Below are a couple RAW files: 1Ds3 with 50mm 1.4 and the Phase p30+ on a Phase One/Mamiya body with the 80mm. There were shot at a dealer in order to compare shooting speed, tonality, shadow detail, etc. A crude test at best.
*Because the viewfinder mask for the Phase was not in use, the bodies are effectively at different distances from the subjects, so these files are really no good for comparing resolution.*

The sales guy will be coming to a model shoot with me in the next week or two for a more effective test of the Phase camera…

Canon:
https://download.yousendit.com/96779E3E5C9A34A4

Phase:
https://download.yousendit.com/961BB1660EDA6194

As I wrote in another thread:
Other notes on the Phase:

Lens feels much better than older mamiya 645 as far as I rememher, AF is good – way better than the Contax AF in a dimly lit room. The Contax hunted repeatedly…

Viewfinder seems about as large as the Canons, though the Canon is brighter. (50mm 1.4 on the Canon, 80mm on the Phase.)

Phase kit is a little heavier than the Canon with 50 1.4.

Operation of the Phase is pretty fast – seems like about a second between exposures

Zooming and scrolling on the Phase is awkward I.M.H.O.

(this is a continuation of my comments in the thread about stitching...)

Ron,

Thanks for doing this. I'm interested in your take on the P30+. I have one, its great and all, but the camera part of the equation can be disappointing.

What was your take on the shutter lag, if any, of the Phase camera?

T

I don't know if it was lag, or what...but I did ask the guy if second shutter sync was on, as the camera would click twice, once at first, and then maybe a half of a second later a second click and then the strobe would fire...
It seemed odd, but I didn't have time to delve into the menus to see what was up. The sales guy didn't know what it was either.

I'll see it that still happens when I test the camera with daylight in a week or two.

Sounds like he was using mirror lock up, but I don't know for sure of course.

Nice work Godman. I like your landscapes and the scratchy portraits. (I like your name too.)

I had a look at the Canon file. Was that ISO 100? It was noisier than expected.

How did you convert it? DPP has the least noise it seems (with NR turned off). C1 4.1 gets rid of the jaggies, but it's noisier.

The two shots are quite good comparable; the Canon's exposure is about 1/3 stop lower than that of the P30+. That's very good (in a test in controlled situation one would have to make several shots with both cameras 1/3 stop apart and select those, which are the closest).

Note: the histogram shows, that ISO 100 on the P30+ is "fake", it is the numerical derivative of whatever. I don't know if this fact had any influence on the result. One would have to create a serie of images to find out, which is the base ISO.

It is important to know, that in ACR

- the P30+ shots gets adjusted by -1 EV,

- the 1DsMkIII shot is adjusted by +0.35 EV.

These adjustments are not visible on the "Exposure" slider. In order to go back to the origin, the P30+ image has to be adjusted by +1 EV and the 1DsMkIII image by -0.35 EV.

After this reverse-adjustment and picking WB on the white or grey card, the RGB values on this card reflect the small true difference in the exposure; replacing the -0.35 EV by -0.15 EV equalizes the exposures.

Anyway, Rawnalyze is more suitable for analyzing them. It shows, that the level of noise in terms of standard deviation is virtually identical in these images, measured on very dark, smooth areas. However, I find the P30+ better on such areas subjectively.

Screen captures of the histograms and of different displays are in this, 3MB large layered TIFF

Note, that Rawnalyze displays the *raw* data. There is no noise reduction, etc. except for what is unavoidable, namely black level compensation on the 1DsMkIII image. I picked WB on the white card in both images; that changes the RGB values, and it was necessary to make the standard deviations comparable. The raw values are not comparable directly, as the pixel values of the 1DsMkIII are between 0 and 14256, and of the P30+ between 0 and 65535.

I explain the captures only if someone is really interested for that.

Panopeeper, I am personally not that interested in all the heavy duty science, but I do have a variety of exposures of these scenes, so let me know if you want the Canon at 1/3 brighter...

Thanks, but it is not necessary. This small difference is all right, as the comparison shows.

I miss something else, but you can't deliver that (until another test occasion?): very fine structures in very dark areas, to judge the reproduction of details. For example in these shots there is a blue strip around a cuchion or whatever beside the battery (?). This shows, that the P30+ captured details *much* better than the 1DsMkIII (shown in the captures), but

a. it is not dark enough,

b. more importantly, it is 50% larger on the P30+ (in pixels), thus it is not reasonable to compare the details on them. You can not expect the same level of details in much less pixels.

Thank you.

The files are fairly closely matched for FoV in the vertical direction, although the P30 image is a bit closer. But this is all right if one is testing the DR of the 1Ds3. Give it a slight disadvantage and see if it holds up. It does, in my opinion.

This result is surprising for me. I've always been willing to cut the MFDB crowd a lot of slack. Bigger sensors with more pixels have to produce better image quality. Same number of pixels on bigger sensors have to produce lower shadow noise.

But these shots from #203 seems to smash the myth. Both shots seem to be equally and fully exposed to the right; if anything, they are both overexposed.

Exposure is the same for both shots. In the 100% crops below, I've lightened the shadows equally and, this time, I've upressed the smaller file (the 1Ds3) to the same size as the P30 file.

Both images, for all practical purposes, are on a par, as I see it. This is a very surprising result indeed.

Click to view attachment

Oops! The 1Ds3 image is on the right. Did you notice?

Is anyone able to assess dynamic rance differences from these files?

Why is the file on the left magenta?

I'm interested in how well the 1ds3 performs. I tested one when it came out and frankly thought it was too much of an incremental improvement over the 1ds2. Its a leap from a 5d, but not worth the upgrade.

Perhaps Michael could have a "Technical" forum that was devoted to MFDB v. 1ds3 and lens tech? Seriously, it would streamline things.

WHY?
We are medium format shooters in here...
I think Dpreview is good for 35mm dslr's..
People in here may be comparing Medium format cameras with one another, but not 35mm...
Snook

Yeah Snook, I think there should be a different forum for this stuff. But then again, I can just ignore it. I am interested in the 1ds3 for when I need a 35mm solution, not as a comparison to MFDB. I don't think the question, at least for me, is whether a 1ds3 is on par with a P30, its whether a 1ds3 is usable for my work in a pinch. I'm shooting a catalogue next week and might rent one for some of the atmospheric/environmental shots.

Yawn. I'll check back in a few weeks to see if things are any better.

The last two layers in the file I posted show a selection on the very darkest, smooth area: the black scaffold. The selected spot's pixels are in the 9th and 10th stop in both shots.

The noise level is the same in both shots, but I find the P30+ reproduction much better. Unfortunately there is nothing there to compare the fine detail reproduction because of the different magnification. For proper comparison, the magnificationmeasured on pixels should be the same; the total field of view it totally irrelevant from this point.

Note, that it required +5 EV to lift this spot out of the darkness.

My impression is, that the DR of the P30 is somewhat better, although it is partly balanced by the slightly higher exposure of the P30+ shot.

WHY?
Because anyone, professional or otherwise, who has an interest in the technology they use to make photographs will surely also be interested in the way the formats have slowly begun to converge over the last year or so. I don't think they will meet....but it's getting closer with every upgrade.

I personally use both formats, both have their particular place; but the times when I use my Mk3 instead of my Blad are becoming more frequent than they were last year with my Mk2.

Unlike a number of posts in this thread, I don't agree that the Canon files stand up against the Phase with very close scrutiny. But for a great many images that are destined to be crucified by a litho press in a commercial world... it really does do a remarkable job.

I think your "I think Dpreview is good for 35mm dslr's.." comment is simple, petty snobbery which serves no useful purpose. The tread is related to medium format photography, if it's specific content doesn't interest you....why comment at all? A number of people "in here" do seem interested.

Bitching about formats is just so tiresome...

Thanks Ron for posting the samples.... I appreciate a chance to look at this.

Also thanks Gabor for posting the photobola stuff.

I'd be surprised if anyone can really get meaningful DR data from the samples but in practical terms
just push up the exposure and lift the shadows in both files and see what kind of detail you get. I see clear banding in the 1DsIII file when pushed hard and none in the P30. But in the p30 file there is moire in the bag when you push everything up (so the detail is still there). So the P30 seems to have more reach.

Also from a purely esthetic look the P30 file has a more real look - check out the red apple on the left side - color is more real in the p30 as there is more detail.


lastly we don't need another forum for DSLR vs MFDB we just need another forum for Ray.

Here are a couple more files, this time from the H2 with the same Phase P30+ back, and with more comparable cropping.

