Next Gen Alpha FF DSLR

[hardloaf]

I'd say that if you see difference between your G11 and MF files on screen, numbers say that you should see difference, but prints are close to identical it's time to start questioning your printing workflow, i.e. it's actual resolution, profiles and involved software.

 

[peterv]

Interesting discussion.

Georg, as a serious question: What do you look for in prints?
I understand at A3 there may be enough resolution in a GF1 or G10.
There is a difference between the optical quality of for example the 45-200 kit lens on the G1 and the CZ 135 on the a900.
Of course than there's the different DOF with the various formats, etc.

All these things will help determine which is which. But apart from all that, what reveals the difference in print?

 

[Bill Caulfeild-Browne]

I suspect that the resolution of an inkjet printer is the lowest denominator. Average lenses and superb lenses will be dumbed down by the printer. (Remember 2880 dpi has to be divided by R, G and B!) Small sensors and large sensors will likewise become somewhat indistinguishable in a "regular" sized print.

HOWEVER, once you start printing "big" (say bigger than 24 inches) then the printer resolution becomes less important than the lens/sensor combination. I have printed a900 files of exactly the same scene, same time, same place as p45+ files and the results are clearly distinguishable. And that's just in terms of resolution. There's also gradation and DR to be considered.

And finally - I crop a lot! The P&S just isn't going to permit that, the a900 will allow a fair amount, but the MFDB gives lots of leeway.

But my message is still - if you don't need big prints you don't need big sensors.

Just my 0.02 worth....
Bill

 

[fotografz]

You are right, if you do not print larger than A2 you will definitely be sufficiently happy with 35mm FF DSLRs, even with 4/3 or M4/3.

I am very close to buy a GF1 as complement to my H3D39 in order to have a high quality P&S which will allow me to print at least A3+ in sufficient quality.

Not sure where my A900 will go then???? Maybe no longer needed? As well as my Leica M system, as I am not so impressed with the M9 and can produce many of the images also with a GF1 - at least for my quality expectations.

If your M9 is a black one, PLEASE sell it to me if you decide to sell it.

Thanks,

Marc

 

[fotografz]

I must say that I am grateful for the A900 for my primary application in wedding work. It is a perfect companion to the FF M9 for this type of work. I've never liked cropped frame cameras or small sensor cameras for all the reasons mentioned above ... depth-of-field, dynamic range, tonal gradations, ability to crop at will ... not to mention squinty little viewfinders or none at all.

The difference between the A900 and MFD is not as dramatic, but is still similar. The difference now being that with 25 meg full frame with the Sony, there is less reason to bring a MFD camera to a wedding anymore. I still do in certain circumstances, like when asked to shoot everyone at a wedding in one shot for a large print or something like that. However, that's a rare need anymore.

When traveling I rarely even consider dragging even an A900 kit around with me. The A900 and one lens barely can fit in a bag that can carry most all of my M9 system. I even plop a little Leica P&S in the M bag, to put in my pocket when going to dinner while on vacation.

In the studio, or on location for commercial work, the A900 and Leica stay in the gear closet.

Forgive the following explanations as to why MFD is often the tool of choice that non-commercial photographers may not understand: The MFD provides everything needed for all the diverse client applications often asked of an image. Images for magazine print can range in use from double page spreads in tabloid sized publications all the way down to digest size. This often involves leaving massive amounts of background area for type, and additional area around an image to allow for a multitude of bleed configuration requirements ... so the actual image size can quite often be less than 50% of the actual taking area. The demand here is the highest level of data going into the final image for application that can range from bus side banners, wall posters at point-of-purchase, severe crops and repurposing any given image from billboard sizes all the way down to internet use. In this case it is the "highest common denominator" that is the criteria going in, not the lowest.

This doesn't even take into account that digital backs can be mounted on a view camera to provide access to lenses not even dreamed of in 35mm DSLR work, and perspective control or DOF control with every lens possible.

Many folks on this forum are landscape shooters ... and as Bill says, for larger prints, the MFD rules. Large prints are often the goal of landscape shooters. If I where a landscape shooter I'd own nothing but a MFD and something like an ALPA using HRD view lenses.

In the end the arguments are moot ... and it all comes right back to "Horses for Courses".

-Marc

 

[gogopix]

It is interesting that this Sony forum discusses "S2 as having only 10-20% IQ edge over A900"

In another part of GetDPI, someone said "the X1 images to me look better than the Sony A900!"

Interesting world we have ...

 

[hardloaf]

Bill,

I suspect that the resolution of an inkjet printer is the lowest denominator. Average lenses and superb lenses will be dumbed down by the printer. (Remember 2880 dpi has to be divided by R, G and B!) Small sensors and large sensors will likewise become somewhat indistinguishable in a "regular" sized print.

HOWEVER, once you start printing "big" (say bigger than 24 inches) then the printer resolution becomes less important than the lens/sensor combination. I have printed a900 files of exactly the same scene, same time, same place as p45+ files and the results are clearly distinguishable. And that's just in terms of resolution. There's also gradation and DR to be considered.

And finally - I crop a lot! The P&S just isn't going to permit that, the a900 will allow a fair amount, but the MFDB gives lots of leeway.

But my message is still - if you don't need big prints you don't need big sensors.

Just my 0.02 worth....
Bill

Unfortunately it's not that simple. In MR test we had G10 vs. P45+, i.e. 4416 x 3312 vs. 7216 x 5412 images. His Epson definitely can print at 288ppi (forget about those marketing 2880). Print size is 13x19", so for P45+ 7216/19 = 380ppi, 4416/19=232ppi. In optimal conditions these numbers would mean that G10 shot has to be upsampled and P45+ downsampled. In reality we have to cut numbers about 1.5 times more because image source is a Bayer patterned sensor and the pattern shaves of resolution even more. So we have 380/1.5=253, 232/1.5=155. Now it's clear that P45+ actually should be barely enough to make bleeding 13x19" and G10 is nowhere near that. Now MR mentioned that he didn't run output sharpening, which is a very bad thing to do if you really want to squeeze all resolution out of your printer - ink squashing eats resolution and must be compensated. There are also other factors to consider and all of them affect output resolution. F.e. proper printer profile is important, applying it with minimal loss even more important. Essentially this all means that there is pretty much only one proper printing workflow sequence which will produce best resolution, it's rather complex and not as simple as clicking Print menu in Photoshop after color and contrast corrections.

 

[Hank Graber]

This often involves leaving massive amounts of background area for type, and additional area around an image to allow for a multitude of bleed configuration requirements ... so the actual image size can quite often be less than 50% of the actual taking area.

I always create backgrounds in post. What's the point of wasting massive amounts of resolution on background only to find they wound up retouching more in anyway when they changed the layout. I rarely see an image in a national ad that has not been practically painted over in post no matter how it was captured. If the client wants more background on top or one of the products shifted left, no problem it's all ready in layers and ready for changes.

That's not to say that MF is not the ultimate studio tool but if you don't have a client looking over your shoulder you can shoot a pro DSLR with a top line lens and if you have a handle on capture, lighting and post-production you can produce a 2 page spread competitive with anything produced on any equipment. If you think an ink jet is a big equalizer a web press printing on coated stock can reproduce a grand total of about 1500 colors. A tiny fraction of what you captured.

 

[philip_pj]

This thread would benefit from a link to Michael Reichmann's G10 - MFD comparo, apologies if I missed it above.

Hardloaf, if you revisit, it would be very interesting to hear your thoughts on the comments made by theSuede in the thread linked by Douglas (8.04am, 8.30am, dec 21) re Canon color. I have seen a RGB response showing the 5D and 5DII - the graph of the 5DII looks eerily like his opinion re peak separations - closer together than the 5D, and with a big 'stockmarket' type dip in the leading red 'hillside'. Is there anyone doing this kind of comparison across DLSRs, as it seems to be more illuminating than the color charts used in most reviews.

I agree re print workflow in that I think these semi-facetious articles should be accompanied by enough information about total workflow to enable the interested observer to make a fair judgment on the validity of the test. Not a PhD but a para on tripod, aperture, lenses, all post work, print prep, paper.

I came across a comment by Agorabasta over at DPR recently to the effect that print examination lags screen 'pixel peeping' badly for assessing IQ. Finally, some subjects are not too demanding of system quality, and I do envy those who shoot that material.

 

[fotografz]

I always create backgrounds in post. What's the point of wasting massive amounts of resolution on background only to find they wound up retouching more in anyway when they changed the layout. I rarely see an image in a national ad that has not been practically painted over in post no matter how it was captured. If the client wants more background on top or one of the products shifted left, no problem it's all ready in layers and ready for changes.

That's not to say that MF is not the ultimate studio tool but if you don't have a client looking over your shoulder you can shoot a pro DSLR with a top line lens and if you have a handle on capture, lighting and post-production you can produce a 2 page spread competitive with anything produced on any equipment. If you think an ink jet is a big equalizer a web press printing on coated stock can reproduce a grand total of about 1500 colors. A tiny fraction of what you captured.

This may not be the proper place for this specific discussion ... but it's started, so ....

Hank, I'll give you a perspective from the other end of the process ... since I spent 40 years as an art director then Executive Creative Director for places like Young and Rubicam, as well as the last few years as a partner in my own ad agency.

