[kenafein]

So the DP1m does #1 and the Q7 does the rest

Sounds like a good combo, unless you want to shoot interior architecture. I am looking forward to a low light, wide prime, for the Q. I don't know if they'll be able to get 18mm equivalent wide, but I'd like a 24mm equivalent f/2.

---

Last edited by kenafein; 08-10-2013 at 06:58 AM.

[Rondec]

Nice write-up...

Just curious, i know that people like to say that Pentax DR is so good that we can pull out a lot of details from the shadow. But i'm just curious, Pentax cam seems to love underexposing an image. Could this be a pseudo-dynamic range advantage?

And yes, i shoot a K30, just curious.. Could you try your very best and shoot similar images with the histogram as close to each other as possible and compare again if the shadow does really hold more details compared to the other?

PS: Asking this since i tend to shoot at +0.7 to obtain a middling histogram and without having to keep pumping the exposure value in LR for the RAW images.

Just looking at DXO Mark comparison between the 5D Mk III and the K5 II, the K5 II actually betters the Canon's sensor score and in particular at base iso has dynamic range of 11.7 EVs compared to 14.1 for the K5 II. There really is more shadow detail in a K5 II shot, whether or not you are shooting a little underexposed. I personally prefer to shoot a little underexposed, because highlights are basically not recoverable on the K5 (5D MK III is a little better at this) while I can easily pull up exposures in post processing without problem.

[pinholecam]

Sounds like a good combo, unless you want to shoot interior architecture. I am looking forward to a low light, wide prime, for the Q. I don't know if they'll be able to get 18mm equivalent wide, but I'd like a 24mm equivalent f/2.

Exactly!.
If they have one at 24mm f2.8 equivalent or 20mm eq (f4), I'd get one.
I presently thinking of using the 21mm wide converter for the GR (same 49mm thread mount).
Will try it out as a friend has it.
Even if I can crop out bad edges for a 24mm eq and if the remaining 24mm FOV is good, I'd be happy enough to buy it.

Currently, the plan is to use a pano head. (I have that)
I need a pano head anyway for may wide shots on whatever lens I carry (even the 10-20).
Not ideal for sure but I guess it will have to do for now.

[SyncGuy]

Just looking at DXO Mark comparison between the 5D Mk III and the K5 II, the K5 II actually betters the Canon's sensor score and in particular at base iso has dynamic range of 11.7 EVs compared to 14.1 for the K5 II. There really is more shadow detail in a K5 II shot, whether or not you are shooting a little underexposed. I personally prefer to shoot a little underexposed, because highlights are basically not recoverable on the K5 (5D MK III is a little better at this) while I can easily pull up exposures in post processing without problem.

Ah... Ok!
Thank you so much for the confirmation!

[Winder]

I agree with the point that the flagship Pentax ASP-C from 2013 is every bit as good as the mid-grade Canon FF from 2009. That is what you are demonstrating. The Sony APS-C sensor is 4 years newer, and Sony makes better sensors than Canon in general. With the exception of the current 20MP Sony APS-C sensor, I'm not aware of any bad or even below average sensors from Sony.

How can we do a comparisons where we control for the age of the technology used in the sensor. The OM-D sensor is better than the Canon 5D so I guess we can say that there is no difference between 4/3 sensors and FF sensors...... It's the same logic. Or since the K-5 is actually much better than my old Canon 5D we can say that APS-C has surpassed FF in terms of image quality.....

Its pretty pointless to compare cameras of different formats and different generations.

My next question would be what monitor and printer are you using?

When you say "whether you’re pixel peeping at 100% on your monitor or closely examining 20 x 30 prints side-by-side under bright light, there is no difference in quality."

Most monitors only have an 8 bit display, but I assume you are using something with at least a 10-bit display and a 16 bit LUT. 8-bit displays homogenize the images displayed on them and make many differences disappear. We have the same issue with printers when dealing with the color and dynamic range. Differences captured are lost in final output.

[philbaum]

I agree with the point that the flagship Pentax ASP-C from 2013 is every bit as good as the mid-grade Canon FF from 2009. That is what you are demonstrating. The Sony APS-C sensor is 4 years newer, and Sony makes better sensors than Canon in general. With the exception of the current 20MP Sony APS-C sensor, I'm not aware of any bad or even below average sensors from Sony. ....