Canon:

https://download.yousendit.com/781BD1FF45987C57



Phase:

https://download.yousendit.com/12343CDC4F46AFAB

There is a serious problem with the back on the H2. Look at the fine histograms; the first one is the P30+ on the H2, the second one is the P30+ on the Mamiya.

I have reason to believe, that this is caused by a faulty connection somewhere. The consequence of this error can be posterization.

Added: and as it is visible, this defect adds horrendeous noise.

Here's a C1 conversion (100% crop) of both files, with the 1Ds resized larger to match the Phase...so this is the equivalent of viewing a 9.5 foot print from 12" away - as rendered on your 72 ppi monitor:

I don't think the Phase file is sharpened at all. When looking at the sharpened Phase file last night I got Moire in the texture of the lizard's skin. Much, much sharper that these crops.

That being said, the ds3 crop looks a little over sharpened, but holds up well.

Yes I saw moire in the Phase too, in the sandbag on the floor, so I didn't want to over sharpen anything. I also found that halos show up pretty quickly, so I backed off the sharpening. (thus the RAW files are posted...) And I think part of the reason the 1Ds look over sharpened is because it has been enlarged to this ridiculous level so the diagonal lines can go a bit jaggy, and we are pixel peeping at 100% ;-)

Anyway, I am doing this testing only because my business model dictates that before I invest in another system, I should see some proof. Proof that it will make my images better somehow, in the real world. Is the resolution or sharpness actually better that my old Leaf back, or my Canons? Will that show in print? How about the D.R. – is that better? How much better? Would $30K be better spent flying around the world shooting images for my portfolios?

Yeah, I actually think it would.

No question the MFDB files are better than 35mm dig files, to me anyway. But, CMYK web press magazine printed, who cares. DS1, ds2, ds3, 5d, P45 whatever. 14 stops of DR is great but you are always limited to what, a three stop range of ink black and (crappy) paper white in a magazine? I looked at some images I shot with the P30 that were printed on a web press and, well, they could have been shot on 645 film, or maybe good scans of 35mm chromes. What MF offers, and is reproduced on a web press, is the look of MF, the look of the lenses and the larger sensor. That is what's important with an MFDB, to me. (Shoot a girl under hard, contrasty lights and retouch it. That's where you'll see a big difference in file quality). So, in short, I would take the cash and finance some editorials for Mixte or L'Official or French or Dazed etc., have a web site redesign and a party kicking it off at Lit/Fuse, Mars Bar, or 124 Rabbit Club or something. That will make you more money than a back.

Since I was pressed for time, I just converted both images in ACR using auto settings and 'as shot' white balance. I've addressed only shadow noise in those crops. I noticed the magenta hue in the P30 crop and I would definitely correct that when assessing the entire image for resolution and other qualities, which I will do later. I would also like to downsample the P30 file to the same size as the 1Ds3 file to see if the noise difference remains the same.

The 1Ds3 image does have slightly more banding in those crops, but it's not significant. I find it necessary to engage in a bit of pixel peeping in order to determine if differences are of pixel-peeping proportions or not. In my view the shadow noise differences in these shots are of pixel-peeping proportions and would not be noticeable even in a large print. A crop at 100% magnification on the avarage monitor is representative of a really huge print.

I matched levels across both images (the raws appear to have different tone curves applied to the *raw* data), making white backdrop, grey card, and deep shadows the same brightness.

The DB was obviously shot hotter, *and* had more blacks in the histogram.
That would mean it has LESS DR.

I also think the Canon has a larger DR, since it can go to 3200 ISO, with cleaner shadows. I suspect Canon can easily outspend and outtech the DB makers.

Look at the far right (mac?) keyboard. The Canon is much sharper. Is the Canon backfocused?

I see a slightly more saturated, slightly different hue in the P30, but not much:
Click to view attachment

Agreed, and seldom mentioned.

The real advantage for medium format, in my work, is the difference in "look" or the perspective of the lenses. The differences in resolution and color are not important to me -- most of my work is for print (magazines or advertising) and has plenty of post-production.

I routinely use my 1DsII and the H2/P30+ in the same shoots -- their strengths are in their differences. Different tools, different applications.

The Canon is great in a certain type of morning or late afternoon window light, or open shade. It is predictable and performs extremely well, handles well, and allows me to more more freely with the subject.

