I have tested the 7D and 20D side by side

Hi Wendy,

I hope Dave doesn't mind me piggy backing off his thread, but back focus can be attributed to either camera or lens. The initial batch of Nikon D70 had the dreaded back focus issue as well. I have encounted back focus with my D70, but it was due to a miscalibrated lens. I wanted to buy the Tamron 25-75mm f2.8 lens as I heard that it was an excellent lens. I tried a total of 5 from two different stores. They all had back focus. I tested at f2.8 which gives the narrowed DOF that will bring out any BF issues. While shooting at f5.6 the images were sharp, but the larger DOF only masked the BF problem, and why buy a f2.8 lens if you have to shoot a f5.6?! I do not have any BF issues with my Nikkor lens. I ended up going with a Sigma 24-70 f2.8 lens that was spot on. So, I would try other lenses before concluding that it's the camera, not to say that it's not the camera, but to confirm. Contact me offline, if you would like for me to email you the back focus testing chart. -Norm

David Martin said:

wendyorg said:

SNIP

Click to expand...

Member said:

Lastly, the only big problem I have with the 20D is that the one I
have appears to backfocus in certain situations. In my noise tests
(informal) I had the camera on a tripod with my husbands face
behind the AF point and he sat still while I shot at different
ISO's. At 200 ISO on the Canon, the shot looked to be backfocused

• his ears were sharp, his face was not. ISO 100 and ISO 400 etc

were all fine. I know little about the problem, but now that I
have seen it for myself, I realize it is something I have never,
ever seen before with any camera I have used - it's a very weird
phenomenon.

I think the image noise at high ISO's was what has swayed me
towards the 20D, but on the other hand, unless I can get one that
doesn't back-focus, I would not be happy worrying about
unpredictable backfocusing results.

Does anyone know if it is a hit and miss thing within a camera model
or if one does it, they all do it? I have heard they can
re-calibrate the camera, but you would think they would start
turning out new lots with the corrections made by now. I did not
notice the KM doing any back-focusing, although in my tripod tests,
one KM picture was drastically out of focus when nothing had been
moved or changed.
SNIP
Wendy

Click to expand...

Unlike on the 10D, there is no systemic problem with back-focussing
on the 20D.
All the Canon cameraa models can have individual issues in their
calibration with their lenses, and theway out of it is to send it
in together with the lenses to Canon service, who turn it around
very quickly, free of charge, and the problem is solved.
If you check on the 20D forum, you will see that this has sorted
out the few who have had any issues.
I guess it may be inherent in the whole idea of having motors in
the lenses, as it occurs across the Canon range, and some of the
pros with a wide range of lensews and several bodies just routinely
send in their gear when they buy any new equipment - it's just a
bit of a PIA in not having the camera, but the tech problem is
rapidly solved.

--
Regards,
DaveMart
Please see profile for equipment

Click to expand...

I own a Canon 20D and have used it for 8000 shutter actuations and almost two months and find the focus on my camera to be 94 to 95% accurate with AF. Although I would like it to be 100% accurate, 95% is realistically good for this level of DSLR. However, the Canon forums are replete with focus complaints, which apparently, a small number of cameras have. Sending it in for calibration often helps some of those situations.

I am most distubed by you report of Anti-Shake not making much of a difference. I will test it for myself and report back and see if I concur with this. But I thank you very very much, Wendy, for you extensive test which is of interest to a great many of us. I look forward to your follow up.
--
Vance Zachary
http://www.pbase.com/photoworkszach
http://www.photoworksbyzachary.com

Wendy's extensive review confirms what I have been saying comparing my 20D with 7D sample images I have seen online (I do not have my 7D yet), at 1600 and 3200, 20D is better than 7D by at least one stop. What is amazing to me about 20D after living with it now for two months, is that at 1600 the in camera noise reduction is so good that loss of detail is present but minimal and one can then apply noise reduction software and get better results at 1600 than 10D + Neat Image (and probably 7D + Neat Image). 7D is still better than 10D and most other DSLRs at overall high iso performance.

Of course, there are many other issues: color balance in mixed lighting situations, AF speed, low light AF accuracy (20D excels here), effectiveness of AntiShake when handholding, and many other issues many of which are up for subjective assessment.

My suspicion is that 20D excels in several ways over 7D (and 7D over 20D in just a few areas) but that 7D is better than 10D (the camera it was suppose to compete against). In subsequent camera releases, K-M will catch up to Canon and perhaps even surpass it.

For most of us heavily invested in Minolta gear, this is all a mute question. We are just thrilled to have a good quality DSLR to use our accessories and talents. We are happy to be able to stay home with Minolta. I can't wait to get my 7D.

When I get my 7D, I will report my findings based on my long experience with 10D and 20D and my experience with Minolta Maxxuum 9 and 7HI. I will compare the two cameras using Minolta G lenses with Canon L lenses. Although I am a Minolta guy, I won't hesitate to call a spade a spade.

But I am grateful to Wendy for starting the discussion

Vance Zachary
http://www.pbase.com/photoworkszach
http://www.photoworksbyzachary.com

JackRiley said:

Not according to the new Imaging Resources review. Noise reduction
is NOT at the expense of fine detail in the 20D.

Click to expand...

NR as a specific function (dark exposure or hot pixel detection) should not be, but the general NR incorporated in de-Bayering certainly does lose fine detail. It can't really do otherwise. Canon's images have a characteristic look which is very successful.

JackRiley said:

JackRiley said:

In theory the
Canon 20D should outresolve the Minolta A2 just on the grounds of
larger sensor; in practice, it's about 5 per cent worse than the
EOS 1D Mk II, and about equal to the A2.

Click to expand...

According to whom? Phil's tests show identical resolution to the 1D
Mk II.

Click to expand...

FOTO Sweden, whose res tests I would trust above any current on-line review. 1800-1850 as opposed to 1900-1900 for the 1DS Mk II. They show the A2 at 1600-1650 but this was from the original before firmware upgrades; my experience after doing the firmware upgrade, and especially when CS RAW 2.3 became available, was a quantum leap in resolved detail.

Like Canon, KM have been their own worst enemies with their bundled software. It always seems to be a third party which can bring out the best in the cameras.

DK

looking forward to your comments Vance. When do you think you'll have your 7D?

Vance Zachary said:

Wendy's extensive review confirms what I have been saying comparing
my 20D with 7D sample images I have seen online (I do not have my
7D yet), at 1600 and 3200, 20D is better than 7D by at least one
stop. What is amazing to me about 20D after living with it now for
two months, is that at 1600 the in camera noise reduction is so
good that loss of detail is present but minimal and one can then
apply noise reduction software and get better results at 1600 than
10D + Neat Image (and probably 7D + Neat Image). 7D is still better
than 10D and most other DSLRs at overall high iso performance.

Of course, there are many other issues: color balance in mixed
lighting situations, AF speed, low light AF accuracy (20D excels
here), effectiveness of AntiShake when handholding, and many other
issues many of which are up for subjective assessment.

My suspicion is that 20D excels in several ways over 7D (and 7D
over 20D in just a few areas) but that 7D is better than 10D (the
camera it was suppose to compete against). In subsequent camera
releases, K-M will catch up to Canon and perhaps even surpass it.

For most of us heavily invested in Minolta gear, this is all a mute
question. We are just thrilled to have a good quality DSLR to use
our accessories and talents. We are happy to be able to stay home
with Minolta. I can't wait to get my 7D.

When I get my 7D, I will report my findings based on my long
experience with 10D and 20D and my experience with Minolta Maxxuum
9 and 7HI. I will compare the two cameras using Minolta G lenses
with Canon L lenses. Although I am a Minolta guy, I won't hesitate
to call a spade a spade.

But I am grateful to Wendy for starting the discussion

Vance Zachary
http://www.pbase.com/photoworkszach
http://www.photoworksbyzachary.com

Click to expand...

Paul Hancock said:

Vance Zachary said:

Wendy's extensive review confirms what I have been saying comparing
my 20D with 7D sample images I have seen online (I do not have my
7D yet), at 1600 and 3200, 20D is better than 7D by at least one
stop. What is amazing to me about 20D after living with it now for
two months, is that at 1600 the in camera noise reduction is so
good that loss of detail is present but minimal and one can then
apply noise reduction software and get better results at 1600 than
10D + Neat Image (and probably 7D + Neat Image). 7D is still better
than 10D and most other DSLRs at overall high iso performance.

Of course, there are many other issues: color balance in mixed
lighting situations, AF speed, low light AF accuracy (20D excels
here), effectiveness of AntiShake when handholding, and many other
issues many of which are up for subjective assessment.

My suspicion is that 20D excels in several ways over 7D (and 7D
over 20D in just a few areas) but that 7D is better than 10D (the
camera it was suppose to compete against). In subsequent camera
releases, K-M will catch up to Canon and perhaps even surpass it.