Its pretty pointless to compare cameras of different formats and different generations.

I don't think comparisons are pointless. The APS DSLR owner of today is constantly bombarded with rampant consumerism with the message that their cameras are practically obsolete. The Canon 5D is a popular FF camera that is a well known icon of camera technology and achievement. To show that the K5 series is competitive against this icon, clearly demonstrates a more rational message concerning the need to purchase a full-frame camera. Also, the K5 series was announced in Sep 2010, altho admittedly the resolution and AF was bumped up in recent models.

I sympathize though with being the owner of a older camera that is often involved in comparisons with newer cameras. Also, the K5 in its comparison with the 5D is a much less expensive camera which is an advantage to the 5D.

Think i'll get back to taking pictures :-)

[bkpix]

This is a comparison of two popular cameras -- no more and no less. Like many people on this forum, I have been interested in the idea of shooting with a 5D2, so I borrowed one to try out.

As for output, the prints I make are on an Epson 7600. Image quality differences that can't be seen in a print are of no interest or consequence to me.

[Winder]

This is a comparison of two popular cameras -- no more and no less. Like many people on this forum, I have been interested in the idea of shooting with a 5D2, so I borrowed one to try out.

That's not the way the conclusions of your last 2 paragraphs read. You close out with statements about crop-sensors being just as good as full frame and the need to upgrade is just "playing on consumers’ insecurity".

There is nothing in your closing paragraphs that says this is simply a comparison of two popular cameras. The conclusion you are writing is that APS-C is just as good as full frame and "All that most of these folks will upgrade is their credit card debt."

There is a 4 year difference in the cameras (sensors & image processors) you are comparing. So statements like "But now crop cameras have leveled the playing field and are superior for a lot of kinds of shooting." If you mean that NOW crop cameras have caught up with 4 year old full frame cameras, then you are correct.

As for output, the prints I make are on an Epson 7600. Image quality differences that can't be seen in a print are of no interest or consequence to me.

And what monitor are you using?

When I upgraded to a 2560 x 1440 Eizo I quickly realized that the camera was capturing a lot more detail than I was seeing on my old monitor. It changed the way I was processing images and it changed my prints.

[Rondec]

I agree with the point that the flagship Pentax ASP-C from 2013 is every bit as good as the mid-grade Canon FF from 2009. That is what you are demonstrating. The Sony APS-C sensor is 4 years newer, and Sony makes better sensors than Canon in general. With the exception of the current 20MP Sony APS-C sensor, I'm not aware of any bad or even below average sensors from Sony.

How can we do a comparisons where we control for the age of the technology used in the sensor. The OM-D sensor is better than the Canon 5D so I guess we can say that there is no difference between 4/3 sensors and FF sensors...... It's the same logic. Or since the K-5 is actually much better than my old Canon 5D we can say that APS-C has surpassed FF in terms of image quality.....

Its pretty pointless to compare cameras of different formats and different generations.

My next question would be what monitor and printer are you using?

When you say "whether you’re pixel peeping at 100% on your monitor or closely examining 20 x 30 prints side-by-side under bright light, there is no difference in quality."

Most monitors only have an 8 bit display, but I assume you are using something with at least a 10-bit display and a 16 bit LUT. 8-bit displays homogenize the images displayed on them and make many differences disappear. We have the same issue with printers when dealing with the color and dynamic range. Differences captured are lost in final output.

Honestly, DXO Mark ranks the K5 II sensor as better than the 5D Mark III sensor and that is a comparison of recent generation cameras. The issue isn't really a full frame issue, but more that Canon's sensor tech has lagged behind Sony's (and I guess Toshiba and Aptina as well), and if you want a camera that has higher dynamic range, you probably need to go with Nikon or Sony or Pentax at this point.

[Winder]

Honestly, DXO Mark ranks the K5 II sensor as better than the 5D Mark III sensor and that is a comparison of recent generation cameras. The issue isn't really a full frame issue, but more that Canon's sensor tech has lagged behind Sony's (and I guess Toshiba and Aptina as well), and if you want a camera that has higher dynamic range, you probably need to go with Nikon or Sony or Pentax at this point.

This is true, but DxO uses base ISO to compare DR and Color bit depth. At base ISO the OM-D has more DR and color depth than can be reproduced with even high end inkjet printers or most monitors. How the camera manufacturer optimizes the sensor is also very important. To use the Nikon D3s as an example. Nikon optimized the D3s for high-ISO work and it excelled at that at the cost of lower ISO quality.

Both Canon and Nikon optimize their pro cameras for higher ISO field work that sports and PJs need at the expense of lower ISO quality. This is reflected in the DxO scores that use base ISO for DR and color depth. Pentax has done an excellent job with the K-5 line and optimized the sensors for shadow detail and given it a base ISO of 80, but if you look at the DR of the K-5 you see it clips highlights sooner than most of the competition. It excels at the DxO test because of this. Yes, it still has a wider DR than most, but that DR is shifted to the left. It is a trade off for better shadow detail and lower shadow noise.

[Rondec]

This is true, but DxO uses base ISO to compare DR and Color bit depth. At base ISO the OM-D has more DR and color depth than can be reproduced with even high end inkjet printers or most monitors. How the camera manufacturer optimizes the sensor is also very important. To use the Nikon D3s as an example. Nikon optimized the D3s for high-ISO work and it excelled at that at the cost of lower ISO quality.

Both Canon and Nikon optimize their pro cameras for higher ISO field work that sports and PJs need at the expense of lower ISO quality. This is reflected in the DxO scores that use base ISO for DR and color depth. Pentax has done an excellent job with the K-5 line and optimized the sensors for shadow detail and given it a base ISO of 80, but if you look at the DR of the K-5 you see it clips highlights sooner than most of the competition. It excels at the DxO test because of this. Yes, it still has a wider DR than most, but that DR is shifted to the left. It is a trade off for better shadow detail and lower shadow noise.

I don't really buy that. The difference in dynamic range is a couple of stops. Plus, the D800 does better, both in low iso and high iso situations. Comparing the 5D Mk II to the 5D Mk III sensor, you see minimal improvements from one to the other. An improved high iso score by a quarter of a stop and that's it. The big improvements in the 5D Mk III are supposed to be in the autofocus department, not the sensor.

[jsherman999]

I don't really buy that. The difference in dynamic range is a couple of stops. Plus, the D800 does better, both in low iso and high iso situations. Comparing the 5D Mk II to the 5D Mk III sensor, you see minimal improvements from one to the other. An improved high iso score by a quarter of a stop and that's it. The big improvements in the 5D Mk III are supposed to be in the autofocus department, not the sensor.

That said, it's important to keep in mind that most FF cameras will move ahead in DR after you move a bit away from base ISO - even Canon, who doesn't implement this feature as well as Sony/Nikon:

.

.

We should be matching our shooting wants/needs to the format, not picking a single DxO summary-master score to determine what camera we should buy

.

[Rondec]

That said, it's important to keep in mind that most FF cameras will move ahead in DR after you move a bit away from base ISO - even Canon, who doesn't implement this feature as well as Sony/Nikon:

We should be matching our shooting wants/needs to the format, not picking a single DxO summary-master score to determine what camera we should buy

.

DXO mark shows the K5 II equalling or surpassing the 5D Mark III dynamic range to iso 1600. Even after that, there is only a half stop between the two cameras. As I said before, it just shows that Canon is falling behind in certain respects. Both the D600 and D800 surpass the K5 at all isos except for iso 80.

[Winder]

I don't really buy that. The difference in dynamic range is a couple of stops. Plus, the D800 does better, both in low iso and high iso situations. Comparing the 5D Mk II to the 5D Mk III sensor, you see minimal improvements from one to the other. An improved high iso score by a quarter of a stop and that's it. The big improvements in the 5D Mk III are supposed to be in the autofocus department, not the sensor.

When I'm talking about the "pro" cameras I'm referring to the 1DX and D4. Both of which have been optimized for higher ISO work at the expense of lower ISO. Sensors don't have variable ISO settings. They only have a base ISO and everything else is image processing.