Yes, there are instances that a good 35mm DSLR can be, and are, used for commercial work ... even work published larger. Depends on the subject and creative objectives going in. Lots of lifestyle subjects and action work is best done with a 35mm DSLR and is actually preferred. In fact, a very good friend of mine in NYC shot a global wide lifestyle campaign for a cell phone company with a P&S.

The last full page, full bleed national ad campaign I created for Unilever was shot by an editorial type "kids" photographer with a Canon 1DsMKIII and 300/2.8L IS ... the intent was more editorial in look and feel ... so pixel peeping resolution wasn't the objective. Since this was an environmental shot with a full bleed background, there was no adding background without incurring huge retouching costs ... an unnecessary line item the Unilever cost consultant and Brand Manager would NOT have been pleased with ... so we shot the key subject loose enough for various print bleed demands. In this specific case, emotion was more important than resolution.

Other subjects and creative objectives demand other tools be used ... and it doesn't matter how much the CMYK printing process dumbs down the result ... despite expert lighting and top talent doing it, if you can't fully capture the chrome on a car, or the specular sparkle of jewelry or faint whiff of steam rising from a baked potato going in, then it'll never be there without spending a fortune in retouching ... which almost always looks a bit fake compared to the real thing. The age of willy-nilly cost over-runs for retouching are long gone, and photographers that force me to the retoucher's studio unnecessarily piss me off :angry:

Put simply, I'm not a fan of "fix it post" photographers, and neither are my VERY cost sensitive clients anymore. You may not have any idea what a cluster "bleep" it is to go back to a client for a cost over-run these days.

As to MFD applications, many of the food shots and images involving fabrics that I bought in the past few years weren't just shot on MFD ... they were multi-shot MFDs! Minimum retouching was required in post thanks to the incredible fidelity of the original captures.

The last job I personally shot myself (with no art director looking over my shoulder since I was the AD ) was a new dyno room for American Axel ... the use was intended for 9' wide print in their R&D building as well as for print applications and their web site. A 28mm on a H3D-II/39 was used, and I stitched 4 shots for the final lobby display print. It was barely enough for their eagle eyed V.P. Marketing Manager.

Horses for courses. If you shoot one basic type of work, the tool can often be just one type of system ... be it a rangefinder, 35mm DSLR or a MFD system. If you do a variety of work it's nice to be prepared for different demands ... (some of which is better to rent if it's an infrequent demand). Then there are those like Guy Mancusio who would rather use a MFD for everything and not worry about it.

As for me ... I can't wait to get my hands on a H4D/60. :thumbs:

-Marc

 

[Hank Graber]

The age of willy-nilly cost over-runs for retouching are long gone, and photographers that force me to the retoucher's studio unnecessarily piss me off :angry:
-Marc

I do all my own post. The retouching is invisible to the client. I choose on a shot what combo will get me to the end result fastest and with best IQ. I won't spend 2 days trying to set up the impossible shot, renting extra lights and equipment (and trying to get the client to pay for it) when I can capture it in an hour and finish it in 3 more on the computer. It's not a question of fixing it in post. It's planning the final image. If the client wanted raw captures like they used to want slides I'd have to own MFD no question to do almost anything. That's like hitting every tack with a sledgehammer because you have no control how the capture will be used or processed downstream.

The one time I came close to pulling the trigger on an MFD system was not because of IQ but workflow requirements. I was quoting on what was essentially a massive catalog job that would have required a large amount of new products up on the fashion manufacturers web site weekly. Didn't need MF for web resolution pictures but in order to get a seamless capture to web workflow with no post involved. Didn't get the gig so didn't buy the equipment.

MF will always be better then 35mm digital as far as capture quality is concerned but that's just part of the equation. I need to capture enough to make some component of the final image. Sometimes that might require MFD and sometimes that might require swings and tilts, etc., Not enough times in my case for me to own that equipment.

A lot of times what photographers 'need' is not driven by dollars and cents but by gear lust. That's OK. That's how I wound up with a Leica M. I remember Guy jumping through some hoops to use the M on everything I'm sure he is having an easier time with his MFD set up. I just got the Alpha 900. Other then the incredible viewfinder (finally a decent 35 DSLR finder) it's about as lustworthy as a copy machine but it's more practical for work.

It's more fun to have a Ferrari then a Toyota and the Ferrari will always be faster and cooler but that doesn't mean it's a business requirement. Depends on your situation and work.

 

[fotografz]

" ... it's about as lustworthy as a copy machine but it's more practical for work."

+1

See, we do agree on some things.

