K5 and high ISO in the real world (our kitchen)

ManuH said:

Jono Slack said:

The image quality seems excellent (perhaps I'd have liked a lighter AA filter, but it's okay).

Click to expand...

I think the K-5 has the right balance. I've seen comparisons with the D7000 where the Nikon is prone to moiré and false colors, artifacts due to a too weak AA filter.

Here's one example:

Click to expand...

I'm really glad that you posted these comparison shots. People who complain about loss of sharpness due to AA filters seem to forget that there is good reason for having them. Also, to a large extent, the softness that AA filters produce can be corrected with digital sharpening.

Rob

HI Rob and Manu

While I understand what you're saying, I'm used to using cameras without an AA filter at all - I find that false colours and moire are an issue in about 1% of photos - and it's always fairly easy to deal with. On the other hand without an AA, the other 99% of photos simply have more detail.

Of course, it's only my feeling, and I guess that camera companies would rather have people wishing for a lighter AA filter than people complaining about moire!

all the best

--
Jono Slack
http://www.slack.co.uk

Jono Slack said:

HI Rob and Manu

While I understand what you're saying, I'm used to using cameras without an AA filter at all - I find that false colours and moire are an issue in about 1% of photos - and it's always fairly easy to deal with. On the other hand without an AA, the other 99% of photos simply have more detail.

Click to expand...

That's maybe the case for what you are shooting (landscapes?). But in my short experience, shooting human constructions (cities, cars, cloth fabric, etc.) you may get them more than necessary, here is an example of mine (K-m + DA 40mm Ltd), magnified at 400% (demosaiced by LR 2.x).



http://www.flickr.com/photos/ensh/3688100009/in/set-72157609956959021/

--



http://flickriver.com/photos/ensh/popular-interesting/
http://www.flickr.com/photos/ensh/
PPG: http://bit.ly/cQhegL

ManuH said:

Jono Slack said:

HI Rob and Manu

While I understand what you're saying, I'm used to using cameras without an AA filter at all - I find that false colours and moire are an issue in about 1% of photos - and it's always fairly easy to deal with. On the other hand without an AA, the other 99% of photos simply have more detail.

Click to expand...

That's maybe the case for what you are shooting (landscapes?). But in my short experience, shooting human constructions (cities, cars, cloth fabric, etc.) you may get them more than necessary, here is an example of mine (K-m + DA 40mm Ltd), magnified at 400% (demosaiced by LR 2.x).

Click to expand...

Hi Manu

there is DEFINITELY a fault in the demosaicing on that shot - it's nothing to do with an AA filter.

Those square aberrations are a dead giveaway - I didn't know that some Pentax cameras did it, but there was certainly a similar problem with the raw conversion of some Olympus cameras with LR2. Not to say that doesn't occur sometimes, but that's not it.

Me - I shoot landscapes, travel and . . . . weddings, and I still would prefer not to have an AA filter.

I'd be interested to see that file processed with different software, as I promise you there is another serious software related problem there (nothing to do with the camera).

all the best

--
Jono Slack
http://www.slack.co.uk

Jono,

Everyone is always comparing Pentax to Nikon and Canon, but for me looking for a compact kit for general use (not sports or wildlife), I wonder if you'd share your opinions on the K-5 and lenses compared to your M9.

Thanks in advance for your thoughts.

--
Jeff Kott

Sauna is heating right now. Just put some logs in to kiuas (oven) and ahh!

It is also rather cold in here. -10 degrees and snow storm rising...

--
K20 / K10D/*istD/MZ5n/Z70/SF7/MESuper/MX/Spotmatic F
Pentax since Spotmatic F

Jono Slack said:

HI Rob and Manu

While I understand what you're saying, I'm used to using cameras without an AA filter at all - I find that false colours and moire are an issue in about 1% of photos - and it's always fairly easy to deal with. On the other hand without an AA, the other 99% of photos simply have more detail.

Of course, it's only my feeling, and I guess that camera companies would rather have people wishing for a lighter AA filter than people complaining about moire!

all the best

Click to expand...

Jono,

It would seem that we are splitting hairs here. The K-5 certainly produces plenty of detail, and we have all seen that with proper sharpening and a good lens, its images are super sharp. How much more detail is necessary depends upon the output. If you are printing at huge sizes, perhaps more would be better, but few us print really big. For that, a 645 would be preferable. OTOH, moire and color artifacts can be ugly even at smaller sizes.

Life is full of trade-offs, and this is one of them.

Rob

Jono Slack said:

ManuH said:

Jono Slack said:

HI Rob and Manu

While I understand what you're saying, I'm used to using cameras without an AA filter at all - I find that false colours and moire are an issue in about 1% of photos - and it's always fairly easy to deal with. On the other hand without an AA, the other 99% of photos simply have more detail.

Click to expand...

That's maybe the case for what you are shooting (landscapes?). But in my short experience, shooting human constructions (cities, cars, cloth fabric, etc.) you may get them more than necessary, here is an example of mine (K-m + DA 40mm Ltd), magnified at 400% (demosaiced by LR 2.x).

Click to expand...

Hi Manu

there is DEFINITELY a fault in the demosaicing on that shot - it's nothing to do with an AA filter.

Those square aberrations are a dead giveaway - I didn't know that some Pentax cameras did it, but there was certainly a similar problem with the raw conversion of some Olympus cameras with LR2. Not to say that doesn't occur sometimes, but that's not it.

Me - I shoot landscapes, travel and . . . . weddings, and I still would prefer not to have an AA filter.

I'd be interested to see that file processed with different software, as I promise you there is another serious software related problem there (nothing to do with the camera).

Click to expand...

No that is the trademark of a weak AA. Sure another demosaicing algorithm will produce different patterns, some will even make a nice blur of it but that will also cost detail, like a stronger AA would have done in the first place.

Gmbal said:

Jono Slack said:

ManuH said:

Jono Slack said:

HI Rob and Manu

While I understand what you're saying, I'm used to using cameras without an AA filter at all - I find that false colours and moire are an issue in about 1% of photos - and it's always fairly easy to deal with. On the other hand without an AA, the other 99% of photos simply have more detail.

Click to expand...

That's maybe the case for what you are shooting (landscapes?). But in my short experience, shooting human constructions (cities, cars, cloth fabric, etc.) you may get them more than necessary, here is an example of mine (K-m + DA 40mm Ltd), magnified at 400% (demosaiced by LR 2.x).

Click to expand...

Hi Manu

there is DEFINITELY a fault in the demosaicing on that shot - it's nothing to do with an AA filter.

Those square aberrations are a dead giveaway - I didn't know that some Pentax cameras did it, but there was certainly a similar problem with the raw conversion of some Olympus cameras with LR2. Not to say that doesn't occur sometimes, but that's not it.

Me - I shoot landscapes, travel and . . . . weddings, and I still would prefer not to have an AA filter.

I'd be interested to see that file processed with different software, as I promise you there is another serious software related problem there (nothing to do with the camera).

Click to expand...

No that is the trademark of a weak AA. Sure another demosaicing algorithm will produce different patterns, some will even make a nice blur of it but that will also cost detail, like a stronger AA would have done in the first place.

Click to expand...

I'm sorry - that's a demosaicing fault.

Absolutely no question - it's nothing to do with a a weak AA filter - whoever has been telling you this is quite wrong .

I did some work with a well known Raw Converter company with respect to this, (and well respected). Weak AA filters can give you colour moire .. . . .but not block shapes like that.

Added to that, over 60,000 shots over the last 4 years with cameras with no AA filter . .. and not a single shot with that kind of error.

I DID get that error with an Olympus camera - and that was where the dealing with the Raw converter company came up.

Believe me!

--
Jono Slack
http://www.slack.co.uk

Robgo2 said:

Jono,

It would seem that we are splitting hairs here. The K-5 certainly produces plenty of detail, and we have all seen that with proper sharpening and a good lens, its images are super sharp.

Click to expand...

Absolutely - I've no argument with that!

all the best
--
Jono Slack
http://www.slack.co.uk

Voikkari said:

Sauna is heating right now. Just put some logs in to kiuas (oven) and ahh!

It is also rather cold in here. -10 degrees and snow storm rising...

Click to expand...

MMMmmmmm - it's only +2C here, and no time for a sauna tonight . . . but tomorrow I'll be out there again (we have snow forecast).


--
Jono Slack
http://www.slack.co.uk

Hi Jeff

Jeff Kott said:

Jono,

Everyone is always comparing Pentax to Nikon and Canon, but for me looking for a compact kit for general use (not sports or wildlife), I wonder if you'd share your opinions on the K-5 and lenses compared to your M9.

Thanks in advance for your thoughts.

Click to expand...

I'm really pleased with the K5, and I also like the pentax lenses I've tried, but I think there are two points which really need to be made (of course, you can disagree )

1. An M9 with a Leica lens (well, most Leica lenses) at ISO up to 800 is really really good.

2. The K5 shots are fine at normal ISO (but they don't really compete with those from the M9). Above 800 ISO they are so much better than the M9 that it isn't funny.

For me there isn't much of a conflict. If I'm concentrating hard, especially if I'm expecting it to be black and white, and there are people involved, I'm likely to pick up the M9.

On the other hand, if I'm doing something with others, or else being relaxed, it seems I can rely on the Pentax to focus / expose correctly in almost any conditions I can see, and be sure that it'll produce good results, even at silly high ISO.

So, it would seem I have two compact lightweight kits! They are really really different, but they both have real advantages. On the other hand, it's worth pointing out that I have financed the purchase of the K5 35 macro, 16-50 and 60-250 zooms together with a Zeiss 50 f1.4 on the sale of a couple ofnot much used Leica lenses . . . . .

all the best

Jono Slack said:

I'm sorry - that's a demosaicing fault.

Absolutely no question - it's nothing to do with a a weak AA filter - whoever has been telling you this is quite wrong .

I did some work with a well known Raw Converter company with respect to this, (and well respected). Weak AA filters can give you colour moire .. . . .but not block shapes like that.

Click to expand...

Yepp - thats right. That kind of problem you get if the demosaicing software do a too local and too aggressive gradient adaptation. I.e. To enhance vertical and horizontal edges you do a gradient detection. If the gradient is vertical - you do a horizontal interpolation - and vice versa.

But - if the software is too eager at applying that technique for enhancing sharpness - then you might get meandering patterns and other very strange artefacts.

I have done some experiments with programming Bayer converters - and that problem is very easy to get if you are too eager at trying to get the image sharp.

Now - of course - its also correct that a weak AA filter might enhance the problem - as a strong AA filter might blur away the details that the Bayer conversion turns into that mess.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

ManuH said:

That's maybe the case for what you are shooting (landscapes?). But in my short experience, shooting human constructions (cities, cars, cloth fabric, etc.) you may get them more than necessary, here is an example of mine (K-m + DA 40mm Ltd), magnified at 400% (demosaiced by LR 2.x).

Click to expand...

Woa! If you get that - choose another demosaicing program for that image. Those patterns may look fancy - but they generally dont belong on a photo.

Some years ago I had exactly the problem when using the Photoshop RAW converter. It did thus. As far as I know - newer dont. And Silky Pix (the standard converter for Pentax) dont.

Those patterns are the result of a way to naïve approach at sharpening the image when doing the conversion.

I know nothing about LR. Does the 2.x use an old or new Adobe RAW converter?

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

Jono Slack said:

there is DEFINITELY a fault in the demosaicing on that shot - it's nothing to do with an AA filter.

Click to expand...

You maybe right and I'm not an expert in raw conversion. But I just tried in Raw Therapee that has different algorithms for demosaicing and still see the streaks of false colors. It would be nice to have the opinion of raw conversion experts on this subject.

--



http://flickriver.com/photos/ensh/popular-interesting/
http://www.flickr.com/photos/ensh/
PPG: http://bit.ly/cQhegL

Roland Karlsson said:

I know nothing about LR. Does the 2.x use an old or new Adobe RAW converter?

Click to expand...

I tried the LR 3.x with the new "2010 processing" and get the same result but with sharper mazes...

If you wish to play with it I can provide the original DNG file.

--



http://flickriver.com/photos/ensh/popular-interesting/
http://www.flickr.com/photos/ensh/
PPG: http://bit.ly/cQhegL

ManuH said:

Roland Karlsson said:

I know nothing about LR. Does the 2.x use an old or new Adobe RAW converter?

Click to expand...

I tried the LR 3.x with the new "2010 processing" and get the same result but with sharper mazes...

Click to expand...

LOL
Well, at least something has changed!

I think you have to accept that the mazes are to do with the software . . . but colour changes may be to do with a light AA filter

ManuH said:

If you wish to play with it I can provide the original DNG file.

Click to expand...

I'm using Aperture - and I'll certainly give it a look . . . . but time is a bit of an issue right now!

all the best

--
Jono Slack
http://www.slack.co.uk

Jono Slack said:

I think you have to accept that the mazes are to do with the software . . .

Click to expand...

The maze itself yes.

Jono Slack said:

but colour changes may be to do with a light AA filter

Click to expand...

That's the worst thing because it's visible even at small size...

here is the file if someone wants to play with it:

http://www.manhuy.com/test/IMGP2751.dng

--



http://flickriver.com/photos/ensh/popular-interesting/
http://www.flickr.com/photos/ensh/
PPG: http://bit.ly/cQhegL

ManuH said:

That's the worst thing because it's visible even at small size...

here is the file if someone wants to play with it:

http://www.manhuy.com/test/IMGP2751.dng

Click to expand...

I opened it in Photoshop - and got the maze!!!

I tried to open it in Pentax Digital Camera Utility - but it couldnt.

Cant my K-7 compatible program open K-m? Its version 4.11

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

HI Manu

Well, I had a go with your DNG file (it was rather small, was it straight from the camera).

I think there are two distinct issue here
1. the Maze effect

I'm wondering whether this (the MAZE effect) wasn't introduced in the DNG creation in the Pentax (assuming you haven't resaved the file from elsewhere?)

2. the Moire (light AA filter)
which is pretty easy to deal with.

I hadn't noticed before that we were looking at a 400% crop, which does slightly beg the question of the significance of the Maze (but not the moire).

I spent less than a minute dealing with the moire, and didn't bother with the Maze effect.

Not sure how big you were planning to print this, but I would have thought that it would be pretty big before you needed to worry about these artifacts, and as a web image I don't think they signify at all!

Anyway - here is the whole picture:



and here is a 100% crop of the area you showed:

ManuH said:

Jono Slack said:

I think you have to accept that the mazes are to do with the software . . .

Click to expand...

The maze itself yes.

ManuH said:

but colour changes may be to do with a light AA filter

Click to expand...

That's the worst thing because it's visible even at small size...

here is the file if someone wants to play with it:

http://www.manhuy.com/test/IMGP2751.dng

--



http://flickriver.com/photos/ensh/popular-interesting/
http://www.flickr.com/photos/ensh/
PPG: http://bit.ly/cQhegL

Click to expand...

--
Jono Slack
http://www.slack.co.uk