s90 vs s95 high iso comparison with my sloppy, not-very-good pics

Started Aug 29, 2010 | Discussions
PaulRivers
Veteran MemberPosts: 7,414
Like?
s90 vs s95 high iso comparison with my sloppy, not-very-good pics
Aug 29, 2010

s90 vs s95 high iso comparison with my sloppy, not-very-good pics

Short Version

Picked up my new s95 last night and shot some comparison pics vs my older s90. Conclusion - Low light performance between the cameras is pretty much the same .

For a number of reasons, the pictures I took could definitely be better, but...they aren't (more details below if you're interested). Let's just say my test pictures are rather crappy, and I hope no one will mistake these for photos that are actually good. I missed a couple of minor things I should have done for the comparison - but I figure rather than spending more time on it I'll just wait for the professional reviews from professionals.

Long Version

Last night the s95 I preordered came in and I picked it up (National Camera Exchange in Minneapolis). Someone posted a thread about "dissapointing" image quality from their s95 at higher iso's, and I thought I would take some comparison pictures.

Unfortunately, it turns out I didn't do a very good job. First I did "reset all" on both cameras. It was outside in the late evening and I was nearly getting eaten alive by bugs. I forgot to turn RAW on the s95 and only got jpegs...plus I realized looking at the pictures later that the reason some looked better than others was simply that the sky was really, actually getting darker.

Then I tried to take pictures indoors in poor indoor lighting. This time I was like "I'm going to do it right!". My indoor lighting wasn't going to change in brightness. I set the camera to:

  • aperture priority with f2.0

  • set the iso manually

  • put the camera on a solid surface

  • used a timer and took 2 shots in case 1 turned out blurry

Looking through my pictures, though, I realized I still missed a couple of important points.

  • Apparently both cameras underexposed by about 2/3rds of a stop, resulting in more muddied colors. They look nicer when I use RAW brightness to brighten them, but I didn't want to do that for a straight-from-the-camera comparison. Factory settings apparently use "evaluative" metering rather than "center weighted average".

  • And apparently the s95 meters differently than the s90. I let the camera decide because in the past I tried to do a comparison with identical settings between a Nikon and a Canon and the Nikon shots and the exposure was noticeably different.

  • But unfortunately, the s95 choose ever-so-slightly brighter settings, if you look at the exposure settings the s95 chose slightly longer shutter speeds. For example, in the iso800 shots the s90 used 1/30, but the s95 used 1/25. So if the s95 shots look a little brighter - they are, they used a longer shutter speeds.

  • The table I used to set the camera on moved a bit between some of the shots so you can't do an exact A/B comparison between 2 entire pictures.

So I was thinking of redoing them, but then I realized - they're good enough for me to tell there's little or no difference between the s90 and the s95. I'll let the real, professional reviewers write a real, professional review to compare them.

Here's my gallery with s90 shots. I took them in RAW, then converted them to jpg at 10/10 quality in dpp using the latest version of dpp that came with my s95.

http://www.dpreview.com/galleries/8483028053/albums/s90-vs-s95-raw-converted-to-jpg

If you're really interested in the original RAW files send me a message, but be aware that you can't actually open them unless you have the latest version of dpp which comes with the s95 - and which wasn't available from the Canon website yesterday.

Here's some side-by-side crops at 100%. dpreview won't accept png files on it's gallery so I posted them on my own server.

iso800 s90 (left) vs s95 (right)

iso1600 s90 (left) vs s95 (right)

iso3200 s90 (left) vs s95 (right)

Again, I think they're pretty much identical - the s95 shots are slightly brighter but that's what happens because it chose slightly longer shutter speeds. But if you look at the noise and detail, they're pretty much the same. There's a slight difference in the white balance the cameras chose so a slight difference in color. The text on the seat in the s95 pic is slightly clearer, but as I suspect it's just because the exposure was brighter to begin with.

You might ask "well what would you consider to be a big difference then?". Here's a side-by-side at 100% of the s95 at iso800 and the s95 again at iso1600 - iso1600 is definitely noisier -

For anyone who's not familiar with the image comparison game - pictures from most modern 10, 12, 14...megapixel cameras, especially at high iso's like this, all look pretty crappy like these pictures. If you're thinking "wow, that camera takes terrible pictures" I know what you mean, but go over to imaging resource and use their side-by-side comparison tool and you'll find other cameras smaller than digital slr's or micro four thirds take even worse pictures when you use high iso's like 800 and 1600 then blow them up to 100%. They look a lot nicer at 50%...a lot of that noise goes away.

Psych0path
Regular MemberPosts: 280
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to PaulRivers, Aug 29, 2010

To my eye, the S95 is better at color retention and detail in these pics - but I would gladly take either camera.

Reply   Reply with quote   Complain
ijustloveshooting
Senior MemberPosts: 1,890Gear list
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to Psych0path, Aug 29, 2010

it's clear that, s95 shows much more detail,,,however can we count on these shots ... can you please post more shots, daylight , i really wonder if s95 shows artifacts in blue skies, like s90.

 ijustloveshooting's gear list:ijustloveshooting's gear list
Sony Alpha NEX-5N Fujifilm X-A1 Fujifilm XF 35mm F1.4 R Fujifilm XF 23mm F1.4 R
Reply   Reply with quote   Complain
Richard Osborne
Regular MemberPosts: 292
Like?
s95 picture with my sloppy, not-very-good pic - ? barrel distortion
In reply to PaulRivers, Aug 29, 2010

I like your idea of throwing together some quick shots for general comment, so here's one of mine, shot at f8, 1/160, iso80

I can assure you the left-hand wall in this picture is vertical in real life, but in the picture it is bent over to the left at the top.

I am struck by the fact this new camera appears to distort verticals badly and the in-camera adjustment (see s90 DP Review analysis) seems inadequate to my eyes.

This raises a couple of questions - is the s95 lens and software the same as the s90? Have I got to stop expecting vertical lines?

Reply   Reply with quote   Complain
rayman 2
Veteran MemberPosts: 3,226
Like?
Re: s95 picture with my sloppy, not-very-good pic - ? barrel distortion
In reply to Richard Osborne, Aug 29, 2010

Richard Osborne wrote:

I like your idea of throwing together some quick shots for general comment, so here's one of mine, shot at f8, 1/160, iso80

I can assure you the left-hand wall in this picture is vertical in real life, but in the picture it is bent over to the left at the top.

I am struck by the fact this new camera appears to distort verticals badly and the in-camera adjustment (see s90 DP Review analysis) seems inadequate to my eyes.

This raises a couple of questions - is the s95 lens and software the same as the s90? Have I got to stop expecting vertical lines?

You got to stop expecting vertical lines when you point a camera down in general.
Thats the effect of perspective.

But if you want to get away with it its easily done in PP and you can even do it with a slider in raw converters like Dxo 6
The s90 is a very good camera you just have to know how to use it and most
of the s90 and s95 bashers dont!
Peter

Reply   Reply with quote   Complain
PaulRivers
Veteran MemberPosts: 7,414
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to ijustloveshooting, Aug 29, 2010

ijustloveshooting wrote:

it's clear that, s95 shows much more detail,,,however can we count on these shots ... can you please post more shots, daylight , i really wonder if s95 shows artifacts in blue skies, like s90.

Well...as the OP who took the pictures and took the full size pics, I really don't agree that the "s95 shows much more detail" - it they might show a tiny bit more detail, but as I explained the s95 chose slightly faster shutter speeds which gave it more light, and I think that's the reason. In the full size photos I felt there were areas where the s90 was ever so slightly sharper than the s95, and other areas where it was the opposite.

I retook the pictures last night, only this time I let the s95 meter the shots, then I put the exact same settings into the s90 and took the shot again so they have the exact same exposure settings. I also used "tungsten" as the white balance for both photos - I don't know if the tungsten setting has changed, though...hmm, maybe I should use a custom setting in ddp and apply it to the raw for both....

I don't have time to post them right now, but I'll try to post them within the next couple of days.

Reply   Reply with quote   Complain
rayman 2
Veteran MemberPosts: 3,226
Like?
I dont expect them to be different but.......
In reply to PaulRivers, Aug 29, 2010

PaulRivers wrote:

ijustloveshooting wrote:

it's clear that, s95 shows much more detail,,,however can we count on these shots ... can you please post more shots, daylight , i really wonder if s95 shows artifacts in blue skies, like s90.

Well...as the OP who took the pictures and took the full size pics, I really don't agree that the "s95 shows much more detail" - it they might show a tiny bit more detail, but as I explained the s95 chose slightly faster shutter speeds which gave it more light, and I think that's the reason. In the full size photos I felt there were areas where the s90 was ever so slightly sharper than the s95, and other areas where it was the opposite.

I retook the pictures last night, only this time I let the s95 meter the shots, then I put the exact same settings into the s90 and took the shot again so they have the exact same exposure settings. I also used "tungsten" as the white balance for both photos - I don't know if the tungsten setting has changed, though...hmm, maybe I should use a custom setting in ddp and apply it to the raw for both....

I don't have time to post them right now, but I'll try to post them within the next couple of days.

I dont expect them to be different but.......
I expect the vibration reduction to be better then on the S90 !
So could you make some tests with the new hybrid VR.
That could lead to pics with 1 stop advantage making the 400 iso shots

possible with 200 iso on some pictures leading to much better pictures in general...
Please try it out for us...my S95 isnt here yet i cant do it myself yet..
Peter

Reply   Reply with quote   Complain
pittguy53178
Contributing MemberPosts: 779
Like?
S95 Pics look a little bit better
In reply to PaulRivers, Aug 29, 2010

Maybe Canon turned down the NR slightly on the S95?

Reply   Reply with quote   Complain
Maxime B
Regular MemberPosts: 323
Like?
Re: S95 Pics look a little bit better
In reply to pittguy53178, Aug 29, 2010

Look that way to me to. That would be quite welcome.
PS: PaulRivers, thanks for all the work you've been doing recently.

pittguy53178 wrote:

Maybe Canon turned down the NR slightly on the S95?

Reply   Reply with quote   Complain
TechOutsider
Senior MemberPosts: 2,979
Like?
Re: I dont expect them to be different but.......
In reply to rayman 2, Aug 29, 2010

About the distorted building - was that a jpg or a RAW processed to jpg?
--

Overall I can't say enough positive things about this camera. Sure it would be nice if it had HD video or a higher resolution screen but those aren't required to take fantastic pictures.

Reply   Reply with quote   Complain
vincent filomena
Veteran MemberPosts: 5,106Gear list
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to PaulRivers, Aug 30, 2010

Interesting: Thanks for doing the work !

Vjim

 vincent filomena's gear list:vincent filomena's gear list
Panasonic Lumix DMC-LF1 Panasonic Lumix DMC-G5 Olympus PEN E-PL5 Olympus M.Zuiko Digital 45mm 1:1.8 Olympus M.Zuiko Digital ED 60mm 1:2.8 Macro +1 more
Reply   Reply with quote   Complain
powerslave
Regular MemberPosts: 282Gear list
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to Psych0path, Aug 30, 2010

Psych0path wrote:

To my eye, the S95 is better at color retention and detail in these pics - but I would gladly take either camera.

Agree, less noise reduction I think. A better move by Canon.

 powerslave's gear list:powerslave's gear list
Canon PowerShot S2 IS Canon PowerShot S3 IS Canon PowerShot G9 Fujifilm FinePix F20 Zoom Fujifilm FinePix F60fd +15 more
Reply   Reply with quote   Complain
PaulRivers
Veteran MemberPosts: 7,414
Like?
Re: I dont expect them to be different but.......
In reply to rayman 2, Aug 30, 2010

rayman 2 wrote:

PaulRivers wrote:

ijustloveshooting wrote:

it's clear that, s95 shows much more detail,,,however can we count on these shots ... can you please post more shots, daylight , i really wonder if s95 shows artifacts in blue skies, like s90.

Well...as the OP who took the pictures and took the full size pics, I really don't agree that the "s95 shows much more detail" - it they might show a tiny bit more detail, but as I explained the s95 chose slightly faster shutter speeds which gave it more light, and I think that's the reason. In the full size photos I felt there were areas where the s90 was ever so slightly sharper than the s95, and other areas where it was the opposite.

I retook the pictures last night, only this time I let the s95 meter the shots, then I put the exact same settings into the s90 and took the shot again so they have the exact same exposure settings. I also used "tungsten" as the white balance for both photos - I don't know if the tungsten setting has changed, though...hmm, maybe I should use a custom setting in ddp and apply it to the raw for both....

I don't have time to post them right now, but I'll try to post them within the next couple of days.

I dont expect them to be different but.......
I expect the vibration reduction to be better then on the S90 !
So could you make some tests with the new hybrid VR.
That could lead to pics with 1 stop advantage making the 400 iso shots

possible with 200 iso on some pictures leading to much better pictures in general...
Please try it out for us...my S95 isnt here yet i cant do it myself yet..
Peter

I don't know that there's really any way for me to test the vibration reduction. Not consistently, at least...

Reply   Reply with quote   Complain
Richard Osborne
Regular MemberPosts: 292
Like?
Re: I dont expect them to be different but.......
In reply to TechOutsider, Aug 30, 2010

Jpeg straight out of the camera. (Taken looking straight at building - not downwards)

Reply   Reply with quote   Complain
ZoranC
Veteran MemberPosts: 5,510
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to PaulRivers, Aug 30, 2010

Unfortunately these shots illustrate nothing. This kind of difference could have easily been due to sample variation or difference in focus.

Reply   Reply with quote   Complain
PaulRivers
Veteran MemberPosts: 7,414
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to ZoranC, Aug 30, 2010

ZoranC wrote:

Unfortunately these shots illustrate nothing. This kind of difference could have easily been due to sample variation or difference in focus.

I suppose it was only a matter of time until someone trolling showed up.

Reply   Reply with quote   Complain
ZoranC
Veteran MemberPosts: 5,510
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to PaulRivers, Aug 30, 2010

PaulRivers wrote:

ZoranC wrote:

Unfortunately these shots illustrate nothing. This kind of difference could have easily been due to sample variation or difference in focus.

I suppose it was only a matter of time until someone trolling showed up.

Sorry that my statement touched the sensitive nerve by not heaping praise but I stand by it because I consider it on topic and accurate. Care then to explain what in it you find trollish and inaccurate? How you focused and assured focus point is accurate and that both cameras focus accurately? How you assured metering will be using same point? How you assured light is same between shots? Or it was just put the camera down, press the shutter and try to present that randomness as accurate indicator for people to make their decisions on?

Reply   Reply with quote   Complain
PuzzleGal
Contributing MemberPosts: 929
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to ZoranC, Aug 30, 2010

ZoranC wrote:

Unfortunately these shots illustrate nothing. This kind of difference could have easily been due to sample variation or difference in focus.

Since the OP's conclusion was "the two cameras seem to do about the same", I don't see how you are disagreeing with him at all. You both agree that the differences between the photos are within the normal variation you'd expect from shot to shot when you take pictures "in the wild", and not under controlled, studio conditions.

:shrug:

I'm quite happy with the S90's low light performance, and consider "looks about the same" to be good news about the S95.

Reply   Reply with quote   Complain
PaulRivers
Veteran MemberPosts: 7,414
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to ZoranC, Aug 30, 2010

ZoranC wrote:

PaulRivers wrote:

ZoranC wrote:

Unfortunately these shots illustrate nothing. This kind of difference could have easily been due to sample variation or difference in focus.

I suppose it was only a matter of time until someone trolling showed up.

Sorry that my statement touched the sensitive nerve by not heaping praise but I stand by it because I consider it on topic and accurate. Care then to explain what in it you find trollish and inaccurate? How you focused and assured focus point is accurate and that both cameras focus accurately? How you assured metering will be using same point? How you assured light is same between shots? Or it was just put the camera down, press the shutter and try to present that randomness as accurate indicator for people to make their decisions on?

Uh-hu.

1. You came into a thread titled "s90 vs s95 high iso comparison with my sloppy, not-very-good pics"

2. Clearly didn't read through the first post or you wouldn't have bothered asking questions which were already answered

3. Got all aggressive and offensive about information I offered for free - "Care to explain"? Really?

4. The questions that I hadn't already answered...are really obvious. How did I assure the light is the same between shots? You're kidding right?

If anyone has questions they can handle asking that I did not already answer, and without being all aggressive and fighty about it, I'll be happy to answer them. Anyone who appears to just be looking for a fight, from this post forward I'll just ignore those posts.

Reply   Reply with quote   Complain
ZoranC
Veteran MemberPosts: 5,510
Like?
Re: s90 vs s95 high iso comparison with my sloppy, not-very-good pics
In reply to PuzzleGal, Aug 30, 2010

PuzzleGal wrote:

ZoranC wrote:

Unfortunately these shots illustrate nothing. This kind of difference could have easily been due to sample variation or difference in focus.

Since the OP's conclusion was "the two cameras seem to do about the same", I don't see how you are disagreeing with him at all.

Please notice I was not commenting on conclusion but on "quality" of "data" used to reach that conclusion. And when you have "non-quality" data as an input then output (conclusion) is not worth looking at as it will have same quality.

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads