Ad
Ad
Ad
Pages: [1] 2 3 ... 13 »   Bottom of Page
Print
Author Topic: 645D vs D3x  (Read 79994 times)
mhecker*
Contributor
Jr. Member
*
Offline Offline

Posts: 93


WWW
« on: February 04, 2011, 03:06:32 PM »
ReplyReply

I've done some more testing outdoors with my 645D vs Nikon D3x

Daylight  landscape, ISO 100  RAW

D3x,   24-70 zoom lens     @ f8 live view manual focus
645D, 45-85mm zoom lens @ f11  autofocus

The focus point was the tree on the center of the image.

After capture daylight color balance in LR, no sharpening.
Sharpening in CS5 D3x shot USM 300,0.4,1   645D USM 300,0.4,1

Some adjustment in CS5 to match the color.
ACR does a poor job with D3x color as rumored.


This was shot near Angels Camp, CA in the Sierra foothills and it really  was very, very green.

Links.

D3x      http://wyofoto.com/Pentax_645D/Nikon_D3x_field_test.tif
645D    http://wyofoto.com/Pentax_645D/Pentax_645D_field_test.tif

Warning big files D3x is 44MB,  645D file is 77MB

Comments of course are welcome.   Wink

Enjoy....  

« Last Edit: February 05, 2011, 04:25:07 PM by mhecker* » Logged
nazdravanul
Jr. Member
**
Offline Offline

Posts: 81


« Reply #1 on: February 04, 2011, 03:31:41 PM »
ReplyReply

First thing that comes to my attention is the fact that the D3x is cropped, so maybe you should look into that first ? The 2 formats are not identical even with the crop, the d3x is in a 5x4 format  (while the 645 appears to be in it's native 4x3 full format) - check the settings on your camera (D3x), you may have shot it like that.
If you want to crop for total similarity , you should do it both ways, just to be fair - crop the 645 to a 3x2 format, then crop the d3x to a 4x3 format if you feel it makes sense for your final output, but anyway show both comparisons, make both pairs.

Logged
mhecker*
Contributor
Jr. Member
*
Offline Offline

Posts: 93


WWW
« Reply #2 on: February 04, 2011, 03:55:19 PM »
ReplyReply

The D3x was shot in 5x4 format as that is standard for the owner.

The 645D was cropped to match the 5x4 format on purpose.
The shots also don't align perfectly, that doesn't void the test for me, YMMV. 





Logged
TMARK
Sr. Member
****
Offline Offline

Posts: 1843


« Reply #3 on: February 04, 2011, 04:05:52 PM »
ReplyReply

Confirms how good the D3x is.  I see a bit more rez, a touch smoother tones in the 645 file, but that's about it. 
Logged
eronald
Sr. Member
****
Offline Offline

Posts: 4121



« Reply #4 on: February 04, 2011, 04:20:27 PM »
ReplyReply

Nikon killed the D3x by overpricing it. People went for the 5D2 instead, which doesn't deliver the same quality, and then they "know" that fullframe cannot compete with cropped 645.

Edmund
Logged
uaiomex
Sr. Member
****
Offline Offline

Posts: 1005


WWW
« Reply #5 on: February 05, 2011, 12:24:57 AM »
ReplyReply

The 645D is clearly superior in IQ. Again, the most of subtle foliage areas in 645D look "distinct". In the 3DX they turn to mush as it happened with the 5D2 file. No contest, sorry. I don't think the next versions of Canikon top guns will be able to match the 645D quality and they definitely overpriced now. Didn't they ever?
Eduardo
« Last Edit: February 05, 2011, 12:28:03 AM by uaiomex » Logged
ErikKaffehr
Sr. Member
****
Online Online

Posts: 7656


WWW
« Reply #6 on: February 05, 2011, 01:26:27 AM »
ReplyReply

Hi,

Interesting comparison. I would like Miles to elaborate a bit on color rendition. No question that Pentax wins hands down, at least in my humble opinion. The way I analyze pictures is that I enlarge both images to 70x100 cm at 200PPI and look at actual pixels. That is a gargantuan view but it really shows the differences. In print I'd presume the differences would be far less noticeable. Looking at the same images at 50% reduces the differences dramatically.

