Jump to content
Wumpus

Outstanding BW Open issue for Sigma Merrill RAW (1.6x, 1.7.x, 1.8.x)

Recommended Posts

Resolved:
Can confirm 1.7.1.404RC2 has full control over noise removal, especially chroma noise (working since recent betas: thank you so much), and is fully usable now on Sigma Merrill cameras. Also, only as of 404RC2 is EXIF and focus reliably consistent.

Outstanding:
BW images still open as colour: no benefits from in-camera BW captured images. Additionally, some BW images shot at higher than base ISO open like linear exposure (super dark).

Quick test images (apologies for quality):
6745, ISO100, BW: in 1.7.1.367 and .404RC2 has EXIF and focus info, but recent prior betas were partially missing EXIF & focus metadata. In all versions, these open as colour (should be BW).

6742, ISO800, BW: Very difficult trying to extract a workable image from these RAWs (shot higher than base ISO). Opens fine in other RAW editors.

Why Not Use Normal Colour to BW (Post-processing Conversion):
The key issue behind this is that the Foveon Merrill can capture the full spectrum at every single 'pixel' (photosite), and output it in BW (like a Leica M Monochrom). No demosaicing is required (or ever wanted). When these are shot as BW in the camera, Affinity Photo opens them as normal colour images, and loses all their benefits. For instance, Foveons also normally capture full spectrum at every 'pixel', and when all treated as BW, you can get 3 times as much sampling per pixel as the monochrome Leica, for improved noise.

See earlier post (Issue #1): https://forum.affinity.serif.com/index.php?/topic/80333-med-resolution-images-auto-upsampled-to-high-bw-images-import-as-colour/

SDIM6745.X3F

SDIM6742.X3F

Share this post


Link to post
Share on other sites

That's super kind and professional to follow-up, @Chris B! Really, really appreciated.

I just shot a parade, and had to go beyond what the camera can normally manage for colour photos (don't laugh: 400-800 ISO!), and so was forced to fall back to BW (unexpected blockages stopped me from using my tripod, so I had to deal with 450mm and motion and smoke bombs, hand-held on my toes!).

I'm getting better at AP's workflow, and I hope its BW gets improved so I can stay within AP for my BW images. I've attached a better one to play with. Congrats on the new platform that lets you dynamically switch Affinity apps according to the context of your work. That (from the Affinity Live video) was impressive, and will really help the brand's success.

AP-BW-6813.jpg

SDIM6813.jpg

SDIM6813.X3F

Share this post


Link to post
Share on other sites

Still no change for current 1.7.2.414 beta. In fact, BW images still missing all quality benefits, and still an outstanding issue for all versions since January 2019.

Nudges, bumps or escalations all appreciated.

Share this post


Link to post
Share on other sites

BW Images still open as colour, and even if converted to BW in AP, it loses all the Foveon benefits (not very usable).

There's two ways to obtain BWs from Foveon RAWs, and neither works. This has been occurring for six months of production and beta releases, and is not a newly reported issue.

Share this post


Link to post
Share on other sites

BW Images still open as colour, and even if converted to BW, lose all the Foveon benefits. There's two ways to obtain BWs from Foveon RAWs, and neither of them work. Images taken with 1:1:1 Foveons (such as those with the Merrill series sensors) are able to deliver BW images without the mushiness of a colour filter array, or the heavy resolution loss from an antialias filter, like a Leica Monokrom, yet are not prone to strong aliasing artifacts like other cameras. All test images were all captured in BW, latest firmware, and open in many other applications as BW (some open in colour, and two of those use LibRaw).

Manually forcing them to convert to BW in Affinity Photo for the last nearly 8 months (all  production and beta versions) results in severely downgraded noise and detail and tonal range.

Share this post


Link to post
Share on other sites

BW Images still open as colour, and even if manually converted to BW, they lose all the Foveon benefits. There's two methods to obtain BWs from Foveon RAWs, and neither of them work. Images taken with 1:1:1 Foveons (such as those with the Merrill series sensors) are normally able to deliver BW images without the mushiness of a CFA (colour filter array, such as Bayer), or the heavy resolution loss from an antialias filter, like a Leica Monokrom, yet are not prone to strong aliasing artifacts like other cameras. All test images were captured in BW, latest firmware, and open properly in many other applications as BW (however, some LibRaw demosaic engine apps open as colour, suggesting a possible source for the problem, for someone within Appian to register this bug to the LibRaw community).

Forcing BW RAWs to convert from colour to BW in Affinity Photo for the last 8 months (all  production and beta versions) results in severely downgraded noise and detail and tonal range. Thx.

Share this post


Link to post
Share on other sites

Thanks so much for testing this, @Mark Ingram

How can I support getting the issue posted to the LibRAW devs, so that it gets added into their mix of stuff 'in the queue' to be prioritized? I'd really prefer if it had the gravity of a respected integrator (Affinity) verifying that it's not some lone user that hasn't done any due diligence, and also, the LibRAW folks might not otherwise appreciate that this affects all Merrill sensor Foveons (not all Sigma cameras, but a large percentage of them).

Happy to assist. DC RAW engine doesn't have the same issue.

Share this post


Link to post
Share on other sites

BW Images still open as colour, and when manually converted to BW, they lose all their Foveon benefits. There's two methods to obtain BWs from Foveon RAWs, and neither of them work in AP. Images taken with the Merrill series cameras are normally able to deliver BW images without the mushiness of a CFA (colour filter array, such as Bayer), or the heavy resolution loss from an antialias filter, like with the a Leica Monokrom, yet are not prone to strong aliasing artifacts like many other cameras. All test images were captured in BW, latest camera firmware, and open properly in some other applications as BW (however, some LibRaw demosaic engine apps open them as colour, suggesting LibRaw as a possible element for the problem. Note however, that digiKam which uses LibRaw 1.19 builds these image thumbnails correctly in BW, and it is configured not to extract the embedded jpeg - it is building a downsampled version of the raw).


Forcing BW RAWs to convert from colour to BW in Affinity Photo for the nearly 9 months of production and beta AP versions results in severely downgraded noise and detail and tonal range.


How can I support Affinity support getting the issue posted to LibRAW's devs, so that it gets added into their mix of stuff 'in the queue' to be prioritized? I'd really prefer if it had the gravity of a respected LibRaw integrator (Affinity) verifying that it's not some lone user that hasn't done any due diligence. As well, the LibRAW folks might not otherwise appreciate that this affects all Merrill sensor Foveons (not all Sigma cameras, but a large percentage of the current user base).

Share this post


Link to post
Share on other sites

Thanks so much, @DWright

The cameras with the problem all have "Merrill" in the title, plus their predecessor model, just called the "SD1": they all use the identical sensor, and have the same output, but slightly different firmware. That comprises around 5 or 6 cameras.

I personally tested it with my "SD1 Merrill" (a re-issue of the original SD1), as well as my "SD14", and some different Merrill user's provided images. My SD14's images behave very nicely in AP, but all the Merrill-type sensor images exhibit the same problem.

HTH

Share this post


Link to post
Share on other sites
On 9/2/2019 at 9:55 PM, Wumpus said:

...How can I support Affinity support getting the issue posted to LibRAW's devs, so that it gets added into their mix of stuff 'in the queue' to be prioritized? I'd really prefer if it had the gravity of a respected LibRaw integrator (Affinity) verifying that it's not some lone user that hasn't done any due diligence. As well, the LibRAW folks might not otherwise appreciate that this affects all Merrill sensor Foveons (not all Sigma cameras, but a large percentage of the current user base).

  • SD1
  • SD1 Merrill
  • DP1 Merrill
  • DP2 Merrill
  • DP3 Merrill

Share this post


Link to post
Share on other sites

Forcing BW RAWs to convert from colour to BW in Affinity Photo for one full year now, for all production and beta versions results in severely downgraded noise, detail and tonal range.

On 9/2/2019 at 9:55 PM, Wumpus said:

How can I support Affinity support getting the issue posted to LibRaw's devs, so that it gets added into their mix of stuff 'in the queue' to be prioritized? I'd really prefer if it had the gravity of a respected LibRaw integrator (Affinity) verifying that it's not some lone user that hasn't done any due diligence. As well, the LibRaw folks might not otherwise appreciate that this affects all Merrill sensor Foveons (not all Sigma cameras, but a large percentage of the current user base).

Camera bodies affected:

  • SD1
  • SD1 Merrill
  • DP1 Merrill
  • DP2 Merrill
  • DP3 Merrill

Just trying to keep this topic alive after a year of continual requests. Please see the significantly reduced ask (in red) as of last Fall. I basically would just like the devs to register this problem with the LibRaw team. It has long been confirmed by Affinity to be a problem. No delivery required, other than filling out a form.

Thx very much

Share this post


Link to post
Share on other sites

@Wumpus - I have merged all the threads you have created regarding this. There is no need to create any further posts about this. The development team are fully aware of what you are requesting.

Thank you.

Share this post


Link to post
Share on other sites

Yes, I appreciate that the dev team all knows about this, and they have identified it as a LibRaw issue, and then the issue just died in terms of support at that point. No one from Affinity will notify LibRaw that there is an issue, and so it will never be handed-off and get resolved by LibRaw. Since months passed with no Affinity notification of any form of action, I decided to create the lightest-effort plan of action I could request: could Affinity please notify LibRaw. That's all I'm looking for.

I haven't enjoyed posting these probably as much as Affinity hasn't enjoyed reading them for a year either. I'm just trying to motivate to the point of action, then I'll gladly stop. Thank you, and my apologies for persisting in my quest for product support.

