Pages: 1 2 3 [4] 5 6 ... 13   Go Down

Author Topic: 645D vs D3x  (Read 148084 times)

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: 645D vs D3x
« Reply #60 on: February 11, 2011, 05:44:30 pm »

We all know what are the strenghs and wicknesses of each system (35mm and MF), we know that for ages, it has been discussed a zillion times and it has never changed since.....

How can you compare 2 cameras with such a difference in sensor size and philosophy, aimed to different applications and needs (and in the case of the D3x, if needs can be similar the shooting style belongs to different tastes)?....


If you'd find silly to open a thread like "Canon G12 vs Canon 5D2", I don't get why don't you find as silly to compare a D3x to a Pentax 645D or any other big sensor's camera.....


No we don't all know what the strengths and weakness are, Fred. Some people still seem to think that all MFDBs have a 4 stop DR advantage over all 35mm DSLRs.

Then other people seem to think that differences between an MFDB and the 35mm format is like the difference between a Canon G12 and a Canon 5D2, apparently unaware that the 5D2 sensor is about 20x the area of the G12 sensor, whereas the 645D sensor is only 1.7x the area of full frame 35mm.

Just as it's quite sensible to compare one MFDB with another, and one 35mm DSLR with another, it's also sensible to compare a top-of-the-line DSLR (the D3X) with the next step up in MFDB format (the 645D) at a no-so-outrageous price.

A better analogy than the G12 compared with the 5D2, would be the 10mp Canon 1D3 with the 16.7mp Canon 1Ds2. The difference in pixel count and sensor size is almost exactly the same as the difference between the D3X and the 645D.
Logged

kers

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4388
    • Pieter Kers
Re: 645D vs D3x
« Reply #61 on: February 11, 2011, 07:35:50 pm »

Just my humble expression: i have seen Diglloyd testing  lot of things.... i am sure he makes a lot of money with what he does. IMHO

In that case I guess Hasselblad did not pay enough...
Logged
Pieter Kers
www.beeld.nu/la

eronald

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6642
    • My gallery on Instagram
Re: 645D vs D3x
« Reply #62 on: February 11, 2011, 07:50:22 pm »

In that case I guess Hasselblad did not pay enough...

Digilloyd is pretty honest, the problem is he is good at doing what he usually does.
Digilloyd is a landscape photographer skilled at dSLR work; he doesn't do the stuff which dSLRs do best, and he doesn't really know how to use MFDB at its best.

Edmund
Logged
If you appreciate my blog posts help me by following on https://instagram.com/edmundronald

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: 645D vs D3x
« Reply #63 on: February 11, 2011, 08:27:34 pm »

Hi!

Prices may vary, but I'd say that Pentax is like 30% more expensive than the Nikon D3X. To that comes lenses.

Best regards
Erik

IMO the Nikon looks rough and nasty in comparison to the Pentax image. Color, detail and tonality all superior in the 645.

What are the price differences on these cameras?
Logged
Erik Kaffehr
 

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: 645D vs D3x
« Reply #64 on: February 11, 2011, 08:52:58 pm »

Hi!

Prices may vary, but I'd say that Pentax is like 30% more expensive than the Nikon D3X. To that comes lenses.

Best regards
Erik


And as we all know, Erik, a camera without a lens is useless. The cost, quality, range and type of lenses available (or in the pipeline) has to be part of the equation.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: 645D vs D3x
« Reply #65 on: February 12, 2011, 11:33:11 am »

Hi,

The Pentax 645 was quite popular so there are a lot of Pentax 645 lenses around, some at very attractive prices. Many of those lenses seem to be excellent but not all. So we have at least three scenarios:

- Someone already having 645 lenses of good quality
- Someone moving to 645 for improved quality and trying to find lenses on EBay
- Someone building a 645 system from scratch and buying new lenses

The costs of these alternatives are very much different. Would anyone invest into the D3x a similar situation would apply.

Best regards
Erik


And as we all know, Erik, a camera without a lens is useless. The cost, quality, range and type of lenses available (or in the pipeline) has to be part of the equation.
Logged
Erik Kaffehr
 

paratom

  • Full Member
  • ***
  • Offline Offline
  • Posts: 205
Re: 645D vs D3x
« Reply #66 on: February 12, 2011, 02:27:00 pm »

.....

A better analogy than the G12 compared with the 5D2, would be the 10mp Canon 1D3 with the 16.7mp Canon 1Ds2. The difference in pixel count and sensor size is almost exactly the same as the difference between the D3X and the 645D.

The difference between the MF-digitals and DSLRs is not only about sensor size and MP but also about ccd vs cmos, and 16 bit vs 12 or 14, and about AA-filter vs no AA filter.
IMO its just a different look, even if the sensor would be same size and resolution.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: 645D vs D3x
« Reply #67 on: February 12, 2011, 02:53:18 pm »

Hi,

We recently got some good comparison images from Miles Checker (for P645D and D3X) and also from marcmccalmont (for P45+ and Pentax K5) really indicate that the differences between MFD Digital and CMOS based DSLRs is less than suggested by conventional wisdom.

The Pentax 645D images posted by Miles are without doubt sharper at a given print size than the D3X images and very much different in color. That difference in color is disturbing. Marc has posted good comparison images from his P45+ and Pentax K5 and there is no doubt that the APS-C camera delivers better shadow detail than the once mighty P45+.

The P645D has much better resolution, at given print size, but also needs about the same level of sharpening as the Nikon. The result is essentially what I would expect. A size advantage for the bigger sensor.

Comparing color rendition is difficult. They are different but which one is right?! The Nikon image has better separation between yellow-green and yellow. I prefer the color rendition of Pentax but Nikon might be more correct.

Best regards
Erik

In my view this really puts things in perspective

The difference between the MF-digitals and DSLRs is not only about sensor size and MP but also about ccd vs cmos, and 16 bit vs 12 or 14, and about AA-filter vs no AA filter.
IMO its just a different look, even if the sensor would be same size and resolution.

Logged
Erik Kaffehr
 

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: 645D vs D3x
« Reply #68 on: February 12, 2011, 07:47:47 pm »

Hi,

The Pentax 645 was quite popular so there are a lot of Pentax 645 lenses around, some at very attractive prices. Many of those lenses seem to be excellent but not all. So we have at least three scenarios:

- Someone already having 645 lenses of good quality
- Someone moving to 645 for improved quality and trying to find lenses on EBay
- Someone building a 645 system from scratch and buying new lenses

The costs of these alternatives are very much different. Would anyone invest into the D3x a similar situation would apply.

Best regards
Erik

Erik,
These are issues which need further investigation. The Nikon cameras are famous for their older-lens compatibility. However, one tends to lose certain functions with older lenses, such as autofocus maybe, and/or recording of full EXIF data.

In this day and age it is surely a basic requirement that all cameras have full auto-metering and autofocus.

It doesn't seem sensible to me to make a camera-body purchasing decision on the basis that one already owns, or can get, old lenses of mediocre quality when such resolution improvements of the new camera have been demonstrated with a top quality lens.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: 645D vs D3x
« Reply #69 on: February 12, 2011, 08:08:51 pm »

The difference between the MF-digitals and DSLRs is not only about sensor size and MP but also about ccd vs cmos, and 16 bit vs 12 or 14, and about AA-filter vs no AA filter.
IMO its just a different look, even if the sensor would be same size and resolution.


As far as I'm aware, the advantages and disadvantages of no AA filter tend to canel each other, on balance. The initial impetus to produce a camera without an AA filter was probably due to the cost savings for cameras which were already too expensive for many.

16 bit versus 12 bit may be of some significance. However, the current latest cameras that have the best DR, at the pixel level, without doubt and by a wide margin, are the D7000 and Pentax K5, which are both 14 bit.

Interestingly, the other IQ parameters such as SNR at 18% grey, tonal range and color sensitivity are about the same; slightly better than the 645D, but slightly worse than the P65+, at the pixel level. Not an issue for me.

Logged

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Re: 645D vs D3x
« Reply #70 on: February 13, 2011, 12:11:51 pm »

As far as I'm aware, the advantages and disadvantages of no AA filter tend to canel each other, on balance.



I wouldn't agree with that statement. Besides mitigating moire and muddying the image even when moire isn't present what advantage is there for the AA filter?
Logged
Rolleiflex USA

Dennis Carbo

  • Full Member
  • ***
  • Offline Offline
  • Posts: 163
Re: 645D vs D3x
« Reply #71 on: February 13, 2011, 12:57:37 pm »

"As far as I'm aware, the advantages and disadvantages of no AA filter tend to canel each other, on balance. The initial impetus to produce a camera without an AA filter was probably due to the cost savings for cameras which were already too expensive for many.

16 bit versus 12 bit may be of some significance. However, the current latest cameras that have the best DR, at the pixel level, without doubt and by a wide margin, are the D7000 and Pentax K5, which are both 14 bit."

I have been shooting with a D7000 since november - and while the DR is amazing - it certainly doesnt appear to beat my MFDB by a "wide margin" or at all for that matter.  The D7000 files fall apart quickly when recovering detail from shadows, the 16 bit MFDB file holds up much better. The DR seems pretty close between the D7000 and the MFDB.....usable files the edge still goes to the MFDB for me. As far as the AA Filter - I find it a huge advantage when fine detail is involved - branches and grass for example seem mushy even with the D3X. 16 bit and no AA filter are a great advantage to me that is why I own a MFDB.  Shooting in low light at ISO6400 and producing a clean file with great DR is also a great advantage, which is why I own the D7000.  Horses for courses i guess

Between the D3X and 645D......tough choice....I think I would go D3X for an all around system as I own a ton of nikon glass.

nice to have choices...I love what i have seen of the 645D too.....awwww hell....I need them both !.....oh...and a Phase IQ180 too !

Regards,

Dennis
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: 645D vs D3x
« Reply #72 on: February 13, 2011, 03:27:30 pm »

Hi,

This article from Schneider Optics describes the issue pretty well:

http://www.schneideroptics.com/pdfs/whitepapers/optics_for_digital_photography.pdf (on page 8).

The screendump below is taken from that document. It shows aliasing. A lot of detail, but all fake.

In my view the negative effect of the aliasing filter is exaggerated. From any number of tests it's quite obvious lenses tend to be diffraction limited around f/5.6 - f/8, clearly indicating that the effect of the AA-filter is relatively small. I'd also assume that increasing the fill factor would act a bit like AA-filtering, making the pixel larger increases the probability that a single point affects more than one sensual.

Best regards
Erik



I wouldn't agree with that statement. Besides mitigating moire and muddying the image even when moire isn't present what advantage is there for the AA filter?

Logged
Erik Kaffehr
 

Dennis Carbo

  • Full Member
  • ***
  • Offline Offline
  • Posts: 163
Re: 645D vs D3x
« Reply #73 on: February 13, 2011, 03:38:31 pm »

"In my view the negative effect of the aliasing filter is exaggerated. From any number of tests it's quite obvious lenses tend to be diffraction limited around f/5.6 - f/8, clearly indicating that the effect of the AA-filter is relatively small. I'd also assume that increasing the fill factor would act a bit like AA-filtering, making the pixel larger increases the probability that a single point affects more than one sensual."   


The effect of the AA Filter is quite noticeable see here:

http://www.maxmax.com/hot_rod_visible.htm

these guys do AA Filter removal on DSLR'S

May not like the look but it is a noticeably sharper image - granted moire can be an issue  ;D

Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: 645D vs D3x
« Reply #74 on: February 13, 2011, 05:25:02 pm »

The effect of the AA Filter is quite noticeable see here:

http://www.maxmax.com/hot_rod_visible.htm

these guys do AA Filter removal on DSLR'S

May not like the look but it is a noticeably sharper image - granted moire can be an issue  ;D

This has been discussed before, but moire is not the only problem. From a theoretical standpoint, it has been discussed several times on this forum that perfect sharpness with a discrete sampling system is a myth.

What this means is that most of the micro detail you can feel in AA less images cannot be guaranteed to be real. They might be more pleasing to the eyes of some people, but the "sharpness" of AA filter less images is in fact a bigger departure from the reality of the scene than the "softness" of AA filter images. Increasing the resolution reduces the relative scale of these artifacts but the problem remains the same. These images are built on a foundation of false data.

This is a well accepted fact in most domains like high end audio where companies typically boast about their dithering algos (the equivalent of AA filters), the difference being that you cannot do a 100% pixel zoom on an audio file. Perhaps pixelpeepers need to invent the concept of "decibelpeeper" to generate (artificially again) support from their audiophile friends?

I don't really get why we still need AA filter less imaging sensors with 80 megapixel class sensors. The only reason is to please pixel peeping photographers who have been used to the artificially sharp looking images of their 40 megapixel sensors, but all things considered it would be better to move back to AA filters for the higher resolution backs.

Why do I somehow I feel that this post will not be received well... :)

Cheers,
Bernard

Dennis Carbo

  • Full Member
  • ***
  • Offline Offline
  • Posts: 163
Re: 645D vs D3x
« Reply #75 on: February 13, 2011, 06:05:11 pm »

"What this means is that most of the micro detail you can feel in AA less images cannot be guaranteed to be real. They might be more pleasing to the eyes of some people, but the "sharpness" of AA filter less images is in fact a bigger departure from the reality of the scene than the "softness" of AA filter images. Increasing the resolution reduces the relative scale of these artifacts but the problem remains the same. These images are built on a foundation of false data."

Ok - I will admit, I dont understand this...the "micro detail" seems real to me ..I certainly find the AA Filtered, mushy look of fine branches and twigs a "bigger departure from reality" than the crisp version from a Non AA Filtered sensor. Are you saying the AA filtered image is actually more accurate ?

No disrespect Bernard...just trying to understand  ;D

I know it is a bit off topic for the thread but....Does a Microstep Multi-shot back only address the moire issue or do they also "build the images on false date" as you stated because they use no AA filter ?



Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: 645D vs D3x
« Reply #76 on: February 13, 2011, 07:10:36 pm »

Quote
"As far as I'm aware, the advantages and disadvantages of no AA filter tend to canel each other, on balance. The initial impetus to produce a camera without an AA filter was probably due to the cost savings for cameras which were already too expensive for many.

16 bit versus 12 bit may be of some significance. However, the current latest cameras that have the best DR, at the pixel level, without doubt and by a wide margin, are the D7000 and Pentax K5, which are both 14 bit."

I have been shooting with a D7000 since november - and while the DR is amazing - it certainly doesnt appear to beat my MFDB by a "wide margin" or at all for that matter.  The D7000 files fall apart quickly when recovering detail from shadows, the 16 bit MFDB file holds up much better. The DR seems pretty close between the D7000 and the MFDB.....usable files the edge still goes to the MFDB for me. As far as the AA Filter - I find it a huge advantage when fine detail is involved - branches and grass for example seem mushy even with the D3X. 16 bit and no AA filter are a great advantage to me that is why I own a MFDB.  Shooting in low light at ISO6400 and producing a clean file with great DR is also a great advantage, which is why I own the D7000.  Horses for courses i guess

Between the D3X and 645D......tough choice....I think I would go D3X for an all around system as I own a ton of nikon glass.

nice to have choices...I love what i have seen of the 645D too.....awwww hell....I need them both !.....oh...and a Phase IQ180 too !

Regards,

Dennis


Dennis,
I really think you are confusing issues here and attributing certain obvious advantages of the larger sensor, with its higher pixel count, to the 16 bit pipeline.

My claim is that the D7000 has a far better DR at the pixel level, than any MFDB that DXO have tested, whilst also maintaining similar tonal range and color sensitivity etc. This augurs well for the future of FF 35mm. There is no reason to suppose that Nikon (or Sony) will not produce a 40mp FF 35mm sensor in the near future, comprised of D7000 pixels. I would consider such a camera an irresistable upgrade, if the price were right.

Nevertheless, I don't wish to appear arrogant after your claim you are using both a D7000 and an MFDB, because clearly you have the opportunity to make comparisons that I don't.

I would be very interested to see a comparison at the pixel level, between your D7000 and whatever MFDB you use.

In order to exclude influences such as the DB's larger sensor and greater pixel count, so we can see just what effect might be attributable to the 16 bit pipeline, (but including also any advantages of the lack of an AA filter and the differences of the CCD design in general, which are impossible to exclude), I think it would be necessary to shoot the same high-DR scene choosing appropriately different focal lengths so that same FoV crops are the same file size.

On the basis that the D7000 has greater pixel density than your DB, this would  entail using a slightly longer focal length with your D7000 than you would use with the MFDB, instead of the usually shorter FL to equalize the FoVs of the entire frame. (I think I got that right. Too early for a glass of wine  ;D  ).
Logged

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Re: 645D vs D3x
« Reply #77 on: February 13, 2011, 08:39:20 pm »

Hi,

This article from Schneider Optics describes the issue pretty well:

http://www.schneideroptics.com/pdfs/whitepapers/optics_for_digital_photography.pdf (on page 8).

The screendump below is taken from that document. It shows aliasing. A lot of detail, but all fake.

In my view the negative effect of the aliasing filter is exaggerated. From any number of tests it's quite obvious lenses tend to be diffraction limited around f/5.6 - f/8, clearly indicating that the effect of the AA-filter is relatively small. I'd also assume that increasing the fill factor would act a bit like AA-filtering, making the pixel larger increases the probability that a single point affects more than one sensual.

Best regards
Erik





I've seen that article, but that doesn't mean that Canon Nikon et all are using the ideal AA filters tuned as describe in it.   Anecdotally, my old leica DMR without AA filter could get as much fine detail - real detail - as my canon 5d2 could using the exact same lens fit with an adapter. Taking the lens out of the equation how could the DMR get the same detail with half the pixels?  Explain that then - if it isn't the AA filter, what then is robbing the DSLR's of fine detail?

Logged
Rolleiflex USA

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: 645D vs D3x
« Reply #78 on: February 13, 2011, 09:57:49 pm »

..I certainly find the AA Filtered, mushy look of fine branches and twigs a "bigger departure from reality" than the crisp version from a Non AA Filtered sensor. Are you saying the AA filtered image is actually more accurate ?

There can be a slightly mushy appearance to AA filtered images, depending on the RAW converter used, the adjustments mades, and the degree and type of sharpening applied.

The detail slider in ACR is very good at removing mushiness.

Below are conversions I've made of the 645D and D3x examples from Miles, using ACR 6.3. I used the same sharpening for both images; detail at 100%, pixel radius 0.8, and sharpening amount 50.

I downsampled the 645D image to the same file size as the D3X image using 'Bicubic Sharper' to give the 645D image the full sharpness benefit that might flow from downsampling.

I also made an attempt to equalize tonality and color hue in both images, but haven't quite succeeded. The 645D image still retains a more pervasive green.

Comparing 100% crops from the central area of both images, I see no mushiness in the D3X image. But you be the judge.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: 645D vs D3x
« Reply #79 on: February 13, 2011, 11:50:25 pm »

Hi,

Did you focus bracket or used live view for focusing on the 5D?  I checked Photozone tests on the Canon 5DII using the Carl Zeiss Macro Planar 100/2 and the Canon 135/2.

The Zeiss is clearly diffraction limited at f/4 and it achieves higher MTF 50 than the Canon 135. So it seems less likely that AA-filtering is a major culprit. Would AA filtering dominate over diffraction there would be no difference between the Canon and the Zeiss.

Another issue is that the microlenses also act against against aliasing. Reducing fill factor results in higher apparent resolution but it's still fake.

I don't argue with you, but it is obvious that the effect of diffraction is clearly significant at relatively large apertures on the 5DII, clearly indicating that resolution is not significantly reduced. Fine detail contrast may be reduced, however. It is generally seen that AA-filtered images need sharpening with a small radius but high amount.

Could you post some raw example images? Using the same lens but different sensors?

When I was looking at the Pentax 645D images I essentially found that they needed exactly the same amount of sharpening as the Nikon D3X images. That may depend on sensor, lens and microlenses. And yes, also on focus. The Pentax has an advantage of more pixels but the pixel size on both cameras is similar. So everything else being equal the Pentax would deliver better resolution and it does indeed.

It may be very hard to tell true and false resolution apart. The best example of false resolution I have seen is on the Sigma DP2 in DPreview,
It's quite obvious that the all resolution beyond Nyquist is fake, the white and dark lines switch place. Sigma is top left quadrant. Sigma seems to have superior resolution but actually all the others resolve better.

Best regards
Erik





I've seen that article, but that doesn't mean that Canon Nikon et all are using the ideal AA filters tuned as describe in it.   Anecdotally, my old leica DMR without AA filter could get as much fine detail - real detail - as my canon 5d2 could using the exact same lens fit with an adapter. Taking the lens out of the equation how could the DMR get the same detail with half the pixels?  Explain that then - if it isn't the AA filter, what then is robbing the DSLR's of fine detail?


« Last Edit: February 13, 2011, 11:56:13 pm by ErikKaffehr »
Logged
Erik Kaffehr
 
Pages: 1 2 3 [4] 5 6 ... 13   Go Up