An IQ test to blow your mind

[PaulC]

This should blow your mind, especially if you are from the school that says the old folding cameras are just toys.

I decided to give my 1935 Bessa (the one in my avatar) with its f7.7 Voigtar 105mm lens a real test, at the same time testing my own skills in using it to try to get the best possible output. I took 16 shots (it is 6x9cm, of course) and this is the first one I have scanned, I hope the others are as good.

The scan produced a file of about 120MB, which is roughly 38MP. The first sample is a 100% crop (72dpi) from the 38MP file, which is almost twice the size of the output of the top DSLRs:



This would be exactly what the print would look like if I blew the neg up so that it was nine feet wide and six feet tall (2.8m x 1.7m) and you were looking at it from the same distance that you look at your computer screen.

The next thing was to look at a crop when the file is reduced in size to match a Canon 1Ds MkIII or 5D MkII



At this size, the 72dpi print would be six feet wide. At 300dpi glossy magazine quality you would get an 18x12 inch print - a centrefold spread - and it would be four times sharper than this crop.

So, what does the more modest 5D Mk1 size look like? Here is the crop from the same area of the image reduced to old 5D size, which is more than enough for most uses:



And, finally, here is the full frame so you can see what area of the 6x9 neg I cropped down to in this test.



The shot was a 20 second exposure at f11 on Fuji Sensia 100, scanned with an Epson V500. The red line in the foreground is a car light trail.

There are several points to make: If a 75-year-old 6x9 folder with the lowest spec uncoated lens can produce this quality, then is there really any justification in the inflated prices of folders with famous lenses? How many people who splash out them are getting quality as good as this, let alone better? The lens won't do anything for you if you don't use it carefully (though I guess a coated lens would have improved these a bit).

Secondly .... I still believe in tripods for folders and (with respect to others) I don't believe 1/25s handheld will generate this quality on any 6x9 ... and this is the quality I want.

Thirdly, yes, uncoated lenses do flare but not necessarily catastrophically (unfortunately, I don't have a hood for this camera).

Fourth ... using film it is possible to get images that challenge the quality of a top DSLR for an outlay of only about ten quid.

I hope you found the outcome of this test interesting.

[Attila]

Awesome!! Many thanks for sharing! I hope this post greatly help to keep alive these cameras.

[GrahamNR17]

One of the many reasons I have returned to film - I can afford much better old cameras than new ones

[poilu]

not bad from a 75 year old lens
but modern lens get even better result with 24x36 negative at 200 asa
click here to read more about
40MP from 24x36

[Attila]

and you will see soon for old camera prices what you have it you can buy new ones too

[Nesster]

Thank you for putting up the examples. I haven't ever scanned at that resolution, as I haven't needed it for my usual purposes. But the lesser lenses are indeed capable of excellence, as you show here. Tweaking the infinity point is sometimes required, as production tolerances and age may have caused slight drift.

The main difference between my Bessa 66 Skopar and Bessa 1 coated Vaskar perhaps comes down to how out of focus is rendered, where the Skopar is better. However, these cameras aren't desgined to be narrow dof bokeh monsters.

I agree: especially with the 105mm folders good support is essential.

[no-X]

Very convincing. Thanks for presentation

[LucisPictor]

An impressive result indeed!

But don't forget the film and developing costs.
6x9 is great for some high quality shots but definitely not for everyday shooting.

Each format has its place.

[PaulC]

Your thread is interesting, Poilu, but I don't think it really makes the case for 35mm at 40MP. That crop looks really unsharp to me. I'm inclined to agree with what Alessandro said about limits on resolution.

The voigtar lens is a triplet and although its output would certainly be usable on a wall poster the samples I posted are really a bit soft above 13MP. I think I may have pushed Alessandro's resolution limit for triplets up a little because my 13MP sample really does look OK to me - but I imagine there is a fair amount of variation in these old lenses and I may just have been lucky.

It's worth noting that this is a flatbed scans from a $200 scanner (and I'm not an expert at it) so the IQ could probably be further improved with a better scanner operator and a professional quality scanner.

[PaulC]

[greg]

The images are a true joy to behold.

RE: film development costs.

In our area it seems that each and every roll developed pays for a large part of the lab owners yacht docking fees. I had a single roll of Kodak 120 developed last Fall and was amazed at the +$20USD cost.

That, and locating vilva's website, set the stage for my adapting old lenses to my DSLR's. I realize I am accepting some shortfalls in using digital - crop factors, tradition, and large format results. Being retired determined that I would be unable to finance additional equipment AND pay for film developing, other than an occasional roll of B&W developed in the cellar for old times sake.

[Nesster]

That $20 probably included prints? Develop only is very cheap indeed, and many minilabs will do this. Unfortunately the one I was using in town closed down, and the CVS drug store stopped their send-out service.

[LucisPictor]

You get it developed for a really low price. It's much more expensive where I live.
A roll of good film is about €4,- often more and for the development I have to pay about €10,- without prints!!

Still one could argue that the price of a DSLR with a nice lens would pay for about 100 rolls. But then you need a scanner etc.

But I did not only mean the developing costs, I doubt that the practicability and usability of 6x9 folder would comply with my kind of everyday shooting.

This is not supposed to reduce your results at all. They really are amazing.

For me it's just as I said: Each format has its place...

[nemesis101]

And who amongst us hasn't had a good time in a cellar? lol

Doug

[Orio]

Very interesting test Paul, great read.
Your image speaks for yourself.
However in my mind I already had not doubt that a 60x90 mm negative or slide would beat a 24x36 mm digital. In spite of the undisputed (I think) better signal to noise ratio at low ISO settings of the digital medium, the difference that the size implies just can not be matched in any way by other technical factors.
I think that if a digital back for say a Contax or Pentax 6ox70 mm camera was compared with the Bessa, we would not talk about such obvious results anymore.
In any case, I am inclined to think like Carsten in this issue. Every medium has it's strenghts and weaknesses and it's uses. For very large prints, for either wall posters or high quality books, the medium and large format film cameras still deliver the best.
However, when good image quality has to be coupled with portability and when post processing is an important part of the workflow, the digital negative format of RAW and the digital reflex cameras are the better choice.

[PaulC]

Here are a few more from last night's test. So far, everything I've looked at has been OK, at least close to the quality of the test if not as good as it.

I'm not arguing that everyone should go and get folding cameras - just that they are a damned sight better than most people think. Personally, I find the more contemplative approach that they force on you (expensive film and setting up time) is helping me to take better pictures which has got to be good for my long-term development as a photographer.

And I do like the look I am getting from this old beast.

One definite advantage they have is that nobody thinks you are a serious photographer. In one of the following shots there are two Muslim Arab women sitting on the wall, one of them looking straight at the camera. No way would she have stayed if it had been a 5D with a 24-70 zoom sticking out.

I pay the equivalent of 1.5 euros per film for developing at a little shop/mugshot studio around the corner from where I live and I imported a bulk supply of film from ebay - including a couple of hundred rolls that were a few months out of date.

As for portability - surely nothing beats the portability/image quality ratio of these little folders? That interests me because I do a fair bit of hill walking on holiday and a folder with a gorillapod would be much easier to lug around than a 5D, if only I can satisfy myself that I can get the quality I want out of it on a consistent basis.

[F16SUNSHINE]

Paul
Excellent work on this presentation.
It is really nice the way you compare the sizes down for us.
Thank you

[Farside]

I bet the original owner of that Bessa never thought that fifty years later, somebody would be knocking out good images from it.

[PaulC]

[Farside]

[PaulC]

So where are the shots from yours?

I took the same picture with my 5D Mk2 as the original one I posted here and I've got trouble because some of the blacks are totally filled in, maybe I can rescue them in RAW (I still have to find out) but the jpgs are a bit of a disaster with areas of zero detail. Is film better than digital for night photography?

The 5D Mk2 seems to be prone to this blocking in of solids in a way the Mk 1 isn't (or at least isn't so much). Presumably the tighter packing of the pixels is something to do with it. What is particularly worrying is that I am also seeing it in areas of dark solid colour on shots with ordinary lighting. Maybe RAW conversion is the only option for the Mk2.

Because of the digital blocking problem, I can see myself ending up using a 6x9 folder as my prime camera for night photography, even though removing dust is a real PITA.

[Orio]

Paul,
judging the 5D Mark II on the jpgs it makes is not really fair to the potential of the camera, which deserves to be evaluated on the raw files. The jpgs are there only as a commercial strategy to sell the camera to non expert photographers. If you compare a jpg shot with a raw shot you will notice a big difference.

The shake issue with medium format cameras is the reason why I referred to the 5D as being more portable. You can take still pictures with a 5D also hand held. You can take them with a 6x9 also, but it's a lot more difficult. The 6x9 really needs a tripod.

[F16SUNSHINE]

[PaulC]

Thanks for the info. I have just been experimenting with that file in the raw converter and I can save the shadows by sacrificing the highlights, so now I have to TIFFs, one for shadows and one for highlights, and not much clue of how to merge them into one ... Just have to mess about in PS I suppose.

[greg]