Comparing the M8 to The M7 (Mamiya 7II that is!)

[gdi]

I have spent some time today doing a basic test of the results of an M8 and scanned Mamiya 7II files. Maybe I can sell my MF kit and free up money?? !! Specifics:

• Tripod and outdoors shots
• M8 - ISO 160, Hexanon Limited 50mm at F8
• M7II - Fuji Acros B&W ISO 100, 150mm lens @F8

• M8 files processed in Lightroom and adjustments in PS CS3
• M7II Negs scanned at 4000DPI on Nikon 9000, adjustments in PS CS3

I cropped a section of the files, in center, and upsized the M8 files using PS Bicubic Smoother in 10% increments. I have used GF and can see no difference in it and PS upsizing (in fact I think upsizing in 10% increments is a waste of time, IMHO). A print would be 24x30 at 360DPI.

Anyway, if interested, here are the results...

The boring test setup:

A crop from the M7II file:

And here is the upsized M8 crop:

 

[gdi]

Now - it would appear that the M8 file looks bad compared to the Mamiya file. But having printed similar files, it is surprising how good a file like that will actually print.

But in an attempt to even things up, I played around with the M8 files more to try to make them look closer to the 7II files. Here are a couple of tweaks.

I added a low to moderate level of noise to the file and it looks like this:

Then I tried adding some sharpening before added the noise to see what would happen...

I think the results are pretty darn good for the M8. Now, of course my next step is to print samples of the 4 files to compare. I think the M8 may hold up well at the native size of the 7II (of course, there is a lot more headroom/buffer with the Mamiya for even larger prints).

 

[sam_m]

Interesting results, I'm especially surprised at the effect of adding noise to the digital file, coupled with the sharpening it certainly has the effect of increased apparent sharpness/acuity (?) Is this just basic photoshop "noise", or something else?

[edit]
I just ran a little test, take the crops into photoshop and view them at 25%, on most monitors this equates pretty close to the printed resolution (take a tape measure and measure the image size on the screen, it should be pretty close to the 5.8cm the crops would relate to if printed at 360ppi - 24x30inch print).

My take, using this quick and dirty comparison, is that the straight upscaled m8 crop is very soft in comparison to the m7ii crop, the m8+grain crop looks no better, and the m8+sharpening+grain holds up the best of all, though still noticeably inferior to the film.

Reducing the crops to 66.6% and further to 50% (that is, down-sampling), and still viewing at 25%, simulates print sizes I am more likely to use (16x20 and 12x15 inch @ 360), and evens the playing field considerably, to the point where I struggle to see any difference.

Prints of course may tell a different story!

 

[mfogiel]

This speaks only about the resolution, and misses the main point for the B&W, which is the tonality. Try to shoot a portrait in B&W with the two cameras and compare the results.

 

[gdi]

Thanks for the comments (and tests Sam).

Sam, this has been my experience - prints will reduce the distance between the two formats significantly (though I never measured my screen!) I'll have to try them in the next few days.

Also, I did a simple "Add Noise" filter in PS.

Ideadog, I doubt anyone has ever used the same lens on both these cameras. The Mamiya lenses have electronically controlled leaf shutters.

MFogiel, I think even the small upsized crops show the difference pretty clearly. The blotchy spots on the original M8 file do illustrate this. The M8 file is just not as smooth in its tonal transitions as the 7II.

 

[Pherdinand]

I think you took the wrong way. The good think in the M8 image, IMO, is its smoothness. I dont' see that increased "apparent sharpness" by the noise added - it just looks more weird (...noisy).
I would have tried to smoothen the grain of the mamiya instead, to get it as smooth as the M8, and then compare what is the sharpness that is still retained...
Further, you used a traditional black and white film with traditional grains in it. Maybe a comparison with C41 film, or even slide, would be more appropriate (unless you only want to figure out how good you can get with the M8 for black and whites)

Also, agree wityh michiel, resolution is not everything - and it also should depend on the lens you used, as well as the film.

