dSLR Face-Off. Advice needed! [Archive] - Digital Camera Resource Page Forums

Cyberwlf

11-06-2007, 05:29 PM

Long time digital camera owner looking to become first time dSLR owner seeking advice!

Budget
* What budget have you allocated for buying this camera? Please be as specific as possible.
Around $1000US, some flexibility there if justified.

Size
* What size camera are you looking for? Or does size not matter at all to you?
It's not the size, it's how you use it ;). But urm, no, don't want anything too big I suppose.

Features
* How many megapixels will suffice for you?
10mp+

* What optical zoom will you need? (None, Standard = 3x-4x, Ultrazoom = 10x-12x, Other - Specify)
Around 28-200mm desired minimum range (can be over two lenses, not three), ie 18-55, 55-200.

* How important is “image quality” to you? (Rate using a scale of 1-10)
9-10

* Do you care for manual controls?
Yes

General Usage
* What will you generally use the camera for?
Travel (inside buildings, landscapes, outdoors (people, animals, plants). Outdoor music events (people, decorations). General photography (no purpose other than exploration of ideas)

* Will you be making big prints of your photos or not?
Not always, but at times yes. Must be capable of it.

* Will you be shooting a lot of indoor photos or low light photos?
Not the majority, but maybe around a 1/3 - 1/4. Would like as 'noise free' as possible upto ISO 800 minimum, and preferably only starting to become visible @ 1600.

* Will you be shooting sports and/or action photos?
Not buying for sports photography, but action does come into it at times.

Miscellaneous
* Are there particular brands you like or hate?
Like Panasonic, I own two, but not impressed with their noise levels.

* Are there particular models you already have in mind?
Nikon D80, Olympus E-510, Canon 400D, Nikon D40x, Pentax K10D, Panasonic L1/10, Sony Alpha

* (If applicable) Do you need any of the following special features? (Wide Angle, Image Stabilization, Weatherproof, Hotshoe, Rotating LCD)
Image Stabilization essential, preferably in body but lens could be acceptable too.
Wide Angle desired

----------------------------------------------------------

From my research so far I have made the following observations:

Olympyus E-510

- 'Best bang for buck'. Massive feature set, Liveview (but limited in usefulness), IS onboard, dust cleaner, good lens kits - 4/3 camera.
- IQ is a bit soft, no external AF/MF switch, AF a 'bit slow', low-light AF requires flash opened, AF only goes upto 3 points, stuck with Olympus lenses only (limited, but good quality). Had WB/EV issues but fixed with firmware update.

Canon 400D aka Rebel XTi

- Best low light quality, good reputation, dust cleaner, solid build, large range of available lenses - 3/2 camera.
- No IS onboard meaning more expenses lenses, kit lens not good as it could be, new IS lenses can be costly, no support for Spot focusing.

Nikon D80

- Plenty of positive feedback, good IQ, SD support, multiple exposure support - 3/2 camera
- Most costly out of the lot, no IS onboard, noise starts to become noticable from ISO 400 in some reports

Nikon D40x

- New version of D40 with 10mp, Excellent IQ.
- Doesnt have lens flexibility or feature set of D80, can be a bit slow to focus, no sensor cleaning, no built in AF motor!

Panasonic L1/10
- Has Leica lenses, Intuitive, L10 has FZ-like body - 4/3 camera.
- (L1, L10?) Noise issues similar to that of their FZ/TZ/etc ranges.

Pentax's K10D - Has some good reviews (better than K100D), but has some noise issues and other things.
Sony Alpha has on-board IS, and some mixed reviews.

Currently own Panasonic Lumix FZ20 (+ Tele / WA adaptors, Filters, Macro, etc) and Lumix TZ3.

I am new to the world of dSLR and even though ive spent a lot of time researching this now i still feel unsure which direction to go in.

Much appreciate any assistance provided!

Thanks

coldrain

11-06-2007, 05:53 PM

A few points about the Canon 400D:
You mean spot metering, not spot focusing. Of the cameras you mention, it has the fastest AF, and also very accurate (D80 is good in that respect too).

It does have "spot metering", just a bit bigger spot. If you can't get to terms with the 400D's bigger spot in metering, you will mess up metering with a smaller spot too.

The improved 18-55 f3.5-4.5 IS USM has Image stabilization, for a very affordable price.
The to be introduced this month 55-250mm IS also is interesting, and affordable, and will perform quite well.

About the D80, noise does not become very apparent, but Nikon applies noise reduction, noticable from ISO 400 and up, losing some detail. Nikon also does that (image processing) in RAW... with the D40(x) and D80, and the D50/D70. This makes Nikon RAW oddly not really RAW.
And t explains why the D40x and D80 apparently have a lot less noise than the Sony A100 with same sensor.

The E-510 has the least image quality (looking at the sensor), because it is more noisy and less dymanic than the Canon and Nikon. The noise reduction makes the results soft. One possible solution is to shoot RAW, and use 3rd party noise reduction like noise ninja.
But still the smaller sensor has other impacts... like the 4:3 image proportions, the lack of shallow depth of field possibilities. And then there is the annoying fly by wire manual focusing action, when MF is wanted/needed.

I would choose between the Nikon D80 and Canon 400D. The EOS 40D is a bit too expensive for your budget.

swpars

11-06-2007, 07:35 PM

As a D40 user, I'd like to provide my perspective on the D40X.

- Lens selection: I haven't regretted not having an internal AF motor. Sure -- there are some older-design Nikon lenses like the 50mm f1.8 prime lens (which is a great bargain low-light lens). All the Nikon lenses that should matter to a beginning DSLR shooter are AF-S lenses with in-lens motors.

-AF speed: A friend has a Rebel XTi and with the Canon kit lens (17-55mm f3.5-5.6?) I couldn't discern any noticeable speed difference using center point AF between his camera and mine.

-Image quality: The D40X is better at high ISO than the D80. I've gotten some very usable images out of my D40 at ISO 1600.

-The included kit lens has really good optical quality for a kit lens. I don't plan on ditching mine any time soon (I may get a Sigma f2.8 lens that will AF on the D40 with similar zoom range to the kit lens down the road).

My advice for putting a kit together around $1000 would be:

Nikon D40X at $669 (B&H Photo) with 18-55mm f3.5-5.6 kit lens.

Nikon 55-200mm f4-5.6 VR (Nikon's name for image stabilization) Nikkor at $229 from B&H Photo. Yes, $229 for an image stabilized telephoto lens with totally decent optical quality.

Nikon 50mm f1.8D at $114 (B&H Photo). While you'll have to manually focus on a D40X, this is easier than it sounds (the D40X has a focus confirmation light in the viewfinder).

You could also put together a Canon kit for about the same price using the Rebel XTi & kit lens, the upcoming Canon 55-250mm IS lens, and the Canon 50mm f1.8 (which will AF on a Rebel XTi). I bought the Nikon because I liked the LCD display & camera ergonomics better, and at the time Canon had no IS telephoto lenses in the price range of the 55-200 VR.

I would avoid the Four Thirds SLRs from Olympus and Panasonic if you are looking for really good high ISO performance, with the exception of the (expensive) Olympus E-3.

The Pentax K10D is weatherproof and a really durable body for the price, but according to many Internet reviews (including DCRP's own) the out of box JPEG quality leaves something to be desired. You'll have to shoot RAW and convert with a RAW converter to realize the K10D's IQ potential.

The Sony Alpha is noisier than it should be at high ISO, which eliminated it from my consideration when I was camera shopping. The body has a really nice feel -- and it will mount all Minolta AF lenses, some of which are pretty cheap on www.keh.com -- like this one:

http://www.keh.com/OnLineStore/ProductDetail.aspx?groupsku=MA07009008069N&brandcategoryname=35MM&Mode=&item=40&ActivateTOC2=&ID=19&BC=MA&BCC=1&CC=7&CCC=2&BCL=&GBC=&GCC=

fionndruinne

11-07-2007, 12:50 PM

The D40 is not slow to focus. I've taken about 12,000 shots now and the focus speed is not a whit slower than another camera. Only in low light, low contrast situations will you have a problem, just like any other camera.

Besides, with the use of an AF-S lens, focus is going to be faster than a non-AF-S lens on another camera like the D80.

I still don't understand the appeal of the D40x over the D40, because the money you spend on the higher MP count is money that cannot go towards high-quality glass. I personally would start a system out with fast-aperture lenses to begin with (like the Sigma 18-50mm f/2.8 macro HSM); they do everything the other lenses do, with the added advantage of more light-gatering capability and shallow depth of field for that artistic look.

Cyberwlf

11-07-2007, 01:38 PM

Thanks for the replies so far. Learnt a variety of interesting points I was unaware of in this thread. Read some interesting comparison of D40 vs D40x today, also seems the D40(x) is seen in ways as a better camera than the D80 as mentioned above (but also elsewhere ive read today too), but not quite as good as the D200. It is concerning though if there is truth to there being PP in Nikons RAW format though.

Don't like the idea of going for a lower MP camera than my most recently purchased one, especially as one of my motivations for upgrading (one of many i should add), is a photographic sales service required a minimum of 8mp for various size prints they offer customers. Seen this with other similar services too. It also correlates to another reason for me going dSLR, RAW format, the ability to manipulate lossless files and retain highest IQ, so going for lower MP gives less to 'work with'.

Olympus seems to be worth ignoring currently based on all the things ive now read. Pentax K10D I will investigate further, as it is the one i read least on (read more on K100D which i was not so impressed with based on reports). Canon 400D seems to be a bit of a tall poppy, and seems to have people steering for the Nikon or Pentax over it.

Anyhow any more thoughts on this welcomed, and appreciate those who've replied so far!

fionndruinne

11-07-2007, 02:15 PM

No, the D40x has some slight improvements in sensor performance but shouldn't be seen as a "better camera than the D80" for that. There's a lot more to camera function than just sensor output. Unless small size is really something you want, I'd recommend the D80 over the D40x.

