April 18, 2014, 02:35:53 PM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Aglet

Pages: 1 ... 28 29 [30] 31 32 ... 55
436
EOS Bodies / Re: Which is better? 5D MKII or 6D?
« on: January 11, 2013, 12:17:11 AM »
You misunderstood what I meant (I think), Canon can  with several more rows of black / covered pixels offset the accruing noise already  at the readout.(CDS)  Nothing to do with PP work

I did misunderstand what you meant, i read too fast.
Thanks for clearing that up for me.

Yes, that's a perfectly feasible method and may even be doable even in firmware as-is with the existing masked pixels?

437
EOS Bodies / Re: Which is better? 5D MKII or 6D?
« on: January 10, 2013, 05:41:47 PM »
Read noise and fixed patter noise   can easily been level out / removed with black frames,..
true, but that's more PP work than I want, or any of us should have, to do to get a decent low ISO image

There are also banding  where  the read out levels are  uneven calibrated  to the ADC , like the  problem with the first 7d.
Sold my 7D because of that, tired of those 8 pixel wide strips in the shadows because the dual readout channels were not properly matched

If Canon would  they can limit the read-out noise, others with the same type of readout and signal path to the ADC can do it. ..
That's what I'm thinking. There are other sensor systems out there that are made more like Canon's than Sony's and they still perform better than Canon.  Obviously the technology exists for Canon to improve their read noise, and their read noise variation, even without spending a lot on developing a sensor as good as Sony's, etc.

The 6D is the best for reduced low ISO pattern noise I've seen since the 50D came out.  Their earlier cameras did not seem to suffer from FPN nearly as much.  Seems that when they added video capability, that's when things started to go sour.  Even the live-view capable 40D and 400D didn't have the FPN isssues as bad as later models like the 7D, 5D2, etc.

438
Arghh!
All this juicy sounding Fuji talk is making me want to buy one to see what it can do!
Not that I wouldn't anyway..  Was just a matter of when. but by the enthusiasm you guys have for it I might boost it up my priority ladder.  I already know the IQ can be very impressive from them but the overall handling sounds to be pretty good and I'd like that.
Gheez.  And I'm still waiting for the silly Pentax K-01 I ordered for dirt cheap.
Guess I'll wait and see what the new X20 performs like.  I'm not keen on spending midrange SLR prices for the bigger X bodies as I don't mind carrying a small SLR around most times as it is.
If that little X20 really performs for IQ and overall handling in its slightly smaller package that could be something I'd enjoy.  I already like the IQ from my G11 and G12 but the ergonomics and overall handling are really quite poor IMO.

Arghh.  And there's that Olympus OMD EM-5 too!
another very good overall package, altho a bit more than I'd care to spend right now as I'd want the grip and extra battery layer too.

439
EOS Bodies / Re: Which is better? 5D MKII or 6D?
« on: January 10, 2013, 12:24:57 PM »
here is what happening" since 2008" sensor  and  todays 6D sensor
improved  QE, higher FWC, but still a lot of read out noise  and low DR at base iso,
improved high iso, and less banding at base iso

+1
however, the slight reduction in 6D's read noise figure does not tell the whole story.

We don't know HOW the read noise is calculated.  To me it seems they're taking an AVERAGE read noise value instead of a peak-to-peak read noise value.

The difference being you can have 2 levels of average read noise that are very similar, as in the 5d2 and 6d, but they can have different peak-to-peak values.
The peak-to-peak read noise levels are more relevant because the peaks are what become more visible in shadow areas.
In the case of the 5d2 the peak read noise occurs repetitively when going across the sensor horizontally, which results in a pattern of vertical noise stripes visible at lower ISO settings.  These stripes get obscurred by larger amounts of relative random noise as you increase ISO.
The 6d's low ISO readout noise is much more uniform from pixel to pixel, resulting in a smoother looking tone in the shadow areas.  This sort of noise also responds better to NR software so you lose less actual image detail because you don't need to smear the image so much to obscure the banding structures.

In this respect alone, the 6D is quite far improved over the 5d2 and even the 5d3.

As an overall package, I'd still consider the other improvements and refinements of the 6D over the 5d2 as significant.
I'm selling my 5d2 to fund a 6d. The 6d is Canon's best IQ per cost FF camera and if you don't need the performance features of the 5d3 or 1dx it's a no-brainer.
I don't care about video so haven't compared it.

440
I seriously doubt that the RX-1 will have much better IQ, but we'll have to wait for the reviews.


Posted yesterday by DxOmark - the RX-1 kicks serious butt!

I'm not crazy about their mysterious overall score number but the raw sensor scores still tell the story.  It's a very technically capable little camera, as it should be for that price!

www.dxomark.com/index.php/Publications/DxOMark-Reviews/Is-the-RX1-the-compact-photojournalists-are-waiting-for

441
I love all kinds of cameras and have been lusting for one of these new Fujis for a while.  Altho I've seen an earlier model best my 5D2 for IQ at times I was just not familiar enough with them. The cost was high, and the raw workflow would be another direction I'd have to change; too many little factors held me back from purchasing.  I did like the way the X100 felt in my hand, if a tad "light."

The new X20 is hitting a price point where I could justify yet another camera in my collection and figure out what niche it best fills later.

Definitely not going to forget about Fuji now!

442
EOS Bodies / Re: What Canon Body is threatened buy the Nikon D5200?
« on: January 09, 2013, 01:02:28 AM »
all canon crops.
it not a competition anyway. its an outright flogging..

+1  ;)
and at a good price point too.
Let's see how the new, Toshiba-made, sensor performs.  It's supposed to be an improvement over the D3200's.
I'm certainly watching this one closely but won't pounce on one until the reviews are out, AF performance shows its worth in this body.. and maybe even the price drops a little.
LOTS of features in a compact body!

443
Nikon  has now another partner besides Sony, Aptina Renesas and it's Toshiba. The sensor in the camera is a Toshiba manufactured sensor.

I was just about to post that. :)
Toshiba's been out of the limelight in electronics for a while.  I hope Nikon was not exaggerating when one of their people said the d5200's sensor will not only be different from the d3200 but will also be measurably better.
Toshiba has the fab and tech to produce pretty much anything from heavy power devices to the best sensors so hopefully this is another source of top quality silicon fabrication for Nikon to tap into with their patents & licensing.

Which now leaves us wondering, what sensor and what will be the performance of it when it comes to Nikon's pro-level crop bodies; D400? D7200?..  They should better the consumer bodies by a little bit.
And is Nikon waiting for Canon to announce 7D Mark II or 70D before they release these new crop bodies?
Waiting game for all of us now.

Meanwhile, if anyone's looking, Pentax is selling off their Q and K-01 stock in north america and likely around the planet at really low prices.  I just ordered a K-01 with 40mm kit lens for a pittance.  Looked forward to finding out what it's good for; video perhaps.

445
Got to play with one a bit at the shop today, shot some dark frames and pushed them as i usually do to better visualize the pesky noise patterns.

The 6D is considerably improved over the 5d2 and even 5d3.  Although the total amount of dark level noise is not so greatly reduced, the subjective annoyance of its patterning is noticeably less.

5d2 could produce some fairly harsh noise lines (hor. & vert) in shadows at 100 ISO, they soften a bit from random noise at 400 ISO and up.

6d, at the same ISO settings, produces far less obviously structured noise lines.  THIS IS A GOOD THING! :)
Canon has actually made some significant improvement in this area over the older cameras!

6d's overall noise is more random and uniform, tho still exhibits some fine vertical stripe structures at 400 ISO, similar to 7d's low iso noise but not as severe.  Overall there is still some coarse but smooth transition variations in horizontal and vertical axis, almost like large blotchiness.  This would not show up on large prints anywhere near as bad as 5d2 or 7d noise stripes.

6d's subjective total noise at 400 iso is similar to 5d2 at 100 iso.  And the 6d's noise would clean up better with NR software.

6d's high iso performance is impressive!
I could print a decent 6x9" from a reasonably exposed 25600 iso shot! (larger if not too fussy)

So, altho 6D's dynamic range will not be significantly better than 5D bodies at low ISO, its raw files should provide more post-processing leeway because of the reduced severity of patterned noise. Its extra performance at high ISO is a definite bonus.
Fussy shooters who do a lot of post-processing should definitely consider the 6D over either 5D body if low ISO performance is critical to their work.

 SoNikon sensored cameras still do blow the 6D into the weeds with far cleaner low ISO data.
Hopefully Canon's (soon to be released?) next generation sensors will further improve on this dark/read noise issue.

As it is now, I am very tempted to sell my 5D2 to fund the 6D, it's that much more usable to me.

446
So would it make you feel better and as though you've accomplished something with your life if I say, "Ok, you win."
only if you actually understand, otherwise we're wasting time
Good job.  I actually have important things to do, like my career.
  . . .
  By the way, I bought the fan boy hype of the 5D2 and bought early and I wasn't disappointed.  Maybe photography is what you should focus on, rather than camera models?
35 years of photography and hybrid electronic, data-acquisition and mechatronic systems design experience, all apparently wasted.
Maybe I'll try again when I retire. ;)

447
@ RLPhoto - nothing personal but this near diatribe I've written here is a culmination of of seeing the same hash from the same multiple sources (pg7, pg8), you just happened to have been under the quote button.  :)
I believe you are over blowing the noise issue. All my landscape photos I submit to istock photos came from canon cameras and they have some of strictest standard for files. They do quite well for me.

If you get your exposure right, there should be no issues.


sigh... again... ???

I'm glad istock likes your stuff but that doesn't trump what I and a few other people here have said, repeatedly, ad nauseam about FPN (fixed pattern noise) problems with Canon cameras.

