DPR Screwed-up the Review of the 50D!

I am just curious, how many here open their RAW files as Smart Objects in ACR and take advantage of them and if you do, would you be willing to give that up for a small increase in conversion IQ?

I know I could get slightly improved IQ our of either the Canon (for the 40D we have) or out of the Nikon (for the D300 we have) but you'd have a hard time convincing me it would be worth giving up the final IQ quality benefits I get using ACR and going back to manipulate the RAW files after extensive Photoshop editing when opening and using them as Smart Objects.

So sure, all review results can improve INITIALLY with the manufacturer's RAW converter - is that where your image IQ processing results stop?

Regards,
Mike

-

Polaroid Swinger; Kodak Instamatic 126 Ricoh 500G; Canon FTb; Nikon F2AS; Nikon F3HP; Hasselblad 501CM; Pentax 67II, Nikon 990; Nikon D1x; Nikon D300; PhaseOne P65+ (in my dreams ;-)

Michael Firstlight said:

How is replacing ACR with a variable - like each manufacturer's own
RAW converter, an improved constant? Please be detailed in your
response to help us better understand your point.

Click to expand...

ACR is itself a variable depending on which camera one is testing. That is equivalent to using a different converter in itself! At least with the manufactures converter you can reasonably expect a "best" result since they should best know their product. I would like to make my buying decision based on the best pq the camera is capable of. Not how well ACR converts it. (Especially, not a yet optimized ACR).

Michael Firstlight said:

Mike

--
Polaroid Swinger; Kodak Instamatic 126 Ricoh 500G; Canon FTb; Nikon
F2AS; Nikon F3HP; Hasselblad 501CM; Pentax 67II, Nikon 990; Nikon
D1x; Nikon D300; PhaseOne P65+ (in my dreams ;-)

Click to expand...

--
Fred

fwampler said:

OleThorsen said:

fwampler said:

OleThorsen said:

Alan87 said:

I'm a long time canon user (currently I have, on the digital side, a
1Ds and a 400D). I've not been intending to buy into the XXD line, so
I have been watching all of these discussions on the 50D with some
neutrality (in that I don't own one and I wasn't intending to buy
one).

From this perspective, my only observation would be that the three
reviews I've now read (DPR, Imagining Resource Test and two now in
Amateur Photographer) are all, essentially, saying the same thing
about the camera.

Click to expand...

I agree with you.

Have you seen another very serious review from Digitalcamerainfo,
they have both advanced tests and very detailed transparent
scorecards. They agree with DPreview.

http://www.digitalcamerainfo.com/content/Canon-EOS-50D-Digital-Camera-Review-19359.htm

So it seems that Imaging Resource, DPreview, Digitalcamerainfo,
Amateur Photographer all say the same and rate the 50D pretty equal -
they are all corrupt and paid by the competition ; )

Click to expand...

That was never the issue. You said it not me. This is about the best
method for testing the camera. For all I know, the other sites used a
flawed converter as well.

Click to expand...

No you didn't say it, but there has been hundreds of attacks on
DPreview for being biased against Canon.

Click to expand...

It's actually immaterial.

fwampler said:

ACR is not more flawed for 50D than any other cameras.

Click to expand...

The 50D profile is just out and apparently not optmize for the 50D yet.

fwampler said:

If we imagine a future were the most important review sites has
chosen to use the Manufacturers RAW conversion applications, like DPP
for Canon, CNX for Nikon, Image Data Converter for Sony, the next
question will be, should the reviews be based on
A control based on the manufactures converter plus may ACR is people want that.

Click to expand...

fwampler said:

1. standard settings
2. 0 settings
3. the full PP package

In the case of 1. this can end with any result depending on the
standard settings the manufacturer choose as standard, just like
JPEGs.

In the case of 2. this will drive the manufacturers to do even more
automatically than today, not giving us a free choice of how we want
our RAWs converted, and will end up beeing no better than in-camera
JPEGs. Both DPP and CNX applies NR behind the scenes at 0 settings
today. You can see this if you dare to open your 50D RAW files with
open source top notch RAW converters.

In the case of 3. do we really want a DSLR comparison to depend on
how good a reviewer are at postprocessing with the different
Manufacturer RAW converters? For sure this will result in even more
scewed results than you can imagine. I don't want this.

So if you try to think a littlle deeper about the consequences of
your approach and not just outside your 50D box, I don't believe you
will like the result at all. I'm certain I will not. I find
DPreview's (and other serious review sites) choice of the de facto
third part standard RAW converter of choice as the least evil and
most neutral of all possibilities.

Click to expand...

That approach does not make for a definitive review.

Click to expand...

That was a short and well thought through answer from you ; ) Here is a even more well thought through answer from me based upon the my previous list of consequences with your approach:

Your approach will be the most skewed of all testing methods.

Results
Ole Thorsen
http://www.pbase.com/ole_thorsen

• OMNISCIENCE

Knowing what
thou knowest not
is in a sense
omniscience.
(Grook by Piet Hein)

So now that they updated the review using the production version of ACR and results are the same, what are your thoughts? I agree, they should not have used beta software - ever.

Mike

--

Polaroid Swinger; Kodak Instamatic 126 Ricoh 500G; Canon FTb; Nikon F2AS; Nikon F3HP; Hasselblad 501CM; Pentax 67II, Nikon 990; Nikon D1x; Nikon D300; PhaseOne P65+ (in my dreams ;-)

Look - the science is straightforward - it is the collection of photons on different color sensitive electronic elements. If the data is clean then the result will be easy to generate. If the data is a mess (i.e. noisy) then the result will be hard to glean. DPP could be optimized for resolution charts and that might actually be 100% appropriate, whereas ACR might optimize DR and that would also be appropriate. At some point the reviewer has to make an informed decision on which to base the analysis. In addition the reviewer noted they used other converters and got similar results.

And what do you base the notion that the manufacture's RAW converter is always the best for demonstrating maximum IQ? What if PhaseOne produces even better IQ results, then what?

Mike

--

Polaroid Swinger; Kodak Instamatic 126 Ricoh 500G; Canon FTb; Nikon F2AS; Nikon F3HP; Hasselblad 501CM; Pentax 67II, Nikon 990; Nikon D1x; Nikon D300; PhaseOne P65+ (in my dreams ;-)

OleThorsen said:

fwampler said:

OleThorsen said:

fwampler said:

OleThorsen said:

Alan87 said:

I'm a long time canon user (currently I have, on the digital side, a
1Ds and a 400D). I've not been intending to buy into the XXD line, so
I have been watching all of these discussions on the 50D with some
neutrality (in that I don't own one and I wasn't intending to buy
one).

From this perspective, my only observation would be that the three
reviews I've now read (DPR, Imagining Resource Test and two now in
Amateur Photographer) are all, essentially, saying the same thing
about the camera.

Click to expand...

I agree with you.

Have you seen another very serious review from Digitalcamerainfo,
they have both advanced tests and very detailed transparent
scorecards. They agree with DPreview.

http://www.digitalcamerainfo.com/content/Canon-EOS-50D-Digital-Camera-Review-19359.htm

So it seems that Imaging Resource, DPreview, Digitalcamerainfo,
Amateur Photographer all say the same and rate the 50D pretty equal -
they are all corrupt and paid by the competition ; )

Click to expand...

That was never the issue. You said it not me. This is about the best
method for testing the camera. For all I know, the other sites used a
flawed converter as well.

Click to expand...

No you didn't say it, but there has been hundreds of attacks on
DPreview for being biased against Canon.

Click to expand...

It's actually immaterial.

OleThorsen said:

ACR is not more flawed for 50D than any other cameras.

Click to expand...

The 50D profile is just out and apparently not optmize for the 50D yet.

OleThorsen said:

If we imagine a future were the most important review sites has
chosen to use the Manufacturers RAW conversion applications, like DPP
for Canon, CNX for Nikon, Image Data Converter for Sony, the next
question will be, should the reviews be based on
A control based on the manufactures converter plus may ACR is people want that.

Click to expand...

OleThorsen said:

1. standard settings
2. 0 settings
3. the full PP package

In the case of 1. this can end with any result depending on the
standard settings the manufacturer choose as standard, just like
JPEGs.

In the case of 2. this will drive the manufacturers to do even more
automatically than today, not giving us a free choice of how we want
our RAWs converted, and will end up beeing no better than in-camera
JPEGs. Both DPP and CNX applies NR behind the scenes at 0 settings
today. You can see this if you dare to open your 50D RAW files with
open source top notch RAW converters.

In the case of 3. do we really want a DSLR comparison to depend on
how good a reviewer are at postprocessing with the different
Manufacturer RAW converters? For sure this will result in even more
scewed results than you can imagine. I don't want this.

So if you try to think a littlle deeper about the consequences of
your approach and not just outside your 50D box, I don't believe you
will like the result at all. I'm certain I will not. I find
DPreview's (and other serious review sites) choice of the de facto
third part standard RAW converter of choice as the least evil and
most neutral of all possibilities.

Click to expand...

That approach does not make for a definitive review.

Click to expand...

That was a short and well thought through answer from you ; ) Here is
a even more well thought through answer from me based upon the my
previous list of consequences with your approach:

Your approach will be the most skewed of all testing methods.

Click to expand...

Actually not, since it tells my what I need to know. That is, what pq can I expect from my camera using the best converter available. (If there is one better than the manufactures, then that is ok too).

OleThorsen said:

Results
Ole Thorsen
http://www.pbase.com/ole_thorsen

• OMNISCIENCE

Knowing what
thou knowest not
is in a sense
omniscience.
(Grook by Piet Hein)

Click to expand...

--
Fred

Michael Firstlight said:

And what do you base the notion that the manufacture's RAW converter
is always the best for demonstrating maximum IQ? What if PhaseOne
produces even better IQ results, then what?

Mike

--
Polaroid Swinger; Kodak Instamatic 126 Ricoh 500G; Canon FTb; Nikon
F2AS; Nikon F3HP; Hasselblad 501CM; Pentax 67II, Nikon 990; Nikon
D1x; Nikon D300; PhaseOne P65+ (in my dreams ;-)

Click to expand...

Then, that would be wonderful. But, I want one as good as comes free with the camera. If there is one better, then I might buy it.
--
Fred

I checked just 3 cameras - the 40D, the 50D and the 450D using DPR own test images and saw some interesting things:

(a) Items closer to the frame like the Elmar lens are very sharp in the 40D and 50D pics and slightly blurred in the 450D sample

(b) The Queen of Hearts looks sharpest on the 450D followed by the 40D and slightly fuzzy on the 50D

If various such near and far objects are compared, on an average the 40D looks sharp for all objects but for the distant elements the 450D image is sharper and for near objects the 50D image is sharper.

Now considering that all these were shot at f8 I would not have expected such DOF issues with the 3 test shots (for that is what they look like!

So to me it looks like the 50D image is focused to the front so that the back panel is fuzzier and the 450D is focused to the back making the near objects a bit fuzzier when pixel peeping.

Thats my 2 bits for now

(wish i could have posted the crops. Maybe i will attach a flickr link)

Michael Firstlight said:

I am just curious, how many here open their RAW files as Smart
Objects in ACR and take advantage of them and if you do, would you be
willing to give that up for a small increase in conversion IQ?

I know I could get slightly improved IQ our of either the Canon (for
the 40D we have) or out of the Nikon (for the D300 we have) but you'd
have a hard time convincing me it would be worth giving up the final
IQ quality benefits I get using ACR and going back to manipulate the
RAW files after extensive Photoshop editing when opening and using
them as Smart Objects.

So sure, all review results can improve INITIALLY with the
manufacturer's RAW converter - is that where your image IQ processing
results stop?

Click to expand...

I just want to get an idea of what a camera is capable of. Probably the raw converter provided best does that (these reviews come early on so that Adobe is probably not yet optimized). It really is all about trying to determine the "potential" of the camera being considered.

Michael Firstlight said:

Regards,
Mike

-
Polaroid Swinger; Kodak Instamatic 126 Ricoh 500G; Canon FTb; Nikon
F2AS; Nikon F3HP; Hasselblad 501CM; Pentax 67II, Nikon 990; Nikon
D1x; Nikon D300; PhaseOne P65+ (in my dreams ;-)

Click to expand...

--
Fred

So who determines which converter is equal to or better than the manufacturer's? What are the criteria for picking one for testing. See, you are treading down a path with les and less constants other than its from the manufacturer and that's not the RAW convert what most buyers use. Sure, they can test for the small minority of sophisticated users, but I think these reviewer's target market is geared to the resuls most people will actually see.

Mike

--

Polaroid Swinger; Kodak Instamatic 126 Ricoh 500G; Canon FTb; Nikon F2AS; Nikon F3HP; Hasselblad 501CM; Pentax 67II, Nikon 990; Nikon D1x; Nikon D300; PhaseOne P65+ (in my dreams ;-)

fwampler said:

OleThorsen said:

fwampler said:

OleThorsen said:

fwampler said:

OleThorsen said:

Alan87 said:

I'm a long time canon user (currently I have, on the digital side, a
1Ds and a 400D). I've not been intending to buy into the XXD line, so
I have been watching all of these discussions on the 50D with some
neutrality (in that I don't own one and I wasn't intending to buy
one).

From this perspective, my only observation would be that the three
reviews I've now read (DPR, Imagining Resource Test and two now in
Amateur Photographer) are all, essentially, saying the same thing
about the camera.

Click to expand...

I agree with you.

Have you seen another very serious review from Digitalcamerainfo,
they have both advanced tests and very detailed transparent
scorecards. They agree with DPreview.

http://www.digitalcamerainfo.com/content/Canon-EOS-50D-Digital-Camera-Review-19359.htm

So it seems that Imaging Resource, DPreview, Digitalcamerainfo,
Amateur Photographer all say the same and rate the 50D pretty equal -
they are all corrupt and paid by the competition ; )

Click to expand...

That was never the issue. You said it not me. This is about the best
method for testing the camera. For all I know, the other sites used a
flawed converter as well.

Click to expand...

No you didn't say it, but there has been hundreds of attacks on
DPreview for being biased against Canon.

Click to expand...

It's actually immaterial.

fwampler said:

ACR is not more flawed for 50D than any other cameras.

Click to expand...

The 50D profile is just out and apparently not optmize for the 50D yet.

fwampler said:

If we imagine a future were the most important review sites has
chosen to use the Manufacturers RAW conversion applications, like DPP
for Canon, CNX for Nikon, Image Data Converter for Sony, the next
question will be, should the reviews be based on
A control based on the manufactures converter plus may ACR is people want that.

Click to expand...

fwampler said:

1. standard settings
2. 0 settings
3. the full PP package

In the case of 1. this can end with any result depending on the
standard settings the manufacturer choose as standard, just like
JPEGs.

In the case of 2. this will drive the manufacturers to do even more
automatically than today, not giving us a free choice of how we want
our RAWs converted, and will end up beeing no better than in-camera
JPEGs. Both DPP and CNX applies NR behind the scenes at 0 settings
today. You can see this if you dare to open your 50D RAW files with
open source top notch RAW converters.

In the case of 3. do we really want a DSLR comparison to depend on
how good a reviewer are at postprocessing with the different
Manufacturer RAW converters? For sure this will result in even more
scewed results than you can imagine. I don't want this.

So if you try to think a littlle deeper about the consequences of
your approach and not just outside your 50D box, I don't believe you
will like the result at all. I'm certain I will not. I find
DPreview's (and other serious review sites) choice of the de facto
third part standard RAW converter of choice as the least evil and
most neutral of all possibilities.

Click to expand...

That approach does not make for a definitive review.

Click to expand...

That was a short and well thought through answer from you ; ) Here is
a even more well thought through answer from me based upon the my
previous list of consequences with your approach:

Your approach will be the most skewed of all testing methods.

Click to expand...

Actually not, since it tells my what I need to know. That is, what pq
can I expect from my camera using the best converter available. (If
there is one better than the manufactures, then that is ok too).

Click to expand...

You just want to know how good your 50D are and not how it compares to other cameras....I can help you with that - your 50D is very good - just not as good as some other cameras ; )

Regards
Ole Thorsen
http://www.pbase.com/ole_thorsen

• OMNISCIENCE

Knowing what
thou knowest not
is in a sense
omniscience.
(Grook by Piet Hein)

Michael Firstlight said:

So now that they updated the review using the production version of
ACR and results are the same, what are your thoughts? I agree, they
should not have used beta software - ever.

Click to expand...

Frankly, I'm puzzled why such emphasis and analysis on what the camera does not do rather than what it is capable of doing. Not everyone buys the camera for ISO3200 and up pixel peeping. Some of us want to capture kids and grandkids at play, not low light fabric swatches. Others are interested in bird detail; how high can that go? How good is the AIServo vs. other cameras? That has been something that is never reviewed. But if I spend half my time in AIServo, for me it's very important. The typical DPR review does not tell me anything about that. I have to rely on other users and my own experience to determine that. How about ALO? I never had that in any of my other Canon cameras, it would be nice to know how that works. Once again I have to rely on my own experiments and experience; not from the review.

OTOH I understand the need to debunk all the Canon marketing fluff which were exaggeration at best, if not outright lies.

