Inside this issue
Big Camera Comparison – Comments
Feedback from our epic review
Amateur Photographer who plays with big cameras and film when in between digital photographs.
Having been involved in the testing process I was a little nervous whether our 'work' in the field and studio would stand up to close scrutiny. Previous internet-published tests often seem to have a hidden agenda, (possibly to prove that the tester is right having invested in a particular digital workflow!). As a reader it would be helpful to understand what that agenda is, if any. Tim, no regular reader of this magazine will doubt that you are a fan of film, and I know you were always intending to 'redress the balance' in this mammoth effort but I am confident that you have been as objective as possible. I think this examination has been incredibly thorough and is something of a revelation.
What it proves unequivocally is that technology and marketing cannot conceal the laws of physics. Film 'engineering' is a great deal more mature than digital, and modern films are simply superb (digital is also superb, but we can all agree that a great deal of additional development lies ahead). Relatively low-tech, inexpensive lenses render huge amounts of detail on film. The vast size of the film capture area means that the images are subject to far less enlargement stress (let us not forget that 8x10inch film is nearly 25 times the image area of the IQ180's CCD sensor). The fantastic detail and colour rendering of the negative films was striking. The remarkably usable detail of the IQ180 was, in its own way, equally impressive.
Overall, the pervasive degrading effect of diffraction is an issue on all devices, but because of the greater degree of enlargement required, far more significant on the MF sensors than on the large format film. Subsequently I have tested most of my film lenses with the IQ180, and that proves beyond doubt that even film lenses are diffraction-affected from F/11 and smaller. None of this is new, but the relative importance of it can be nicely glimpsed in this test. As a committed 'fence-sitter' who uses both digital and film these were fascinating results. It is important to remember that just because diffraction is noticeable at 100% on a high resolution digital back does not make smaller apertures unusable for real world print application. In theory there is a clear trade off: depth of field vs fine detail. But the fine detail 'sacrifice' is in practice insignificant for most print scales.
Resolution is an important part of photographic quality, but only one part of the photographic experience. The opportunities and the ideas created by a workflow inevitably influence our approach. The scale and wonder of a large format ground-glass encourage very 'deep and resolved' approaches to composition (but possibly a risk-averse attitude due to the 'running costs'). The low cost and instant feedback of a digital system encourages a more experimental attitude, and gives us many more opportunities to explore timing relationships with fleeting and elusive subjects, such as clouds, light beams, waves etc (but it can also lead to an undisciplined 'I can fix it later' mentality). These apply equally, whether the work being done is personal or professional.
As a matter of interest, one of the opportunities created by digital (and not covered here) is stitching; once we start putting a few files together (for example in a panorama) then they obviously become much larger files and relatively speaking higher resolution 'devices'. Additionally, I was so impressed by the quality of the IQ180, and with its ease of use that I ended up part-exchanging my nearly 4-year old P45+ with Chris Ireland's demo unit, the very one we used in these tests.
As far as the whys and wherefores of which medium to adopt, the arguments advanced by David and Andrew above cover all the ground I would wish to visit. In reality, most will have already decided, and may simply use the results to prove that they were right. Especially if they shoot 10x8inch colour neg! However, in the end all of this is beside the point. For 99% of photographic applications, all of the devices tested will provide excellent quality. As has been said so many times before, the best camera is the one you have with you.
I think it was Niall Benvie who described me as an ‘evolutionary denier’ (that’s someone who denies things, and nothing to do with hosiery) because I’m still emulsionally attached to film. (Enough puns for now…) But reading Tim’s incredibly in-depth and fastidious examination of the differences in quality between digital capture and film gives me, on the one hand, some reason to believe that I have made the ‘right’ choice but, on the other, shows me that in real world situations some of the quality differences aren’t that great.
So if there’s not that much to choose in terms of quality, why don’t I convert? It seems to me that there are two main reasons, one economic and the other based on quality, but not just a narrow definition of quality relating to resolution.
Let’s get the economic argument out of the way first.
It has to be said that looked at dispassionately on purely economic grounds I would have to be mad to want to convert to digital capture when I only make around three hundred images a year on 5x4. I calculate that at around £5 per sheet processed (and that’s Quickload so things will be a lot cheaper when I run out) my film and processing bill is around £4,500 per year. If I wanted to swap to an IQ180 not only would I have to invest in the back but also change the camera body, a number of my lenses and get some new gear, such as the Linhof sliding back. It’s interesting to note that I failed to find a price for the IQ180 during a quick look around the net. My mother always told me to be wary of shops that don’t put prices on their goods… In search of answers, I rang Paula at Linhof and Studio and asked her what I’d need to put on my shopping list and how much this would come to. The basic price for me to convert to digital camera - including Phase 1 back, Techno body and other gubbins - would be £35,283. On top of this I would need to change some of my lenses (a further £3,600) and perhaps upgrade my computer to handle the large file size (let’s say £2,800 for a suitable Mac Pro). So the total bill is in excess of £41,000 – or more than nine year’s worth of film and processing. Even given that I will have scanning costs on top of the film and processing, it’s easy to see that there’s a strong economic disincentive for me to make the move.
But for someone shooting product (who would have spent a far larger annual sum on film than I do) or in an industry that demands digital output such as advertising (where the day rates are also much higher) converting to digital capture is a no-brainer. And we should not overlook the reassurance, that digital brings, the comfort of knowing that the shot is in the bag. When working in a pressured, professional environment this is a huge psychological benefit and one that simply cannot be overestimated.
And so to the issue of quality, my second reason for preferring 5x4.
One look at the maximum print sizes for critical sharpness will confirm that few of us will ever need to approach the limits of resolution that even the Mamiya 7 is capable of. So why would anyone want to shoot on an IQ180 or, as I do, on 5x4 film? Surely in real world situations we simply don’t need the quality that these cameras can achieve? Perhaps a more pertinent question is how relevant a measure of overall quality is this analysis of acuity?
Whilst it’s fairly easy to define the technical results for this aspect of quality there are other indefinite, even ‘unseen’ qualities which using a view camera brings to photographs. For me there are two main reasons for shooting 5x4. The first is that a view camera allows me to manipulate perspective and plane of focus in ways that are simply not achievable on a rigid camera. Though it may not always be readily apparent in the finished photograph, these manipulations are critical to my image making. It’s wonderful to have a large ‘original’ to make a print from but my overriding reason for shooting 5x4 is very definitely not simply one of quality as expressed by the resolution of the picture. But, I hear you say, you can have the view camera experience with a Techno and an IQ180.
It’s true that many of the manipulations that I use are available on other ‘platforms’ (why aren’t they called cameras when a digital back is attached?) but a medium format view camera offers, in my opinion, a diminished experience - especially when one has used a 5x4 for as long as I have. The 6x4.5 ground glass feels cramped and it’s hard to focus accurately in the corners. As the format size diminishes with a view camera so the accuracy with which one needs to use movements such as tilt increases. Yet it’s harder to see if one has applied these ideally. The IQ180 helps out by showing the ‘in focus’ regions on live view. But this only works really well on subjects with reasonable contrast and granularity. In focus, smooth, evenly lit surfaces won’t be identified so well.
The second reason, and perhaps the more compelling one for me, is that I still feel that film gives a better result than digital – both in terms of sharpness and other rendering qualities. It’s interesting looking at the image resolution target images for the Provia 5x4 and the IQ180. There are some very weird artefacts appearing for the latter, presumably because the algorithms for calculating what happens with a curved line are failing to cope very well. This rendering problem is also apparent in the image of the Nikon lens. On the 5x4 Provia it is possible to read the figures on the barrel but these become meaningless blobs on the IQ. Now it can be argued that this doesn’t matter in real world situations as this is an extreme enlargement and having readable text in an image is not of huge importance to landscape photographers. This is a little like the CD versus LP debate in hi-fi circles. The CD manufacturers claim that the audio information that is lacking in comparison to an analogue recording is of no importance, as most people can’t hear the difference. I’m just not sure that this argument really holds up under close scrutiny.
Of more significance to me is the colour rendering of film, the feel for want of a better term. Each film has a particular feel as each responds in a prescribed way to the differing wavelengths of light. I really like working with the fixed palette of a film and treat what might be considered a limitation as a positive advantage. I am convinced that after a while one begins to see like one’s chosen emulsion and that this helps with the creative process. With digital there are no such fixed points. Whilst this might seem to offer boundless opportunities I feel that it can actually limit people’s creativity as they may become paralysed by the choice. Colour is a very personal matter, with each individual seeing things slightly differently – or even, in the case of some well known artists, drastically differently – but I believe that 5x4 transparency’s colour rendition is both more appealing and natural than digital. I know that a film like Velvia can hardly be described as neutral but the superb way that it handles fine colour detail in my opinion far outstrips the digital equivalent.
So to summarise… Perhaps the economic question is a red herring and a more interesting question to ask is would I convert to digital if there were no effective cost barrier? Like many others I wistfully imagine the day when someone from Camelot phones to say, “It’s you!” but, come the day, I can’t quite envisage me spending a large chunk of dosh on a digital back. Having examined my motives for this view quite carefully I can honestly say that I don’t hold it because of any Luddite tendencies I may have, nor because I’m prone to parsimony. The real motives for my wanting to stay in the photographic Iron Age are to do with those qualities that I have mentioned. They may be hard to quantify but, nevertheless, I feel they are essential to my image making.
There are going to be some people who will see this extensive test as fuelling the fire of a ‘digital vs film’ debate. This is unfortunate as both have different strengths and for me it simply comes down to using the one best suited for a given situation.
I think the test does debunk some of the misinformation that is peddled about the resolving power of digital compared to film. Over the years I have grown tired of reading that each new generation digital medium format back betters 5x4 sheet film; some people have been claiming that since the days of 22mp backs. I can even remember reading many years ago that a 6mp Canon was supposed to be ‘better’ than medium format film.
If the goal is maximum resolution then at last we have the opportunity of an unbiased comparison of the best possible performers from each camp - drum scanned 10x8 film and an 80mp IQ180 digital back costing around £28,000. We can also see how the smaller formats scale up.
To my eyes, and my own experience, the overall ‘winner’ today is 10x8 Portra, both in resolution and dynamic range.
But looking forward, and being realistic, there is unlikely to any significant further development of film emulsions and we won’t be seeing any ‘new’ drum scanners or new film scanners of any quality. The next ‘phase’ (excuse the pun) of medium format digital backs will probably beat the best that film can realistically achieve in resolution but I think it is going to take a few more generations in sensor development before they can better the dynamic range of negative film. And, they are not exactly going to be cheap when they do!
There is an argument that says the best camera is the one that you have with you and the one that allows you to get the shot you want in the circumstances.
The following applies to me and how I work today:
My personal landscape work
For some of my landscape work I struggle to achieve with digital (regardless of price) what I can with film. That is to produce an exhibition print the size I want to be able to make with my style of working.
My commercial work
I cannot match the versatility of digital compared to film in today’s marketplace. To put it simply, I couldn’t work with the limitations of film in a commercial environment.
I have been shooting colour negative film for 25 years, moving from hand printing to scanning my negatives when the Nikon LS8000 scanner became available. I found that by scanning my negs I could make exhibition prints which were far superior to anything I could achieve in the darkroom. I wrote an article for the late Chris Dickie’s AG magazine in 2004 about this.
With my personal work the only deadlines I have are the ones I set myself. There is no client demanding pictures the next day so I can wait for films to be processed and the one hour plus it takes me to scan a frame of 6x6 at 4000ppi. This file equates to a file size of around 430mb and can produce a print size of just around 29 inches square at 300ppi. Although I have produced prints around 36 inches square I generally downsize that file and make 20x20 inch exhibition prints. What is sobering is that I can achieve this with cameras that can be bought for a few hundred pounds today.