I'm somewhat surprised that the Pentax 645D has that much better color. The house shot was really similar colors were much better on the Pentax 645D. Would I have a truckload of Pentax 645 lenses I'd be on my way to the bank on monday... Actually I have a bunch of Pentax 67 lenses, but they may not be the best ones ever built for that camera.

Best regards
Erik





I've done some more testing outdoors with my 645D vs Nikon D3x

Daylight  landscape, ISO 100  RAW

D3x,   24-70 zoom lens     @ f8 live view manual focus
645D, 45-85mm zoom lens @ f11  autofocus

The focus point was the tree on the center of the image.

After capture daylight color balance in LR, no sharpening.
Sharpening in CS5 5D shot USM 300,0.4,1   645D USM 300,0.4,1

Some adjustment in CS5 to match the color.
ACR does a poor job with D3x color as rumored.


This was shot near Angels Camp, CA in the Sierra foothills and it really  was very, very green.

Links.

D3x      http://wyofoto.com/Pentax_645D/Nikon_D3x_field_test.tif
645D    http://wyofoto.com/Pentax_645D/Pentax_645D_field_test.tif

Warning big files D3x is 44MB,  645D file is 77MB

Comments of course are welcome.   Wink

Enjoy.... 


Logged

TMARK
Sr. Member
****
Offline Offline

Posts: 1843


« Reply #7 on: February 05, 2011, 02:32:14 AM »
ReplyReply

I'm not convinced.  Better, yes, but to my eye not clearly superior after color was adjusted in the D3x file to something that I found pleasing.  I look at files at 50%, and yes I saw some aliasing (mush) in both files.  I'm not a landscaper, but in my line of work it would be a toss up between the two.  And yes the Nikon is stupidly over priced.  I think i'd take an M9 over either of these files!

The 645D is clearly superior in IQ. Again, the most of subtle foliage areas in 645D look "distinct". In the 3DX they turn to mush as it happened with the 5D2 file. No contest, sorry. I don't think the next versions of Canikon top guns will be able to match the 645D quality and they definitely overpriced now. Didn't they ever?
Eduardo

Logged
ErikKaffehr
Sr. Member
****
Online Online

Posts: 7656


WWW
« Reply #8 on: February 05, 2011, 03:50:00 AM »
ReplyReply

Miles,

Thanks for sharing! The images that you posted are very good test images and that also applies to the previous posting. It would be nice if you could also share the raw files.

I just made some prints from 30% crops. Before printing I opened both images and adjusted color on the Nikon to match Pentax. The crops were made on A4 and it was approximately 33% crops corresponding to 53x83 cm. Looking at the prints from close the difference in resolution is astonishing, but looking at arms length they are pretty close. Color is pretty close after matching color. You cannot really look at detail, because the Nikon image would fall apart.

To me it seems that the Pentax is an impressive camera, much preferable to the Nikon. I'm not prepared to jump the wagon, because I don't have the lenses and also because 10 grand is still considerable money.

Best regards
Erik

I've done some more testing outdoors with my 645D vs Nikon D3x

Daylight  landscape, ISO 100  RAW

D3x,   24-70 zoom lens     @ f8 live view manual focus
645D, 45-85mm zoom lens @ f11  autofocus

The focus point was the tree on the center of the image.

After capture daylight color balance in LR, no sharpening.
Sharpening in CS5 5D shot USM 300,0.4,1   645D USM 300,0.4,1

Some adjustment in CS5 to match the color.
ACR does a poor job with D3x color as rumored.


This was shot near Angels Camp, CA in the Sierra foothills and it really  was very, very green.

Links.

D3x      http://wyofoto.com/Pentax_645D/Nikon_D3x_field_test.tif
645D    http://wyofoto.com/Pentax_645D/Pentax_645D_field_test.tif

Warning big files D3x is 44MB,  645D file is 77MB

Comments of course are welcome.   Wink

Enjoy.... 


Logged

ErikKaffehr
Sr. Member
****
Online Online

Posts: 7656


WWW
« Reply #9 on: February 05, 2011, 04:27:36 AM »
ReplyReply

Hi,

The files that Miles posted have a significant advantage in resolution. I don't think Leica or any smaller format can match that. A Leica M10 with 40 Mpixels may do, but it's not here yet.

Best regards
Erik


So do I. M9 files are so far the ones I like most from digiland.
But I have trouble to get the focus spot-on with the M. Maybe it's just me. With the Nikon no probs. Doubts with the Pentax for subjects in movement if they really improved.
Logged

jduncan
Sr. Member
****
Offline Offline

Posts: 296


« Reply #10 on: February 05, 2011, 11:02:20 AM »
ReplyReply

Nikon killed the D3x by overpricing it. People went for the 5D2 instead
Edmund
I have never knew why they priced it like that, even today, years after the fact, is still at 7.5K.
Even if they had and issue with sensor yields it should be resolved by know.
Other option was something about the relationship with Sony, but right know, this seems unlikely.
That pricing don't look smart even in terms of making money. As I Nikon shooter I hope that they know something that we don't.
The alternative is terrifying.
Best regards

 
Logged

english is not my first language, an I know is shows
mhecker*
Contributor
Jr. Member
*
Offline Offline

Posts: 93


WWW
« Reply #11 on: February 05, 2011, 11:14:50 AM »
ReplyReply

I will be making 24"x30" and 30"x40" prints of these files and using them in a blind test.
The photographers viewing them will not know what camera the files came from. I suspect the difference in printed IQ will be small at 24"x30" and much greater at 30"x40", but we will see.

As I have been told the ACR RAW converter in Lightroom sometimes produces horrible color with the D3x files.
I worked on the image in the test and manually color balanced it to get it close to the 645D file.
If someone has a color profile better than the Adobe one for the D3x I would gladly use it.

My take on things so far is as follows.
The analysis is for landscape use at low ISO only.

The 5Dmk2 is good value for the money. A $3000 camera for $2500.
Good color, very good resolution, mediocre autofocus, a lot of shadow noise when pushed, very little when the exposure is spot on.

The D3x is a the best 35mm digital camera available, but not a good value. A $5500 camera for $7500.
Good color with the proper RAW converter, a tad better than very good resolution, excellent autofocus, very little shadow noise even when pushed.

The 645D is slightly superior to either in image quality.
Very good color, a notch better resolution than any 35mm SLR,excellent autofocus, very little shadow noise when pushed.
It is a good, but not great value. An $8500 camera for $9995. I believe the price has already fallen to $9000 in Japan and will do so in the USA by years end.

Logged
bcooter
Guest
« Reply #12 on: February 05, 2011, 11:24:11 AM »
ReplyReply

I have never knew why they priced it like that, even today, years after the fact, is still at 7.5K.
 

I don't think anyone expected the 5d2.   Actually I'm not sure if Canon planned on making it 20 something mpx and a video game changer, it's just Sony came in with 20 something megapixels cheap, is sitting there with all those Ziess lenses and I think kind of put the scare in Canon, so we have the 5d2, which makes almost any camera in the world look way overpriced.

Think about a non 5d2 world and the d3x makes sense.   Nikon's competition was Canon though Nikon has better autofocus, better resolution, a sharper file than the 1ds3 and only sells for a grand more.  The only glitch of the Nikon is their in-house software is awful and when you tether the lcd on the camera goes blank.  If it wasn't for that in a non 5d2 world it would almost be the perfect camera.

