Author Topic: Lets pixel peep some JPEGs... (Read 3188 times)

lisandra · « **on:** June 07, 2011, 03:38:24 PM »

There's no shame in it. Sure I might say the contrary some time in the future (or past) but really, its your money. And its not 15-20$ were talking about, cameras cost anywhere from 600-1000$ so you should be able to scrutinize a product as much as you want.
That said, dpreview posted the studio scene comparison of the G3 and to me it was so amazing I deemed it worthy of a topic. These are still JPEGs but amazing nonetheless. I know I know, dpreview kinda sucks but all were doing is looking at photos here.
Before going there, know this. First you have to choose the G3 from one of the tabs, if you want my suggestion use the tab thats using the 60D and keep the D7000 and the A55. Second change the position of the 100% crop square to either the old lady face (best), or the adonis looking statue face. These are areas with lots of detail and the coins are a bad choice. Switch over to ISO 1600 (switching on top switches all of them simultaneously). The GH2 and the G3 look pretty much alike, and they absolutely murder their APS C counterparts (A55, D7000). Switching over to ISO 3200 is where things got "wow" for me. The G3 clearly does better in retaining detail tan the GH2 and the D7000 (the A55 blurs everything to hell by now) so I boldly changed the A55 to a K5. I still preferred the G3. But what about ISO 6400? surely m4/3s cant keep up with that K5 at this ISO...it does, to me it looks better, even if its by a bit (its more obvious in the adonis face). Go see for yourself and post your amazement here, Ill wait. THIS is the camera that should have had ISO 12k (or higher).
http://www.dpreview.com/reviews/PanasonicDMCGH2/page18.asp
P.S. just for kicks, after doing all of that, go ahead and change one of those cameras for a 5D mk II (at ISO 3200-6400). Make youre own judgement on which looks more detailed, but me, I still preferred the G3

rparmar · « **Reply #1 on:** June 07, 2011, 06:04:30 PM »

OK, you convinced me to have a look. Unfortunately comparing JPG is useless to me, since that just compares the built-in tweaks and settings that could be nothing like the actual potential of the sensor or how I would use it.

Second is the issue of lenses. If these shots use the kit lens I am uninterested, since such lenses do not show off the cameras.

The third problem is the implementation. The Pentax K-5 seems misfocused. Or something else is wrong, but it is definitely not working to par.

Between all these factors I am sorry to say the comparison is completely useless to me. The softness of the K-5 even at ISO 100 in RAW tells me something is seriously wrong. And even so, the E-P1 is distinctly sharper and more contrasty than the G3.

cosinaphile · « **Reply #2 on:** June 07, 2011, 06:46:59 PM »

lisandra , im right there with you , i take exception with much at dp review , illogical scoring and sometime an agenda in a review from where i stand

but still a useful recource especially with the updated comparison tool
so dp review is a guilty pleasure for me
i love to pixel peep, simply to understand better a sensors
potential for better images
dynamic range and high iso behavior as well as differing jpeg approaches and cooking of various makers

i would love that sensor or something jusyt as good in my next pen or panny
rangfinder formfactor body
my ep1s and gf1s are gettin old,still lovem though

lisandra · « **Reply #3 on:** June 07, 2011, 06:57:39 PM »

Quote

Unfortunately comparing JPG is useless to me, since that just compares the built-in tweaks and settings that could be nothing like the actual potential of the sensor or how I would use it.

A JPEG is a processed RAW file in camera, so yeah, to me it shows the sensor potential. I seriously doubt you can do better processing a high ISO RAW file with software .

Quote

Second is the issue of lenses. If these shots use the kit lens I am uninterested, since such lenses do not show off the cameras.

They use the best prime their respective company has to offer. On pannys/olys the use the 50mm f2.

Quote

The third problem is the implementation. The Pentax K-5 seems misfocused. Or something else is wrong, but it is definitely not working to par.

The issue was brought up in the K5 review, its not a misfocus because they manually focus. They say Pentax "learned" from the K7 and now implements a hefty amount of noise reduction. Even to RAW files. This is not the only review that found this, they tried to fix it via firmware but it didnt do much.

Quote

And even so, the E-P1 is distinctly sharper and more contrasty than the G3.

At ISO 100-200. I was going for the whole high ISO thing. If you set it at ISO 3200 at the adonis face the G3 looks like the E P1s ISO 100. The PEN at ISO 3200 is just a big blur.
Its either that or my monitor favors panny files. lets see... More opinions please!!

Ray · « **Reply #4 on:** June 08, 2011, 01:26:05 AM »

I have choosen and bought my first digital camera after studying lots of reviews of different cams on dpreview and by following their 'conclusions'

I was foolish enough to repeat it all a second time when I didnt end up happy with that first cam...

So the camera after those two I tested personally at the local camerastore before buying, those guys must have thought me to be a real pain (and I am really

) bringing my own memcards/laptop shooting the hell out of those demo cams pixelpeeping for hours, on different days even

But it really paid of as from then on I was happy with the cams I ended up with but nevertheless I do hope dpreview have it spot on concerning these G3 samples this time

Panther · « **Reply #5 on:** June 08, 2011, 04:08:48 AM »

One thing is certain, Panasonic should give all of you free gear for being so dedicated, regardless of all else!

watkins · « **Reply #6 on:** June 08, 2011, 04:22:19 AM »

To me, someone must be wearing Panasonic glasses to even think it compares.

Nikon wins as usual when it come to high iso

adash · « **Reply #7 on:** June 08, 2011, 04:27:22 AM »

Quote from: watkins on June 08, 2011, 04:22:19 AM

To me, someone must be wearing Panasonic glasses to even think it compares.

Nikon wins as usual when it come to high iso

I wear Polaroid polarized sunglasses and these seem the same at 6400 RAW tests:

http://www.dpreview.com/reviews/PanasonicDMCGH2/page20.asp

watkins · « **Reply #8 on:** June 08, 2011, 04:38:29 AM »

Quote from: adash on June 08, 2011, 04:27:22 AM

Quote from: watkins on June 08, 2011, 04:22:19 AM
To me, someone must be wearing Panasonic glasses to even think it compares.

Nikon wins as usual when it come to high iso

I wear Polaroid polarized sunglasses and these seem the same at 6400 RAW tests:

http://www.dpreview.com/reviews/PanasonicDMCGH2/page20.asp

Seeing as the g3 only has jpg, put the d5100, g3, and the claimed 5d mkii that are all the same and look over the image.

Nikon's budget d5100 or the d7000 wins to me for noise. You may have differing opinions to me and that's fine.

I'm happy to admit the downsides of the m4/3 cameras compared to the larger bodies,this is clearly always going to be one.

adash · « **Reply #9 on:** June 08, 2011, 04:47:38 AM »

Quote

I'm happy to admit the downsides of the m4/3 cameras compared to the larger bodies

The theoretical advantage is 1/2 of a stop (that's the size difference between the sensors) but then again, softness in D7000 and a55 astonishes me more.

Quote

Seeing as the g3 only has jpg

I wasn't pixel peeping G3. GH2 seems to be a good performer.

rparmar · « **Reply #10 on:** June 08, 2011, 09:36:33 AM »

Quote from: lisandra on June 07, 2011, 06:57:39 PM

A JPEG is a processed RAW file in camera, so yeah, to me it shows the sensor potential. I seriously doubt you can do better processing a high ISO RAW file with software .

The opposite is true. One can always do better than in-camera JPG by processing RAW files. I would have though this was common knowledge, else why would anyone shoot RAW? In particular, many manufacturers have hyped JPG processing designed to look good in reviews but which do not translate to accurate images.

Quote from: lisandra on June 07, 2011, 06:57:39 PM

They use the best prime their respective company has to offer. On pannys/olys the use the 50mm f2.

Thanks. Good to know.

Quote from: lisandra on June 07, 2011, 06:57:39 PM

The issue was brought up in the K5 review, its not a misfocus because they manually focus. They say Pentax "learned" from the K7 and now implements a hefty amount of noise reduction. Even to RAW files. This is not the only review that found this, they tried to fix it via firmware but it didnt do much.

They are wrong or misled. First, no K-5 images I have seen looks this bad except by user error.

Second, Pentax does not implement a "hefty amount" of noise reduction in RAW. What they do in JPG I cannot say or care about, as previously stated.

Perhaps the reviewer is confused by the fact the K-5 implements a stronger AA filter than in previous models.

Quote from: lisandra on June 07, 2011, 06:57:39 PM

At ISO 100-200. I was going for the whole high ISO thing.

Ah. High ISO is not representative of image quality to me. Instead it is representative of noise reduction. I apply my own noise reduction after the fact and so can properly tune it for subject and effect. I prefer a camera to give me the best possible untouched images so I can use my photographic discretion. The overly-smoothed high ISO images on display here are exactly the sort of thing I do not want.

Panther · « **Reply #11 on:** June 08, 2011, 10:22:44 AM »

Quote from: rparmar on June 08, 2011, 09:36:33 AM

[Ah. High ISO is not representative of image quality to me. Instead it is representative of noise reduction. I apply my own noise reduction after the fact and so can properly tune it for subject and effect. I prefer a camera to give me the best possible untouched images so I can use my photographic discretion. The overly-smoothed high ISO images on display here are exactly the sort of thing I do not want.

Quite a few people shoot in RAW with NR off....but this doesn't degrade inherent high ISO performance. A Nikon D3s with NR off is still a killer high ISO performer...

adash · « **Reply #12 on:** June 08, 2011, 10:23:49 AM »

Quote

A Nikon D3s with NR off is still a killer high ISO performer...

I wonder if anyone would use that camera with NR on....

Panther · « **Reply #13 on:** June 08, 2011, 10:25:54 AM »

Quote from: adash on June 08, 2011, 10:23:49 AM

I wonder if anyone would use that camera with NR on....

If you want,
I'll pose the question to D3s owners in another forum that I belong to....

adash · « **Reply #14 on:** June 08, 2011, 10:29:38 AM »

Quote from: Panther on June 08, 2011, 10:25:54 AM

Quote from: adash on June 08, 2011, 10:23:49 AM
I wonder if anyone would use that camera with NR on....

If you want,
I'll pose the question to D3s owners in another forum that I belong to....

Please do, if you feel like it. Note that I only care about in-camera NR, what users do when importing to LR is irrelevant to the discussion.

Panther · « **Reply #15 on:** June 08, 2011, 10:30:59 AM »

In camera is exactly what I'm talking about...

pretzston · « **Reply #16 on:** June 08, 2011, 11:00:21 AM »

I had never seen that comparison before! Neato.

I have to say though, at 160 vs 100 the GH2 blows most of them away with one of the exceptions I saw being the EP1 and EP1, very impressive as well. At the upper spectrum it easily gets crushed by the heavy hitters. Impressive though nonetheless!

rparmar · « **Reply #17 on:** June 08, 2011, 11:06:05 AM »

Quote from: Panther on June 08, 2011, 10:22:44 AM

Quite a few people shoot in RAW with NR off....but this doesn't degrade inherent high ISO performance.

Yes, I realise that.

I was referring to the obvious in-camera NR in the G3 images under discussion.

lisandra · « **Reply #18 on:** June 08, 2011, 11:56:02 AM »

Instead of getting all defensive and stuff, lets establish some personal parameters on what were talking about. To me a good high ISO performance is not the one that shows the least amount of noise (in JPEGs), its the one thats does so without loosing an exorbitant amount of detail or blurring the image to hell. The closer to detail as the original to me the better. I dont have the RAWs, I dont know where to get them, so lets stop being redundant and concentrate on what we have. Let me illustrate my point (you HAVE to click on them):
G3 ISO 160

Details

G3 ISO 3200

Details

Here on the G3 the ISO 3200 is not far of from the 160. Where? the pattern in the adonis face is still intact and although somewhat worse on the old lady its still there
D7000 ISO 200

Details

D7000 ISO 3200

Details

Unlike the G3 (to me) the D7000 ISO 3200 looks blurrier and duller than its ISO 200 counterpart. The pattern on the adonis nose is gone (retains it on the cheek) and the pattern on the old ladys face is completely gone. Is there less noise than the G3? yes, but again, at the cost of these details.
A55 ISO 200

Details

A55 ISO 3200

Details

Tell me this is better than the G3...ok no sarcasm... the pattern on both the adonis and the old lady are completely gone, the image is squeaky clean, but its blurry as hell and the detail is all gone.
Those are my criteria when evaluating in camera JPEGs. If that makes me have panny glasses then so be it, but I prefer detail and pixel sharpness above all else. Extra noise I can always get rid off, but details I cant get back.
And those of you who think processing is not applied to RAW files are being fooled. Look at the A55s RAWs, there's clearly something going on behind curtains (based on detail loss).

rparmar · « **Reply #19 on:** June 08, 2011, 12:08:41 PM »

I am not getting "defensive" and wonder why the insinuation? But I have plainly stated why comparing JPGs is pointless. I have seen no counter-argument. Furthermore comparing images from an obviously flawed testing regimen is doubly pointless. I will bow out of this conversation so as not to cause more upset by pointing out the obvious!

Ray · « **Reply #20 on:** June 08, 2011, 12:50:57 PM »

If I may be so bold to step in here its obvious both of you are standing your ground, and with good reason, because from my point of view the both of you are right in your own way. My english most unfortunately isnt as good as my own native language so please excuse me for being brief about it, it would take me twice as long to explain myself compared to you, but both of you are stating the obvious to me. As far as I'm concerned you cant tell camera's apart solely by comparing their Jpeg output however the camera with the best (in-camera) processing results is the better camera, but just in the Jpeg algorithm department, Raw output is an entirely different beast and clearly will have its best peers as well.

lisandra · « **Reply #21 on:** June 08, 2011, 02:14:36 PM »

Quote

The opposite is true. One can always do better than in-camera JPG by processing RAW files. I would have though this was common knowledge, else why would anyone shoot RAW? In particular, many manufacturers have hyped JPG processing designed to look good in reviews but which do not translate to accurate images.

This is right 95% of the time my friend. But then people started noticing they couldnt get the look of oly EPL1 at high ISO from RAW no matter how they tried.

Quote

I prefer a camera to give me the best possible untouched images so I can use my photographic discretion.

Theres no such thing, I use to think teh same, then panther pointed out that even in RAw files in the analog to digital conversion manufacturers dont leave at as it is. Im with you on this actually, I prefer the GH2 over the A55 (I have both) because at high Iso the GH2 has every glorious detail in RAW (noise included) and the A55 has a bit less noise, but detail (in RAW) is not all there.

lisandra · « **Reply #22 on:** June 08, 2011, 02:18:04 PM »

Quote

I am not getting "defensive" and wonder why the insinuation?

Im not saying you are, I was referring instead of me getting defensive and replying that way Id rather demonstrate why I think like I think.

Quote

But I have plainly stated why comparing JPGs is pointless.

I dont compare JPEGs either usually, I shoot RAW only. But JPEGs are important to most people and right now its all we have. Most people reserve RAW for important stuff and mostly shoot JPEG. I cant wait for the RAW files to be posted too.

lisandra · « **Reply #23 on:** June 08, 2011, 02:21:51 PM »

Quote

If I may be so bold to step in here its obvious both of you are standing your ground, and with good reason, because from my point of view the both of you are right in your own way.

Thing is, Im not disagreeing with the dude, Im a RAW shooter myself, but I do disagree with him in the statement that comparing JPEGs is useless. To me the better the JPEG the better the processor (a good thing) or the easier it was to process (a cleaner file, a good thing too). And besides, Im comparing JPEGs because there are no RAW files available.
Lets do JPEGs for now and when RAWs are up I promise Ill post them.

lisandra · « **Reply #24 on:** June 08, 2011, 02:25:22 PM »

Quote

I will bow out of this conversation so as not to cause more upset by pointing out the obvious!

You dont have to man! Im not attacking you or taking offense. In fact I rather like people not seeing eye to eye with me, gives me a different perspective.

Author Topic: Lets pixel peep some JPEGs... (Read 3188 times)

Panther

Panther

Panther

Panther

Related Topics

Recent Topics

Recent Gallery