For most of us heavily invested in Minolta gear, this is all a mute
question. We are just thrilled to have a good quality DSLR to use
our accessories and talents. We are happy to be able to stay home
with Minolta. I can't wait to get my 7D.

When I get my 7D, I will report my findings based on my long
experience with 10D and 20D and my experience with Minolta Maxxuum
9 and 7HI. I will compare the two cameras using Minolta G lenses
with Canon L lenses. Although I am a Minolta guy, I won't hesitate
to call a spade a spade.

But I am grateful to Wendy for starting the discussion

Vance Zachary
http://www.pbase.com/photoworkszach
http://www.photoworksbyzachary.com

Click to expand...

Click to expand...

--
Vance Zachary
http://www.pbase.com/photoworkszach
http://www.photoworksbyzachary.com

Hello Wendy,
Hello All,

Thank you very much Wendy for your comprehensive feed back about side by side 20D & D7D comparison !

I just have two small observations to say and/or ask (it is way I wish to contradict or criticize what you have said Wendy, not at all!)

1) about the “shutter” noise : this might be a quite subjective issue because on my side (and for most people I have discussed here and/or who have tried it with me) :

=> I definitely find the EOS20D shutter noise really and clearly higher then the D7D shutter noise : As you said, the D7D shutter is “longer” (it gives you the impression that you are shooting at “low shutter” speed : actually the D7D shutter noise is “longer” but much (far much) softer : it is due to the fact that the mirror moves (a little bit) slower, and is better amortized at both end (up and down).

On the other hand the 20D shutter noise is (to me at least, and to many people I have tried it with) much louder, but, you’re quite right for this, much “shorter”…
Another metaphoric comparison would be to say that

• Eos 20D shutter noise is “dryer” (but louder)
• Dynax 7D shurtter noise is “wet” (but softer and less loud)

I’m pretty sure that if I had the opportunity to physically “measure” both sounds and analyze them, I would obtain what I ascertained above

• for the 20D : something very short, very quick, and roughly said, quite like “aggressive” I would even say : like a gunfire
• for the D7D : something softer, longer, and quiet… hence something that can appear too long and/or too soft for some people

Now about the AF noise, you are definitely right : as far as you use USM ring enabled lenses, the AF is not only very quiet (more than D7D non SSM AF), and also sometimes (not always…) faster…

But then the comparison would be unfair, because using Minolta’s SSM lens can be as much as silent as the Eos 20D…

Now, however it’s also right that for canon, USM lenses are much more common…

2) About the noise now (under and above ISO 800):
=> I quite agree that until Iso 800, both camera are quite similar

(I have noticed though, that at Iso 100 and 200, on very big crop magnification, I could see that the D7D are (very very [...] very very slightly) cleaner than the 20D images… but AGAIN, please consider that both camera produce practically no noise until ISO 400 (and very acceptable (non evidently visible) noise at ISO 800

Now at ISO 1600 and 3200, I agree that eos 20D pics are “more esthetical”

But If you look at it carefully, you notice that both camera produce noticeable noise, but not the same way

• for the D7D, the noise appears as dust flying all over the picture
• for the eos 20D, the noise is much “esthetically” managed : it appears like small (and soft) flakes of snow : that’s optically and artistically “better” to see as the noise (like dust) from the D7D

Actually, if we speak strictly physically , the 8 Mpix CMOS Sensor produces natively MUCH MUCH MORE noise at ISO 3200 equivalent amplification…

But due to the very great job, and very judicious algorithms designed by Canon’s Engineers, the out-of the camera results are better…

Now, again because I did test it also, I can tell that for either cameras, if using an adequate neat image profile, it can end up cleaning the ISO 1600 and ISO 3200 images very efficiently, making the pictures very usable !

Something more important : An official info stated that here in France, at mid November, a 2000 unit D7D shipment (that was about to be dispatched in France for retailers) had to go back to japan because of a “noise test” done by a well know French magazine laboratory.

After this laboratory had reported the amazing noise issue to KM headquarters, KM engineers detected a problem in the noise management algorithms and wisely decided to call back this (production) shipment back for revision :

This partially explains how and why there is so much delay in France for the D7D on retail

It can be quite clear that you (Wendy) and some other people in this forum that I have read complaining about the noise) have had the (“un-luck”) to be buying these first defected units

Cheers

hdophan

...

Very good and thorough evaluation Wendy and obviously very personal. However your comments on comparing 20D vs 7D at higher ISO's noise are very much different of those on the tests at Luminous Landscape website.

... Lucas

Hi Henry, I took them with and without NR on, and saw small differences only at the 1600 and 3200 levels. At 1600, there was a slight improvement in noise, but many people would not notice it, and at 3200 there was slightly more noticeable improvement, so I would suggest having NR ON with these 2 ISO's, especially 3200. With NR on in either case, the images were still noisier than the 20D at the same ISO's.

Wendy

I saw this too so never bothered to worry that NR was on. But when I did these tests, I saw a difference at high ISO's, and the exposures were not over 1 sec or even close.

Wendy

pramarko said:

Henry Richardson said:

You mention that the 7D's ISO 1600 and 3200 were very noisy. In
another place you mentioned that you recommend turning off noise
reduction. I'm wondering if the reason ISO 1600 and 3200 were so
noisy was because you turned off noise reduction? I ask because
high ISO examples we have seen from the 7D have looked pretty good.

--
Henry Richardson
http://www.richardson.photoshare.co.nz/
http://www.printroom.com/pro/intrepid

Click to expand...

--
Help us build 7D community: http://www.dyxum.com
7D, lens and flash reviews

Click to expand...

Hi, yes, I thought of this, but the test pictures I did did not seem to differ too much with the exception of 3200 ISO on the Canon which for some reason was brighter than the rest. For my own knowledge, I did some levels on it to make sure they basically looked similar, not very scientific and maybe totally wrong, but the best I could do - and after levels, it didn't look any noisier in the dark areas.

Wendy

distant said:

I would just like to point out one thing, considering the noise
levels. I was surprised that there would be a big difference att
iso1600, when examples I've seen have indicated that there isn't
any very big difference, although the d20 seems to be a little bit
better in that aspect.

The thing I was thinking about was if you took into account the
difference in under- and overexposing. You said that the canon
overexposed and the minolta underexposed. This would give the
canonimges a higher signal and thus less noise. It would perhaps be
better to judge noise on similarly exposed pictures. I don't know
if that was the case conserning the your high-iso pictures, but
just a thought.

An iso1600 example from the d7 that i found :



-Thomas

Chris1943 said:

Just one point, according to the manual, noise reduction is only
supposed to affect exposures longer than 1 second, and isn't
applied to continuous frame shooting. From a software point of
view, even if on, but not applicable, some processing in the chip
must take place, if only to recognise the set of conditions and to
decide whether to apply or not.

Only the software designer would know if this affceted other
porcessing, even if the set of conditions were negative.

I certainly shall try and see, when I have a little spare time, but
the soft images get very decently sharp with only a little PPing,
and I do like the control that this gives me.

If you were shooting fast moving images, I guess your shutter
speeds would be high, so perhaps AS doesn't have a great deal of
impact. A lot of my photos are slower, big depth of focus, small
aperture, and often in tricky light, so my experience to date is
rather more positive. But no-one said that the 7D was the camera
for every situation!

Good luck with your final choice.

wendyorg said:

Well, I felt I had to test both of these cameras in a number of
real-world situations so I could make my final decision on whether
teh KM 7D was for me or not, so I got my hands on a 20D and shot a
dogs show, toured a bit of Washington DC, the National Aquarium,
the Reagan Building and did ISO noise tests using Program mode and
mainly default settings with both cameras. I just wanted to see
how each camera interpreted the exposure and WB out of the box,
also wanted to see noise and focus accuracy and speed.

I will post some images tomorrow.

Wendy

Click to expand...

Click to expand...

Click to expand...

Norm is absolutely right for Nikon, I am sure, but you really do not have to bother for Canon - just send the gear in, and it will be fixed

nboyer said:

I hope Dave doesn't mind me piggy backing off his thread, but back
focus can be attributed to either camera or lens. The initial
batch of Nikon D70 had the dreaded back focus issue as well. I
have encounted back focus with my D70, but it was due to a
miscalibrated lens. I wanted to buy the Tamron 25-75mm f2.8 lens
as I heard that it was an excellent lens. I tried a total of 5
from two different stores. They all had back focus. I tested at
f2.8 which gives the narrowed DOF that will bring out any BF
issues. While shooting at f5.6 the images were sharp, but the
larger DOF only masked the BF problem, and why buy a f2.8 lens if
you have to shoot a f5.6?! I do not have any BF issues with my
Nikkor lens. I ended up going with a Sigma 24-70 f2.8 lens that
was spot on. So, I would try other lenses before concluding that
it's the camera, not to say that it's not the camera, but to
confirm. Contact me offline, if you would like for me to email you
the back focus testing chart. -Norm

David Martin said:

wendyorg said:

SNIP

Click to expand...

nboyer said:

Lastly, the only big problem I have with the 20D is that the one I
have appears to backfocus in certain situations. In my noise tests
(informal) I had the camera on a tripod with my husbands face
behind the AF point and he sat still while I shot at different
ISO's. At 200 ISO on the Canon, the shot looked to be backfocused

• his ears were sharp, his face was not. ISO 100 and ISO 400 etc

were all fine. I know little about the problem, but now that I
have seen it for myself, I realize it is something I have never,
ever seen before with any camera I have used - it's a very weird
phenomenon.

I think the image noise at high ISO's was what has swayed me
towards the 20D, but on the other hand, unless I can get one that
doesn't back-focus, I would not be happy worrying about
unpredictable backfocusing results.

Does anyone know if it is a hit and miss thing within a camera model
or if one does it, they all do it? I have heard they can
re-calibrate the camera, but you would think they would start
turning out new lots with the corrections made by now. I did not
notice the KM doing any back-focusing, although in my tripod tests,
one KM picture was drastically out of focus when nothing had been
moved or changed.
SNIP
Wendy

Click to expand...

Unlike on the 10D, there is no systemic problem with back-focussing
on the 20D.
All the Canon cameraa models can have individual issues in their
calibration with their lenses, and theway out of it is to send it
in together with the lenses to Canon service, who turn it around
very quickly, free of charge, and the problem is solved.
If you check on the 20D forum, you will see that this has sorted
out the few who have had any issues.
I guess it may be inherent in the whole idea of having motors in
the lenses, as it occurs across the Canon range, and some of the
pros with a wide range of lensews and several bodies just routinely
send in their gear when they buy any new equipment - it's just a
bit of a PIA in not having the camera, but the tech problem is
rapidly solved.

--
Regards,
DaveMart
Please see profile for equipment

Click to expand...

Click to expand...

--
Regards,
DaveMart
Please see profile for equipment

alex904 said:

Since it was said that KM shots were underexposed I wonder if it
was a cause of noise. It may be actually.

Click to expand...

It's may be responsible for some but not all of the noise I don't think. You can be the judge when I post the images.

alex904 said:

Wendy, did you try to turn off shutter simulation sound? Sorry, for
stupid question.

Click to expand...

I don't think there is a shutter simulation noise like on non-DSLR's - only the beep confirmations you can turn off. It's just the actual sound of the shutter you hear.

Wendy

splugnut said:

The shutter noise is very quiet, probably inaudible at 10 feet.

Click to expand...

This is one of those things you have to test for yourself, some may feel it's quiet, some may not. However, try it next to the 20D for comparison's sake.

Wendy

David, thanks, that sounds great, I am happy that it is a problem that can be solved.

Wendy

Thanks Norm, I'll do that.

Wendy

I know, I LOVE everyone on here!! It does not feel the same over at Canon as it does here ))

Wendy

Hehe, only my husband, hope that will do.

Wendy.

In previous generations of Canon, I did notice the plastic look faces had - and I HATED IT! That is why I didn't even want to consider Canon. The dog show photographers were using them and the images were positively AWFUL, I got to see lots of them in ads (and had to buy) a number of them myself. But the 20D seems to have solved this problem for me, at least what I was seeing looks to be gone. Again, so much depends on tastes and how much one is willing to sacrifice in PP. I am sure all your info about the NR algorithms are valid.

Wendy

My test images were not in low light, but I made sure to have a black area to check for noise. Please remember, I don't have testing equipmant like Dave Etchells here, I did a very simple test in a real world situation and only using on-board flash because that is what I personally needed to know. Needs will be different with each person and there is lots more testing that could be done. Again, we are talking mainly about the higher ISO's having issues with noise and if one will rarely use them, noise may not be an issue for you at all. You really cannot see it with either through ISO 800. Also remember that I obviously cannot use the same lens on both, so some differences in lenses will factor in. I asked my husband not to shave for a few days so I could get a decent amount of stubble to shoot , and what I found was that the the Canon lens I was using, a new 50mm 1.4 resolved the detail in the beard much better than my KM 50mm macro did - to my eye.

Wendy

John Shelton said:

Wendy,

Lots of good info, but can't believe that a rigorous low light test
would favor Canon.

In fact, AS and the quality of the User Interface, is what draws me
to this camera. Any comments on the UI, KM vs Canon?

John

Click to expand...