 

[jaapv]

Umm - those crops. How wide would the equivalent print be? 1.50 meter? I'll get a Mamiya when I want to make one, or rather if ever....

 

[jplomley]

Interesting test. Not to be too critical, as I appreciate the effort, but try a scan on a Cezanne Elite, Creo IQ3, or Tango drum scan and the difference between the M8 and Mamiya 7 will be dramatic. The 9000 is a pro-sumer scanner and simply does not have the resolution nor Dmax required to extract every last bit of information out of the neg. As well, you will see significant improvements if you oil mount. You will find plenty of information and comparisons of dry mount vs wet mount littered throughout the inter-web. Just for the record, I am not film biased. I shoot with an M8, Mamiya 7, and Arca Swiss 4x5. I simply pick the right tool for the job. It is interesting that for the Mamiya test you chose their sharpest optic, in fact reputed to be the sharpest MF telephoto ever designed. To be fair, it would have been prudent to do the same for the Leica (i.e. 50/1.4 Lux Asph).

 

[gdi]

Thanks for the comments - I was confident tha my technique would be "beta tested" here to my benefit! So here is my perspective about some of the comments.

Pherdinand..
My goal would not be to make the 7ii files look like the M8 - if so a decision to sell the 7ii would be easy! I did a quick scan of the film shot and yes, after sharpening you can see the grain. A scan of a really good transparency would raise the bar even higher for the M8, I'm afraid.

As for C-41 B&W, I never shoot it so a comparison to that would not be very meaningful to me.



Jappv...

The prints would be less than a meter. It is meaningful to me since a use a 24" printer and 24x30s are not that uncommon a goal. If I didn't want to print that size, I agree, no problem with the M8.


JPLomley...

Sorry, my point wasn't clear, I suppose. I know that there is more available in the 7ii files - but I think this is a relevant test given that there are a lot more people with a 9000 than a drum scanner or high end flatbed. This is, basically, my workflow with film and I can create excellent prints with the current setup at the max sizes I can print (24"). I do believe that high end scans would show a significant gap between the two systems.

I have not tried fluid mounting due to my impression of it being messy and toxic - but I could be wrong!

About the 150mm - I didn't choose it to put the M8 at a disadvantage, promise! 🙂 I used it because it was pretty close in fov of the 50mm on the M8 (though it may be a bit longer effective FL than the cropped Hex) I don't think the Hex is a hindrance at all at F8 - the M8 is not really suffering from a lack of sharpness at these enlargements- more from a lack of sensor "horsepower"!

BTW, do you have a Tango or Creo scanner? I'd be happy to send you a shot for a comparison of the 9000 dry mounted and a top notch scanner! 🙂 I have looked but I don't see any direct comparison samples on the web....

Thanks again all...

 

[jplomley]

gdi,

The scanners I mention are very $$$ and so I chose to outsource. As a 4x5 landscape photographer primarily, I use those $$$ for travel and the ever increasing costs of sheet film. I used to have my Cezanne and Creo scans done by Ted Harris (editor of View Camera magazine) until his unfortunate passing. This was most convenient since Ted was only 3 hr away and so we could spend the entire weekend just scanning. Now I send my flatbed scans to either JaincoTech (Tanya Stafford is the guru operator at that organization) or West Coast Imaging (WCI). For drum scans WCI is terrific. If you shoot color neg, you might be better going with Danny Burke as he has an optimized scan workflow for handling negs. Danny shoots 4x5 C-41 exclusively and is one the top fine art landscape photographers in the US, so he knows how to handle the unique characteristics of color neg scanning. I might mention that you can get Imacon 949 scans done at JaincoTech for very reasonable prices and I have done so in the past, but the success really depends on the image. Both the flatbeds have higher resolution than the Imacon (5500 ppi for the flatbeds vs 2040 ppi for the Imacon) and much better Dmax.

Hope this helps.

Cheers,
Jeff

 

[RObert Budding]

I'll take the Mamiya 7, thank you. It's less expensive, even with the cost of the scanner added in. And the resolution is clearly better.