Cyberwlf

11-07-2007, 02:29 PM

It just seems that in various discussions ive seen comparing the D40 vs 40x vs 80, the 80 seems to virtually get written off by some at the expense of the 40/40x. At the end of the day im happy to spend a bit more to get both quality and something which will last me past my initial stages of making this jump to dSLR, so I dont have to rush quite as quickly to a Canon 40D/5D, Nikon D200/300, etc... ;)

Rooz

11-07-2007, 02:44 PM

the IQ of the 10mp cams are all much of a muchness with each one having slightly different characteristics. the d80 is feature rich which is why it costs more. to say a 40x is a "better camera" is quite a bizarre statement imo. what makes it so ? :confused: buggered if i know. it's a d40 with 10mp. i think for what it is, its quite overpriced.

if you can afford a d80 you may even be able o afford a 30d on some runout sale. so, if you're prepared to spend around the $850-$950 mark on body only then i'd skip the xti and look at the 30d/ d80 or even k10D. the oly of course is a very good option aswell is 4/3rd's is something you're comfortable with.

if you don;t want to spend that much on the body then the d40 is a great starter kit for unde $500 with a kit lens. or the k100d is exceptional value.

erichlund

11-07-2007, 05:26 PM

About the D80, noise does not become very apparent, but Nikon applies noise reduction, noticable from ISO 400 and up, losing some detail. Nikon also does that (image processing) in RAW... with the D40(x) and D80, and the D50/D70. This makes Nikon RAW oddly not really RAW.
And t explains why the D40x and D80 apparently have a lot less noise than the Sony A100 with same sensor.

The noise reduction here is called High ISO noise reduction. It has two settings on and off, but off only really means later. On my D200, turning it off means it will be applied over ISO 800. so if you set 1/3 stop ISOs (default), it will be in affect at ISO 1000 and above. If you turn it on, it will be in affect at ISO 400 and above. This does point out that you do have some control over when the noise reduction starts. You can also specify Low, Normal or High amount of noise reduction if you turn it on.

If you are going to nail your exposures, turn it off. If you are going to be underexposing, turn it on, because it is effective, though it does have a slight deleterious effect on detail capture. If you have a good noise reduction application like Noise Ninja, turn it off and leave it off, because then you will only get it at or above ISO 1000.

There's also a long exposure NR for exposures over 8 seconds. This is to control amp noise, and you have no control over it. Trust me, you want this.

The D80 is exactly like the D200 for noise control.

On the D40, it's just called Noise Reduction, and is only applied above 800 with exposures over 1 second, above 1600 if turned off, and then only minimal.

On the D40x, same as D40 except use ISOs 400 and 800, like the D200.

I don't have access to the manuals for the D300 or D3 yet, so I cannot paraphrase those.

As far as their RAW file format not being RAW, that's bunk. RAW file formats are proprietary. All contain some image processing, because all are converted from sensor analog to digital format. The manufacturers try to minimize the effect of the processing, but they do what they need to do, and they don't necessarily tell you what they are doing. Adobe would like you to think that they have created the "Standard" for RAW format, but that's not accepted by the industry. What is in Nikon's RAW format is Nikon's RAW format and is there to provide the best image quality they can. It's still RAW. It still acts as the digital negative, so to speak. If you use Nikon CaptureNX and ViewNX to process your RAW images, the only copies you need to make are archival, because the RAW data never gets changed by the RAW converter, and the original EXIF is retained as long as the file does not get corrupted by some other application. You can always revert to the camera settings that you shot, or any alterations that you saved as a version (Versions are really cool, and they have started popping up in other applications. Corel PSP even has versions of JPG files, but it's just saving multiple copies. Nikon saves all the versions in the original RAW file).

coldrain

11-08-2007, 12:26 PM

Cyberwlf

11-08-2007, 01:19 PM

The PP in RAW doesnt seem to be such a major issue here im gathering, as ive not seen people commenting on it having any noticable difference (for the effects the PP does do), but correct me if im wrong.

Confusing anyhow. Contrasting opinions. E-510 good... E-510 not so good. K10D good....K10D not so good. Nikon D40(x)/80 good... not good...etc.

Anyhow, interesting, as DC now has its Pana L10 review, so will be giving that a read too, but im guessing the L10 even if it gains a good review will probably not still match the Nikon D80/40x, Pentax K10D... but given Olympus and Panasonic working together on their dSLR's, from what i recall the E-510 is like a brother to the L10. But will know more once ive read this review ;)

Review of L10 here.. http://www.dcresource.com/reviews/panasonic/dmc_l10-review/index.shtml

Rooz

11-08-2007, 02:42 PM

and the more you read rather than look/ feel the cams the more confused you will be. analysis by paralysis.

as i said before each one of them is a great camera and there is no PERFECT camera out there. as i also said before you need to think more about what features are important to you and what budget you finally have; then try and get a camera body that fits those features and budget.

ANY dslr you mentioned will produce excellent images which are more limitede by your ability rather than their specification.

erichlund

11-08-2007, 02:48 PM

RAW is not really RAW with most nikon models, Eric.
Even if you turn off NR some processing is applied, like some noise reduction (this is probably why the D200 seems to lack chrominance noise, where the Sony A100 and Pentax K10D show a healthy amount of chrominance noise?) and in some models maybe even some lens adjustment (like vignetting?).

Please post the ISO standard definition of RAW image format that every manufacturer must follow before they can call their format RAW. I presume you are referring to an ISO format, because you are not American and so I would not expect you to specify an ANSI standard, but use that if you cannot find an ISO standard. I personally don't even choose to try and look this up, because I'm pretty certain there is no such standard, and RAW is RAW as defined by each manufacturer. Prove me wrong.

You can see this "behavior" in the infamous unofficial "mode 3" NR setting, that has been used by some Nikon shooters since the Nikon D70.
Mode 1 being "noise reduction off", mode 2 being "noise reduction on" (for longer exposures, it uses a black frame to cancel out hot pixels and such), and mode 3 being that you shoot in mode 2, but turn the camera off while the black frame is being taken. The actual RAW data will be written to your card still, but the black frame will not be used to alter the photo, not will other processing be applied.

When you now compare a photo taken at mode 1, and at mode 3, you will notice how Nikon actually processes the actual RAW data before it is stored in the NEF file. The mode 1 photo will be processed, the mode 3 photo is not processed.

Not only the D70(s) shows this filtering of RAW, the D40(x) and D50 do too. Probably the D80 and D200 too, but I have no D80 or D200 to try that out of course.

Try it out on longer exposure photos, it will be interesting to see what the D200 does.

The rest of this "stuff" coldrain goes on about is not important because what goes into the processing of any companies RAW processing is irrelevant. Every digital image from every commercially acceptable camera is processed to some degree from the analog sensor data to that companies acceptable concept of what an image should look like.

Canon dSLRs provide a consistent "look" to their images, yet the cameras have different sensors and different processing engines (yes, they even name the processing engines, so does Nikon). Why, then, do they all look the same? Because Canon engineers design the post processing of the sensor data so that they achieve the Canon "look". Some people go absolutely gaga in love with the look, others absolutely detest it, and most are somewhere in between. But, make no mistake, it is not just this voltage becomes that number with no relation to the rest of what's going on in the image. It is carefully crafted with very close attention to detail. The end result is a set of analog data converted to digital image data that represents an image.

This digital data is considered the RAW output from the sensor. If that image is going to be further converted to jpg, it then goes under significant further processing to impose the camera settings on the data. If it is to be stored as RAW, the camera settings are attached to the file as what is called EXIF data.

This is no less post processing than what coldrain is talking about concerning Nikon images. I don't deny in any manner that Nikon post processes their images, or that some of that processing is to control image noise. They would be utterly stupid not to do so. The reason for it is to put your best foot forward, within the company's determination of what makes a good image.

If Nikon's RAW is not RAW, then nobody produces RAW, because all images have some forms of post processing going on. NONE of the major manufacturers (and probably not the minor ones either) are going to tell you what they do: That is a closely guarded secret. However, what I am delivered is Nikon's version of RAW, whether coldrain wants to call it that or not. I will continue to call it RAW, because that's the way I use it. It makes no difference whether one version or another of some limited post processing was performed. The user controllable data is still attached as separate data, and does not affect the image in any way that I don't want it to.

Rooz

11-08-2007, 03:18 PM

erichlund

11-08-2007, 04:00 PM

FWIW...i had always thought RAW was just that....RAW and unaltered. so there you go, we learn something new everyday. but really...this is a bit of a redundant technical argument for a noob looking to buy their first camera isn't it ?

the fact that i never knew and more to the point, never noticed any of that technical mumbo jumbo, speaks volumes. and i don;t mean just my own ignorance about RAW. :p

Absolutely. Coldrain just has a way of tweaking that last nerve. He's used this logic before. "Sure the Nikon is a nice camera but..." and there's always a but. It doesn't even have to be relevant.

fionndruinne

11-08-2007, 04:33 PM

You're looking at the decision wrong if you're looking for "good/not good" in evaluating. Evaluate based on what you need from the camera - what type of shooting you are most likely to be doing, and what camera(s) supports this with the right feature-set. You're buying a tool, so focus on buying the right tool for the job. Not the "perfect" tool all-around. Such does not exist.

And don't let coldrain confuse you. His esoteric argument, wile interesting perhaps to those who really know this stuff in depth, has no bearing on your decision.

Cyberwlf

11-08-2007, 05:51 PM

I wont deny the desire to buy the 'perfect' camera... When i bought my FZ20 a few years ago, for a Prosumer/P&S it was the closest thing to perfect which was on the market, without issues, no, but it became a tool I loved/love.

The more I research the more I realise, even if i increase my budget, I won't get everything I would like, but I can get much like the FZ20 was, something which becomes perfect.

Nikon D80 is probably now first in my mind, and Pentax K10D second, with Oly E-510 in third... but i'm not sure if i want a 4/3 camera or not.

