Pages: [1] 2 3   Go Down

Author Topic: Measuring Megabytes  (Read 36206 times)

Ben Rubinstein

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1822
Measuring Megabytes
« on: May 16, 2006, 07:51:10 am »

Didn't realise I'd be the first!

Couple of points from reading the article:

Firstly thanks for doing it, having done many pixel peeping tests I know how hard it is to get everything 'equal' in the studio.

Secondly and I feel this is important. The on screen jpgs are worthless. Totally. Applying the same sharpening to canon cameras with AA filters and different strengths at that (the difference between the AA filters of the 5D and 1Ds mkII is readily apparent) and medium format backs, never mind film, cannot tell you ANYTHING! The only way to evaluate these pictures is by seeing the RAW files and sharpening to your taste for print, period. I'm suprised that the gentlemen put up these jpgs, they show an area where significant differences in resolution are hard to see (not that much really fine detail, I was looking at the hair of the doll to try and see differences) and are only going to cause trouble through the uneducated and fanatics of the internet. Flame Retardent Suits everyone!

Thirdly it's interesting to see that the differences between the original 1Ds and 5D which I've owned and shot tens of thousands of frames on agrees with what I've seen, the 1Ds files are 'chunkier' than the 5D for the same level of resolution. Almost as if there are artifacts in the 1Ds at 100% in comparison. Don't know if this is connected to the noise difference which even at iso 100 is significant, could be.

Fourthly and this came as a suprise to my own eyes and MR's own conclusion. 'And to my eyes images from the 16Mp Canon 1Ds MKII are awfully close to those from drum scanned 645 format Velvia.' I assume you mean the RAW files not the onscreen jpgs. OUCH you are going to get murdered for that given the original 1Ds vs 6X& drum scanned comparison way back then. 1Ds mkII only just equals 645, oh dear.
Logged

michael

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5084
Measuring Megabytes
« Reply #1 on: May 16, 2006, 08:47:44 am »

I said it in the article, Bill said it, and so did Charles.

Get the disk, do your own sharpeneing, vodoo, or whatever, and then draw your own conclusions.

It's only $10.

Inevitably with anything like this someone (everyone) will have a different opinion. Bitching about how we did the jpgs is Monday morning quarterbacking. However we did them someone would have found fault.

Michael
Logged

Julian Love

  • Guest
Measuring Megabytes
« Reply #2 on: May 16, 2006, 08:56:23 am »

Thanks Michael for sharing this test with us. However, I can't help feeling that the subjects used were not really detailed enough to show any meaningful difference between the cameras. Looking at the jpegs I can make out improvements between the 1Ds/5D and the 1DsMII/645 velvia, but after that they are pretty much all the same unless you are a pixel-peeper extrordinaire. A dollar bill at that maginifcation really doesn't pose these sensors much of a challenge. Maybe it is more obvious in the RAWs.

I agree with Pom that stating the 1DsMII is very close to 645 Velvia is quite a retraction from the 1Ds vs 6x7 Velvia test from a few years ago!

Julian
Logged

tsjanik

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 720
Measuring Megabytes
« Reply #3 on: May 16, 2006, 09:37:21 am »

Gee, lighten up people.  Take the free!(or $10 if you want all the details) information for what value it is to you and be thankful someone has gone to the trouble of posting it.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Measuring Megabytes
« Reply #4 on: May 16, 2006, 10:07:32 am »

I'm very skeptical about the idea that subtle differences in resolution cannot be transmitted over the internet. If you crop enough and the jpeg is 100% quality, you should get 99% of the effect.

First, I should thank this merry band of middle-aged photo gurus for taking the time to make a thorough comparison which is believable and credible. It's an interesting comparison and highlights from my, perhaps slapdash, big picture approach, the small differences between the 5D and the Betterlight Super 6K, or the P45.

At a pixel-peeping level, there's no doubt that the 1Ds2 has a very,very slight 'accutance' advantage over the 5D, and so on to the Betterlight which appears to have the detail edge over all the others.

However, the big message for me is that the image improvement from the 1Ds to the Betterlight is relatively trivial in the broad scheme of things.

For anyone primarily interested in creativity, originality, expression and communication, I don't see any big advantage in these super expensive and cumbersome super-cameras.

I think I'll stick with whatever the 35mm format has to offer.

But I could change my mind some time in the future   .
« Last Edit: May 16, 2006, 10:16:06 am by Ray »
Logged

michael

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5084
Measuring Megabytes
« Reply #5 on: May 16, 2006, 10:35:49 am »

Which goes to show the difficulty of such comparisons. Change even the slightest thing and you get differing results.

We entered into this project to answer questions that we ourselves had about this equipment. We set up the tests are rigerously as we could and tried to remove as many variables as possible.

Only afterward did we decide to make the results and the raw file publicly available.

Inevitably people will find fault with our methodolgy, choice of dolls and dollar bills, jpg presentation, and anything and everything else that can be imagined.

Such is life.

Michael
Logged

michael

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5084
Measuring Megabytes
« Reply #6 on: May 16, 2006, 10:37:25 am »

Ray,

If you obtain the disk, and print out some of the real-world files shot by Bill, Charlie and me in the Redwoods, you might think otherwise.

Many thousands of photographers around the world aren't spending their hard earned money on these tools just on the off chance that they may provide slightly better images.

The differences are quite real, and pros make their livelyhoods based on these differences.

Tests like ours only tell a small part of the full story.

Michael
« Last Edit: May 16, 2006, 10:39:21 am by michael »
Logged

eleanorbrown

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 637
    • Eleanor Brown Photography
Measuring Megabytes
« Reply #7 on: May 16, 2006, 10:41:20 am »

Michael, do you still think these digital backs are "lens limited" when using the high end lenses that you all did your testing with?  Thanks, Eleanor
Logged
Eleanor Brown
[url=http://www.eleanorbro

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Measuring Megabytes
« Reply #8 on: May 16, 2006, 10:44:03 am »

Quote
For anyone primarily interested in creativity, originality, expression and communication, I don't see any big advantage in these super expensive and cumbersome super-cameras.
[a href=\"index.php?act=findpost&pid=65655\"][{POST_SNAPBACK}][/a]

There you go.

To be fair though, I feel that we would see significantly more difference if these files were all printed at A1 or A0 size.  

Even then, no need to spend fortunes. The only think really missing from this fascinating comparison is a 3*2 stitched image shot with the 5D with a 200 mm lens... that would have shown with breathaking clarity that a 600 US$ spherical pano head + 3000 US$ DSLR can close the gap very easily with 40.000 US$ gear.  Granted, it doesn't work for all subjects and the productivity won't be as high... but is it really a problem for fine art work?

Cheers,
Bernard
« Last Edit: May 16, 2006, 10:46:32 am by BernardLanguillier »
Logged

michael

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5084
Measuring Megabytes
« Reply #9 on: May 16, 2006, 11:14:33 am »

Eleanor,

It's something that Charlie, Bill and I have discussed, and to my mind it's a quibble. In other words, I think that with the best lenses, we're very close in the 39MP world.

With lower resolution backs and DSLRs I think this is likely the case, except maybe, again, with the very finest lenses and technique applied.

Michael
Logged

madmanchan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2115
    • Web
Measuring Megabytes
« Reply #10 on: May 16, 2006, 11:14:46 am »

I just wanted to offer my thanks to Michael and the others for conducting this test, but especially for making the RAW data available (at a very accessible price).  It ensures not only that people can come to their own conclusions, but also that the test will still be useful going forward as RAW development software improves.

Eric
Logged
Eric Chan

mtomalty

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 541
    • http://www.marktomalty.com
Measuring Megabytes
« Reply #11 on: May 16, 2006, 11:43:00 am »

Michael,

Thanks for making the effort to post your groups efforts regarding these comparisons.
While they certainly won't address many (most?) issues on the subject of 'what is best' it does an
excellent job of giving a visual record of how the different contenders compare.
There are an almost infinite number of variables that will allow one system to extract more
'quality' from a given situation than another on any given day.

Looking forward to getting the DVD and comparing the landscape RAWS to see how the
various camera/lens combos perform when printed to a large size.

Mark
Logged

Ben Rubinstein

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1822
Measuring Megabytes
« Reply #12 on: May 16, 2006, 11:43:17 am »

Michael, I wasn't getting at you at all, just trying to reinforce your point and wondering at your bravery of putting up the jpgs in the first place [/b]given that those are going to be the point of argument over the various forums by as I said, the uneducated and the fanatics.

The point is that from what can be seen so far the jpgs are essentially worthless, not that you may have different opinions based on the RAW files, no opinion should even be allowed to form until the RAW files have been seen!. It was with this point that I wanted to start the thread.

The title of the front page image 'Threatening Storm' is rather apt!  
« Last Edit: May 16, 2006, 11:44:37 am by pom »
Logged

theophilus

  • Full Member
  • ***
  • Offline Offline
  • Posts: 155
Measuring Megabytes
« Reply #13 on: May 16, 2006, 11:44:42 am »

Quote
Even then, no need to spend fortunes. The only think really missing from this fascinating comparison is a 3*2 stitched image shot with the 5D with a 200 mm lens... that would have shown with breathaking clarity that a 600 US$ spherical pano head + 3000 US$ DSLR can close the gap very easily with 40.000 US$ gear.  Granted, it doesn't work for all subjects and the productivity won't be as high... but is it really a problem for fine art work?

[a href=\"index.php?act=findpost&pid=65664\"][{POST_SNAPBACK}][/a]

I've been meaning to bring up this very subject.  How close can you get with a stitched 5D/1Ds2 shot compared to the MF digital backs?

It's certainly not a viable alternative in every circumstance, I am just curious if with the right subject (Horseshoe Bend?) how similar a large print of 30x40" would be.
Logged

David Mantripp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 826
    • :: snowhenge dot net ::
Measuring Megabytes
« Reply #14 on: May 16, 2006, 12:07:53 pm »

Quote
Even then, no need to spend fortunes. ... that would have shown with breathaking clarity that a 600 US$ spherical pano head + 3000 US$ DSLR can close the gap very easily with 40.000 US$ gear.

You don't think even 3600 US$ is a fortune ?  I'm sure many would disagree....

Photography - in the artistic sense - is rapidly losing grip on economic reality.
Logged
--
David Mantripp

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Measuring Megabytes
« Reply #15 on: May 16, 2006, 12:49:33 pm »

Quote
Ray,
If you obtain the disk, and print out some of the real-world files shot by Bill, Charlie and me in the Redwoods, you might think otherwise.
[a href=\"index.php?act=findpost&pid=65662\"][{POST_SNAPBACK}][/a]

Michael,
Okay! I've ordered it. I'll report back to you later.
Logged

Ben Rubinstein

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1822
Measuring Megabytes
« Reply #16 on: May 16, 2006, 12:53:56 pm »

Quote
Photography - in the artistic sense - is rapidly losing grip on economic reality.

Don't think so, it's just becoming more 'up front' as opposed to years of buying and processing film, nevermind scanning. A DSLR and especially a med format system has to be seen as an investment that will prove worthwhile and more profitable than the equivelent film usage, after a year or two not straight away.

Where it falls down is for those who are buying the camera or back then upgrading it before it has paid for itself in film savings. As with any business expense it has to be done sensibly. If you speak to your accountant it may be far more prudent to go digital, I'm sure that once the quality was there, that is the reason that the big boys are changing over, not necessarily because they like digital better though it may be an added bonus.

To put it into perspective, I bought a 2nd hand 1Ds for £2500. It was paid for in film _savings_ within 3 months of a wedding season. That means that for the amount of frames shot during that time, the same amount spent on film or processing would have bought me that 2nd hand 1Ds. And I was charging the clients the same amount of course. Yes I had to sit infront of the computer, but I'm very fast in ACR, I can process 200 pics in an hour and actions do the rest. For all that film would be 'easier' there is no way I can afford to lose that level of profit by going back to film period.

Then I sold the 1Ds for £2200 and bought a new 5D +grip in December for approx (I bought it as a package and sold the 24-105L) for £2050. It has dropped in price quite seriously by now, but I've shot over 20,000 frames with it, weddings, portraiture and commercial. It was more than paid for when I bought it, it is now coining money for me. The trick is to charge the same under the title 'Digital Capture Fee', that is how to pay for your DSLR and how it will make you money fast! The cost difference in capture mediums is your business not the clients instead of paying for the film, now they have to pay for the percentage of the cost of your digital equipment, plus your time. Streamline your workflow and they end up paying for your equipment for you. As it should be.
« Last Edit: May 16, 2006, 12:57:04 pm by pom »
Logged

David Mantripp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 826
    • :: snowhenge dot net ::
Measuring Megabytes
« Reply #17 on: May 16, 2006, 03:58:56 pm »

I see your point, Pom, but it only adds up if you're a commercial photographer with a very high capture rate.  If you're working as what people call a "Fine Art" photographer, then your time to amortize that up front payment is going to be very long indeed.  Ok, people like David Noton to name but one have switched, so I guess the economics do stack up even for a landscape photographer, but for amateurs it is a totally different situation.

Some time back, an amateur could at least aspire to buy pro-level gear, say Fuji or Mamiya 67, without "paying up front" a huge premium for something he or she was ever going to use.  Even a prolific shooter, with a day job, isn't going to burn through $20K's worth of film within the lifetime of the gear (which seems under 1 year, if the turnover rate on PhaseOne backs is anything to go by :-) )

The digital revolution has very much widened the gap between the amateur who is in it simply for pleasure / fulfillment, and the pro who earns a living (or the very rich, of course).  I think this is unfortunate...
Logged
--
David Mantripp

Ben Rubinstein

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1822
Measuring Megabytes
« Reply #18 on: May 16, 2006, 05:47:32 pm »

David, I first decided to go digital, with the D60 then, when I came back from a two week intensive shooting trip (landscape) and having shot 250 rolls of velvia 120. The cost of the film and the processing itself would have paid for that 2nd hand D60, by the time I sold the Mamiya 645 and bevy of lenses I paid for a 10D which had then only just been replaced by the 20D. Not printing larger than 18X12" for sales and exhibition the 10D gave me incredible prints as well as earning my bread and butter shooting weddings/events.

Seriously, these days, even a years worth of shooting 120 film for fine art will easily buy a 5D with pretty much the same quality as well as being far more versatile in that you can shoot most everything else with it too. Then the year after that you're laughing. What I mean is that a body like the 5D or older 1Ds can replace a med format system and a film 35mm system so the saving is not just in the 120 film.

If you stick to primes as most people shoot with med format anyway, even top quality lenses aren't that expensive. If you don't mind manual hyperfocal focusing (as I always did with med format when shooting landscape) then a range of even the great (and here in the UK at give away prices) Contax 35mm lenses is opened up.
Logged

Ben Rubinstein

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1822
Measuring Megabytes
« Reply #19 on: May 16, 2006, 06:55:52 pm »

Woops, just realised that this thread should have gone in the new forum that I hadn't noticed, then again considering it's gone OT anyway....

All in the name of fun!  
Logged
Pages: [1] 2 3   Go Up