The Pentax 645z (I’ll call it the Z) houses a new cmos sensor and generates jpg’s that are slightly different to it’s old self, the 645d or I’ll call the D. During the first week I received the Z I wasn’t shooting to compare and contrast, but as I shot more with the Z along side the D, my inner urge told me to start comparing the two at base iso 200. Why at iso200? because that is where most 645D users stay at to get the best quality file from the D. So it’s natural to know if there is a difference at base iso and whether upgrading to the Z makes any sense if you remove the high-iso equation from these two cameras. How much resolution advantage does the Z really have over the D? Are there any differences in color? What about dynamic range? And should I keep the D because of it’s ccd sensor “look” is the question I keep asking myself. Although my camera is now on ebay I am thinking about keeping the D as my backup. But if I can make a real comparison between the two cameras and see that the Z can produce similar or better files at base iso 200 then I can comfortably let go of the D and bid it farewell. All these example files were produced using Pentax Digital Utility software v5 and/or are ooc (out of camera jog’s). As I address these questions from this post, I will also revisit these important topics again as I get more familiar with the new Z.
SKIN TONES AND COLOR
Here is the first comparison. Skin Tones — the D likes magenta and generally produces jpg’s with a magenta cast. The Z got rid of this and now you will see a warm cast in your images. And it’s not the kind of warm that’s orangy (ala Nikon) but closer to Canon’s warmth (with more red/brown tones) but not exactly the same. I welcome this change very much as I generally prefer warmth. Look closely and you can see it on the model’s skin. White balance set to AWB.
Although not as apparent, the picture on your left (D) has a cooler magenta cast.
Below you can really see the difference in skin tone. (D left with a cooler cast, Z right with warm cast). The D was shot with a 75mm/2.8 lens and the Z with a 200mm/4 lens — although you might suspect that the coatings might have played a role in influencing the color, I assure you this was not the case – both are Pentax lenses and I’m sure they use the same coatings to match all their lenses. Also as a side note, to get the best looking face from your subject, use longer lenses. You can see a slight face distortion with the 75mm lens with this model. The 200mm “compressed look” looks so much better on him. Now I’m not saying that the 75mm can’t be used for headshots — in fact I use it all the time for candid closeups and photojournalism it’s just that beauty is not a concern for those type of applications.
* I noticed that the D is producing brighter images by a half stop or more than the Z given the same settings. I don’t know why this is happening.
Another difference between the two cameras is the shadow color. On the D (left) the darks or shadow’s tend to be more on the green side, whereas the Z is more red/brown. This is something I’ve never noticed about the D and I have been using this camera since 2012. I have set the white balance to daylight and used a daylight balanced continuous light source.
The reds from the Z are prettier and more alive and it looks good and not overdone. By default, this will make the Z more vibrant-looking than the D (assuming both cameras use the same color-modes, etc.). However the reds from the D are closer to the original rendition of the picture below which brings me to another point. In general, the overall tonal rendition of the D is more natural looking than the Z. This opinion might change as I get more familiar with the Z but this is my initial impression.
When I first started using the D, I was excited by the possibilities of doing natural light portraits outdoors on a medium format sensor. After countless shoots I am of the opinion that the D is not very good at sunny/contrasty conditions, especially if the subject is backlit and you are metering for the person’s face. You will see that almost every highlight will be blown out. It’s a very frustrating experience and I have learned to shoot in shady areas instead, which are safe options for the background. Another option is to use the camera’s matrix meter which will underexpose your pictures (because it’s accounting for the highlights) and you will have to do some post work afterwards to bring up the foreground. Or as many do, use a second light like a flash to control the brightness of the background. Anyway, a good digital picture is when all highlighted pixels have color/brightness information intact.
As you can see from this example (both 645D pics), I metered for this girl’s face but then the highlights from the trees, grass, and her hand are completely blown out. I tried to recover highlights by underexposing 3 stops in post but as you can see I lost most of them.
Ok now the comparison with the new Z. Here I metered for the boy’s face. It has been said that the Z has 2 more stops of highlight information than the D (the 645z has 14 stops and the 645d has 12-12.5 stops). Here on the right picture, you can see the Z highlights are not blown out on the green neon shirt and tennis court behind the subject and retains color information but on the D (left), all of this color information is gone and you will probably, at best, recover only about one stop of highlight information using software, or most likely you will not recover anything at all. Again using the -3EV technique, color information is gone on the D. The Z holds up pretty well here. If there is a reason to get the Z, this would be on top of my list.
This is a topic I will have to explore more than once because it’s not an easy topic to talk about. Setting up the test is probably the hardest part. But so far, this is what I have observed. The Z shadow’s are contrastier and darker than the original D at iso 200. As you can see below, the D is starting to show grain at iso 200. Moreover, the D has a matte look to it’s shadows giving it the appearance that there is more detail there. The Z is darker and looks noise-less. Because it is darker there appears to be less detail. However don’t be fooled. There is alot of information in the Z’s shadow that is being hidden from us.
* Disregard the sharpness of the two pictures below – the brick on the Z is clearly sharper but that was due some focusing issues during the test. Noise reduction was turned off for both cameras and views are 100 percent.
As you can see the Z is much darker in the shadows than the D. Out of curiosity, I pushed the EV on the Z by 2.5 stops in post and discovered two things. First there is much more detail in the Z’s shadow than the D at iso 200. Much more. Second if you like to lift shadows in post, it takes about 2.5 EV’s of Z’s noise to match D’s noise level at iso 200. That would mean the grain on the D’s iso200 is about iso 1250 on the Z (let’s see from iso 200, iso 400 is 1 stop, iso 800 is 2 stops and another half stop would make it iso 1250).
* I had to match the brightness of the +2.5EV Z to match iso200 on the D so it looks somewhat the same. There’s clearly see alot more information on the Z.
Here I went a little overboard and I set the EV at +3.0 on the Z to see what it looks like. There is alot more noise on the Z. So I’ve concluded that the noise structure on the Z is somewhere between 2.5 – 2.75 EV’s at iso 200 on the D.
So am I saying that iso 1250 on the Z is the new iso 200? From these examples it may seem like that but that is farther from the truth. The noise/grain structure may be similar but if you shoot at iso 1250 on the Z you will automatically lose clarity. Although the Z was designed to be a high-iso machine, the more you go up the iso scale the more things get smeared up. It’s just the way it is. Just like the D, if you want ultimate image quality or retain as much detail as possible you must stay closer to iso 200. There is no way around it. The question is which iso on the Z does smearing become acceptable to you as a photographer since noise becomes less of an issue with the Z. This is a question I will address in another post. View’s below are 100 percent.
Here the D looks so much better at iso 200 than the Z at iso 1250 although the noise structure is similar. Observe the flower pattern on the Z and the blurry details become obvious compared to the D.
Resolution and micro-contrast
Well just how much more detail are you getting from the Z? Well on the image below the Z on the right has a bit more resolution than the D but it’s not much at all. If one were going to buy the Z soley on the increased resolution, I would just buy a D because there really is not much of an advantage here. We all know that the Z has 10MP more information than the D – but how does that resolve in the micro-contrast department and is the Z sharper than the D because it has 10MP more resolution? The answer to this question is not straightforward. Although we can see from this example that the Z has slightly better detail (from the extra 10MP) than the D, the the D looks to have slightly better micro contrast even though it has less resolution. Or one might argue it looks the same. But you have to look really close because I can see that the edges on the D pop a little more than the Z. Perhaps the D’s CCD sensor is able to produce more acuity and sharper pixels? Or perhaps Pentax needed to put 10mp more to match the acuity and sharpness of the D to even the playing field? Anyway I will have to explore this micro-contrast topic again as I was surprised by these initial findings.
** I repeated this test again with the same aperture, shutter speed, iso, and infinity focus and noticed that the D is always producing a brighter image by as much as a half a stop. I don’t know why this is happening. But the result is the same. I feel like the D has a little bit more presence than the Z at iso 200. But it’s not much and the extra 10MP does help in offsetting any difference in acuity or sharpness with the D.
There you go. Hopefully this is some helpful information. In the meantime, I am still investigating the differences so there will be mini-reports coming up.