The Lumix L10 seems to be better than the Oly E-510 (it's 'brother') in various ways, but worse in others, including significantly more $$$

coldrain

11-09-2007, 05:19 AM

The reason I mentioned the Nikon "RAW" was because the OP mentioned two things:
According to him, according to some sources, the D80 does not perform all that well at ISO above 400.
I said its performance actually is quite good. And that the K10D according to some sources is not great in that area.

That I then explain why probably the D80 performs a bit better than the Sony A100 and Pentax K10D in regards to high ISO noise performance has myabe no relevance (I just thought it might be interesting to know why the same kind of sensor gives different results in different cameras), and maybe it is relevant. Because if the D80 performs a bit better than the K10D in this area, logic dictates that the same results can be reached with a K10D with post processing noise reduction.

And Eric, with doing some tests and looking carefully at the results, it is fairly safe to say that the Canon RAW files are not processed/filtered by the camera's processor.
There are not enough Pentax K10D owners for such detailed information to be found on internet in Pentax' case.

I did say I would choose between the Canon 400D and Nikon D80.
I would choose the Canon because of the lens line-up, others might choose the Nikon because the D80's body feels more comfortable.

erichlund

11-09-2007, 08:18 AM

And Eric, with doing some tests and looking carefully at the results, it is fairly safe to say that the Canon RAW files are not processed/filtered by the camera's processor.

Careful tests will not tell you diddly about what voltages are occuring at each photosite and what math algorithms are used to ensure that things like crossover errors and other image defects are not being corrected as part of the analog to digital conversion. You simply don't have access to that data. The fact that Nikon made an error in their processing that let some people see their underwear, so to speak, simply indicates that software is not perfect. Being a software engineer, myself, I can understand this.

I can also understand the difference between black box and white box testing. You have an end product. You do not have access to the internal routines used to convert analog to digital, so you can only do black box testing. Now, if you somehow acquire a lab, take the camera apart, download the software, intercept the signals at the sensor and compare them to the individual routines to do analog to digital conversion, and prove that nothing but straight AD conversions are taking place, then you will have performed white box testing. You will also have destroyed your camera, something I don't think you are prepared to do.

There is no black box test you can perform that will prove anything other than the camera works as advertised (or not). In the business, this is called functional requirements testing. It tells you nothing about how the thing works, only that it does or doesn't meet requirements.

swpars

11-09-2007, 08:22 AM

And Eric, with doing some tests and looking carefully at the results, it is fairly safe to say that the Canon RAW files are not processed/filtered by the camera's processor.
There are not enough Pentax K10D owners for such detailed information to be found on internet in Pentax' case.

I did say I would choose between the Canon 400D and Nikon D80.
I would choose the Canon because of the lens line-up, others might choose the Nikon because the D80's body feels more comfortable.

Coldrain, can you direct me to some of these tests showing that Canon RAW is more "unfiltered" than Nikon RAW? I'd like to see these results and see if there's any discernable result to anyone other than a hardcore pixel-peeper.

Regarding the Canon vs. Nikon lens lineup - the only real gap in the Nikon lineup that the original poster would care about is that is Nikon has nothing really equivalent to the 70-200mm f4 L lens. Not a problem - I'd rather have a Sigma 50-150mm f2.8 anyway for speed or a Nikkor 70-300mm f4-5.6 VR for length. There's no substitute for aperature, and there's plenty of f2.8 Nikkors that are the equal or better of similar Canon L lenses.

Rooz

11-09-2007, 02:42 PM

the RAW issue is one of complete irrelevance, lets not start an argument over something so trivial and redundant.

RE: the lens line up. that will entirely depend on what you shoot and the focal range you shoot at. since i use both nikon and canon systems extensively, there is only 1 lens missing from the line up as far as i'm concerned. the 17-55 f2.8 IS. yes there is the 17-55 2.8 nikkor but NO VR. that is a complete and utter joke for the price the nikkor sells at. there is also to a lesser degree the 17-40 f4L which i fell in love with when using. but the range and f4 means its a pretty significant comprimise that i don;t know if i could live with.

the 70-200 f4 is a nice lens cos its smaller, lighter, cheaper and more manageable than the 2.8 versions. a nice lens option to have but not a deal breaker like the 17-55.

Phill D

11-10-2007, 01:33 AM

Cyberwlf I have a lot of synergy with your dilema. I could have posted almost an identical post to yours having had an FZ20 for a few years now so this thread has been very interesting. Even the RAW diversion was fascinating. Have you tried out any of the suggested combinations in a shop? It seems to me that that is how you will make your final decision. I decided a while ago that there were two reasons why I found it difficult to decide what dslr to get, first was convincing my wife that I needed one :( and second because the FZ ended up doing most of what I needed. Lower noise especially in low light images is the one thing I'd change for. I agree entirely with your comments on the Oly E510 vs the Pana L10 again neither is quite right if only someone could mix the best bits from them both & add in a canon sensor!
So I offer just another thought for you to consider as it is one going round my head at present. What would be the best way to get a dslr to replace the FZ20 & retain as closely as possible the zoom ability from wide angle to tele without changing lenses. (because I just wonder how much I'd miss the convenience of just being able to zoom quickly at a moments notice). Options image stabilised of course well how about;
- the L10 with the new leica 14-150 IS lens (pretty pricy I guess)
- the Oly E410 with the above lens, cheaper maybe
- the Oly 510 with oly 18-180 (may be a bit pricy too)
- the canon 350D (or 400D if funds stretch) with a sigma 18-200 DC OS lens
- the Nikon 40D (or DX/D80 if funds stretch) with the same sigma lens (will this AF on the 40?)
- the pentax K100D (or super or K10D if funds allow) with the tamron 18-250 lens
- the Sony alpha with the tamron 18-250 lens

I have not priced all these to check they are within your budget but if you were to go this way you could add better low light gathering ability later perhaps with something like a 50mm f1.8 lens canon has one that seems pretty cheap & to be well regarded. I'm sure the others do too.

I have a friend who has the canon 350D plus sigma 18-200 & is very pleased with it as a combination.

Let us know whatever you finally decide on.

coldrain

11-10-2007, 08:18 AM

Coldrain, can you direct me to some of these tests showing that Canon RAW is more "unfiltered" than Nikon RAW? I'd like to see these results and see if there's any discernable result to anyone other than a hardcore pixel-peeper.

Regarding the Canon vs. Nikon lens lineup - the only real gap in the Nikon lineup that the original poster would care about is that is Nikon has nothing really equivalent to the 70-200mm f4 L lens. Not a problem - I'd rather have a Sigma 50-150mm f2.8 anyway for speed or a Nikkor 70-300mm f4-5.6 VR for length. There's no substitute for aperature, and there's plenty of f2.8 Nikkors that are the equal or better of similar Canon L lenses.
It is pretty hard to find RAW comparisons, most comparisons you find are from JPEG. And it also is hard to get people to post RAW samples, simply because they are so defensive about "their" brand.

When you look at D80 and D200 photos, at higher ISO, you may find some detail lacking compared to lower ISO photos. Partly due to noise, partly due to "filtering"?

Here is one link to how Canon and Nikon differ in image capturing, from the D70 and 10D. This overall difference still holds true today I think, but it is hard to prove that since not a lot of people will be prepared to actually have a look into what their camera actually does and help figure things out.
http://www.astrosurf.com/buil/d70v10d/eval.htm

About the lens lineup, and why I do prefer the Canon lens lineup:

Lenses I find interesting and that I want to have:
Canon EF 35mm f1.4 L. Great build quality, very nice lens to have.
Canon EF 135mm f2.8 SF. Very sharp, nice for longer portrait lens, very affordable.
Canon EF 70-200 f4 L USM. Has no equivalent on Nikon or any other DLSR platform. Half the weight (and price) of f2.8 70-200 lenses.
Canon EF 70-200 f4 L IS USM. Has no equivalent on Nikon or any other DSLR platform. Sharper than any other 70-200 lens, great colour and contrast, and 4 stop IS (and even nicer bokeh than the 70-200 f4 non-IS).
Canon EF-S 10-22 f3.5-4.5 USM. Wider than most competition, less distortion at the widest end too (except for the Olympus 7-14mm).
Canon EF 17-40 f4 L USM. Very good build quality, affordable, handles very nicely. No Nikon equivalent.
Canon EF 17-55 f2.8 IS USM. Very good optically for its kind, has IS, quite a bit cheaper than the Nikon 17-55 f2.8 DX.
Canon EF 100 f2.8 USM macro. Sharper, better contrast than Nikon counterpart. Great handling and very fast AF for a macro lens. Nikon's 105mm has VR though.
Canon EF 180mm f3.5 L USM. Very high on my want list. With Tamron 180mm f3.5 the top of the tele macro lenses optically, but much better build and fast silent AF.
Canon EF 70-300 f4-5.6 IS USM. Sharper and better contrast than any 70/75-300mm class lens.
Canon MP-E 65mm f2.8 1-5x macro. A quite unique lens I will get one day for artistic extreme macro photography.
Canon TS-E 24mm f3.5 L. Shift lens... nice stuff, when i get too much money maybe...

Of course, all of above lenses may not be of interest to other photographers, but that is why I said I prefer the Canon lens lineup, not that everybody should prefer it. And the most expensive lens of above list is the Canon 180mm f3.5 L USM macro, at around $1200.

swpars

11-10-2007, 10:39 AM

Of course, all of above lenses may not be of interest to other photographers, but that is why I said I prefer the Canon lens lineup, not that everybody should prefer it. And the most expensive lens of above list is the Canon 180mm f3.5 L USM macro, at around $1200.

True, every lens maker offers lens lengths and speeds that other manufacturers do not offer. I bought my Nikon because at the time, Canon didn't offer a budget telephoto lens with IS under $300 USD and I also wanted the option of being able to use old Nikkor manual focus lens from Ebay or keh.com for inexpensive prices.

Plus, of all the budget DSLRs, I preferred the D40's out of the box JPEG quality and high ISO noise performance to anything else on the market around the same price range, and in-body image stabilization seemed like a future major mechanical failure point in my camera body. But what worked for me may not work for others.

The website regarding the D70 vs 10D performance when shooting RAW for astrophotography is interesting. Thanks for posting that.

coldrain

11-10-2007, 10:47 AM

True, every lens maker offers lens lengths and speeds that other manufacturers do not offer. I bought my Nikon because at the time, Canon didn't offer a budget telephoto lens with IS under $300 USD and I also wanted the option of being able to use old Nikkor manual focus lens from Ebay or keh.com for inexpensive prices.
For older Nikkor manual focus lenses you did not make a good choice, though.
A nikon D40 can not meter with a manual focus Nikon mount lens (unless it is one of the newer ones with CPU inside that tells distance information or something). To be able to actually meter with one on a Nikon DSLR, you need to look at a Nikon D200. Anything below (D40/50/70/80) will not meter.

Any Canon EOS DSLR will do better with older Nikkor lenses. You need to get a Nikon to EOS lens mount adapter, and then any Canon DSLR will allow you to meter while using the older Nikkor lenses.

I do not know why Nikon prevents you from metering with the older lenses.

erichlund

11-10-2007, 12:43 PM

The problem with the work done by Silvain (I use Silvain Rondi, since it's the only name I found in the entire article) is not the technical work. That's fine. It's still a black box test, as I described earlier. The fact that you can fool a Nikon camera into doing something it does not ordinarily do is a functional defect of the software/hardware. Apparently, for astrophotography, it is a beneficial defect.

I will not argue that the specific camera tested, the D70, is perhaps not the best choice for astrophotography. I won't even argue that Nikon's "median" filter, as it is described, does a poor job on long exposures. The evidence there is clear. If I decide to take up astrophotography, I won't go looking for a D70. I may even consider another brand of camera, if it does a better job. It's not like I'll be using it hand held. It's going to be mounted on a telescope.

However, where I do have an issue is with his assumption that because he is able to expose an issue on the D70 that results in a degraded image quality in some portions of the camera's useability, that this is not RAW. Why? Because he does not define what is RAW for a Nikon camera. Nikon does. It is proprietary. All RAW formats in existance are proprietary. Even Adobe's DNG is proprietary, though they would like others to accept it as the standard.

The fact that Nikon's RAW format is degraded by the median filter is a problem for certain users. However, his definition of lossless is incorrect. Once the image is stored in the RAW file, from that point, it is lossless. I can make any number of manipulations of the image. I can store many different versions, both in the original file, and by saving copies. The RAW data stored in that file will be the same in all versions.

Every RAW format is converted from analog photosite voltages to digital pixel values. The act of conversion from analog to digital is not lossless, since it implies a limited number of steps, where an analog voltage has an infinite number of points in its potential amplitude. So, if we are going to say that the definition of a RAW file is that nothing is lost from sensor to file, then we have already concluded that a RAW file cannot exist unless you actually store the actual voltages from the sensor in the file. There is no camera that does this, and even if one did, it would be limited in accuracy by the number of decimals allowed in the stored value.

Nothing in this article exposes what Canon has done within their files. They may, in fact, also have a "median" filter. They may simply have a much better one (in the cameras represented in the study). However, because Canon has done a better job of hiding their underwear, we have no way of knowing what they do to their images. We cannot say that, even if they do have a median filter, that it is or isn't lossless (prior to storage), because we have no way to expose the image processing that is done to achieve the Canon look.

So, when Silvain says that the Nikon RAW format is not truly RAW, I must disagree. However, I will say that the Nikon D70 RAW format (each camera has it's own, it's not just by company), without modification, is not an acceptable format for astrophotography. It is modifiable in a manner that makes it more acceptable, but the resulting format is not actually the Nikon D70 RAW format, since it does not contain all the elements that Nikon imposes on their RAW format. This is the case, even if it is still readable as a RAW format.

Now, if the American National Standards Institute or ISO (I can't remember what the name of the organization is, and it's not International Standards Organization, much as that would make sense) want to get the industry together and publish a standard for RAW, I'm all behind that. That would certainly make writing software for RAW manipulation easier. They could even use Adobe's format as a starting point. But they have not done that, so RAW is proprietary, period.

swpars

11-10-2007, 01:43 PM

For older Nikkor manual focus lenses you did not make a good choice, though.
A nikon D40 can not meter with a manual focus Nikon mount lens (unless it is one of the newer ones with CPU inside that tells distance information or something). To be able to actually meter with one on a Nikon DSLR, you need to look at a Nikon D200. Anything below (D40/50/70/80) will not meter.

Any Canon EOS DSLR will do better with older Nikkor lenses. You need to get a Nikon to EOS lens mount adapter, and then any Canon DSLR will allow you to meter while using the older Nikkor lenses.

I do not know why Nikon prevents you from metering with the older lenses.

I didn't want to spend over $1000 on a camera body so a D200 (even used) was not an option.

You might want to take a look at this Flickr member's site. This is proof that you don't need a D200 that meters with MF lenses to shoot pictures with old Nikkor MF glass. He shoots a D40 and a bunch of old pre-AI Nikkor MF lenses.

http://www.flickr.com/photos/kukkurovaca/

Besides, I wouldn't want to risk trashing a Canon with a 3rd party Canon to Nikon adapter, which is possible (read below). If someone did this to a 5D like the link mentions - ouch.

http://www.kenrockwell.com/tech/roxsen-converter.htm

coldrain

11-10-2007, 02:09 PM

For all the writings of eric, RAW still is generally considered to be the actual DATA the sensor delivers. The sensor being the chip, not the photo diodes only.

So I do not see where a confusion about what constitutes real RAW is about.
The "filtering" and/or "processing" Nikon seems to apply to many models is done by the camera's computer, not the sensor. So, where they speak of "real RAW" with so called "mode 3" long exposure NR photos seems to be correct: Data which the sensor delivers.
Mode 3 RAW/NEF can be made with ANY ISO setting, but only with longer exposure times.

All nice, about "we do not know what Canon does", but it just is a fact that Canon RAW files actually deliver the data from the sensor, and studying the actual RAW files in a RAW convertor does show that, you do see normally which pixels have wrong values and which are right. And you also quite clearly see actual hot pixels just one pixel in size, without any of the noise reduction speckling (which you see when colour noise is filtered out Nikon style) and softening of details.

RAW: data from the sensor. Semi RAW: Nikon NEF.
Blurred line with RAW and processing: Sony's A700, which apparently has on-sensor noise reduction processing?

coldrain

11-10-2007, 02:21 PM

I didn't want to spend over $1000 on a camera body so a D200 (even used) was not an option.

You might want to take a look at this Flickr member's site. This is proof that you don't need a D200 that meters with MF lenses to shoot pictures with old Nikkor MF glass. He shoots a D40 and a bunch of old pre-AI Nikkor MF lenses.

http://www.flickr.com/photos/kukkurovaca/

Besides, I wouldn't want to risk trashing a Canon with a 3rd party Canon to Nikon adapter, which is possible (read below). If someone did this to a 5D like the link mentions - ouch.

http://www.kenrockwell.com/tech/roxsen-converter.htm
There are very reliable adapters available from real manufacturers.... so there is little risk in that. Like the novoflex adapter:
http://www.camerahacker.com/Novoflex/EOSNIK.shtml
(much cheaper than the $200 quoted in that article by the way).

I will buy a Nikon to EOS adapter somewhere in future, it will be funny and fun to use the Nikon lenses from my Nikkormat FTn on my 350D. I have a 35mm, 50mm, 85mm, 135mm and a 50mm micro.

The only way to make correctly exposed photos with your D40 and older lenses is with a light meter. Have you ever used one, and/or do you have one? It does complicate matters a lot, so the EOS route for this particular use would have payed off. Yes of course it is possible with a D40, not easy though, and needs knowledge in the use of light meters.

swpars

11-10-2007, 06:24 PM

The only way to make correctly exposed photos with your D40 and older lenses is with a light meter. Have you ever used one, and/or do you have one? It does complicate matters a lot, so the EOS route for this particular use would have payed off. Yes of course it is possible with a D40, not easy though, and needs knowledge in the use of light meters.

Learning about manual metering doesn't intimidate me - rather, I look forward to it. Enjoy your adapter!

fionndruinne

11-10-2007, 10:59 PM

For all the writings of eric, RAW still is generally considered to be the actual DATA the sensor delivers. The sensor being the chip, not the photo diodes only.

That sentence doesn't disprove anything Eric said, though. And given the amount of modifications which camera manufacturers impose on their sensor (the photo-capture diode part) by means of the chip and processing, going around saying two different-brand-name cameras have "the same sensor" doesn't make a whole lot of sense. As Eric pointed out, there is no absolutely lossless "raw" file, because conversion imposes parameters defined by the manufacturer. So what we call "RAW" should not be seen as "raw", just as retaining a higher degree of information. "Half-baked", if you will.:cool:

Cyberwlf

11-18-2007, 02:36 PM

After visiting a lot of shops and getting to play with different cameras I am still to decide between the Nikon D80 and the Pentax K10D now. Some retailers assure me one is better for my needs etc and others say the same for the other. K10D seems to have some acclaim amongst professionals, but criticised for slow AF and lack of lens options, but liked for its rugged sealed body, higher bit processor, features, etc. I keep seeing Nikon D80 users though when reading photography magazines featuring professional photographers (with the other pros often using one of the Canon EOS range).

The store which recommended the K10D the strongest, and didnt seem to be a 'brand retailer' actually said the lens quality of the Nikkors was not good in general (which is contradictory to reports ive read online). He also commented that focusing too much on which body was a waste of my time and i should spend to research lenses more instead.

I am still doing further reading up on this having bought two magazines this week which covers the cameras mentioned in this thread and does comparisons between them too, but the RAW issue discussed here is not an item of major impact to me right now.

This thread though has been very useful for learning things.