For a while I had a Hasselblad H4D 50. This was capable of achieving very impressive resolution and would produce a print 27x20 inches at its native file size, which, if shot under optimal conditions would appear to have more detail than film. I found I could obtain stunning results by focus stacking when I was working with non-moving images. However compared to film there was a greater deterioration in image quality as I stopped the lens down to its smallest apertures.
The biggest problem I had was the necessity to use ND grad filters under the conditions I often choose to shoot under, which I found very frustrating. The other aspect which I found affected my picture making was the value of the equipment! On a couple of occasions I found myself reluctant to get the camera out of the bag due to extreme weather conditions, £20k + doesn’t buy you weather sealing!
There are, it must be said, frustrations with working with film. I have had a film ruined this year by a lab and the Nikon film scanner I use has minimal depth of field even with a dedicated glass carrier. If there is any curvature in the film the sharpness can drop significantly and this can be visible in 20x20 inch prints. This is not an issue with drum scanners but I don’t think I would be prepared to go through wet mounting every piece of film I wanted to scan and print.
If I was shooting on transparency I would have abandoned film years ago. The reason I still shoot film is that at the moment there is no medium format digital back currently available that can match the dynamic range of negative stock.
Where I think digital has leapt ahead is in the smaller formats and this is down to the use of CMOS chips with the associated high ISO performance and other benefits such as Live View.
I shot two jobs the other week that I couldn’t have shot successfully with film. The first was some ‘editorial style’ portraits in a hospital. I wasn’t sure what the lighting was going to be like and the location prohibited lugging studio lights around. I ended up balancing daylight, artificial light and two wireless iTTL flashguns with my Nikon D3. I was able to adjust the ISO up and down as required as the light changed.
The other job was, I thought, going to be a three hour shoot. The client wanted me there all day. They ran me ragged shooting a mix of interiors, exteriors and even some floodlit sport. I couldn’t have packed enough film of differing ISOs to cover all that was required, or even enough film. And, at the end of the day the client asked for the best high res edits by the next afternoon!
Shooting film is like a drive in the countryside in a classic two seater sports car (that may break down on occasion) but has ‘soul’. Digital is like the Audi that you can rely on to get you from A to B.
In a ideal world you would want both but the ‘Audi’ would win if I could only pick one.
Hans Strand (from an email conversation)
Interesting results. However I would have liked to see the test being made on a typical shooting aperture. With 8x10" you rarely use f/22, since the depth of field is too shallow at that aperture. At least for landscape photography. A more relevant f stop would have been f/64. The same with 4x5" a more relevant f stop would have been f/32 and for IQ 180 it would have been f/16. You would never get this high resolution at f64 on an 8x10" as you got at f/22. Still I am surprised that the digital medium format was so mediocre. My own experience is that I am getting better results than from 4x5". Not that I have made tests like this one , but I can compare scanned transparencies with the digital files from my Hasselblad H3DII-50. When I compare large prints there is a significant difference in resolution in favour of the Digital medium format.
F-stop also have to do with shooting style. If you use big foregrounds, you need to stop down your 4x5" to even f/45 and the 8x10" certainly to f64 if you want everything to be inside the depth of field. If you have a large object in the foreground which is sticking up from the ideal tilt plane e.g. a rock a plant etc. Also when I shoot in a forest and have a tree close to me. That happens all the time for me:-) Today I solve the problem with focus stacking at f/11 and two or even three focusings. At f64 you will just have a fraction of the resolution you get at f22. This can be discussed in a special thread :-) However I am surprised, since my own experiences, from using both 4x5" for more than 10 years and 8x10" for 5 years, is that I am getting better files from my medium format and I would never go back to film again. I fully respect your test though.
My experience is that at f/16 the diffraction is still at an acceptable level whereas at f/64 there is no fine details left. Still I would sacrifice resolution for depth of field. A landscape image with patchy sharpness does not look good to me. With 8x10" you are extremely limited to what you are able to shoot. Depth of field is a big issue. Look at the images by 8x10" photographers like Christopher Burkett and you will see mostly flat surfaces like ponds of waterlilies or field of flowers or he shoots across a valley from one side to the other. This because he is avoiding depth of field problems. This problem was limiting me in the way I wanted to shoot my images. Now with focus stacking and better depth of field I can shoot more or less without any limitations. You also have the aspect of shutter speed. With 8x10" I was always getting several seconds and even minutes. Shooting in a forest without getting blurred branches was almost impossible. Then of course also the slowness of the operation of the camera which made it very difficult to catch the moment. That is another topic I know, but still it makes sense.
Baxter used to use Ebony 45SU and a combination of Velvia or Acros and bought a P45+ back and Phase One 645DF a couple of years ago - ed
The test has clearly involved a considerable amount of work and commitment to bring it to the stage whereby results can be shown and compared.
It appears to me that the test was very much driven by the phenomenal capabilities of 8x10 with B&W film coupled with the 4000ppi scan. This is really pushing the envelope and whilst academically interesting, think that the practical needs and applications for such resolution are in the tiny minority, if they exist at all.
Because of the extreme magnification, I was surprised to see how Grainy/Noisy the film files were also how badly the Alpa/IQ180 fared. This latter combination has pixel-peepers world-wide in raptures with the incredible detail obtainable, likewise the IQ180 with lesser camera/lens combinations too. Jack Flesher’s B&W Passing Storm, Yosemite image impressed me greatly.
Post 208 shows 100% crop of trees on a ridgeline with the full image below
Image quality is not all about sharpness and noise though, tonality also comes into the appearance and perceived quality of the look of a print. Deciding how this could be measured leads to another quandary!
At my level of resolution and experience, I think that the P45+ back with the Phase One 645DF and lenses produces images which are the equivalent of my Ebony 45SU with Velvia or Acros using lenses such as the Schneider 110XL. Either system is more than good enough for the vast majority of situations. It was only my very best 5x4 film images destined for big prints that I scanned at the 2040ppi maximum resolution of my Imacon scanner. The vast majority were scanned at half resolution and they make 40x50cm prints at 240ppi which are not criticised for poor image quality. The same cannot be guaranteed regarding the picture’s content!
Comments on swapping from Film to Digital.
I’ve been using a Phase One 645DF with P45+ back for the last 18 months and prior to this had predominantly used an Ebony 45SU with a variety of film emulsions. There were a number of drivers which made me spend a considerable amount of money in order to sell up my film equipment and buy an MF digital camera system.
There is no doubt that my approach to photography has changed in these 18 months, some of which is by necessity for the new system, some by flexibility, trying new things and getting immediate/short term feedback and I have made more images than when using my view camera. I am very aware that quantity isn’t everything! Waves have featured strongly in my subject matter and they are tricky customers necessitating shooting many images which would be very wasteful of sheet film.
For some situations, I do miss not having camera movements to control both perspective and plane of focus. Should I wish to make another appreciable investment, then either a Linhof Techno or an Alpa STC, both with Digeron lenses would enable me to increase sharpness and have movements albeit the two cameras implement them slightly differently.
Loading Darkslides, emptying them, film developing, drying and scanning film and subsequent post-processing account for far more time than I expend when adjusting a very demanding image from my P45+ in Capture One.
In summary, I do not regret making the switch to MF digital from 5x4 film, which is what I expected, having undertaken considerable research before the making the large expenditure in order to effect the change. Equally, I am sure that there are many photographers who would weigh up their own needs and decide that a film camera provides their best solution.
As someone who spends a lot of time considering how to get the best out of the different mediums I was very interested to see the tests that have been done. Myself, I have concluded that my D3X resolves about the same as a 1200 dpi scan of a 5x4 slide on my V700 but the output is very different. I am not surprised that film continues to impress especially in the larger formats. I believe that all the formats discussed are capable of resolving enough detail for commercial or ‘gallery quality’ work up to 30x20inches so we should concentrate on the look and feel of the image and its intended purpose. We are way past the point where any of the cameras tested cannot be said to produce good enough quality.
For me the key difference is that film is a 3 dimensional medium and I believe that this comes through in the lowest resolution scans in the smallest sizes files, the output has a depth. Digital is a 2 dimensional medium and however good the resolving power of the chip the images always look flat to my eye.
There are too many compromises fighting each other in the digital world for it to be an easy solution to get the quality you get easily from film and these are important in the real world. Diffusion on 35mm digital kicks in at F8/F11 with the D3X and lenses have to be the best in Nikon’s draw to get anything like the output the sensor is capable of. Use of T/S or PCE lenses only gets you so far and you still get issues with DOF. With wideangle lenses or movement on medium format digital sensors then you get colour shifts (magenta mainly) and fall-off issues in the corners so have to add extra steps in the workflow or use reference shots to realise the quality available technically. DOF challenges are even worse than 35mm and movements introduce more problems. For a few of the landscape photographers whose work I particularly admire these would be significant issues where extreme movement is used to great effect.
I am with Andrew on this one, for commercial work where the shot has to be nailed on the occasion then digital is king but for ‘slow time’ work in the landscape arena then if you want the best at most reasonable cost it has to be film and preferably on 5x4 for cost/benefit & portability.
Good to chat with you just now, and thanks for asking me for my involvement.
As I mentioned on the phone, I'm impressed with the testing you've done, and admit surprise at the amount of difference between the IQ180 and the 8x10 results. Here are my comments.
My own background is landscape photography and mainly digital. I have about a year's worth of experience with 4x5 film (using Velvia 50, mainly). I started as a digital photographer and moved into film when my resolution needs were not achievable with the digital equipment I could afford. I will confess that while I enjoy working with medium and large format film cameras, I do not enjoy working with film as a medium. You could say I was spoiled by always being able to review and confirm images in the field, and scanning is a process that I abhor.
For my work, as I don't shoot large volumes in a year, 8x10 would possibly be more cost-effective for me, but due to the point I outline above I much prefer working with the IQ180 on my Arca-Swiss tech camera. There is also the issue of real world practicalities - 8x10 cameras are big and heavy, difficult to transport and are incredibly vulnerable to windy conditions. The IQ180 based system is much more compact, easier to travel with and can be shielded from wind much more effectively. As a result, in real field conditions, it's easier to achieve technical perfection than it would be with the 8x10.
I'll admit to some disappointment to seeing that 8x10 film still trumps digital for resolution at this point, but in honesty for my purposes (and the purposes of any non-scientific photography) the differences (while obvious) are probably minimal in the real world. As you mentioned in the test, we're looking at tiny, tiny areas of these images and pixel peeping to a huge degree.
Did you look at large format prints of these images when testing? I'd be interested to see what sort of subjective results you may have gotten if that had been done? I know from my own work that the IQ180 prints beautifully up to 30x45" - I don't make larger prints than that in general. I'd be very curious to compare an 8x10 image against the 180 in prints of that size.
Thank you for the invite to comment on your very extensive report, very impressive Tim, I really have very little to add, it is very comprehensive and the reasons I went down to Phase One back rout are not technical or due to an extensive study. I had a hotel project in London that ran for 28 months that needed weekly progress reports documented with photographs and the the maths simply added up, it was much cheaper to by a Contax fit back P25 (we had a film based 645 Contax and lens) the immediacy of the output, quality and money saved was enough to upgrade to date to an IQ180. The fact that it easily fitted a view camera, which as you know is my preferred system for personal work clinched it. I still have my 5x4 system which I still use but a lot less since the P45+ came to be, it started to become apparent that the quality was approaching 5X4 and now with the IQ180 I think the images are a bit better that a 5X4 Imacon scan.