Want – Need – Afford

[Bill VN]

Mike,

You make some valid points in your essay, but miss some of the economic realities. Throughout the 1980s and 1990s, the best medium format system around with the highest quality was arguably Hasselblad. A basic 500 series body cost around $2,000 to $2,500 USD new, and lenses were around the same. The best Nikon was roughly half the cost.

Today, a modern Hasselblad H-39 is around $25,000, roughly fives times the cost of a Nikon D3 and equivalent to a decent new car. Now I realize that pro photographers saved on film and processing costs, which justified the switch to digital. But, a huge number of artists and advanced amateurs were left out of the MF digital changeover due to the obscene prices. Some of this is due to the cost of CCD and CMOS chips, but a lot is also due to a monopoly of the market by Danish MF back suppliers.

Even in the world of music, a top-of-the-line Martin D-45 still lists for $9,999 USD (roughly $6,000 street).

Landscape photographers and others have been using 8x10, 4x5 and 120 film formats for very sound technical reasons. Very few of them can afford MF digital equipment today.

Again, you make some valid points, but let's not forget who this equipment is made for. The anger over the cost of previously attainable MF systems has a real basis amongst those who care the most about high quality photography.

Regards,

Bill V.

[michael]

What you say is quite correct.

But, I would argue that on any absolute image quality scale you'll find that one can now produce image quality equal to that from a 1980s Hasselblad with a camera such as a Canon 5D MKII or Sony A900, which are priced well under $3,000, even without figuring in an inflation factor.

Today's medium format is on another plane alltogether, and because it pushes current technology to its commercial limits, is high priced.

Michael

[BernardLanguillier]

Landscape photographers and others have been using 8x10, 4x5 and 120 film formats for very sound technical reasons. Very few of them can afford MF digital equipment today.

Again, you make some valid points, but let's not forget who this equipment is made for. The anger over the cost of previously attainable MF systems has a real basis amongst those who care the most about high quality photography.

I would argue that the new 8x10 is not the P65+, it is stitching based imaging, be it from a high end DSLR of from a MFDB. Stitching is the great equalizer since pixel quality becomes the only meaningful measurement...



Not a solution for all needs, but clearly the best absolute image quality by far for landscape, together with an un-beatable price/performance ratio.

Fine art photographers looking for the best possible image quality should IMHO consider it seriously.

Cheers,
Bernard

[Josh-H]

I would argue that the new 8x10 is not the P65+, it is stitching based imaging, be it from a high end DSLR of from a MFDB. Stitching is the great equalizer since pixel quality becomes the only meaningful measurement...

 )

P.S - Lovely Shot.

[Bill VN]

I would argue that the new 8x10 is not the P65+, it is stitching based imaging, be it from a high end DSLR of from a MFDB. Stitching is the great equalizer since pixel quality becomes the only meaningful measurement...



Not a solution for all needs, but clearly the best absolute image quality by far for landscape, together with an un-beatable price/performance ratio.

Fine art photographers looking for the best possible image quality should IMHO consider it seriously.

Cheers,
Bernard

Stitching is one technique, which I use, by the way, with my Nikon D200:

[attachment=16192:20060208...ountains.jpg]

However, we have been talking from a perspective of someone who moved up to MF from 35mm in order to get better results and differentiate themselves. Many of us moved down to MF as a lighter alternative to field cameras and film holders--just like St. Ansel (Adams). We have large investments in optics and equipment, which are not economically updatable to digital. I have a Phase One H20 back I purchased for $3,500, but it has to be tethered. A P20 is just simply beyond my means, and as a long time Hasselblad owner, I am obviously not poor. Even the least expensive, "student level" Better Light back is on the order of five grand. Vendors have to come up with more economical digital alternatives. Perhaps, a scanning back for the Hasselblad could be done for a couple of thousand. Who knows?

Regarding quality, an 8x10 contact print is way beyond the ability of any current digital capture system; you only have to view one at an Edward Weston exhibition to understand. In the end, digital capture only provides an ethereal abstraction of reality recorded by ones and zeroes based on several million tiny light meters. But in the end, it will never be "writing with light," i.e. photography.

[Pete Ferling]

Stitching doesn't solve issues when things are moving in a scene, or dealing with precise movements when a wave hits the pier, or the sun is setting and you have a brief moment to get it right, etc.  Also, there are times that I simply don't want to 3 stop bracket and pan a scene.  Then process nine images in five steps and round trip to PS twice.  It gets ugly fast.

[BernardLanguillier]

However, we have been talking from a perspective of someone who moved up to MF from 35mm in order to get better results and differentiate themselves. Many of us moved down to MF as a lighter alternative to field cameras and film holders--just like St. Ansel (Adams). We have large investments in optics and equipment, which are not economically updatable to digital. I have a Phase One H20 back I purchased for $3,500, but it has to be tethered. A P20 is just simply beyond my means, and as a long time Hasselblad owner, I am obviously not poor. Even the least expensive, "student level" Better Light back is on the order of five grand. Vendors have to come up with more economical digital alternatives. Perhaps, a scanning back for the Hasselblad could be done for a couple of thousand. Who knows?

Agreed, the entry point is now around 7000 US$, which is much higher than it used to be and this is specially painful for those invested in MF gear. Now you should be aware that older MF lenses will not be fully able to tap into the resolution potential of the latest backs... which is why Mamiya has been redesigning its all lens line up.

So yes, did MFDB's elite pricing hurt very badely photography? My opinion has always been that they did. Michael is right though that a 5dII will get you close though.

Regarding quality, an 8x10 contact print is way beyond the ability of any current digital capture system; you only have to view one at an Edward Weston exhibition to understand. In the end, digital capture only provides an ethereal abstraction of reality recorded by ones and zeroes based on several million tiny light meters. But in the end, it will never be "writing with light," i.e. photography.

IMHO the limitation you are talking about here is that of printing technology, not capture. So contact print is a scenario that is very kind to 8x10. Talk about a normal size enlargement, say B0, and I believe that a properlly executed 300 megapixel pano will fare better than drum scanned 8x10.

Light in itself is a discrete quantity and the way light is materialized by film is also made of grains that are spatially well defined and therefore discrete in nature. There is no philosophical gap between the bits and zeros of digital and film. Only the native processes designed to handle film and digital differ which gives us an illusion of continuity for film.

Cheers,
Bernard

[BernardLanguillier]

Stitching doesn't solve issues when things are moving in a scene, or dealing with precise movements when a wave hits the pier, or the sun is setting and you have a brief moment to get it right, etc.  Also, there are times that I simply don't want to 3 stop bracket and pan a scene.  Then process nine images in five steps and round trip to PS twice.  It gets ugly fast.

This thread is originally not about stitching and I should probably not have brought up the topic as it is deviating a bit, but for what it is worth, it is true that stitching doesn't handle everything perfectly, but it can a handle a lot more than you first think when starting to use the technique.

- movement is OK as long as the size of the moving object is smaller than each frame in the scene (birds,...), or continuous in nature (clouds are a total non issue nowadays),
- movement of water is OK as long as the critical scale of the waves is small enough relative to the size of the scene, or with long shutter speeds. The example you give is indeed challenging,
- Sunsets are not a problem at all,
- since I started to use the d3x, I have basically mostly stopped doing bracketing,but if you have bracketed AutoPano Pro and PTgui will deal with it perfectly thanks to their built-in HDR capability. Just feed in the images optinally adjust a few parameters and wait for the output.

The attached set gives many examples of the above:

http://www.flickr.com/photos/bernardlangui...57600916381270/

Cheers,
Bernard

[BernardLanguillier]

Hi Bernard,

I agree stitching is a partial solution only - there are many instances where stitching is either impossible or virtually so.

Stitching can of course bring a lower MP camera up to or greater than the same number of pixels from a MFDB - but pixel count is not the whole story. There are a lot of other factors owners of High end MFDB's attest to as significantly superior to even high end DSLR's like my 1DSMK3 or your D3X.

I think stitching remains a good viable option - but it isnt a substitute for a high end MFDB in my opinion.

For general landscape work, I very much think that stitching with a high end DSLR is a better option that a single frame from a high end MFDB (and mostly than stitched MFDB as well). There are obviously certain images that I will not be able to capture, but the result will be much better for the 95% I will be able to handle.

As far as these high end backs being vastly superior to a D3x in terms of pixel quality, I believe this to be an urban legend non backed up by facts or real world usage.

This sample, already posted a few days ago, might help you understand why I think so. http://www.flickr.com/photos/bernardlangui...720762/sizes/o/

Cheers,
Bernard

[Christopher]

For general landscape work, I very much think that stitching with a high end DSLR is a better option that a single frame from a high end MFDB (and mostly than stitched MFDB as well). There are obviously certain images that I will not be able to capture, but the result will be much better for the 95% I will be able to handle.

As far as these high end backs being vastly superior to a D3x in terms of pixel quality, I believe this to be an urban legend non backed up by facts or real world usage.

This sample, already posted a few days ago, might help you understand why I think so. http://www.flickr.com/photos/bernardlangui...720762/sizes/o/

Cheers,
Bernard

It is all true, but I think it also depends how you want to work. I stitched a lot and still do it from time to time. However I really don't like the feeling. There is nothing nicer than to set up a large format camera and make one image with a MFDB.

Another factor is what you NEED in terms of pixels. Do I really need 100 or 200MP to print ? I know the biggest I will print is something like 40in by XXXin. Is it nice to have 100Mp to print that big ? Yes, but mostly something around 40, will be enough to do the job. The fact I hate about doing many stitches and large one is files are getting bigger and bigger. HDs are cheap, but still I prefer a 500MB tiff to a 4GB tiff ^^

[Josh-H]

For general landscape work, I very much think that stitching with a high end DSLR is a better option that a single frame from a high end MFDB (and mostly than stitched MFDB as well).

Why?

As far as these high end backs being vastly superior to a D3x in terms of pixel quality, I believe this to be an urban legend non backed up by facts or real world usage.

I dont agree with this statement (and I dont think anyone who has viewed high end MFDB images would either) - I have personally seen and viewed D3X files, (my own 1DSMK3 files) and files from high end MFDB's from Phase and Leaf. There is nothing urban legend about it - the MFDB files possess a depth of clarity, accutance and color that DSLR files simply dont have (and that includes both the D3X and 1DSMK3). Does it translate in an 8x10 print? No.. not really, or not at all to be more accurate (just witness the canon G10 v. P45+ back debate) - does it come across in a much larger print or at 100% on screen? - most definitely.

As good as the D3X and 1DSMk3 are (and they are both very very good - I use and love the 1DSMK3 daily) - they cannot compete with the high end MFDB's on the market today - especially the new P65+. Do you think Jeff S. and Michael R and may other pro's. would be shelling out tens of thousands for these backs if there wasn't something to it?

Michael said it in a nutshell above -

Today's medium format is on another plane alltogether

Thus bringing us back to the whole value discussion - if person 'A' see's value in the MFDB, then its worth the price of admission for their line of work provided their hip pocket can take the hit, and/or provided they can get their investment back. And that is the real key to the lock when it comes to value. As a professional photographer its all about return on investment - if product 'A' costs me 'X' dollars, but makes me 'X+1' then its worth the price of admission. If it just makes me 'X' back - then it isn't worth it. Its a business decision based on the individuals business.

Edit - BTW: Bernard, you made a highly valid point above that is worth re-iterating. Price v. Performance - its hard to go passed a stitched image from a high end DSLR.

[michael]

In the end, digital capture only provides an ethereal abstraction of reality recorded by ones and zeroes based on several million tiny light meters. But in the end, it will never be "writing with light," i.e. photography.

Bill,

If you believe this then you should be shooting with a sensor not film, because film is a digital format and a sensor based system is at its heart analog in nature.

How so? Well, each grain of silver is either exposed or unexposed. There's no such thing as a silver grain that's partially exposed. It's on or it's off. ie: it's a one or a zero. It's therefore totally digital in nature. The reason that film looks continuous tone is because particles of silver halide are not consistent. Some turn black sooner or later than others when exposed to the same amount of light. It's millions of these little digital receptors that record the image. This is what gives film a shoulder and a toe curve, rathar than a linear response to light.

On the other hand a camera's sensor is an analog device. It has a continuous response to light and a variable voltage is generated in response to the number of photons hitting the photo site. We then have to turn that voltage into a digital signal so we can store it, transfer it and control it.

The bottom line – film is the original digital medium. If you don't want ones and zeros recording your image you'd better switch to digital.  

Michael

[dreed]

(let me join two topics in one reply - Want-need-Afford & 17mm T-S/E...)
Related to the topic, but not the MFDB discussion, the ne Canon 17mm T-SE fits firmly into this for me. Whilst my photography is almost entirely based around travel and trying to capture something that my eye likes, what I find is that when it comes to certain circumstances (nearly all of which involve buildings), the pictures suffer.  Putting down $2.5k on a lens that will be used "occasionally" is a challenging decision!

As a lens, this one is twice the price, according to B&H, of the 24mm, but yet is sold out whilst the 24mm is not. How do I justify the 17mm over the 24mm? I can get useful wideangle on FF, APS-H and APS-C from 17mm but not 24mm.

For those that have and use T-S lenses, how are they to use with filters such as polarizers and natural graident? In this regard, the 17mm is something of a conundrum: its design would seem to preclude the use of screw in filters - do the filter holder systems make a difference? (Can they be used?)

When I see what one of these lenses does, I feel like I need one (and perhps more to the point, that every camera/lens should provide this) but do I really need to spend that much cash on something that will probably see as many days of use per year as it has mm?

[BernardLanguillier]

Why?

I dont agree with this statement (and I dont think anyone who has viewed high end MFDB images would either) - I have personally seen and viewed D3X files, (my own 1DSMK3 files) and files from high end MFDB's from Phase and Leaf. There is nothing urban legend about it - the MFDB files possess a depth of clarity, accutance and color that DSLR files simply dont have (and that includes both the D3X and 1DSMK3). Does it translate in an 8x10 print? No.. not really, or not at all to be more accurate (just witness the canon G10 v. P45+ back debate) - does it come across in a much larger print or at 100% on screen? - most definitely.

As good as the D3X and 1DSMk3 are (and they are both very very good - I use and love the 1DSMK3 daily) - they cannot compete with the high end MFDB's on the market today - especially the new P65+. Do you think Jeff S. and Michael R and may other pro's. would be shelling out tens of thousands for these backs if there wasn't something to it?

Michael said it in a nutshell above -

Thus bringing us back to the whole value discussion - if person 'A' see's value in the MFDB, then its worth the price of admission for their line of work provided their hip pocket can take the hit, and/or provided they can get their investment back. And that is the real key to the lock when it comes to value. As a professional photographer its all about return on investment - if product 'A' costs me 'X' dollars, but makes me 'X+1' then its worth the price of admission. If it just makes me 'X' back - then it isn't worth it. Its a business decision based on the individuals business.

Edit - BTW: Bernard, you made a highly valid point above that is worth re-iterating. Price v. Performance - its hard to go passed a stitched image from a high end DSLR.

I don't want to go to far into this as it has been discussed quite a bit already, but let's just say that after normal sharpening, I have never seen a MFDB file that looked significantly better than a d3x file. Resolutionwise also, although this stops to be relevant once stitching is part of the equation.

It could very well be that the MFDB files I saw (my Mamiya ZD files or others) had not be optimally captured, so I would be interested in looking at one of your sample files (a crop would do). I would have no problem changing my mind on this.

I believe that many high end shooters buy a P65+ mostly because they don't want to bother with stitching.

Finally, productivity is a strong advocate against stitching in some domains, but clearly not fine art. So I am personally 100% sure that for the fine art landscape work I am trying to do (whether I am succeeding or not is a different matter), a 40.000 US$ back would have zero of negative value compared to my current kit.

As to why landscape stitching is better done with a D3x than a back, there are many objective reasons like:

- lighter than most MF systems, especially when dealing with the kind of focal lenghts I consider best for stitching (100mm on FX), pancake cameras are the exception but are practically very hard to focus accurately in the field (think low light levels,...),
- access to a very wide array of top quality lenses from 14 to 300 mm (to only mention those that I actually use), the image below was shot with a 300 f2.8 as an example:



- live view enables perfect focusing of the main subject with 100% accuracy 100% of the time,
- much better high ISO image quality enlarges dramatically the range of scenes that can be captures with stitching (windy situations in low light,...), the image below was shot at 800 ISO because of low light level and the need to maintain enough shutter speed to avoid clouds migration:



- more DoF thanks to the smaller format reduces the need to do DoF stacking and reduces the lenght of the exposures significantly at equal DoF, which is critical at sunset and sunrise where skies change every 15 sec of so,
- much better long exposure image quality and lack of dark frame substraction until 8 sec results in much easier low light panorama shooting,
- much longer battery life lends itself well to the large amount of capture induced by stitching,
- no need for color calibration when shooting wide,
- much better handling of cold weather,
- support of panoramic robotic heads,
- SDK enables automation of HDR/DoF when needed,
- total lack of moire reduces the need to check images at 100% pixel magnification in post and does therefore saves time,
- typically better availability of accessories from third party (L brackets,...),
- much lower price makes it actually possible to carry a credible back up body when doing long over seas missions,
- better support from third party raw conversion software from some of the backs give more options,
- the 3:2 vs 4:3 aspect ratio increases the pixel count on the long side of the frame and reduces the need to do multi-row stitching at equal resolution (a 24 MP d3x is about equivalent to a 28MP back from this standpoint)
- ...

Cheers,
Bernard

[Bill VN]

Bill,

If you believe this then you should be shooting with a sensor not film, because film is a digital format and a sensor based system is at its heart analog in nature.

How so? Well, each grain of silver is either exposed or unexposed. There's no such thing as a silver grain that's partially exposed. It's on or it's off. ie: it's a one or a zero. It's therefore totally digital in nature. The reason that film looks continuous tone is because particles of silver halide are not consistent. Some turn black sooner or later than others when exposed to the same amount of light. It's millions of these little digital receptors that record the image. This is what gives film a shoulder and a toe curve, rathar than a linear response to light.

On the other hand a camera's sensor is an analog device. It has a continuous response to light and a variable voltage is generated in response to the number of photons hitting the photo site. We then have to turn that voltage into a digital signal so we can store it, transfer it and control it.

The bottom line – film is the original digital medium. If you don't want ones and zeros recording your image you'd better switch to digital.  

Michael

Mike, thanks for adding back the part of my comment I originally edited out. I am perfectly aware of the analog v. digital irony of digital imaging, but in the end, a traditional photograph is a physical medium, not electrical charges in the ether. Remember that when you back up your files yet again ten years from now on some yet to be invented storage device.

[Jeremy Payne]

Remember that when you back up your files yet again ten years from now on some yet to be invented storage device.

Yes ... every time I make yet another perfect, pixel-for-pixel copy of my captures and 'prints' with absolutely no information loss I will remember how I longed to do that with my film and slides ...

[dreed]

Yes ... every time I make yet another perfect, pixel-for-pixel copy of my captures and 'prints' with absolutely no information loss I will remember how I longed to do that with my film and slides ...

It is not enough to copy files - hardware can corrupt them without you being aware - you need to verify every copy, just like when burning DVDs, if you want to be sure that no bits were lost. That includes the initial one from the camera/camera storage.

All computer storage mediums have non-zero MTBF (mean time between failure) and error rates. Thus "xcopy /v/e" is my tool to copy files from camera storage to pc storage on Windows.

I would be interested to know if those small picture storage units (meant for use in the field where you don't have a laptop) do this verification or just initial copy.

[jasonrandolph]

I thought the "Film vs. Digital" war was over...  

[Schewe]

Mike, thanks for adding back the part of my comment I originally edited out. I am perfectly aware of the analog v. digital irony of digital imaging, but in the end, a traditional photograph is a physical medium, not electrical charges in the ether. Remember that when you back up your files yet again ten years from now on some yet to be invented storage device.

Uh, maybe you don't understand forum posting since you're a bit new in these parts...but when you say something either stupid or outrageous, other posters will take your post at face value (in the time frame you wrote it) and respond. Going back after the fact and editing (whether you note the edit or not) doesn't erase what you wrote nor invalidate somebody else quoting you.

Fact is, I would have responded to the rather ill-informed view you have of what constitutes digital capture but Mike beat me to it. In the end, digital isn't really any more or less fragile than film...you ever have film get wet in a basement flood and dry before you got the chance to take it out of sleeves? Between fixer stains, scratches or wet damage, I've lost more film photography to the elements over the years than I have lost digital.

Your view of what constitutes a photograph is horribly pedestrian and backwards...really doode, wake up and smell the silicon. Photography is, whatever the f$%#&k somebody says it is...but if you can't afford the new toys that doesn't give you free range to claim it ain't "writing with light", ya know? Whatever the heck you think THAT actually means, really, "writing" with light? That's how you translate "photo" & "graph", really?

I would call it a "pictorial device" made with "light". But hey, that's just me...

:~)

[darrylbaird]

...SNIPAGE...

I would call it a "pictorial device" made with "light". But hey, that's just me...

:~)

yeah, I like that, but really, if I was going to have a show in Chicago at Edelman or a "Big-Name" Gallery in NYC, I might be tempted to use his "ethereal abstraction of reality" as a truly obtuse line of verbiage to describe the life and death of light itself... as a metaphor for all photography. [we entomb light as a daily practice]

Hey, I'm still pissed Bill Mitchell (the Chicago art historian) basically called anyone using digital as a potential liar (removal of the witness aspect of a silver-gelatin exposure) compared the the "fact" once revered as "photography." The first photo was more like "hardened" by light, like today's printmaker's asphaltum... but I digress. History will make this argument another triviality in the ongoing (and changing) role of visual capture/imagemaking.