even playing field is key. I found the LX2 to perform as well as the F30 on one night shot at ISO 400. And it beats the snot out of it at ISO 100. Will it work for me? Hopefully. For others? Maybe not if they want an ubercamera that they think can punch out larger than 8x10 prints and still fit in their pocket. But for displaying on a monitor--no problem for most compacts. For printing at 8x10, problem for some, not others. That evens the playing field.

For viewing normal size on your screen or printing smaller sizes I would say it's good enough, but that's not what you buy a 10mp camera for (although of course you do have the advantages of 10mp on lower ISO's).

But for cropping it's a whole different story. Let's stay in line with your analysis and take the same ISO 400 shots and make a 2000x1500 crop of the F30, now take a 2560x1920 crop from the LX2 shot (to compensate for the mp difference. If you now look at both crops at the same normal screen size you'll notice quite a difference.

First of all, thanks for not just trying to flame-on!

I concur. If you want to blow up low-light/400 ISO pictures, you may have to look elsewhere. But two points concerning this:

1) I only said low-light/400 ISO. LX2 appears to be way superior at 100 ISO in all light, in most cases (this after exhaustive snooping around at others' photos--and I always ignore the on-the-spot user pronouncements that don't have photos to back them up--one man's porridge or however the hell the saying goes).

2) I'm not sure you will find another 10mp compact camera this year that will pass this test (or please let me know!). So you already have to throw in the towel and accept

CSPronken said:

For viewing normal size on your screen or printing smaller sizes I
would say it's good enough, but that's not what you buy a 10mp
camera for (although of course you do have the advantages of 10mp
on lower ISO's).

But for cropping it's a whole different story. Let's stay in line
with your analysis and take the same ISO 400 shots and make a
2000x1500 crop of the F30, now take a 2560x1920 crop from the LX2
shot (to compensate for the mp difference. If you now look at both
crops at the same normal screen size you'll notice quite a
difference.

Click to expand...

The two reviews the OP compared were written as separate articles
at different times. The reviews follow a consistent framework that
this reviewer has been using and refining for several years.

The OP's "analysis" hinges upon downsampling the LX2 images to
normalize them with the F30 images, While this does allow the LX2
images to compare more favorabily with the F30 images and it also
points out that there are techniques available to mitigate noise in
digital iamges that can be used to make LX2 images more presentable,
it does not "prove" any type of bias on the reviewer's part. He just
reported what he saw in each of the reviews.

Also, as a point of information, lest you think that this reviewer might
be a "closet pany basher" this same reviewer liked the TZ1 so much
when he reviewed it that he bought one for his father.

In any case, I think that public accusations of bias need to have a far
more substantial basis than what the OP presented here.

--
When a hammer is your only tool, all problems begin to look like nails.

I admit that the title I chose for this thread reads a bit too sensationalist. But I presented ample evidence that the standards for evaluation were different for the two cameras. And I encouraged others to do the comparison themselves. Finally, if you read my first post, you'll see that I discourage the reader from thinking that the reviewer is using lower standards than his peers, that in fact his conclusions are in keeping with those of his peers and apparently justified at that.

As the author of a public accusation of bias, I will now say that I should have labled this thread: "100% crop review has built-in bias against LX2 vs F30", but I expected people to read what I had to say first before responding.

Having said that, another of your points is a good one: standards change over time.

DigitalMaven said:

The two reviews the OP compared were written as separate articles
at different times. The reviews follow a consistent framework that
this reviewer has been using and refining for several years.

The OP's "analysis" hinges upon downsampling the LX2 images to
normalize them with the F30 images, While this does allow the LX2
images to compare more favorabily with the F30 images and it also
points out that there are techniques available to mitigate noise in
digital iamges that can be used to make LX2 images more presentable,
it does not "prove" any type of bias on the reviewer's part. He just
reported what he saw in each of the reviews.

Also, as a point of information, lest you think that this reviewer
might
be a "closet pany basher" this same reviewer liked the TZ1 so much
when he reviewed it that he bought one for his father.

In any case, I think that public accusations of bias need to have a
far
more substantial basis than what the OP presented here.

Click to expand...

Retitled.

If you want to know whether F30 is better than LX2, be careful to downsize LX2 to equivalent size or you will see more artifacts than actually occur for a given display or print size. Reviewers typically say what they see at 100%, but your results WILL differ.

and then it got posted as a new thread.

I give up. The "new thread" is once again a reply.

Now you have it right - it is not an issue of bias, but of consistently
applied methodology. Your choice of cameras to compare also added
to the confusion - the Fuji F30 is a 6mp camera that is widely noted
for having relatively low noise levels at high ISOs, while the Panasonic
LX2 is a 10mp camera that has a somewhat different reputation vis-a-vis
noise.

Comparing the image noise characteristics of this pair of cameras is a bit
challenging because of their significantly different megapixel ratings and
was made even more so because the images you chose are sort of a
worst case scenerio for noise because not only were they shot at high
ISOs, but they were also long exposures.

So what is the "take away" from all of this? I think that it is that the
LX2 images are affected by noise and paradoxically, noise reduction, at
hight ISOs. But, the effect of both of these problems can be mitigated,
as was pointed out by the reviewer at DCRP, by using RAW mode.

--
When a hammer is your only tool, all problems begin to look like nails.

Also worth noting is that the F30 is $290, the LX2 $440

--
http://psygnotist.smugmug.com

well, it seems you are partly right.

When people say how f30 beats the hell out of the lxs in low light conditions, they are basically comparing 100% crops, which is not fair to the LXs.

But, reviewers judge cameras for what they are, and while the F30 is a 6 MP camera, the LX-2 is a 10 MP camera and is judged for that.

That doesn't explain how the reviewer (in DCresource) arrives to the conclusion that at 400 ISO the LX-2 would be suitable only for small prints, while the F30 seems ok....

Yes, there are so many variables in how the two images were captured. A problem in both shots is that they were acually taken at night and not at dusk which is when anyone wanting to photograph city lights would shoot. It's a tough ask for any camera to cleanly capture spots of light against a dark background at a long exposure. I assume both cameras were on a solid tripod!

There seem to be exposure differences that tax the LX2 more but it is possible that some of the smearing seen in the lights low in the panny shot is caused by the lighting being seen through exhuast gasses from the cars on the highway that appears to be in that part of the pic.
--
lek

MrXpress said:

Also worth noting is that the F30 is $290, the LX2 $440

Click to expand...

I was waiting for someone to make that point. But everything's relative until you put your conclusions in print, at which time the great sin is committed and you are held to task for it.

For those who are still entertained by this: What brought this on was a deluge of questions by my fellow shoppers at a Fry's Electronics. They saw me eyeing the LX2 (silver) and wanted to know why I'd buy it instead of X-Y-Z on account of the LX2 being so noisey. Isn't it funny how people with common but very specific "shared" experiences come out of the woodwork?

So, I found myself justifying various cameras for reasons that these fellow consumers had not even considered in much detail. They were just going on "good review" vs. "bad review". "Noisey". "Canon reputation". "BIG zoom". But not things like "useability", "consistency", "natural color", "reliability", or nuances like "detailed under ISO 800", "freezes action over ISO 400", etc.

For those who can't read between the lines, there is no simple message. For those of us who can, we know what works for professionals, what works for serious hobbiests, or duffers (like me), and how to discern when a reviewer is addressing one or the other. But without a lot of deliberation and consideration, I never would have arrived at my own personal methodology for comparing shots, and I doubt many others in the same buying category would be able to split these not-so-fine hairs to the extent that I have without a similar set of techniques.

Maybe in two years I'll get my dream "compact" camera (fixed lens, size halfway between LX2 and L1, performance of L1).

Pretty fair..its clear to me that the LX-2 is using heavy nr to cover up noise, when what it needs is less mp and a better sensor.

Saying that the fuji is far from perfect either....
--

I pulled both images up side by side on my calibrated monitor and enlarhed them. To me the LX2 looked a lot clearer and more detailed than the F30. The LX2 took in more light also. I could even see inside some of the offices in the LX2 images although not with much detail.

Don

In two of the pictures I downloaded them and opened them so I could read all of the EXIF info and here is just a little of what I got.

f30 [5 sec] [f num 5,00] [100iso] [scene capture type night scene]
[exposure program landscape mode] [time taken 5/29/2006 22:08:13]

lx2 [6 sec] [f num 4.90] [100iso] [scene capture type standard]
[exposure program shutter priority] [time taken 9/11/2006 09:20:54]

This is just some of what I got from both of pictures and I say you can't compare two night shots taken 4 months apart with different exposure times 1 second difference which is going to make a difference and also 4MB difference at night OH NO there is no favorites.
--
Joe D
God Bless And Have A Nice Day



my Little Girl Shyanne



S&W 4' 500 Magnum
A Winning Pair

I think the review was fair and I do a lot of photo enhancing so I followed what you said and I came to an opposite conclusion.

The F30 is both resoluting better detail at its resolution EVEN with the LX2 image down sized to its level. And this is most likely attributed to the fact that the LX2 just has too much noise and yes even some at ISO 100 that in Night scenes imparticularly it just can't deal with too well when it comes to detail vs noise. By no means is it just terrible but with the right lighting conditions for the camera, considering its lens and sensor the LX2 I think can undoubtedly do better in well light out door situations.

Here is a pic of the two, F30 on left and LX2 on right and I zoomed in 200% after downsizing both down to the F30's level. If anything the LX2 should have resoluted far better or at least equal clarity and even less noise than this at this level.



(I did not apply any bicubic type of resampling on the way up so there is no induced blurriness from any sort of algorithym only pixellation)

I think this is just another case of people mistakening noise for detail... I try hard to make sure I don't go overboard with Noise Reduction software but I most definitly do not slouch while getting rid of the noise, I see a lot of people even after applying Noise Reduction still leaving fairly visible signs of noise that could have been removed, in particular sky's are the most noticeable areas that too much noise gets left behind.

Also yes the atmospheric conditions between shots are undoubtedly different so is the color balance from within their cameras but with a little bit photoshoping you can get them to match color, contrast, and lightness/darkness pretty easily and it is even more evident of how much sharper the F30 is in this situation.
--
glitchbit: getting rid of the glitch bits