With the Phase, I feel more connected to the process of image-making, feeling the image in the viewfinder as I press the shutter. The file are more flexible in post, and I prefer the perspective of the longer medium format lenses. But I find myself focusing on the camera more than I would like -- the whole system is really heavy, and I have to pay more attention to the autofocus. If only the H body had an extra focus point to focus on faces in verticals...

In the studio, and most cases on location, the difference in quality is worth it. But, most of that advantage is only for me to see -- once printed, the only difference that anyone will notice is the difference in perspectives or "look" between medium format and 35mm.

Just as it's always been, film or digital.

It's hard to tell exactly what you mean by the 'look' of medium format, but since you mention the word 'perspective' twice I'm wondering if you're somehow under the impression that perspective has anything to do with lens or format size. Perspective is defined by camera to subject distance, and nothing else. So unless you're talking about shallow DOF, I'm curious what exactly you mean.

If you are talking about DOF, then I suppose that may be a legitimate difference, but only if there is no lens for the 35mm system that will allow you to achieve the combination of perspective, field of view, and depth of field that you get with your MF lens.

That is a common misconception, but it is more surprizing here, on a "professional" forum, than for example on DPReview.

MF looks different. It is the DOF but more the focus fall off. Its different than 35 or a disc camera or 4x5. Its also in the lenses themselves, in that they are better than most 35mm lenses.

But look, I'm not trying to convince anyone. If you can't see a difference in looks in photographs produced by different formats, then don't sweat it. You won't need to get one.

I think the tone of your comment was uncalled for.

When you closely look at the files and the lizzard you can see that with the MFD the lizzard is more LOOSE from the background while on the other shot it's more part of the background.

This for me is the great appeal for MFD, it gives the pictures a more 3D look and not the flat.

This is even better see when going even bigger to for example a 6x7 camera with film.

I don't know what you guys are seeing but throwing focus with medium format digital, especially the medium format lenses that only go to 2.8 or 3.5 is not that spectacular and compared to the 35mm lenses that go to 1.2, 1.4 and 1.8 medium format doesn't have that much if any advantage.

http://www.dofmaster.com/dofjs.html

Keep in mind that 645 is the smallest of the medium format frames and the sensors we have now aren't even full 645 so you always working at a greater distance from the subject.

If it wasn't for cost, I'm sure there would be a complete rethink of the lenses, given the sensor sizes we have now.

Granted at the same f stop a larger format size will throw more focus, but you can always move the numbers to get a look.

I can give you a lot of reasons to shoot a medium format back, but the "3d look" whatever that means isn't necessarily one of them.



Where medium format does have an advantage is if you shoot to a vertical page, as you don't have to keep walking backwards to fit the page, though this advantage works completely opposite when you go horizontal for double truck.

JR


JR

Just out of curiosity, do any of these look "3-D"?

203 Great Shot's..
Specially the second one, the african? girl, Just beautiful!!
The lighting is great..
Is it window light camera left and fill with CT gel?
Looks very nice..
Snook

Really, really nice images. Congrats. I wish I had a studio with light like that!

Thanks Snook and Gregory.

The second image is just daylight from camera left, in my studio - the blue highlights from the blue sky. I think there was light also coming in from another window and bouncing off a beige wall, creating the highlight on her back. (all of these images are 100% daylight, with a mirror & blue gel for the highlights in the third image.)

We actually had some incredible light last week, when the rays were periodically glancing off the Empire State building and into my space. I like daylight for the randomness. What you see behind the model (above her head) in the below image is some light which was randomly bouncing off of parked cars across the street (we were shooting in a hotel.) Someone asked me if I wanted them to block the reflections, and I said, certainly not. The reflections came and went in different formations throughout the day, making each picture a little different.


And yes, the model in the second image in the previous post is African. She was a runner, training for the olympics when a scout from Ford Models saw her picture in the paper, and that was the end of her running career ;-)

http://nymag.com/fashion/models/anasanyana/ajumanasanyana/

Hi,
I would have had to see the shots next to each other.

The so called 3D effect is caused by MANY factors.

Being:
1. Light/shadows
2. Pose/placement
3. DOF
4. Contrast

A lot of these can be achieved with any camera that is of good quality.
I have made some nice 3D looking shots with a 10D.

What I find however is that when I compare two identical shots next to each other the MF shot gives me more sense of 3D.
When you look again at the two samples, the lizzard is more loose of the background than in the other shot, also the stones are more rounded on that same shot, and the red apple looks rounder than on the other shot.

Again, you can get stunning results with all cameras if you control the scene, but make the same shot with a DSLR and a MFDB and somehow the difference is seen in most cases for me.

I think it's mostly due to the better quality of the pixels (higher dynamic range, no filtering etc.).

What I like further about MF is the compression of the longer lenses vs DOF they give, it just is a different look.

But we can talk ages about it, some people see it some don't.
I have a background in Home Theater and Calibrations (lots of them) and there are people that don't see a difference between a $1000.00 projector 720P and a $6000.00 projector 1080P with 1080P source material.
While for me it's day and night.

203 Thanks for the information..
The girl is VERY beautiful. Photographically speaking to me..:+}
Nice Daylight..
I have not shot Daylight in years..
Actually Have shot much since Digital because I have always been un happy with the results and usually I am doing so many pictures per day that it is impossible to have the light run out..
I can imagine that you have a very short window for shooting and tripod a must..?

Nice shot's...
You mind me asking the shot of the girl.. what camera lens and what speend and aperture more or less... what you developing in etc..
Sorry I just really like it and it inspires me to start shooting more daylight..:+}
Thanks
*Snook

You listed just above the "reasons" for the 3D effect, none of which is partticular characteristic of MFDBs. So, there must be something else.


I wonder how (at which zooming) you are looking at them. I find the size difference in pixels very important in the overall effect.

In order to be comparable, not the field of views should be the same but the size of the objects in pixels. Then the larger image should be cropped to the size of the smaller one, and then it becomes reasonable to compare the effect.


The lack of filtering contributes to the sharpness, but that plays no role in the present case.

However, mentioning the dynamic range in this context is particularly interesting: the P45+ image (the second one, on the H2) is wrecked. Its dynamic range is less than that of the 1DsMkIII, so I wonder what you are seeing there.

I don't know what other people see, but when I look at the shots, the MFDB shot just looks more real to me.

But again forums are written about this effect and NO-ONE ever has given a 100% answer.
I only know what I see, and trust me I don't have money to burn and for what I normally do a 1DsII is more than enough

I made the switch to MF because the pictures looked more real for me.

I did not doubt that; I am looking for a plausible explanation.

My own explanation is, that it is the large field of view, coupled with minimal rectilinear distortion. The same field of view can be achieved on DSLRs only with much shorter lenses, which are everything but stellar. Therefor, photogs with DSLRs use lenses with smaller angle of view to achieve the same field of view, which means taking a farther position, i.e. the perspective changes.

In other words, the wide angle of view with MFDBs allows you to shoot from "the middle of the scenery".

Could be, I tried to explain it to myself many times but I don't have a 100% answer.
The only thing I know that when I compare the MFDB shots with my own 5D the 5D seems to be flat.

In the beginning I did some test setups with three bottles in different distances and there the difference was obvious for me (I searched for the photo but can not find it anymore )

re:
"....I would have had to see the shots next to each other...."

So you can only see this effect if the image is next to another image?



This is why I asked you to comment on the series of JPEGs I posted, as they are taken with a variety of cameras/lenses. Care to comment on the "3D"ness of each of them?
Anyone else?
Also, as you can see there is a 4x5 thrown in there. That should look more 3D than the rest for sure, right?

Snook, you should reaqaint yourself with shooting daylight. It's great. While the light changes throughout the day, as long as you have enough windows you should be fine. I just shot a catalog, all daylight, and we were shooting between around 11:00 am until 5 or 6 pm each day, with no problem at all - see two samples from that shoot below.

Of course you need a tripod, and sometimes shift the ISO around.
There are so many dudes shooting with the same studio lighting look (Profotos, bla bla bla) that using the available light in the studio can make your work stand out a bit.

Plus it's free, and you don't have to lift anything ;-) I say this jokingly, but you can imagine how free it felt when I went to Paris to shoot in April with JUST CAMERAS. No 80 lbs. of strobe stuff. And the pictures end up looking nicer, IMO.

It's more with less.

last week's daylight cashmere catalog:
the one with the bow tie:
1/50s f/5.0 at 85.0mm iso200

the other one:
1/100s f/3.5 at 85.0mm iso400