Incidentally, this isn't anywhere near all the posts regarding this issue in this thread, but please don't bother consolidating the rest. One of the earlier ones pertaining to the same problem, where I thought it might be a weird demosaicing issue (which it now sounds like it is) is linked below. At the time, I was trying to obtain pre-sales support, but couldn't get any, so I subsequently bought the product thinking I might then be supported by Affinity: 

 

Share this post


Link to post
Share on other sites

Hey Wumpus,

As I understand it, we have reached out and also contacted Sigma. There's just no new information to update you with. Sometimes these seemingly easy solutions can take a while to come into fruition. 

Share this post


Link to post
Share on other sites

Thank you so much @Chris B!

OK, I am going to assume from your response that "...we have reached out..." means that someone on the Affinity Dev team contacted LibRaw (Sigma wouldn't be of help) about the bad Merrill BW RAW demosaicing.

My psychic hat's batteries are dead, and so I cannot know whether that is what actually happened, and that's why I was forced to keep asking for this unfathomable length of time. While this response totally relies on all my wishful assumptions holding true, I finally have a sort of answer, so that's monumental progress, and dearly appreciated (seriously).

A good take-away from all this is for Affinity to clearly communicate the status of issues with its customers. I ran a Fortune 500 call centre for a few years, and so learned a little about that, and am happy to share that and other ways to improve response time to shorter than 1 year. That would have saved everyone a great deal of time & effort and reduced client friction.

I know it's not your own fault, so please don't take it personally, but Affinity does still have a lot of serious support issues to work on.

Have a great evening, and thank you very much again.

Share this post


Link to post
Share on other sites

Thanks @Chris B and @Mark Ingram

I reached out on Affinity's behalf to the LibRaw devs, and they cannot find a single communication with this issue raised to them by Affinity, so perhaps Affinity's request got lost in the mail. They also felt that for this issue, they wouldn't be the responsible party to undertake it.

Quote

"I still do not see anything that should be *fixed* on our side..." - LibRaw

Therefore, I'd like to suggest that continuing to wait for new information might be like Waiting for Godot, and only beneficial for elements of our self-reflection (but not necessarily for correcting this issue). I'd really love it if we could wrap up this tech support issue before we hit the year and a half milestone.

Please let me help you to help me (and all other users of those camera models sharing the same sensor: SD1, SD1 Merrill, DP1 Merrill, DP2 Merrill, DP3 Merrill)  Give me some constructive steps to isolate where or how the problem is occurring. We can do this! As you say the Affinity dev team is fully aware of what I am requesting, it would be great if they could clarify what our blockage is, so that we can work together on resolving this.

I remain supportive to help conclude this long outstanding issue.
Just one of many ongoing threads, pertaining to this issue...

Share this post


Link to post
Share on other sites
Posted (edited)
On 4/21/2020 at 6:38 AM, Chris B said:

@Wumpus I've merged these posts together as they are related. 

I'm currently discussing this with the developers. 

for the benefit of the general public - here is the related topic @ LibRaw forum  = https://www.libraw.org/node/2552

The issue clearly is that Affinity developers do not Sigma users any option to treat X3F images as monochrome one (= just sum 3 raw channels channels into 1 mono value, instead of converting them into RGB triplet) when they (users) wish... as X3F files do not require any demosaicking that "...and according to `LibRAW`, your images are colour images. ..." is Affinity's non desire to look one iota further into the matter... 

Edited by deejjjaaaa

Share this post


Link to post
Share on other sites

Yes, thanks @deejjjaaaa it seems that some Affinity customers are also stuck in a "he-said, he-said" problem-ownership conundrum. Other points in summary:

  • Affinity has confirmed that the BW images open as colour, and that they otherwise are fully aware of what we're asking for.
  • LibRaw has essentially confirmed the demarcation between their library and the Affinity application's use of their library. They indicate that the issue has nothing to do with LibRaw.
  • I raised this issue during my trial license, and having challenges at that time getting tech support on the issue, I bought a license, on the assumption that I might then begin receiving proper support. That was nearly a year and a half ago, and I keep raising the issue, and have been asked to stop doing so, but am repeatedly forced to resurrect the requests in light of inactivity.
  • When my post's scope of inaction began to draw attention to that fact, one of the posts concerning a current production version was partially consolidated with other posts, and re-filed under a deprecated beta version thread, where it would attract an appropriate amount of attention. Now this time around, a mid-timeline post happening to include the only partial success to date, was re-sequenced to appear as if it was the OP/first post: in that post (in it's new context), it appears as if the majority of the problems have already been resolved, and so also doesn't warrant much attention.

It doesn't instill a great deal of confidence in the process, but I will wait to see what actions are being taken, and, as requested previously, hope that there will be some form of update communication posted, sufficient for customers to understand what has newly transpired, and some upcoming approaches and timelines. The product has proven impressive in ongoing updates and technical scope, but speaking as someone that ran a contact centre, I think some gaps remain for the problem resolution methodology.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...

Important Information

Please note the Annual Company Closure section in the Terms of Use. These are the Terms of Use you will be asked to agree to if you join the forum. | Privacy Policy | Guidelines | We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.