Ad
Ad
Ad
Pages: « 1 2 3 [4] 5 »   Bottom of Page
Print
Author Topic: Sony DSC-F828  (Read 23235 times)
Scott_H
Sr. Member
****
Offline Offline

Posts: 331


« Reply #60 on: January 02, 2004, 05:31:21 AM »
ReplyReply

When I start thinking about a dslr system, I have trouble getting past the fact that I would need two Canon lenses to cover the focal length of the 14-54 mm Zuiko lens.  Even putting aside the cost, the extra weight and effort of changing lenses is something I would have a problem with.
Logged

Ray
Guest
« Reply #61 on: January 03, 2004, 08:30:33 AM »
ReplyReply

Quote
All I want is a 6-8mp full frame camera for wide/normal and I can keep a backup 1.6x crop factor camera for longer reach. No need for huge megapixel inflation; judging from the 1Ds I'd have to shoot all primes to get the most out of an 11mp full frame sensor anyways.
DaShiv,
You're being too modest.  More pixels are always better, provided you (they) can overcome the inherent disadvantage of more noise and less dynamic range. The remarkable thing about the Sony F828 is that they've crammed 8MP into an area about 1/4 the size of the Olympus 4/3rds' sensor, yet still have acceptable noise levels at 64 and 100 ISO. Not sure about the DR though. That's often a spec that gets overlooked. However, judging by some sample pics I've downloaded, DR doesn't look too bad.

One can only assume that the people responsible for the Sony technology could do a better job in terms of noise and/or dynamic range if the pixels were to be 4x the size.

The 6MP 10D sensor, according to test results at Norman Koren's site, can deliver nothing meaningful beyond 54 lp/mm, and the MTF at that peak resolution is only about 10%.

You don't have to go to Zuiko lenses to get better performance than that in a lens, but it seems you do need more pixels. However, there's no point in providing more pixels if the extra resolution is going to have a contrast of less than 10% MTF. And that, I guess, is the technological challenge.

It's interesting to note, although the pixel density of the E-1 is about the same as the 10D (smaller number of pixels, but smaller sensor), the 10D is less noisy (very significantly so at ISO 800 and above) and has marginally better resolution than the E-1 - according to dpreview. This seems a bit odd, considering the Zuiko lenses are so much better than the Canon lenses Huh
Logged
Matthew Cromer
Guest
« Reply #62 on: January 05, 2004, 04:08:46 PM »
ReplyReply

Jonathan,

Thanks for sharing your pictures from the 35-350.
Logged
Fabio Riccardi
Guest
« Reply #63 on: January 06, 2004, 12:41:50 AM »
ReplyReply

Quote
In my testing the Canon cameras are close to a full stop more sensitive than other cameras in their metering / ISO ratings, but not 2 stops.
If you check dpreview's D300 tests against the G5 you can see that at equal ISO the G5 is 2/3 of a stop more sensitive.

In my tests my the 10D is at least one stop faster than the F828.

This adds up to some 1.7-2.0 EV of difference in sensitivity between the F828 and the G5.

Finally, in dpreview's test of Nikon's D100 you can see that the D100 is exactly as sensitive as a Canon D60, so Canon cameras are not "in general more sensitive", only their digicams.

 - Fabio
Logged
BJL
Sr. Member
****
Offline Offline

Posts: 5137


« Reply #64 on: January 08, 2004, 12:29:16 PM »
ReplyReply

Some comments about the effective system speed of the Sony 828, and comparisons to the only comparable priced DSLR option, the Canon D300 with 18-55 f/3.5-5.6 kit lens.

   Even if the 828 is limited to ISO 100 (probably 200 at least works reasonably well), its lens is about two stops faster than the Canon kit lens or typical amateur level zooms. So in this realm, one might be able to achieve the same shutter speeds as working with 35mm film or the 300D at about four times the ISO: 400 or even 800. that sounds fast enough to go well beyond "either bright light or a tripod".
   Also, due to the far smaller sensor size and hence smaller true focal lengths, working at apertures up to two stops faster on the 828 still gives one the same or greater depth of field, when adequate DOF is a consideration.

