dt 3.6.1 (from obs) on opensuse leap 15.3.
Here’s my export:
Here’s was dt screenshot
The same thing happens when I export:
- file on disk
- rl output sharpen
What happened?
Thanks
dt 3.6.1 (from obs) on opensuse leap 15.3.
Here’s my export:
Here’s was dt screenshot
The same thing happens when I export:
What happened?
Thanks
What your your export settings look like…sometimes unintended entries there can create an issue…
opencl? what GPU?
AMD Ryzen 7 with renoir built-in. No opencl. Renoir only has 512MB dedicated so no opencl. I’ve not even installed it on this laptop
This is a wrong stride size at output. Looking at the diagonal, the row is lacking 1 px, so the next row begins 1 px too early, at the end the previous one.
Please state the whole export settings (format, size) and the original size of the file. Try exporting at 1:1 and report.
It’s a Canon 7D cr2: size: 5202(5360)x3464(3516), exported a jpeg quality 70
Thanks
So there is an inconsistent rounding error in the pixel count somewhere. Please report the bug on Github. I suspect the sizes are rounded several times at several places and one of them is 1 px shorter than the others.
i would bet the problem could go away by just making the export height divisible by 2. so maybe just going to 676 might fix it.
Hide it, not fix it. Although it’s true that no real device uses an odd number of pixels anyway.
No it doesn’t. This is very specific to 675 pixels height.
It’s not specific to this image. I tried some other images, same setting: width: 0 and height: 675. Only the portrait/vertical images gets the crooked/sheared look, not the landscape/horizontal images and only at 675.
Thanks
Like @anon41087856 said , please report it on GitHub so it’s reported as an official bug that can be fixed .
I’d like to, but github won’t let me create my account.
I can provide with all the information (including the files) you need/want to file the bug.
Thanks
FYI I tried reproducing the error using your same export settings but didn’t encounter your issue
I can confirm I can’t get this to happen on my machine…WIndows version…could it be a driver issue??
in DT with the export settings and the modules used. This one is B&W.
the exported image
Only happens on verticals and height of 675px with either the rl sharpening or the straight file on disk.
Sorry, but I don’t know what you mean by rl sharpening or strait file from disk. Perhaps I’m being dense. I do have a Github account, so I can help submit a report if I can reproduce your problem. I had a similar issue a while back where it was clear a problem existed but no one could reproduce it, so I feel your pain.
Perhaps some more details…
That is a Lua script the point is I don’t see it so its not universal or its a 3.6.1 problem…I am long past that version…it could still be something related to the OP drivers or video config…or that version of Linux or … or… but I don’t think its universal unless others chime in so it may not even be a bug…
dt 3.7.0+1514~g9203ea3e1
linux 5.4.0-90-generic
Ubuntu 20.04
export → target storage → “file on disk”
OpenCL : activated
clinfo (...) Platform Name NVIDIA CUDA Number of devices 1 Device Name GeForce RTX 2060 Device Vendor NVIDIA Corporation Device Vendor ID 0x10de Device Version OpenCL 1.2 CUDA Driver Version 460.91.03 Device OpenCL C Version OpenCL C 1.2 (...)
Not able to reproduce …
Did you make sure your issue is not related to your image viewer ?
file on disk
$ inxi -Fzx
System: Kernel: 5.3.18-59.34-default x86_64 bits: 64 compiler: gcc v: 7.5.0
Desktop: KDE Plasma 5.18.6 Distro: openSUSE Leap 15.3
Machine: Type: Convertible System: LENOVO product: 81X2 v: IdeaPad Flex 5 14ARE05
serial: <filter>
Mobo: LENOVO model: LNVNB161216 v: SDK0J40709 WIN serial: <filter> UEFI: LENOVO
v: EECN36WW date: 05/17/2021
Battery: ID-1: BAT0 charge: 31.3 Wh condition: 52.5/52.5 Wh (100%)
model: Celxpert L19C3PD6 status: Unknown
CPU: Topology: 8-Core model: AMD Ryzen 7 4700U with Radeon Graphics bits: 64 type: MCP
arch: Zen rev: 1 L2 cache: 4096 KiB
flags: avx avx2 lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm
bogomips: 31938
Speed: 1397 MHz min/max: 1400/2000 MHz Core speeds (MHz): 1: 1397 2: 1397 3: 1397
4: 1397 5: 1397 6: 1397 7: 1397 8: 1397
Graphics: Device-1: AMD Renoir vendor: Lenovo driver: amdgpu v: kernel bus ID: 04:00.0
Device-2: Syntek Integrated Camera type: USB driver: uvcvideo bus ID: 1-4:4
Display: x11 server: X.org 1.20.3 driver: amdgpu,ati
unloaded: fbdev,modesetting,vesa resolution: <xdpyinfo missing>
OpenGL: renderer: AMD RENOIR (DRM 3.39.0 5.3.18-59.34-default LLVM 11.0.1)
v: 4.6 Mesa 20.2.4 direct render: Yes
Audio: Device-1: AMD Renoir Radeon High Definition Audio driver: snd_hda_intel v: kernel
bus ID: 04:00.1
Device-2: AMD Raven/Raven2/FireFlight/Renoir Audio Processor vendor: Lenovo
driver: N/A bus ID: 04:00.5
Device-3: AMD Family 17h HD Audio vendor: Lenovo driver: snd_hda_intel v: kernel
bus ID: 04:00.6
Sound Server: ALSA v: k5.3.18-59.34-default
Network: Device-1: Realtek RTL8822CE 802.11ac PCIe Wireless Network Adapter vendor: Lenovo
driver: rtw_8822ce v: N/A port: 2000 bus ID: 02:00.0
IF: wlan0 state: down mac: <filter>
Device-2: Realtek RTL8153 Gigabit Ethernet Adapter type: USB driver: r8152
bus ID: 2-1.4:3
IF: eth0 state: up speed: 1000 Mbps duplex: full mac: <filter>
Drives: Local Storage: total: 476.94 GiB used: 88.94 GiB (18.6%)
ID-1: /dev/nvme0n1 vendor: Samsung model: MZVLB512HBJQ-000L2 size: 476.94 GiB
Partition: ID-1: / size: 74.31 GiB used: 8.26 GiB (11.1%) fs: ext4 dev: /dev/nvme0n1p2
ID-2: /home size: 383.81 GiB used: 80.67 GiB (21.0%) fs: xfs dev: /dev/nvme0n1p3
Swap: ID-1: swap-1 type: partition size: 16.44 GiB used: 0 KiB (0.0%)
dev: /dev/nvme0n1p4
Sensors: System Temperatures: cpu: 31.0 C mobo: 21.0 C gpu: amdgpu temp: 23 C
Fan Speeds (RPM): N/A
Info: Processes: 242 Uptime: N/A Memory: 15.01 GiB used: 829.3 MiB (5.4%) Init: systemd
runlevel: 5 Compilers: gcc: 7.5.0 Shell: bash v: 4.4.23 inxi: 3.1.00
$ rpm -q darktable
darktable-3.6.1-lp153.31.1.x86_64
I do not have opencl installed since DT will not benefit from it.
I can provide the original photos and the xmps if needed.
The problem is only with MOST verticals with a height of 675 pixels, but changing it to 676 pixels or 674 solves my problem and 3.8 is just around the corner. Whatever my problem is (which is minor in grand scheme), it could be already fixed in the next version.
Thanks