The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Sigma Quattro series - Enduring issues

Quentin_Bargate

Well-known member
Sigma have just released a firmware upgrade for the Quattro series and a new version of Sigma Photo Pro, version 6.2.1.

Yesterday I upgraded both software and firmware. There does seem to be some real improvement in shadow detail, but so far as I can see, there is no improvement in the clipped highlight department.

This image illustrates the problem well. It is a medium contrast image of a slightly messy kitchen work surface (mine - sorry!). At first sight, it is richly detailed and with excellent colour. But look again at the lemons and the tulip tips. The yellow of the lemon skin is completely bleached with jagged edges, as are the tulip tips, in a very unnatural looking way



If you then look at the following enlarged crop, one sees that there is some kind of etched outline detail in the bleached lemon skins. I have never seen this with any other camera and can only surmise it has something to do with the redesigned sensor layout.



One sees this exact same problem with blown out skies in landscapes, but with the added issue of false colour infill, for example, between tree branches.

You don't see this problem with the Merrill series.

Perhaps someone would be good enough to get hold of Kazuto Yamaki, President of Sigma, and ask him to focus his engineers' efforts on this issue before he wastes time and money developing new cameras based on the re-designed sensor.
 

Quentin_Bargate

Well-known member
Is it the same with iso200? Merrills had some issue at 100
Hi

Not tried as yet. I may try a litle later.

An answer, of course, is bracketing using the excellent auto bracketing feature that the Quattros have - assuming this is an option (tripod required, etc).
 

Stoneage

Member
Hello
I sold my DP2 Quattro because of the limited DR (especially clipped highlights)
but it's not the first time that a Sigma camera has this kind of problems. It was the same with the "AFE" series like SD15 and DP1x.
Unfortunately the problem has never been solved, so i don't think it will be solved with the Quattro either. Except maybe a different metering (underexposing) via firmware.
 

Quentin_Bargate

Well-known member
Hello
I sold my DP2 Quattro because of the limited DR (especially clipped highlights)
but it's not the first time that a Sigma camera has this kind of problems. It was the same with the "AFE" series like SD15 and DP1x.
Unfortunately the problem has never been solved, so i don't think it will be solved with the Quattro either. Except maybe a different metering (underexposing) via firmware.
I have not experienced similar problems with the Merrill series. Quite the reverse, in fact. It could be that the Merrill's were set up intentionally to underexpose, but if that was the case, it does not explain the fact the Quattros are scarcely any better at higher ISO.

One only has to shoot two comparative shots with the Merrill and Quattro to see how much worse the Quatto is at clipping the highlights.

Having solved the problem with the Merrills, why was this not addressed in the development of the Quattro's?

I have not given up with the Quattro's. Their auto bracketing is useful if you have a tripod handy.
 
Last edited:

Stoneage

Member
I have not experienced similar problems with the Merrill series. Quite the reverse, in fact. It could be that the Merrill's were set up intentionally to underexpose, but if that was the case, it does not explain the fact the Quattros are scarcely any better at higher ISO.

One only has to shoot two comparative shots with the Merrill and Quattro to see how much worse the Quatto is at clipping the highlights.
Yes thats right, but the previous 4.7 MP sensor in the SD15 and DP1x had the same issues. And these cameras have an AFE (analoge front end), like the quattro.

The cameras without analog front end (SD14, DP1(s) DP2(s)) and all Merrills are very good regarding highlight recovery.
Maybe it has something to do with that.
 

furtle

Active member
Thank you Quentin for keeping this issue in the forefront. I'm still hoping it might be a firmwear issue but this looks increasingly unlikely. It's a great pity as I would like to move to the Quattro and retire the Merrills. Mostly because the Qs offer some user improvements and are supported by Sigma.
 

digi2ap

New member
Yes thats right, but the previous 4.7 MP sensor in the SD15 and DP1x had the same issues. And these cameras have an AFE (analoge front end), like the quattro.

The cameras without analog front end (SD14, DP1(s) DP2(s)) and all Merrills are very good regarding highlight recovery.
Maybe it has something to do with that.
I had the previous generation DPX cameras with the AFE and when I moved to the Merrill cameras it took me an age to get used to setting my exposure with the extra headroom.:eek:
My first solution was to stop using ISO200 in favour of ISO125 where the headroom is less but still plentiful enough for me. So it may be that this issue is not an issue for me but I am still hanging on to the Merrill's for now.:)
 

Quentin_Bargate

Well-known member
Stoneage

I thought an AFE would only affect higher ISO? However, I'm no expert on signal amplification. If it's x1 at base ISO, would it have an impact?
 

xpatUSA

Member
I have not experienced similar problems with the Merrill series. Quite the reverse, in fact. It could be that the Merrill's were set up intentionally to underexpose, but if that was the case, it does not explain the fact the Quattros are scarcely any better at higher ISO.
There's been some related discussion over on DPR, where I mouth off quite a bit.

ISO12232-2006 doesn't allow a manufacturer to intentionally underexpose shots. Thus a shot at 100 ISO should give a certain exposure with a specified amount of headroom. ISO does, however, allow several different methods of coming up with that number (noise-based, standard output sensitivity and others).

Then there is the matter of ISO's "reporting latitude" which, depending on the method of determining ISO, allows 100 ISO selected on a camera to be actually in the range 80 to 125, about +/- 1/3EV?.

On my SD1M, some recent testing (noise-related) showed ISO 100 to be close enough to the Standard for my purpose (less shadow noise).

No experience with Quattro and will never buy one.

One only has to shoot two comparative shots with the Merrill and Quattro to see how much worse the Quatto is at clipping the highlights.

. . . .
Folks on other sites seem to prefer 200 ISO on Merrills for better "highlight recovery" perhaps not realizing that, on the ISO-less Merrill, that underexposes the RAW capture by 1 EV! Why would anybody do that? ;)
 

adsf

New member
Folks on other sites seem to prefer 200 ISO on Merrills for better "highlight recovery" perhaps not realizing that, on the ISO-less Merrill, that underexposes the RAW capture by 1 EV! Why would anybody do that? ;)
to have more headroom in highlights. Of course you can just use 100 and set exposure compensation down a bit and shift the exposure slider a bit to get it brighter again. But i dont know if it is the same. never tried it. I just now that iso100 clips highlights sometimes.
 

xpatUSA

Member
to have more headroom in highlights.
I use RawDigger to see the RAW exposure histogram. What you get with 200 ISO is still 1EV less exposure at capture. So a highlight that, at ISO 100, might be perfectly ETTR at 4095 will be 2047 at ISO 200 - all other things being equal. Thus the headroom "gained" is fully half of the sensor's linear range. I'm not advocating ISO 100 - just pointing out what doesn't get lost by it's usage.

Being a bit silly, even more headroom can be had of course by shooting at ISO 400, 800 or even 1600 - if more headroom is a "good thing".

Of course you can just use 100 and set exposure compensation down a bit and shift the exposure slider a bit to get it brighter again. But i dont know if it is the same. never tried it.
Correct. All my tests indicate that it is exactly the same.

I just know that iso100 clips highlights sometimes.
With all due respect, it is over-exposure that causes highlight clipping, not 100 ISO. I use 100 ISO all the time with no EC and have to take more notice of the scene with spot metering preferred. And yes, it's much easier to overexpose. A user on DPR goes as far as to use 100 ISO and +1/3 EV compensation which gives an effective 80 ISO! Swears by it on the basis of "ETTR is best".

My posts often read a bit pompous - the last thing I'm doing is to try and tell you what to do. Just hoping that something is gained by sharing experience.

My standing joke for my SD9 is that a squirt of WD40 is needed to shift away from ISO 100 ;)
 

Hulyss Bowman

Active member
I told you Quentin :p Nothing gonna improve anything on the Quattro. Ppl see "improvements" with the differents upgrades of SPP but those improvements are as subjective as bokeh.

The thing is even the DP3Q will not bring concord into this mess. The DP3Q is already a superb purple fringing generator, more than the 2Q or 1Q unified. Personally I never seen that much CA on a DP since I use DPs. Example : http://www.yaotomi.co.jp/blog/walk/DP3Q0155b.jpg

This generation should be skipped, completely. Actual Quattro is unfinished prototype. Next gen will be ironed, I hope.

The SD15 was great but only with his LAST firmware.
 

Quentin_Bargate

Well-known member
Glad I recently bought a DP3M then, Hulyss :D

Can't understand how Sigma could have got is so wrong with the Quattros, having got it so right in many respects with the Merrills
 

Hulyss Bowman

Active member
Its a sort of contrast. On a side you have the lens manufacture giving excellent products since the 35 Art. On the other side you have the camera manufacture, experimentations land, a play ground. They loose money running it, as SIGMA CEO said this morning.

I do not like this contrast. They want to be a real camera manufacturer. Yet, they are ! The SD1 is a very well made DSLR, light, ergonomic, rugged. The DPs Merrill are serious products.

The Quattro, with all accessory, is actually very good. Yes I said it. Design wise, the DP3Q with the loupe and the 90 extender look like a mini Hassy. really. It look like a very good tool. But the most important is that the cosmetic is not accorded to the performances; this is sad.

So all in one, seriously, Sigma should have spent the time accorded to Design on "Electronic and sensor" instead.

Letter to Sigma :


Yes Kasuto San. Beyond your kindness and wonderful communication skills I'm sure you see this contrast. People love you actually. Look at the forums !! You have the chance that most of those people never used any of your DSLR or compact. They buy only your lenses so do not see the contrast. IF those people were also buying into your cameras, comments would have been kind of more ... nuanced. And you know this.

The mass would not be unable to understand this delta between your flawless lenses and outdated camera performances (sensor) and features.

You dropped the traditional foveon for a very good reason. I'm sure. I'm also sure that you (and your staff) had to choose between more than one new chip design and you chose the 1/4/4. Ok ! For your engineers it is the best bet on the future chip developments but actually, it is not.

You should absolutely work more on sensor and SPP; a lot more. If SPP was a very good tool, with cropping and rotation, who allow plug-ins for example (ppl should be able to make scripts or plug-ins that can be installed on SPP, once validated by SIGMA and available for download on SIGMA website...), or just try to make DNG raws !

Trust me, this is why you loose money on cameras. This is also why you disappoint your clients, those other clients who buy into your lenses but also your cameras (pragmatic ones, not the fanboyz).

I'm sure you do not like to disappoint, at all. So you need to erase this contrast. It is always risky to invest but you should allow more funds on cameras, to sort it for ever !

Then, be sure we will be here. Serious reviewers/technicians, professional photographers, art photographers will be here to praise your products, as we did for Merrills at the expense of great efforts. Your loses will transform into profits and, as a CEO, you know that is the way to go.

You increased resolution but you should increase other aspects of IQ such as DR (actually Merrills perform better than Quattro ...!!!) and ISO (and this grain). PERIOD.

Became a legend ! Make every thing I said above real with the upcoming new SD. It should perform BETTER than the Merrill (large).
 

The Ute

Well-known member
Its a sort of contrast. On a side you have the lens manufacture giving excellent products since the 35 Art. On the other side you have the camera manufacture, experimentations land, a play ground. They loose money running it, as SIGMA CEO said this morning.

I do not like this contrast. They want to be a real camera manufacturer. Yet, they are ! The SD1 is a very well made DSLR, light, ergonomic, rugged. The DPs Merrill are serious products.

The Quattro, with all accessory, is actually very good. Yes I said it. Design wise, the DP3Q with the loupe and the 90 extender look like a mini Hassy. really. It look like a very good tool. But the most important is that the cosmetic is not accorded to the performances; this is sad.

So all in one, seriously, Sigma should have spent the time accorded to Design on "Electronic and sensor" instead.

Letter to Sigma :


Yes Kasuto San. Beyond your kindness and wonderful communication skills I'm sure you see this contrast. People love you actually. Look at the forums !! You have the chance that most of those people never used any of your DSLR or compact. They buy only your lenses so do not see the contrast. IF those people were also buying into your cameras, comments would have been kind of more ... nuanced. And you know this.

The mass would not be unable to understand this delta between your flawless lenses and outdated camera performances (sensor) and features.

You dropped the traditional foveon for a very good reason. I'm sure. I'm also sure that you (and your staff) had to choose between more than one new chip design and you chose the 1/4/4. Ok ! For your engineers it is the best bet on the future chip developments but actually, it is not.

You should absolutely work more on sensor and SPP; a lot more. If SPP was a very good tool, with cropping and rotation, who allow plug-ins for example (ppl should be able to make scripts or plug-ins that can be installed on SPP, once validated by SIGMA and available for download on SIGMA website...), or just try to make DNG raws !

Trust me, this is why you loose money on cameras. This is also why you disappoint your clients, those other clients who buy into your lenses but also your cameras (pragmatic ones, not the fanboyz).

I'm sure you do not like to disappoint, at all. So you need to erase this contrast. It is always risky to invest but you should allow more funds on cameras, to sort it for ever !

Then, be sure we will be here. Serious reviewers/technicians, professional photographers, art photographers will be here to praise your products, as we did for Merrills at the expense of great efforts. Your loses will transform into profits and, as a CEO, you know that is the way to go.

You increased resolution but you should increase other aspects of IQ such as DR (actually Merrills perform better than Quattro ...!!!) and ISO (and this grain). PERIOD.

Became a legend ! Make every thing I said above real with the upcoming new SD. It should perform BETTER than the Merrill (large).
Unfortunately Hulyss this seems like a case of unrequited love.
I'd expect the new SD1 to simply incorporate the quattro sensor.
A groundbreaking product would be a full frame, mirrorless design w
live view and an EVF option.
Not happening though. :(
 

Quentin_Bargate

Well-known member
Well, we can dream

I admire Sigma's vision. Their recent line of 35mm lenses are exceptional. I use several with my Nikon D810. They innovate and that takes courage.

However, it also takes courage to admit when you are wrong.

I was also an early evangelist for the Merrill series - exceptional, unique cameras with extraordinary image quality, and i still use them frequently. I thought that Sigma were really on to something and looked forward to more and better - perhaps a full frame Merrill, or similar.

Instead, Sigma have developed the new line of sensors for the Quattro series, not it appears because they are better, but to reduce the data processing load on the camera. It thus appears to be a decision not driven by considerations of quality so much as considerations of data load and processing speed. I really do not understand this, given SD card sizes are increasing and processors are becoming more powerful.

So Hulyss, your proposed message should in my view be a little less respectful and a little more critical. If Sigma can deal with the key issues affecting the current sensors - mainly premature and extreme highlight clipping and cyan/purple shadow blotching - then great, otherwise it should be back to the drawing board - and maybe back to a true X3 Foveon sensor design.
 

capital

New member
The Quattro sensor design does seem to be a disappointment, especially for Merrill fans.

At a bare minimum, I think there are two areas that Quattro needs to improve upon to achieve parity with Merrill:

1) Actually use the full luminance information contained in the top layer in a smarter manner.

2) Actually recover blown highlights with data from lower layers.

If they can't do both of these, I think the new layout has failed the plot line of "why Foveon?"

I feel SPP is a beta product for Quattro, and limiting what we can "see" for Quattro. SPP has always felt like a beta product, but even more so for Quattro.
 

Hulyss Bowman

Active member
Damn :D Just when I decide to be a little more diplomatic !!

But all in one yea, that sucks.

I think it matter of pricing. A better product would be more expensive because the foveon is costly to craft at first.

Well... enough questions; lets go to the facts :

When a DP hit the market, it is priced 900/1000 $ per unit. That mean that the shop get it around 800 $ sumthing. Around 6 to 8 month after launch the price drop, approximately like a VAT cut. That mean the DP is around 750 $ public, 580 sumthing for the shop.

All in one it look like a recursive scheme decided by SIGMA. There is an economical science behind it.

So like the lenses, they produce a first big initial batch of cameras ( DP2 for example) and then complete the production with additional little batches. Actually SIGMA is batching DP0 on the DP line and finishing the first initial big batch of DP3Q (maybe packaging).

I will be gentle, large, and will make an educated guess : 5000

5000 Units might be the size of the initial batch of each DP. Additional batches might just only few hundreds, depending on the demand. Some DPs will have more batches than other DPs.

So roughly, at the end you have around 6500 to 7500 of each DPs around the world.

7500 x 800 $ = 6.000.000 $ income for SIGMA on each models, theoretically. They do not sell all so lets estimate between 12.000.000 $ to 17.000.000 $ income for a whole DP family.

On that you need to pay :

¤ The sensor foundry.
¤ The materials.
¤ Engineers.
¤ Factory workers (on the assembly chain).
¤ Packaging.
¤ World wide shipping.
¤ ... and a lot more stuff.

Then you have the yen. I suspect that every SIGMA outposts, like SIGMA America, Germany ... are independents and buy themselves the DPs to SIGMA JAPAN (according to the hypothetical stock they can flush on a given period) at even better prices. I can't go to much more into details here because it's maybe a bit secret but I spoke with some pll who like to speak, inside a certain SIGMA outpost.

So there is a clever economical plan behind it. There is no way, at the end, that SIGMA loose money over here. I would be stupid for Kasuto to really confirm that. Really stupid.

That mean, as a CEO of a company who try to iron his position inside a competitive market, that he accept with full consciousness to loose money on his business. >>> utter non sense. This is stupid.

An other interpretation can be arrogance:

"I'm the CEO of SIGMA, we are a cool traditional company who respect valour and you know what ? We have even the luxe to play and experiment things. We even loose money lol, we do not care".

... confusion.

They simply do not want to really engage themselves as a real camera company. They like the niche market, the little mass of users, because it is less problem for them to manage. This is laziness OR diffidence.
 

foveon

Member
Sigma doesnt solve problems but create new one.
Instead of bringing improved models of the Merrills ( eg. working live view, improved and tiltable monitor ) they jumped into the Quattro disaster,
 
Top