   The SLR's start to win on speed mainly when one spends more money on faster lenses for them, and the resulting lower depth of field is not a problem.
Logged
Scott_H
Sr. Member
****
Offline Offline

Posts: 331


« Reply #65 on: December 27, 2003, 11:31:11 AM »
ReplyReply

Since you're nitpicking Ray...

One mm is about 0.040 inches, so a micron is about 4x10-5 inches
Logged

Dixon Zalit
Guest
« Reply #66 on: December 28, 2003, 01:23:07 AM »
ReplyReply

First, if Michael says the Sony 828 is great, I believe him. But the samples don't rate the comparisons to the 10d

To my eyes, the 828 seems to use some heavy duty sharpening compared to the 10D. The halos are huge in the 100% samples. The appearance of grain may also be just from the sharpening.

Still, 100% is not the same as a print. A print looks more like the 50% view, so the prints probably look great and that is what counts.
Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8887


« Reply #67 on: December 29, 2003, 04:22:54 PM »
ReplyReply

Quote
On the Sony (with the very clear exception of ISO 800) the noise takes on the appearance of "grain", similar to that of film. Using this as an analogy, on the Sony at ISO 64 it's almost invisible, maybe like Panatomic-X. At ISO 100 it looks like Plus-X. At ISO 400 it's like Tri-X developed in Rodinal.
Michael,
Thanks for that clarification. (For a moment you had me worried  Cheesy )

I guess all those film diehards who have resisted going digital on the grounds that digital images are too clean and plasticky will now have no excuse. The F828 could be ideal for them.  Smiley
Logged
MatthewCromer
Sr. Member
****
Offline Offline

Posts: 411


« Reply #68 on: December 28, 2003, 09:58:47 PM »
ReplyReply

Ray,

I don't like 10D or 300D landscape images at 12x18 inches.  To my eye, they look a bit "plasticky" or "over-enlarged".  Compared to 828 prints at 12x16 (an aspect ratio I prefer) I found the 828 images better due to finer details and less ability to see the constituent pixels.  Showed the prints to multiple others in the office who (blindly) preferred the 828 over 300D and 10D images for the same reason.

I can't imagine a 2"x3" landscape image from the 10D would be at all convincing in terms of detail and minimum feature size.
Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8887


« Reply #69 on: December 30, 2003, 06:04:58 PM »
ReplyReply

Quote
I for one, hope the 4/3 system doesn't turn out to be the "Beta Max" of digital photography, ergo the dead end that you predict.
I think Olympus and the 4/3rds system will be secure for as long as full frame (35mm) sensors are expensive to make. At present Olympus can successfully compete with the Canon APS size DSLRs which are heavier and offer little (and possibly no) quality advantage for the extra weight (excluding Canon's really big lenses, that is).

It's difficult to imagine that a camera of the quality of the 1Ds could ever be as affordable as the current 300D. Yet this is basically computer technology and past experience would suggest that sooner or later a camera that actually exceeds the image quality of the 1Ds will be available at less than $1000 (although build quality and other features might be lacking).

If and when this happens, and considering the further development of cameras along the lines of the Sony F828, Olympus could find itself in a big squeeze with nowhere to go.
Logged
David Mantripp
Sr. Member
****
Offline Offline

Posts: 696


WWW
« Reply #70 on: January 02, 2004, 06:26:50 AM »
ReplyReply

the only weakness I see with the E-1 (the E-1, not 4/3) is the relatively low resolution.  I think that the 4/3 system will find enough customers to be profitable. A system does not have to dominate the market in order to be cost effective (if that was the case we'd have 1 choice: Canon).  Only digital has killed off "niche" markets like the fixed lens Fuji 6x7s (for example).  To be honest, I think that an E-1 Mk2 with a stabilised 8Mp sensor (a la Dimage A1) would be close to perfect as a travel / wildlife system, and good enough for most other things.  The rumours of what Minolta are planning are also distracting, but as Michael is fond of pointing out (with a note of exasperation I imagine) you can wait & speculate or you can get on with photography.

Finally, the VJ review of the E-1 seems a little warmer than the web site review. I think that "recommended" is fairly unambiguous...
Logged

--
David Mantripp
http://www.snowhenge.net
DaShiv
Newbie
*
Offline Offline

Posts: 17


« Reply #71 on: January 02, 2004, 09:03:32 PM »
ReplyReply

Here's my take on full frame: given that larger sensors tend to be cleaner, can you imagine, say, a full frame camera limited to 6MP but with an ISO 400-level clean performance at 800, 1600, or (miracle of miracles) 3200? I don't think this scenario would be terribly likely, but I also do think companies are taking note at how the F828 is being taking some heat in certain quarters for the too-many-megapixels issue (although I'm sure it'll still sell like hotcakes). I can't imagine them trying to cram 20 or 30 megapixels into a 35mm format sensor when the 1Ds's sensor is already outperforming many (if not most) lenses, and if they screw up trying to pull something like that then DSLR owners *will* change their buying habits. I have some hope that Canon can use their full-frame 35mm sensor size to produce better performance (and not simply more megapixels) over smaller APS-sized cameras. They're the current market leader in the DSLR segment, the ball's in their court and I really hope they use the 35mm legacy format well and don't squander their current advantage.

All I want is a 6-8mp full frame camera for wide/normal and I can keep a backup 1.6x crop factor camera for longer reach. No need for huge megapixel inflation; judging from the 1Ds I'd have to shoot all primes to get the most out of an 11mp full frame sensor anyways.
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #72 on: January 05, 2004, 03:43:34 PM »
ReplyReply

Quote
If that 35-350 lens isn't a coke bottle, I'd be very surprised. I'm sure it's slow as molassas too.
Umm, it's the main lens on my 1Ds. Here are som hi-res sample shots:
Self-portrait
Rock Wall

The self-portrait was shot at 100mm, the rock wall at 35mm. Not bad for a "coke bottle".
Aperture: f/3.5-5.6
USM focusing with full-time manual focus ring, AF is very fast.
Some softness is visible in the corners of the rock wall shot; but remember the 1Ds is full-frame, and the soft areas would be off-sensor on a 10D (1.6x) or even a 1D (1.3x). The softness is primarily on the wide end; form 70mm on out, overall performance is quite good, at least with my copy.
Logged

BJL
Sr. Member
****
Offline Offline

Posts: 5137


« Reply #73 on: January 06, 2004, 11:26:05 AM »
ReplyReply

Quote
I'll just try to describe where I'm coming from since we seem to be coming from different directions, so to speak.
...
I've listed below the issues that are important to me in the form of 6 guidelines which I hope might be helpful to others.
Ray,

  I like your idea of setting personal selection criteria for a camera or photographic system, and also like many of your criteria, disagreeing mainly with elements that imply making a very long term or even life-time choice. So here is my slightly more "short-sighted" version. Order is vaguely indicative of importance, and realistically, item 1 takes absolute priority!

(1) Cost, based on an annual expenditure/depreciation allowance This includes the losses involved in replacing existing equipment if the choice is a new, incompatable system.
(Note: I expect current DSLR's to have life spans only comparable to laptop computers, after which either some of the fancy electronics will probably wear out or I will not be able to resist upgrading; thus I figure 20-25% per annum depreciation; film bodies and good lenses typically have far lower depreciation rates.)

(2) System image quality achievable, again within the scope of differences that are visibly significant with the print sizes and viewing conditions that I am likely to use in the short term, or have some significant probability of using in the longer term.
(Note: In the low probability of a radical and unforseen change in my photographic aspirations, I would do what I have done several times before; change systems: I will not significantly hamper either my current photography or my bank balance for the sake of hypothetical compatibility with very unlikely future needs.)

(3) Range and quality of lenses and accessories, within the scope of what I am likely to afford and use in the short term, or have some significant probability of buying in the longer term.
(See note above!)

(4) Weight, bulk, general handling, build quality and other features of a more professional nature.

(5) Company reputation for supporting the needs of my style of photography.


  Assuming only that you are not greatly bothered by the lack of a standard zoom lens for Canon's 1.6x "APS" format that comes anywhere close in zoom range to their 28-105, 28-135 or 24-85 for 35mm format, I expect that my criteria will easily vindicate your choice to stay with Canon. For my purposes as a happy Canon film SLR owner, they are just two or three lenses away from a very good ranking; but for now, the 4/3 system comes out on top, with the Nikon mount system (including Fuji and Kodak body offerings) in second place, due partly to Nikon's pricy but impressive DX lens program. Current fixed lens cameras, even quite impressive ones like the Sony 828, and even assuming a few supplementary lenses, probably fall well short for both of us.
Logged
Karel Geertsema
Guest
« Reply #74 on: January 08, 2004, 06:43:20 AM »
ReplyReply

Interesting replies:

Bobtrips wrote:
"Comparing a small sensor F828/A1 type camera with a dSLR, even the 300D is a compact pickup vs. full-sized, V8, 4x4 monster.  They're just not operating in the same categories."
  
Although I have never driven any pickup or 4x4 ... I dare say that actually I am not interested in comparing sensor sizes. I am trying to compare image quality within the limit of my budget while keeping an eye on the "fun to use factor". Michael's admiration for the F828 falls in this last category and I can imagine that the F828 or an A1 for that matter would be fun to use.
So my quest is "if I spend around 1000 $/Euro; what whould suit me best". I know I can buy a 4x5 camera for that amount of money that offers incomparable image quality, but that would not suit my style of photography. If I were to buy something like an F828, I don't want to be limited to bright days or a tripod because I still value image quality (as in limited noise). When shooting landscapes I am prepared however to use a tripod, but I expect more "flexibility" from something like an F828 and that is what I meant when talking about "intended use": an F828 is simply an entirely different thing than e.g. a Pentax 67 or other "tripod bound" jewels.(ouch, I already feel the flames :-)  )

MatthewCromer wrote:
"You can clean up a small amount of noise in an 828 image or some CA, but you cannot fix the soft corners of the Rebel kit lens at wide angle, or reduce the motion blur caused by mirror slap in photoshop."

I am a bit funny in this respect I must admit: when I buy a 1000$/Euro camera I am not prepared to buy an endless stream of add-on software to cope with flaws of the camera, starting with noise killers. CA removal is quite difficult though, read about it in Uwe Steinmullers diary on the F828.
The motion blur because of mirror slap might be a valid issue. but as far as I remember this occurs only at shutter speeds that are "tripod bound" anyway (can't find where I read about this, someone has figured this out in detail). So if I put a Digital Rebel on a tripod, knowing that is has a very small and light mirror I doubt that it would be a problem. Maybe one of these days I will try to compare my EOS 100/Elan (does MLU) with my Rebel TI/300v (which doesn't). The 300D should have a smaller/lighter mirror than the 300v, don't know about the damping.
Well, I do have some EOS/EF primes and zooms and therefore I would not buy the 300D kit lens, but of course not everyone is in that position.
Still, the absence of a mirror, the wide zoom range of the F828 and therefore less need (and of course no possibility) to change lenses and thus avoid dust is compelling.

Bobtrips wrote:
"You can also buy a real good prime for not a lot of money, use a fast enough shutter speed to eliminate mirror slap motion.  The extra ISO settings of a dSLR may give you the room to cut shutter speed."  
A fast lens and DSLR-high ISO would indeed be a great combo.
My Canon 50mm/F1.8 and a D300/Digital Rebel might be an awesome combination for available light; producing high quality images while being very portable and within (hobby) budget limits and of course trading off 28-200, no dust problems etc.

And now back to "being out there, have fun and actually taking photographs".
Logged
Fabio Riccardi
Guest
« Reply #75 on: December 27, 2003, 11:59:05 AM »
ReplyReply

Matthew,

The AF of the G3 sucks big time compared to that of the F828, but the G3 is way more sensitive than the F828, it actually claims ISO 50 that is almost an ISO 100. I used it at ISO 200 handheld indoors a lot and I was able to shoot at 1/30 or 1/25 in most situations.

Under the same circumstances (ISO 200, indoor) the F828 offers me to shoot at 1/10, which is pretty hard to do handheld. Pushing the sensitivity beyond ISO 200 doesn't produce very pretty images.

 - Fabio
Logged
MatthewCromer
Sr. Member
****
Offline Offline

Posts: 411


« Reply #76 on: December 29, 2003, 07:44:46 PM »
ReplyReply

Ray,

For me, a big part of the impace of a landscape print is fine detail.

If someone doesn't care about fine detail, then a 6MP dSLR blown up to 2" x 3" would look great.
Logged
Ray
Guest
« Reply #77 on: January 02, 2004, 09:39:48 AM »
ReplyReply

Sorry guys! I just can't find any facts to support your arguments. It might be true that it's the E-1 sensor that is currently unable to take full advantage of the fine quality of some of the Zuiko lenses and it might prove to be true that the next upgrade to the E-1 will address these deficiencies, perhaps with an 8MP sensor as David suggests, but what do you imagine Canon will be doing in the meantime?

We can speculate on future developments till the cows come home, but it's the current situation I'm looking at.

Let's look at Matthew's suggestion of 2 lenses and a teleconverter that take you from 28mm to 560mm in 35mm terms. That would include the 14-54mm zoom, the 50-200mm, the 1.4x converter and the E-1 body. Total weight, around 2.34Kg; total price around $3500.

Let's look at the Canon equivalent. We'll exclude IS lenses to try to keep a level playing field. I've included a 300D body, 17-40mm zoom and 35-350mm zoom. Both are L lenses and good quality (although obviously not the best that Canon have to offer). The range in 35mm terms is about the same (27mm-560mm). The total weight is 175gms more, at 2.515Kgs (Hey! Big deal! That's the weight of a small block of chocolate) but the price is about $400 less and for that sort of savings I could get the very fine Canon 50mm F1.4.

I know which system I would prefer, but if any of you guys can point me to some sample photos that demonstrate the Olympus 14-54 or 50-200 lenses are significantly sharper than those Canon equivalents I've mentioned, I might change my mind  Cheesy .
Logged
BJL
Sr. Member
****
Offline Offline

Posts: 5137


« Reply #78 on: January 05, 2004, 02:26:50 PM »
ReplyReply

Quote
Did anybody get the fact that F828 ISO values are wrong? (ISO is a standard, isn't it?)
Unfortunately, the so-called "ISO" settings on digital cameras do not follow any standard: I almost wish the International Organization for Standardisation would sue camera makers to stop their misuse of the name "ISO".
  The closest thing is an ISO standard for digital cameras is "base ISO", which is often close to the minimum ISO setting of a digital camera. To be more accurate, base ISO is the setting that causes a scene of 18% average diffuse reflectivity to have highlights blow-out at 170% diffuse reflectivity. Thus, only when a scene's average reflectivity drops below about 10% will any normal subject of less than 100% diffuse reflectivity be blown out.

  Separately, there is the question of what level a pixel of average luminosity comes out at after amplification and A/D conversion. The ISO standard implies that it should come out at 18/170 of maximum, but Kodak has indicated that with compact digicams, they place average luminosity higher than this, so that highlights blow out at just over 100% instead of 170%; sacrificing that ISO recommended highlight headroom to reduce shadow noise I suppose.

  Given all the further manipulation of tone (contrast) curves and gamma, I find it hard to see how one can accurately compare the "sensitivity" settings of different digital cameras.
Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8887


« Reply #79 on: January 09, 2004, 08:37:32 PM »
ReplyReply

Well, I guess there's no doubt that the F828 beats the socks off the A1. However, to normalise the results perhaps we're back to the title of the thread.
Logged
Pages: « 1 2 3 [4] 5 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad