Just how much resolution you need?

[Benjamin Marks]

I was wasting some quality time on dpreview.com this morning and using their "studio shot comparison" tool to compare the new Pentax K5II to the M9, the Nikon D3 and the Fuji XP-1. At their base ISOs all of these cameras resolve the cross-hatching in the engravings that are part of the studio scene. For laughs I threw in the D800E as well. I am not focused on the test methodology. Subjectively, I would rank the new Pentax on a par with the Nikon D3, next clearest were the M9 and then the Fuji and D800E. These last seemed about equal in resolution, but the D800E image was larger. Caveats about the quality of on-screen jpgs must apply, because I am not sure that my screen is capable of revealing resolution differences at the size files that DPReview can post on its web site.

My own impression as an owner of a D3, an XP-1 and an M9 is that there is no need for me to upgrade at this time (or perhaps ever) == this is based on an assessment that even the lowest resolving camera among these is capable of providing an on-screen jpg that has gobs of resolution and micro contrast.

So how about it? If you could buy a camera that doubled your current camera's resolution for $3,000 would you do it? Could you use that extra resolving power for anything practical?

FWIW, this morning I am spotting a scan from a Tri-X negative from 1985 taken with a Pentax K1000 and a 100/2.8 lens and developed in Sprint chemistry. I can discern eyebrow hairs in the scan; with the D3 -- now apparently my lowest resolution camera -- I would be able to count those hairs, not just discern them.

Thoughts?

 

[bigeye]

I wouldn't buy on resolution - quality over quantity, at this point.

Here's another angle on resolution by Ming Thein that's interesting.

.

 

[Benjamin Marks]

Interesting take: his point is that we have reached a point of tradeoffs with resolution so high that it imposes limitations on how the camera is used (e.g. 1/2x focal length shutter speed for hand-held use rather than 1/shutter speed as the rule of thumb went in the days of film). Thanks for the link -- it was definitely worth the read.

 

[santino]

the resolution of my 5d is enough (wish it was an m9 - but because of the fact that it's rf)

 

[denizg7]

18 megapixels is fine , sensor size is more important for me

 

[V-12]

A lot of people seem to say they need extra resolution so they can print bigger, and good luck to them if they have clients that need vast prints. But if not it is a law of rapidly diminishing returns because if these very large prints are for vanity home decoration (no reason why not) sooner or later they will need a bigger house....and that should make them wonder how big is big enough.😀

But I think resolution is a reflection of the modern age, when just because it is easy to do something then people just tend to do it. But bigger isn't necessarily better, many fine art landscape photographers seek to connect with their viewers on a personal level by not going much beyond 16x20. Turning a delicate landscape into an advertising hoarding doesn't make it better, the scale alienates more than woo's a viewer. The eye can't take it all in, the brain can't reflect on it.

 

[edge100]

It's been a number of years now since I've made a camera purchase based on resolution. My day-to-day work cameras (X-Pro1, X100, and to a lesser extent 5D Mark II) all resolve a sufficient amount of detail for any print I am likely to make. I'm far more concerned about dynamic range and tonality, these being the characteristics that I prefer in film over digital. The X-Pro1, for example, is the first camera I've owned where the DR and tonality are sufficiently "film-like" (a loaded term, if there ever was one) to make me give up shooting 35mm film.

 

[scottsa]

There is no possible answer without context.

Just from a numbers game, the square problem rarely makes it into the conversation. As mentioned by others as 'diminishing returns', it still does not drive home the concept for most. Resolution is relative to surface area, so to double the dimension, it squares the problem. ;-).

I never heard complaints about 20x30 prints from my D30 10+ years ago, though the subjects were never ultra-fine detail. That was 3Mp, effectively only 1/2 the linear pixel dimension of a 12Mp camera.

I was recently commissioned to shoot a 9'x36' (~3m x 12m) mural... the designer spec'd 300dpi (I wish I could laugh) which is, to put it politely as unrealistic and out of touch as it ever was.

But 18~24Mp is not enough to support the task either, even after factoring in viewing distance, etc.

110, APS, half-frame, Minox all produced negs and trannies that could make some MF/LF shooters hurl at 100m, yet they were very popular.

Modern P&S have more than enough resolution to get the job done and please a substantial portion of the market. The G10/11/12, LX5, G1x, M4/3, NEX, etc. are all producing excellent results to a purpose. Or, as I opened up with, contexts of use.

All that is fine, but the conversation was started about interchangeable-lens cameras, to which I only mentioned a couple. In the right circumstances, a camera like the Canon G1x will out-perform a Panasonic GF2 (personal experience).

That last statement, along with my first and one of the OP's, is at the heart of where I look at much of this; and where a portion of my frustration with current offerings resides.

The D30 performed so well because the "magic stovepipe" (Canon 80-200 f2.8) was a superb lens for its day. Without a good supply-chain, the sensor has no chance.

The G1x smokes the GF2, E-PL1 & 2 for some tasks because it has better electronics and DSP, but the latter with the same sensor and different electronics trounce the G1x (not GX1, who names these things?) because of optics in their sweet-spot. The Canon is superb by comparison in low-light, the others at low ISO pull away when better glass is deployed.

I've ranted about this before, elsewhere, and I'll just say my biggest grumble about the otherwise interesting Fuji offerings is its mount. Just like the darn names, do you think they could figure it out together? As an island, Fuji has a better chance of sinking when someone else does the job better.

Its all about balance. Great glass on an under-performing body is equivalent to a the reverse configuration.

This is where sensor size plays a huge role. Identical optics on 5DmII & 7D, same subject just further away with the APS-sized sensor on the latter. The differences are tangible, but does not diminish the usefulness of the smaller sensor.

To sum it up, the context that anyone puts the systems in is the real test.

From a decision perspective, balancing the above and more with the desperate need of manufacturers to drive consumption of their wares. The greatest difference for the foreseeable future is electronic (sensor, onward).

The one constant that should be, while the rest of the details keep changing, is the optics. There will always be innovation there too, but those should last for a multiple of bodies. APS-sized sensors will be here a long time, but for users, such as the OP looking for his own sweet-spot (presently "looks" like a comparison, just give him time), at an ever-increasing 'quality' level, will eventually find full-frame in their bag.

Fuji lost my business because while they have nice optics, it will be useless when their mount fades. Sony has a similar problem with their NEX, as does Samsung. The investment in worthy lenses should not be as disposable as the balance of the systems.

Personal note: I went with M4/3 as one of the five main "systems" I use, each have their purpose. I can share more lenses i.e. between film and digital with those bodies, thus reducing investment AND stuff to carry. One could argue that the others offer that too, or that M4/3 exists because 4/3 is struggling, which is only partially true, but the balance does support my perspective.

Just a few thoughts, I've run out of time to tidy them up, sorry.

 

[Murchu]

All depends on intended output really, with most common small- moderate print sizes catered quite well by your average, run of the mill digital camera sensors.

One thing I really enjoy though, is looking at the large format scans of older/ historic images on Shorpy.com, where with the resolution and detail captured in the original, it is possible to zoom in and discern interesting detail that is not immediately apparent on first glance at the image. It would be nice to be able to do the same in decades to come, with high resolution digital images. Presuming, the files are still readable decades down the line 🙂

 

[mugent]

It's a simple question, do you have any photos that would have been improved by extra resolution? I don't think I do, maybe so I could print a little bigger, but my lack of talent is a far bigger hindrance than resolution.

 

[daveleo]

At first I read this as a generic question, now I interpret it as how much resolution do * I * need . . . meaning me, personally.

I need 16Mp on an APS sensor (my Nikon D5100), and I am very very happy with that, for printing 8x10's.

At the moment.


Footnote: My two most popular printed images were printed at 12"x18", and both of these came out of my Panasonic LX3 ! The fact that they are not razor razor sharp added to their character.

 

[thegman]

6x6 is plenty resolution for me, if I ever shoot LF, and I hope to, it's for movements and the experience, not resolution.

 

[Spanik]

As my dslr is still on 6Mpixels I don't think it will be difficult to upgrade to the double for 3000... Just wish I could find a 22Mp back for my M645 for that price. But I fall into the category mentioned above:

my lack of talent is a far bigger hindrance than resolution.

But 6x4.5 slides are otherwise fine for me. There is still 6x9 if needed 😀

 

[Michalm]

That is not an issue for film shooters (good glass/film etc ) but if you are digital man you don't want your best picture ever to be limited by average resolution of the sensor , if that best shot is ever gonna happen. I guess i would be happy with sigma foveon full frame / 35mm 24mp sensor or similar.

 

[jsrockit]

I think resolution is one of those things that is nice to have. A great photo will still translate in small sizes. The internet and museums are full of them. The galleries might want more though.

 

[eric rose]

A stunning slightly fuzzy photograph will out sell a razor sharp boring image any day of the week.

I have a Nikon D700 as well as a Panasonic LX-5. Both produce exceptional images. Their mega pixel count is similar, both shot RAW images. The D700 gives me images that can be made into bus wraps or billboards, the LX-5 can't. For web images or small prints up to 8x10 I would suspect the average buyer could not tell the difference between the two cameras. Heck my old Nikon D70s at 6 mega pixels still produces great 11x14's.

Spend more time on execution and less time pondering pixel counts, sensor size or all the other useless things the marketing types have people all twisted over.

 

[daveleo]

That is not an issue for film shooters (good glass/film etc ) but if you are digital man you don't want your best picture ever to be limited by average resolution of the sensor , if that best shot is ever gonna happen. I guess i would be happy with sigma foveon full frame / 35mm 24mp sensor or similar.

It's arguable, of course, but "best picture" ? . . . "limited"? . . . hard terms to define, I think.

Some of the "best pictures" I have seen are technically photographic garbage in terms of resolution.

Can't argue about your Mp/sensor choice. Just commenting on your words.

 

[Michalm]

It's arguable, of course, but "best picture" ? . . . "limited"? . . . hard terms to define, I think.

Some of the "best pictures" I have seen are technically photographic garbage in terms of resolution.

Can't argue about your Mp/sensor choice. Just commenting on your words.

Can't argue with that either , but this topic is about ,,resolution'' only , - ideally we want good negative or raw file to start with if possibile and not a garbage one - i don't think the photographs you have mentioned about belong to Ansel Adams portfolio. Personally i admire Ansel Adams technically perfect photography but prefer Josef Koudelka or Elliot Erwitt aesthetic.

 

[daveleo]

Can't argue with that either , but this topic is about ,,resolution'' only , - ideally we want good negative or raw file to start with if possibile and not a garbage one - i don't think the photographs you have mentioned about belong to Ansel Adams portfolio. Personally i admire Ansel Adams technically perfect photography but prefer Josef Koudelka or Elliot Erwitt aesthetic.

I understand.

Actually, I was not familiar with the work of Josef Koudelka until your posting, and I did a quick search. . . . what can I say? . . . I am in awe.

 

[thegman]

I think resolution is one of those things that is nice to have. A great photo will still translate in small sizes. The internet and museums are full of them. The galleries might want more though.

I'll have to respectfully disagree just a bit there. Obviously many photos work at small sizes, but to have a spectacular landscape at postcard size can be doing it a disservice sometimes I think.