 

[Iliah]

MR's experiments are very useful. They vividly show that in a typical workflow printing is the weakest link.

In order to see the difference between prints one needs to be sure the printer/paper/ink combo outresolves at least one of the cameras. This comes from direct print resolution measurements, not from numbers that come from manuals.

To see which camera performs better in print in regards of tonal and colour gradations the best thing is a double-blind test. but again - printing workflow should be very close to optimal before even thinking of such a comparison.

From my practice a demanding customer can see a difference between a print from an A900 or D3X file and a print from a DB on A4 format, proofs being printed on gloss proofing paper, HP B9180, custom colour profiles. As they say, the giveaway is richness of shadows and fine gradations of the print.

 

[Steen]

This thread would benefit from a link to Michael Reichmann's G10 - MFD comparo, apologies if I missed it above. (...)

Philip, you can see the comparison here.

And you can download the two compared jpg-files here.

The two comparison files are: G10.jpg and H2-P45.jpg

I think it is a somewhat problematic comparison.
I have a hard time finding anything sharp / in focus in the "Hasselblad / Phase One" shot.
Everything looks blurry to me, even the focused part "the large knot in the foreground tree".
Maybe the camera was shaken during this 1 sec. capture. Or stirred

 

[nautilus]

Who said that the next generation FF Alpha camera will be a higher model?
There's a 80:20 chance that the next FF camera will be priced lower than the Alpha 850 at the level of a Canon 7D.
A700 customers should go the FF route and buy expensive FF lenses whereas APS is the solution for entry level DRSR customers.

 

[philip_pj]

Thanks Steen, his site search is hard to use and 'kidding' is not much of a HTML title tag, which is what search engines rely on most.

I had a stare at the two 'tree' files I d/loaded and read his write-up. LR was never going to do much of a demosaicing job for the MFDB file; 3 seconds is not enough for any pod to settle for a 1 second exposure for a shot with a lot of micro detail; there is plenty of motion blur in leaves etc. for the H2 shot. Goodness knows what 'beta version of another raw processor' means. The H2 file is dreadfully unsharp at what is purported to be the focal plane; these are jpgs so color tone subtlety is squashed into 8 bit sRGB form. Not to be too critical, we need an open mind, but why not get the best from each if you are going to subject your fellow pros to a print comparo?

Iliah, I agree - yet so many people believe that up to a certain print size, all output is the same quality. Reichmann's viewing team should have been able to readily differentiate impressions based on DOF from true sharpness. Sounds like you have a much more discerning set of clients...and I really like my 9180 also.

 

[roweraay]

Not to be too critical, we need an open mind, but why not get the best from each if you are going to subject your fellow pros to a print comparo?.

I think MR was merely trying to prove a point, a bit tongue-in-cheek. That under certain conditions, and under a certain print size, even a humble product like the G10 can give even a lofty MF product a run for its money. Just to roil the waters a bit than striving for perfect scientific comparison between these two classes of products which effectively are on 2 entirely different spectrums.

The other side of the coin, which is the superiority of the MF back (over a humble G10) is VERY easy to prove. I doubt he was trying to do that.

 

[Theresa]

Yes, EVFs are probably going to displace SLRs, like mp3s have displaced CDs, with lots of loss of audio quality as EVFs will be inferior to SLRs. Its a shame that the "newest thing" always is more marketable than the superior. I have a KM a2 and a200 and I would hate to give up my a850s superior viewfinder. EVFs will be less expensive to produce than a good pentaprism and mirror. What people will settle for always trumps what is best in the marketing world. Who knows, maybe EVFs will become as sharp and detailed as an optical finder, I just doubt it will anytime soon.

 

[Bill_Green]

In hopes of getting this thread back on track, my two requests for the next gen Sony would be true 16-bit color and ability to shoot 1080p video. I also wouldn't mind of the hand grip protruded a little further from the camera. As is, I can almost use the camera comfortably without any kind of strap, a slightly larger grip would help even more.

The A900 doesn't top the categories in any field, but overall it is a winner for me.

 

[Rand47]

In hopes of getting this thread back on track, my two requests for the next gen Sony would be true 16-bit color and ability to shoot 1080p video. I also wouldn't mind of the hand grip protruded a little further from the camera. As is, I can almost use the camera comfortably without any kind of strap, a slightly larger grip would help even more.

The A900 doesn't top the categories in any field, but overall it is a winner for me.

Please, SONY..... no video. Spend that money on refining what is already great w/ the A900.

I'd like to see the MLU separated from things like bracketing for ISO, exposure, WB, etc. so that the sequential shots in bracketing can be w/ MLU engaged.

Love to see ISO indicated in the viewfinder.

More "good resolution" is always a good idea, too.