IMO

BC

Logged
Radu Arama
Full Member
***
Offline Offline

Posts: 138


« Reply #13 on: February 05, 2011, 12:37:20 PM »
ReplyReply


The 645D is slightly superior to either in image quality.
Very good color, a notch better resolution than any 35mm SLR,excellent autofocus, very little shadow noise when pushed.
It is a good, but not great value. An $8500 camera for $9995. I believe the price has already fallen to $9000 in Japan and will do so in the USA by years end.



I beg to differ on some of your points, Miles:

- The price in Japan is pretty much equal to what it was last June at launch, and I bet the slight differences are due mostly to competition amongst dealers and the fact that there are a lot of them. In the US the camera is sold by four outfits (BH, Adorama, Ace and the Pentax own site and none of those has any desire to give up one dollar especially if they don't have to at the current level of demand). In Europe you will soon see better prices and much more competition because the dealers that sell currently at list price keep probably over 2000 Euro for themselves for every camera sold.
- I don't know if you can "price up" a camera without many subjective factors. Not many people know that this camera is not a magnesium shell stuffed with electronics it is made around a stiff aluminum chassis for minimal heat/cold induced deformations hence focusing errors. It also has proven cold resistance besides the highest level of weather resistance (any amount of rain for any amount of time). Thirdly it has a plethora of external controls that needs to be wired inside, a second tripod mount of the side that spares you a L plate and two cold and scratch resistant LCDs. Never mind a working anti dust feature that even D3X lacks if I don't make a mistake.

In the end what I am trying to say is that if the D3x's price is most likely determined by Nikon the (at least US) 645D's price is determined by demand and lack of competition amongst dealers.

Radu   
Logged
mhecker*
Contributor
Jr. Member
*
Offline Offline

Posts: 93


WWW
« Reply #14 on: February 05, 2011, 01:13:10 PM »
ReplyReply

I basically agree with you Radu.  Although I would be surprised if the US price doesn't fall by the end of 2011.

I do think part of the reason the 645D nails focus every time is it's tank like build.
My good friend David Brookover thought it was built a notch above even his 2 D3x's.

It also did just fine in an extended, heavy, wet Yosemite snow storm last Sunday.
I shot alongside David and his D3x as many of the lesser cameras, including a 5dmk2 nearby were fogging up.   Smiley

Also, a dust spot that appeared in the sky early one morning, vanished after the camera was turned on and off a few times the same day.

Having shot a borrowed Hasselblad H3Dii, I found it acquired dust on the sensor and only a cleaning at the end of the day would remove it.
« Last Edit: February 05, 2011, 01:43:19 PM by mhecker* » Logged
Leping
Jr. Member
**
Offline Offline

Posts: 87



WWW
« Reply #15 on: February 05, 2011, 03:06:19 PM »
ReplyReply

The both were oversharpen with the LR's "anti-film" signature "digital flavor".

In the deconvolution discussion thread (currently 13 pages!) people have been into extreme pixel peeping (at 400%) and trying to ignore the fact that better RAW converters, such as Raw Developer, makes much more natural looking results besides "super-resolution" (from iterative R-L deconvolution).  Big reason old film landscape masters like Charles Cramer liked them.

At 75mm, the Pentax-FA SMC 45-85mm/f4.5 is little out of its "sweet zone", which is from 45-65mm.  But, not as bad as it is at 85mm anyway, and stopping down f/11 seemed to help (see digLloyd's tests and other posts).  And, to be fair, the Nikon 24-70mm/f2.8 at 56mm is hardly the best 50-60mm lens at f/8 for the D3x (even my old AIS 55mm/f2.8 Macro does better, f/8 and near infinity).
« Last Edit: February 05, 2011, 03:09:22 PM by Leping » Logged

mhecker*
Contributor
Jr. Member
*
Offline Offline

Posts: 93


WWW
« Reply #16 on: February 05, 2011, 03:25:30 PM »
ReplyReply

Hi Leping,

From my original post.