If you're not pushing the limits of your raw files for any artistic or DR compression purposes (and it's not all abstract, really) then you may as well shoot jpg because you'll not likely notice the difference.

It has NOTHING to do with "getting the exposure right."

It has only to do with FPN weakness of the imaging system.

Just because what you, any many others, do works for you, doesn't mean it works for everyone.
We don't shoot the same subjects, we don't shoot the same way.  If we did, it'd be pointless for us to both be doing it.  What you do is not any more "correct" than what I do, it's just different, and the tool you find adequate does not work well for me.

You want to paddle a canoe with a canoe paddle, sure, works good.
You try paddle a kayak with a canoe paddle, it's not so good any more, is it?

Not all bodies of a given type exhibit FPN at the same level.
I have an early 5d2, I had an early 7d.  They both sucked with serious FPN, and so did many other bodies produced in the same time frame. (& yes, I've complained to my local Canon rep directly)
You (RL) may have lucked out with cleaner versions of these same bodies.  I, and many others, did not.  And the way I want to use the gear I paid good money for is compromised because of these problems.  That initially rendered some very expensive outings and shoots a serious loss because I do not accept images with this kind of flaw and even sophisticated post-processed is unable to adequately ameliorate the problem.
FWIW, my 40D, 60D, 350D, 400D, 450D, G11, G12, and needless to say my recent Nikons and even my new Pentax Q, suffice for the same kind of "extreme" shots the 5d2 and 7d fail at because they don't have FPN to the same extent; so can you still tell me it's my technique?  Part of my fun comes from pushing the limits of low end cameras to get good images.  It's pretty disappointing when "high end" cameras have worse IQ than some very much lower end cameras.

The simple fact is that my 5d2, even with latest firmware, shows FPN in shadows of PROPERLY exposed images, even without any significant shadow lifting. It's not the only lousy 5d2 either.  Plenty of people have noticed this same FPN issue, they've posted it in these forums, they've mostly all been rebuked by the regulars, some of which should have the technical knowledge to know better since I've seen such demonstrated regularly.

I'm still hoping Canon will pull a rabbit out of their hat this yeat with new sensor tech that will drastically improve low FPN and low ISO DR while we're at it.

Have a look at my first post on page 5 of this topic if you missed it.
www.canonrumors.com/forum/index.php?topic=12029.60

This topic has devolved so far from the OP's initial query as to likely have bored them.  It started off with good intentions in the first few pages but here we are again. :-\


448
EOS Bodies / Re: Canon DSLR Body Rumors for 2013
« on: January 04, 2013, 01:08:02 AM »
all of you who are interested in small/big pixels. noise etc.
read this papers by Emil  http://theory.uchicago.edu/~ejm/pix/20d/tests/noise/noise-p3.html

Thanks for the link, MR.
I'll read that and see what else I might learn from it.

I haven't examined pattern noise in detail from too many cameras other than to notice that some banding is fixed (sensor system hardware) and some is inconsistent but repetitive (electronic subsystem interference, e.g. from noisy power supplies)

449
you guys are SO funny.
Humor your way of acknowledging defeat? ;)

Sorry, I just couldn't help myself...   ;D
Its only a problem if you miss your exposure 3 stops.  :o

or just dial back your contrast setting, or try Canon's own built-in LANDSCAPE style. Try both if you like stripes.
boost shadows another stop or 2 if blind or on uncalibrated monitor.

Those of use who bought the Canon fanboy hype about the 5d2 and purchased early were richly disappointed.
Mine sat unused in a drawer for most of a year until I found that the firmware updates actually improved things enough to make the camera usable without obvious MIDTONE banding any more.  MOST times anyway.

If you forgot, have a little refresher.  We're here to educate and elucidate:

www.google.ca/search?=en&q=sky+banding+canon+5d+mark+II

see FIRMWARE topic on the wikipedia page

http://en.wikipedia.org/wiki/Canon_EOS_5D_Mark_II#Firmware_updates

450
I have an Epson R3000 printer and I swear by Red River Paper.  My two favorites are Polar Pearl Metallic and Artic Polar Gloss.  They have sample packs and icc printer profiles on their web site.
www.redriverpaper.com


ditto with a friend of mine
I'm running an R3000 now and have only just started using it.  Epson's semi-gloss has a look I like for many shots, about to try their semi-matte.
Used to have some of my outside printed stuff, especially B&W, printed on a matte surface photorag by Hahnemuhle, not sure the exact one now.  It was roll product and we had tremendous problems with the coating spalling off after printing, leaving annoying white flecks or even large patches!
Images, otherwise, looked fantastic but a 1 out of 3 success ration was far too costly.
Looking fwd to trying it again in sheet form as the depth it produced I have not seen matched by other papers.

Pages: 1 ... 28 29 [30] 31 32 ... 55