Olga

Us Sony users have been living with this factor in the reviews. ACR isn't the best program to convert Sony RAW files, therefore it doesn't give the tbe best results the camera can provide.

However, to be fair, DPR is being consistant whit their RAW conversions when thay test cameras.

fwampler said:

fwampler said:

What we are talking about here is testing methodology. An arbitrary

Click to expand...

raw converter that may/probably treat each camera differently is just
plain unacceptable for a meaningful comparison.

Click to expand...

It would be more work, and it probably wouldn't fly with Phil and the
gang, but why not test with three converters and let the readers
decide which is best? Number 1 would be the bundled converter that
came with the camera; number 2 could be ACR, since it's widely used;
and number 3 could be Capture One, or any other suitable software.

Just an idea. Yes, more work, but more info for the reader. Maybe Phil
could twist some arms at Amazon.

Rob

--
'Don't sneak up on it - surround it'

Where can I find any honest and professional test 40D vs 50D using DPP?

Dpreview seem to be really heavyweight, almost a bible.

Perhaps it's hard for people to get the knowledge that 10-12 Mpix is the limit for APS-C.

Regards J O

--
Jan Olof
http://www.grafik99.com

Good points Olga, thanks.

We all use these tools for different applications; I do studio portraiture so I'd have no issues whatsoever with the 50D in that environment; the 40D performs wonderfully for that as well as does my D300. But my wife and I also occasionally shoot weddings and we find ourselves in ceremonies where flash is not allowed and have to dial up the ISO up to 2500 or 3200 even with fast glass (and wish we could go even higher with impunity like we could with say, a D3 or D700), but since that's not the bread and butter application we don't want to spend that much on all of htese quickly depreciating camera bodies - not yet a least- so for folks like us, high ISO performance matters a lot and for other applications, not in the least. So for folks in our position, we'll stay with the 40D for now and not move up to the 50D as we don't want to take a noise hit and were hoping for an improvement in that area of price/performance - real jobs and customer deliverables depend on these things.

Regards,
Mike

--

Polaroid Swinger; Kodak Instamatic 126 Ricoh 500G; Canon FTb; Nikon F2AS; Nikon F3HP; Hasselblad 501CM; Pentax 67II, Nikon 990; Nikon D1x; Nikon D300; PhaseOne P65+ (in my dreams ;-)

Olga Johnson said:

..and make things look a lot worse or better than they are.

Click to expand...

In other words, you can prove almost anything if you carefully pick your data.

Olga Johnson said:

First, you don't even tell us what ISO you are showing.

Click to expand...

In his other postings of these images it's ISO 3200.

Olga Johnson said:

Second, and most importantly, you don't state that what level of NR
the 50D image or the 40D is subjected to. However it looks like you
chose the "strong" rather than NR "off" image for the 50D and the
"off" option of the 40D.

Click to expand...

Actually, I believe these are simply the JPEG defaults which you get through the "Comparameter". If you compare NR off versions and resize the imags, the results are pretty similar between the 40D and 50D. However, when you compare 40D default to 50D LO, you will see some degradation of the red fabric sample. It looks like Canon's new NR algorithm impacts red the most.

--
Erik

I shoot HS sports.

I NEED ISO 1600, 2000, 2500, 3200, 6400 (but not 12800).

The 40d only gave me two of above.

I shot a whole football game at ISO6400. If you nail your exposures, your noise is minimized.

I also do heavy crops. I don't want to have to up-size in bridge or photoshop, i want to be able to crop with enough megapixel left for nice 8x10's. Not that possible with the 40d. every time I would do a big crop on my 40d, i would end u with like... 2mp! that's ok for 4x6 or even 5x7, but not 8x10 and I offer that size for all my photos, so then i wouldn't be able to post it.

I post photos with at least 3mp, for prints up to 8x10. I've ended up printing a few 3-ish mp shots from the 40d and 50d and they look the same in quality and noise.

So, the higher MP for me, ... makes a difference, and the higher ISOs are another reason I'm happy with my upgrade. I may buy another 40d tho, we'll see.

--

lenses make the image, the camera only records it. my zoom is my feet, not my fingers, don't be lazy, buy primes

And ACR is clearly the industry and user standard for processing RAW images. I find DPP good for a free program, but it is cumbersome and its has FAR fewer adjustments than ACR and DPP's adjustments do not allow for small enough adjustments.