Even if the S/N ratio is the same for a given APS-C sensor and a FF sensor of similar pixel density, when you enlarge for print the FF sensor will yield better results.

A3+ is 158,907mm^2
D800E sensor size is 864mm^2
K-5IIs sensor size is 370mm^2

To print at A3+ the image from a D800E has to be enlarged 183 times.
To print at A3+ the image from a K-5IIs has to be enlarged 430 times.

If DxO shows the S/N and DR are exactly the same between the two, you will still get much better results from the larger sensor when you go to print simply because you are enlarging the noise in the FF image less.

If you don't see the differences then it is because the monitor you did post processing on, or the printer used for final output was not capable of reproducing those differences. A lot of noise and DR is lost in final print. A monitor is illuminated and can show much more than a print that is simply a reflective media. Most inkjets only offer the equivalent of an EV 7. Professional inkjets can only reproduce maybe the equivalent of 9EV with the right paper and ink.

If someone says they can't see a difference in prints then there are probably several good reasons.
1. The scene did not have a huge DR to begin with.
2. The image was captured in regular 8 bit JPEG.
3. The monitor was a normal 8 bit SVGA display and not capable of displaying the detail and DR of the captured image. If the monitor can't display it then you can't process.
4. The printer/ink/paper combination is not capable of reproducing the color depth or DR that is captured by the sensor.

I picked up some 24" x 36" posters from the printer the other day. They gave me the wrong order. I know the photographer and she uses Canon. I don't know if she is still using the 5DII or if she has upgraded to the 5DIII, but the images looked like mush. I don't know her settings, but I know the Canon 5DII or III is more than capable of delivering better quality that what I saw. She is a typical batch process wedding tog who has been in business for 5-7 years. She copied the LR presets from her old mentor and has never learned to really process images for maximum quality.

[Rondec]

When I'm talking about the "pro" cameras I'm referring to the 1DX and D4. Both of which have been optimized for higher ISO work at the expense of lower ISO. Sensors don't have variable ISO settings. They only have a base ISO and everything else is image processing.

Even if the S/N ratio is the same for a given APS-C sensor and a FF sensor of similar pixel density, when you enlarge for print the FF sensor will yield better results.

A3+ is 158,907mm^2
D800E sensor size is 864mm^2
K-5IIs sensor size is 370mm^2

To print at A3+ the image from a D800E has to be enlarged 183 times.
To print at A3+ the image from a K-5IIs has to be enlarged 430 times.

If DxO shows the S/N and DR are exactly the same between the two, you will still get much better results from the larger sensor when you go to print simply because you are enlarging the noise in the FF image less.

If you don't see the differences then it is because the monitor you did post processing on, or the printer used for final output was not capable of reproducing those differences. A lot of noise and DR is lost in final print. A monitor is illuminated and can show much more than a print that is simply a reflective media. Most inkjets only offer the equivalent of an EV 7. Professional inkjets can only reproduce maybe the equivalent of 9EV with the right paper and ink.

If someone says they can't see a difference in prints then there are probably several good reasons.
1. The scene did not have a huge DR to begin with.
2. The image was captured in regular 8 bit JPEG.
3. The monitor was a normal 8 bit SVGA display and not capable of displaying the detail and DR of the captured image. If the monitor can't display it then you can't process.
4. The printer/ink/paper combination is not capable of reproducing the color depth or DR that is captured by the sensor.

I picked up some 24" x 36" posters from the printer the other day. They gave me the wrong order. I know the photographer and she uses Canon. I don't know if she is still using the 5DII or if she has upgraded to the 5DIII, but the images looked like mush. I don't know her settings, but I know the Canon 5DII or III is more than capable of delivering better quality that what I saw. She is a typical batch process wedding tog who has been in business for 5-7 years. She copied the LR presets from her old mentor and has never learned to really process images for maximum quality.

You are talking about resolution and I am talking about dynamic range. They are two very different things. I know that you can print a D800 photo bigger than you can print a K5 photo. That isn't rocket science. The question is how much detail you can pull of shadows/highlights. And that is much more sensor specific than format specific.