 

[jplomley]

Robert, depends what scanner. Current pricing on the Creo IQ3 is 20K. I'd say that is more than an equalizer 🙂 If you are a low volume shooter (which certainly is the case when using a view camera), then you are much better off outsourcing.

 

[Tuolumne]

I scan my Mamiya 7 shots on an Epson flatbed 4990. I don't care what anyone says - clearly better than any digital camera I own, even with that scanner. It might be worth outsourcing for an exhibit, otherwise not. Besides, I like the control I get with my own scanning.

/T

 

[gdi]

I know the high end scanners are expensive - that's why I went for the cheap economy model! 😉

If I ever need more than the 9000 cn deliver, I can always outsource it. I am thinking again about wet scanning though. Thanks for the sources Jeff, I may give it a try just to depress myself..

BTW, If anyone has a fluid 9000 setup, I would appreciate a comparison!

Tuolumne - I agree - I am always amazed at what I see when I look at freshly developed slides from the Mamiya.

 

[gdi]

Also - I rescanned last night and changed multiscan settings and did selective sharpening. Below is a side by side with the first M7ii I posted, no noise reduction.. I believe the difference is noticeable...

 

[TJV]

Besides, I like the control I get with my own scanning.

/T

This is really the point, isn't it?

At the end of the day, your film is ending up being converted to binary code, just like any digi camera. And just like any digi camera, the world of digital workflow and editing is limited only by an individuals knowledge and taste. The problem with shooting film and digitizing is there's a huge bottle neck at the scanner end. Even outsourced pro scans, unless you've got a great relationship with a technician, are hit and miss because tastes and skill levels vary.

These days I pretty much exclusively shoot with my Mamiya 7ii. Through a loop, the trannies look like God him/herself. The scans however are clearly worse, when scanned on my V700, than 35mm trannies scanned on my Nikon Coolscan. They're okay for proofing and small book work, but if you choose to shoot MF, you really want a quality jump up from 35mm to justify the extra money, bulk and time. My M8 files we're better than V700 MF scans considering they gave a more forgiving starting point for editing and we're sharp out of the camera and not degraded during a scanning process. My limited skill working with DNG / Raw files were the problem, not the files themselves.
I've just bought a glass holder for a Nikon 9000 scanner I have access to at work. The scans I'm getting using that from my M7ii trannies are truely stunning and finally the results justify the medium format upgrade.

My honest opinion is I love the process of shooting and especially editing film. I think the M7ii is brilliant quality, clearly better in my experience than 35mm film or digi if scanned well, but the M8, and in fact any good digital camera, has an important place to fill. Like Jaap said, who wants to print that large and spend that much time and money scanning and editing all the time? If I didn't have access to a 9000ED I'd seriously consider getting rid of my 7ii, and I think that's the point of this thread, isn't it? The M8 is certainly good enough, if not overkill, for most situations that don't require you to print huge mural size images.

 

[35mmdelux]

I carry M6 or MP, and Hasselblad 501CM in my backpack. Nothing, as far as I am concerned, will outgun the larger negs, especially MF Zeiss. Two Ms, a Hasselblad, and a D200 digital. Bare bones cannot be a money issue.

 

[gdi]

Good points, TJV,

Just what I was talking about the M8 is pretty darn good and it does take a decent amount of work to get a great scan. I do find that the 9000 with glass holder really shines with transparencies. Of course it doesn't compare with projected 6x7s, but they are nice!

Here is a Ektachrome 200 shot (and the "inferior" 65mm!)...

 

[POINT OF VIEW]

Completely unscientific

Completely unscientific

shot at 8 ft. distance. Lens - very old uncoated Summicron 50 mm. 1.2 / f - 16 ISO 160. Croped twice. All letters are legible. Added contrast, only adjustment.

 

[TJV]

Inferior 65mm? Is the Mamiya 7 65mm not of good reputation?

Good points, TJV,

Here is a Ektachrome 200 shot (and the "inferior" 65mm!)...