>After capture daylight color balance in LR, no sharpening.
>Sharpening in CS5 D3x shot USM 300,0.4,1   645D USM 300,0.4,1

There was no sharpening done in LR.
So, the deconvolution LR sharpening you are seeing is in your mind only.
They were sharpened using USM in CS5 with the settings shown above.
I sharpen for printing, not pixel peeping.

They will look slightly oversharpened at 100% resolution on an LCD.
When printed at 2880x1440 DPI on my Epson 9880 at any normal size say 24"x30" or 30"x40" they will not appear oversharpened.

If I wanted to make them look film like at 100% I would reduce the sharpening  and add noise.

But why would anyone want to prepare a 40MP  image for perfect film like pixel peeping?
What would be the purpose of this?

If you would like the RAW files to evaluate, I could make them available.

Cheers,
Miles
« Last Edit: February 05, 2011, 04:24:44 PM by mhecker* » Logged
Leping
Jr. Member
**
Offline Offline

Posts: 87



WWW
« Reply #17 on: February 05, 2011, 04:12:19 PM »
ReplyReply

Hi Miles,

Thank you for the quick reply.  It's fine, and I understand your point.  Even with 40MP for 24x30 and larger prints we still need to upsample to 300 or 360dpi.

Try the following workflow (all numbers approximate and only as guided guess starts):

1. Resampling Photoshop bi-cubic "More Smooth" to size;
2. USM 2/10/0, fade to luminosity;
3. USM 1/20/0, fade to luminosity;
4. USM 0.5/100/0, fade to luminosity;

For printing you are done here.  For 100% pixel peeping no LCD screen, add:

5. Smart Sharpening "Lens Blurr" (deconvolution) 0.2/10/0 (heavily depending on noise level).  Fade to luminosity (to cope with these crazy 200%, 300%, and 400% peepers, since the difference from the fade step could only be detected barely here in this step).

You can of course put a "clarity" USM, something like 80/12/0, before the step 2.  Fade to luminosity immediately always, and very important here since the 80 pixel radius covers a lot of colors.

Thanks and great to see 645D and the 45-85mm zoom is doing so well, and also more and more people realize adding noise is such a great way to "fix" digital captures.  I work in medical imaging field and we do add noise to the images.  Doctors love it (not the heavily "denoised" ones).  5-7 years ago, when I said this, people just laughed at me.

Leping
« Last Edit: February 05, 2011, 04:22:46 PM by Leping » Logged

Graham Welland
Sr. Member
****
Offline Offline

Posts: 612


« Reply #18 on: February 06, 2011, 12:59:44 AM »
ReplyReply

Now what might be an interesting comparison would be the 645D vs a MaxMax converted D3x without the AA filter & either a 24 PC-E, 45 PC-E or Zeiss ZF prime ...  Wink The D3x is brutal on glass & technique and few lenses, including even the 24-70 won't get the best from it.
Logged

Graham
ErikKaffehr
Sr. Member
****
Online Online

Posts: 7656


WWW
« Reply #19 on: February 06, 2011, 01:16:16 AM »
ReplyReply

Miles,

Posting the raw images would be very helpful.

I'm somewhat confused by color. It would be interesting to have some raw images with a color checker included, or having a color checker in similar light.

I made some test prints, pretty large. Before printing I "matched colors" using Photoshop CS5, Image->Adjustment->Match Color so the Nikon colors were quite close to Pentax. The prints turned out pretty similar when viewed at some distance (50-80 cm) but the Pentax was at great advantage at short viewing distance, like 25 cm. The way I made this I had 1/3 crops I printed on A4 so I made small prints corresponding to about 53x83 cm for the full image (ca 21x32").

I'd add that it's a really nice test image! Great work and thanks for sharing!

Best regards
Erik



If you would like the RAW files to evaluate, I could make them available.

Cheers,
Miles

Logged

Pages: [1] 2 3 ... 13 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad