Skip to main content
Known Participant
September 25, 2017
Question

Performance - From a Computer Science Point of View

  • September 25, 2017
  • 74 replies
  • 2259 views

Like others, I am having a frustrating time with Lightroom editing photos from a performance standpoint.

My workflow typically has me creating a new catalog for a particular photo shoot, importing the (RAW) photos and applying a custom preset on import and then editing the photos one by one in the Develop Module.

After editing a number of photos (say, 10-25) the performance of Lightroom degrades significantly and I have to exit Lightroom and re-start it to continue editing (and continue to do so until I have completed my editing).

I have done significant research into Lightroom performance issues and have tried all of the suggestions and all possible permutations thereof to no avail.  I have also profiled Lightroom with a number of Windows development tools to take a look at it's I/O, threads, memory, etc. and nothing specifically stands out.

I do understand that Lightroom stores it's editing changes (whether via a preset or manually) in a SQLite database and then applies these changes to a photo in 'real-time' to render what you see on the screen.  That leads me to believe that the program is having a difficult time querying the database and applying the changes in 'real-time'... it is almost like there is a 'leak' as one moves from one photo to the next.

Edit: I took a quick look at the Lightroom .lcat file which is actually a SQLite database.  In it, I found 102 tables.  Still looking at the relationships between these tables.

All of the suggestions to add faster disks, larger caches, build previews, etc. are all masking the real culprit which I believe is an internal data structure, database, etc. issue.  My 'evidence' for this is that exiting and re-invoking Lightroom will always provide me great performance until I reach 10-25 photos and then I have to 'rinse and repeat'.

What I would like to know from the Lightroom software development team is whether there is a way for uses to 'peek' inside the program to know what is actually going on from an operating system and computer science perspective (open files, memory allocation, threads, locks, etc.).  I would assume that the Lightroom software development team knows exactly what is going on as the program is likely instrumented.  Knowing what it is that is causing Lightroom to gradually degrade its performance will help users understand how our behavior might need to change when using Lightroom to improve our experience (like perhaps not/not applying presets to hundreds of photos at once rather, do so as you edit one photo at a time).

I do love Photoshop and Lightroom and recommend the products to everyone that ask me for advice but I also believe that Lightroom's performance issues are serious and need to be addressed.

Thank you for your consideration

This topic has been closed for replies.

74 replies

Victoria Bampton LR Queen
Community Expert
Community Expert
September 26, 2017
Oh yes, nothing jumps out there then! 
Victoria - The Lightroom Queen
Inspiring
September 26, 2017
On Import, I apply a Metadata preset and a Develop Preset. The develop changes are black clipping, clarity, vibrance, sharpening, noise reduction, and lens corrections. Then, when I am editing the photo, I do White Balance, Crop, and occasionally slight other changes to sections of the histogram. For this shoot, I have applied a few tiny spot corrections (maybe 20 over the few hundred images I have edited so far) and these are tiny. I have not applied any other changes to any photos. No smart or other collections. I am filtering the filmstrip to 3 stars and flagged.
Victoria Bampton LR Queen
Community Expert
Community Expert
September 26, 2017
You may have already said this Joel, but what edits are you applying to each photo? Especially local adjustments of any kind?
Victoria - The Lightroom Queen
Inspiring
September 26, 2017

So, this morning I am working on (or trying to) these wedding photos again. Same slowdown. I think I have discovered that I can shorten the time until slowdown if I zoom in and out on each image in Develop Mode (on my primary monitor). If I do not zoom in and out as much, I can edit quite a few more images before the slowdown starts (maybe 10 more). 

I tried limiting the processor affinity this morning to only 6 cores (of the 12 total) using the even numbered cores only. This had no effect.

I hope Amit is looking at these posts as he indicated he would!!!

Now I am closing Lightroom and restarting it.

Lightroom version: CC 2015.12 [ 1125239 ]
License: Creative Cloud
Operating system: Windows 10
Version: 10.0
Application architecture: x64
System architecture: x64
Logical processor count: 12
Processor speed: 3.3 GHz
Built-in memory: 65446.8 MB
Real memory available to Lightroom: 65446.8 MB
Real memory used by Lightroom: 2980.4 MB (4.5%)
Virtual memory used by Lightroom: 3362.6 MB
GDI objects count: 866
USER objects count: 2676
Process handles count: 1709
Memory cache size: 5532.0MB / 16105.7MB (34.3%)
Maximum thread count used by Camera Raw: 7
Camera Raw SIMD optimization: SSE2,AVX,AVX2
Camera Raw virtual memory: 1854MB / 32723MB (5%)
System DPI setting: 96 DPI
Desktop composition enabled: Yes
Displays: 1) 1920x1080, 2) 1920x1080
Input types: Multitouch: No, Integrated touch: No, Integrated pen: Yes, External touch: No, External pen: Yes, Keyboard: No

Graphics Processor Info: 
GeForce GTX TITAN X/PCIe/SSE2

Check OpenGL support: Passed
Vendor: NVIDIA Corporation
Version: 3.3.0 NVIDIA 384.76
Renderer: GeForce GTX TITAN X/PCIe/SSE2
LanguageVersion: 3.30 NVIDIA via Cg compiler


Application folder: C:\Program Files\Adobe\Adobe Lightroom
Library Path: J:\Lightroom Catalogs\20170923 - Amanda and Dean Wedding\20170923 - Amanda and Dean Wedding.lrcat
Settings Folder: C:\Users\joelw\AppData\Roaming\Adobe\Lightroom

Installed Plugins: 
1) AdobeStock
2) Canon Tether Plugin
3) ColorChecker Passport
4) Facebook
5) Flickr
6) Leica Tether Plugin
7) LR/Instagram
😎 Nikon Tether Plugin
9) ON1 Photo RAW 2017
10) ON1 Resize 2017

Config.lua flags: 
AgGCCache.mainCacheFactor = 0.1
Develop.AdjustMaximumThreadCount = 0.51
AgNegativeCache.factorOfAddressSpace = 0.01

Adapter #1: Vendor : 10de
Device : 17c2
Subsystem : 29903842
Revision : a1
Video Memory : 12244
Adapter #2: Vendor : 1414
Device : 8c
Subsystem : 0
Revision : 0
Video Memory : 0
AudioDeviceIOBlockSize: 1024
AudioDeviceName: Speakers (Realtek High Definition Audio)
AudioDeviceNumberOfChannels: 2
AudioDeviceSampleRate: 48000
Build: LR5x8
Direct2DEnabled: false
GL_ACCUM_ALPHA_BITS: 16
GL_ACCUM_BLUE_BITS: 16
GL_ACCUM_GREEN_BITS: 16
GL_ACCUM_RED_BITS: 16
GL_ALPHA_BITS: 0
GL_BLUE_BITS: 8
GL_DEPTH_BITS: 24
GL_GREEN_BITS: 8
GL_MAX_3D_TEXTURE_SIZE: 2048
GL_MAX_TEXTURE_SIZE: 16384
GL_MAX_TEXTURE_UNITS: 4
GL_MAX_VIEWPORT_DIMS: 16384,16384
GL_RED_BITS: 8
GL_RENDERER: GeForce GTX TITAN X/PCIe/SSE2
GL_SHADING_LANGUAGE_VERSION: 4.50 NVIDIA
GL_STENCIL_BITS: 8
GL_VENDOR: NVIDIA Corporation
GL_VERSION: 4.5.0 NVIDIA 384.76
GPUDeviceEnabled: false
OGLEnabled: true
GL_EXTENSIONS: GL_AMD_multi_draw_indirect GL_AMD_seamless_cubemap_per_texture GL_AMD_vertex_shader_viewport_index GL_AMD_vertex_shader_layer GL_ARB_arrays_of_arrays GL_ARB_base_instance GL_ARB_bindless_texture GL_ARB_blend_func_extended GL_ARB_buffer_storage GL_ARB_cl_event GL_ARB_clear_buffer_object GL_ARB_clear_texture GL_ARB_clip_control GL_ARB_color_buffer_float GL_ARB_compatibility GL_ARB_compressed_texture_pixel_storage GL_ARB_conservative_depth GL_ARB_compute_shader GL_ARB_compute_variable_group_size GL_ARB_conditional_render_inverted GL_ARB_copy_buffer GL_ARB_copy_image GL_ARB_cull_distance GL_ARB_debug_output GL_ARB_depth_buffer_float GL_ARB_depth_clamp GL_ARB_depth_texture GL_ARB_derivative_control GL_ARB_direct_state_access GL_ARB_draw_buffers GL_ARB_draw_buffers_blend GL_ARB_draw_indirect GL_ARB_draw_elements_base_vertex GL_ARB_draw_instanced GL_ARB_enhanced_layouts GL_ARB_ES2_compatibility GL_ARB_ES3_compatibility GL_ARB_ES3_1_compatibility GL_ARB_ES3_2_compatibility GL_ARB_explicit_attrib_location GL_ARB_explicit_uniform_location GL_ARB_fragment_coord_conventions GL_ARB_fragment_layer_viewport GL_ARB_fragment_program GL_ARB_fragment_program_shadow GL_ARB_fragment_shader GL_ARB_fragment_shader_interlock GL_ARB_framebuffer_no_attachments GL_ARB_framebuffer_object GL_ARB_framebuffer_sRGB GL_ARB_geometry_shader4 GL_ARB_get_program_binary GL_ARB_get_texture_sub_image GL_ARB_gl_spirv GL_ARB_gpu_shader5 GL_ARB_gpu_shader_fp64 GL_ARB_gpu_shader_int64 GL_ARB_half_float_pixel GL_ARB_half_float_vertex GL_ARB_imaging GL_ARB_indirect_parameters GL_ARB_instanced_arrays GL_ARB_internalformat_query GL_ARB_internalformat_query2 GL_ARB_invalidate_subdata GL_ARB_map_buffer_alignment GL_ARB_map_buffer_range GL_ARB_multi_bind GL_ARB_multi_draw_indirect GL_ARB_multisample GL_ARB_multitexture GL_ARB_occlusion_query GL_ARB_occlusion_query2 GL_ARB_parallel_shader_compile GL_ARB_pipeline_statistics_query GL_ARB_pixel_buffer_object GL_ARB_point_parameters GL_ARB_point_sprite GL_ARB_post_depth_coverage GL_ARB_program_interface_query GL_ARB_provoking_vertex GL_ARB_query_buffer_object GL_ARB_robust_buffer_access_behavior GL_ARB_robustness GL_ARB_sample_locations GL_ARB_sample_shading GL_ARB_sampler_objects GL_ARB_seamless_cube_map GL_ARB_seamless_cubemap_per_texture GL_ARB_separate_shader_objects GL_ARB_shader_atomic_counter_ops GL_ARB_shader_atomic_counters GL_ARB_shader_ballot GL_ARB_shader_bit_encoding GL_ARB_shader_clock GL_ARB_shader_draw_parameters GL_ARB_shader_group_vote GL_ARB_shader_image_load_store GL_ARB_shader_image_size GL_ARB_shader_objects GL_ARB_shader_precision GL_ARB_shader_storage_buffer_object GL_ARB_shader_subroutine GL_ARB_shader_texture_image_samples GL_ARB_shader_texture_lod GL_ARB_shading_language_100 GL_ARB_shader_viewport_layer_array GL_ARB_shading_language_420pack GL_ARB_shading_language_include GL_ARB_shading_language_packing GL_ARB_shadow GL_ARB_sparse_buffer GL_ARB_sparse_texture GL_ARB_sparse_texture2 GL_ARB_sparse_texture_clamp GL_ARB_stencil_texturing GL_ARB_sync GL_ARB_tessellation_shader GL_ARB_texture_barrier GL_ARB_texture_border_clamp GL_ARB_texture_buffer_object GL_ARB_texture_buffer_object_rgb32 GL_ARB_texture_buffer_range GL_ARB_texture_compression GL_ARB_texture_compression_bptc GL_ARB_texture_compression_rgtc GL_ARB_texture_cube_map GL_ARB_texture_cube_map_array GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_crossbar GL_ARB_texture_env_dot3 GL_ARB_texture_filter_minmax GL_ARB_texture_float GL_ARB_texture_gather GL_ARB_texture_mirror_clamp_to_edge GL_ARB_texture_mirrored_repeat GL_ARB_texture_multisample GL_ARB_texture_non_power_of_two GL_ARB_texture_query_levels GL_ARB_texture_query_lod GL_ARB_texture_rectangle GL_ARB_texture_rg GL_ARB_texture_rgb10_a2ui GL_ARB_texture_stencil8 GL_ARB_texture_storage GL_ARB_texture_storage_multisample GL_ARB_texture_swizzle GL_ARB_texture_view GL_ARB_timer_query GL_ARB_transform_feedback2 GL_ARB_transform_feedback3 GL_ARB_transform_feedback_instanced GL_ARB_transform_feedback_overflow_query GL_ARB_transpose_matrix GL_ARB_uniform_buffer_object GL_ARB_vertex_array_bgra GL_ARB_vertex_array_object GL_ARB_vertex_attrib_64bit GL_ARB_vertex_attrib_binding GL_ARB_vertex_buffer_object GL_ARB_vertex_program GL_ARB_vertex_shader GL_ARB_vertex_type_10f_11f_11f_rev GL_ARB_vertex_type_2_10_10_10_rev GL_ARB_viewport_array GL_ARB_window_pos GL_ATI_draw_buffers GL_ATI_texture_float GL_ATI_texture_mirror_once GL_S3_s3tc GL_EXT_texture_env_add GL_EXT_abgr GL_EXT_bgra GL_EXT_bindable_uniform GL_EXT_blend_color GL_EXT_blend_equation_separate GL_EXT_blend_func_separate GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_compiled_vertex_array GL_EXT_Cg_shader GL_EXT_depth_bounds_test GL_EXT_direct_state_access GL_EXT_draw_buffers2 GL_EXT_draw_instanced GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_framebuffer_blit GL_EXT_framebuffer_multisample GL_EXTX_framebuffer_mixed_formats GL_EXT_framebuffer_multisample_blit_scaled GL_EXT_framebuffer_object GL_EXT_framebuffer_sRGB GL_EXT_geometry_shader4 GL_EXT_gpu_program_parameters GL_EXT_gpu_shader4 GL_EXT_multi_draw_arrays GL_EXT_packed_depth_stencil GL_EXT_packed_float GL_EXT_packed_pixels GL_EXT_pixel_buffer_object GL_EXT_point_parameters GL_EXT_polygon_offset_clamp GL_EXT_post_depth_coverage GL_EXT_provoking_vertex GL_EXT_raster_multisample GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_shader_objects GL_EXT_separate_specular_color GL_EXT_shader_image_load_formatted GL_EXT_shader_image_load_store GL_EXT_shader_integer_mix GL_EXT_shadow_funcs GL_EXT_sparse_texture2 GL_EXT_stencil_two_side GL_EXT_stencil_wrap GL_EXT_texture3D GL_EXT_texture_array GL_EXT_texture_buffer_object GL_EXT_texture_compression_dxt1 GL_EXT_texture_compression_latc GL_EXT_texture_compression_rgtc GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_filter_minmax GL_EXT_texture_integer GL_EXT_texture_lod GL_EXT_texture_lod_bias GL_EXT_texture_mirror_clamp GL_EXT_texture_object GL_EXT_texture_shared_exponent GL_EXT_texture_sRGB GL_EXT_texture_sRGB_decode GL_EXT_texture_storage GL_EXT_texture_swizzle GL_EXT_timer_query GL_EXT_transform_feedback2 GL_EXT_vertex_array GL_EXT_vertex_array_bgra GL_EXT_vertex_attrib_64bit GL_EXT_window_rectangles GL_EXT_import_sync_object GL_IBM_rasterpos_clip GL_IBM_texture_mirrored_repeat GL_KHR_context_flush_control GL_KHR_debug GL_KHR_no_error GL_KHR_robust_buffer_access_behavior GL_KHR_robustness GL_KTX_buffer_region GL_NV_alpha_to_coverage_dither_control GL_NV_bindless_multi_draw_indirect GL_NV_bindless_multi_draw_indirect_count GL_NV_bindless_texture GL_NV_blend_equation_advanced GL_NV_blend_equation_advanced_coherent GL_NV_blend_square GL_NV_command_list GL_NV_compute_program5 GL_NV_conditional_render GL_NV_conservative_raster GL_NV_conservative_raster_dilate GL_NV_copy_depth_to_color GL_NV_copy_image GL_NV_depth_buffer_float GL_NV_depth_clamp GL_NV_draw_texture GL_NV_draw_vulkan_image GL_NV_ES1_1_compatibility GL_NV_ES3_1_compatibility GL_NV_explicit_multisample GL_NV_fence GL_NV_fill_rectangle GL_NV_float_buffer GL_NV_fog_distance GL_NV_fragment_coverage_to_color GL_NV_fragment_program GL_NV_fragment_program_option GL_NV_fragment_program2 GL_NV_fragment_shader_interlock GL_NV_framebuffer_mixed_samples GL_NV_framebuffer_multisample_coverage GL_NV_geometry_shader4 GL_NV_geometry_shader_passthrough GL_NV_gpu_program4 GL_NV_internalformat_sample_query GL_NV_gpu_program4_1 GL_NV_gpu_program5 GL_NV_gpu_program5_mem_extended GL_NV_gpu_program_fp64 GL_NV_gpu_shader5 GL_NV_half_float GL_NV_light_max_exponent GL_NV_multisample_coverage GL_NV_multisample_filter_hint GL_NV_occlusion_query GL_NV_packed_depth_stencil GL_NV_parameter_buffer_object GL_NV_parameter_buffer_object2 GL_NV_path_rendering GL_NV_path_rendering_shared_edge GL_NV_pixel_data_range GL_NV_point_sprite GL_NV_primitive_restart GL_NV_register_combiners GL_NV_register_combiners2 GL_NV_sample_locations GL_NV_sample_mask_override_coverage GL_NV_shader_atomic_counters GL_NV_shader_atomic_float GL_NV_shader_atomic_fp16_vector GL_NV_shader_atomic_int64 GL_NV_shader_buffer_load GL_NV_shader_storage_buffer_object GL_NV_texgen_reflection GL_NV_texture_barrier GL_NV_texture_compression_vtc GL_NV_texture_env_combine4 GL_NV_texture_multisample GL_NV_texture_rectangle GL_NV_texture_shader GL_NV_texture_shader2 GL_NV_texture_shader3 GL_NV_transform_feedback GL_NV_transform_feedback2 GL_NV_uniform_buffer_unified_memory GL_NV_vertex_array_range GL_NV_vertex_array_range2 GL_NV_vertex_attrib_integer_64bit GL_NV_vertex_buffer_unified_memory GL_NV_vertex_program GL_NV_vertex_program1_1 GL_NV_vertex_program2 GL_NV_vertex_program2_option GL_NV_vertex_program3 GL_NV_viewport_array2 GL_NV_viewport_swizzle GL_NVX_conditional_render GL_NVX_gpu_memory_info GL_NVX_multigpu_info GL_NVX_nvenc_interop GL_NV_shader_thread_group GL_NV_shader_thread_shuffle GL_KHR_blend_equation_advanced GL_KHR_blend_equation_advanced_coherent GL_SGIS_generate_mipmap GL_SGIS_texture_lod GL_SGIX_depth_texture GL_SGIX_shadow GL_SUN_slice_accum GL_WIN_swap_hint WGL_EXT_swap_control

Victoria Bampton LR Queen
Community Expert
Community Expert
September 26, 2017
I doubt they'll share quite that level of detail on how it's all done under the hood (and the goal would be to solve the issues so you don't need to know that anyway), but help in tracking down the cause of the gradual slow down would be appreciated, I'm sure. You've already got Amit (one of the senior LR engineers) on the thread, and I've flagged it for one of the product managers.
Victoria - The Lightroom Queen
fingham1Author
Known Participant
September 26, 2017
Hi Victoria,

First, thank you for all of your work and dedication to the Lightroom community.  I have greatly benefited from your books, articles, advice over the years.

I understand the 'stop  gap' advice and in many cases, I believe it is helping many users.  However, there is a growing community of users who also happen to be engineers, computer scientists, etc. who have been/are responsible for developing systems and software far more complex than Lightroom who are puzzled as to why fixing the performance issues are so difficult.

At the macro level, the development module (for instance) is about browsing {n} photos and rendering a 20-50MB  (if RAW) file to a screen and then manipulating the pixels real-time as the users adjusts white balance, tone, etc.  How doing so causes the program in some cases (after 20-40 photos) to come to a crawl is bewildering.

What we would love to see is the browsing/rendering speed of PhotoMechanic with the functionality of Lightroom/Camera RAW.  I realize that PM is only rendering an embedded .jpg in the case of RAW but it is the speed I am speaking about.

In the meantime, some insight into why the performance issues are occurring would help us tailor our workflow to mitigate the issues until Lightroom can be refactored.  Your explanation of parametric editing is a good one at the macro level - what some of would like now is insight into how those operations are executed within the program itself - how they are stored, how they are applied to the underlying bits, how they are cached, etc.  For instance, given 2,000 photos, is it best to apply a preset at import or to do it one by one as you are editing a photo due to the way that the photo's edits are stored/applied in the program itself.

Additionally, it would be nice to have some additional instrumentation, debugging, etc. to be able to help the Lightroom team improve their product.  The System Info dialog box is not sufficient to diagnose the cause and the dialog itself has an error in creating process handles in Windows.

Again, thank you for your tireless work.

Regards
Victoria Bampton LR Queen
Community Expert
Community Expert
September 26, 2017
> all of the great advice from Victoria and others are outstanding but they are covering up a fundamental issue with the underlying program

The team are aware and working on a whole bunch of performance issues. Some are easier to identify than others, but there's no illusion that my and other's suggestions are a long term solution, just a stop gap to keep things rolling for now.
Victoria - The Lightroom Queen
fingham1Author
Known Participant
September 26, 2017
Seems like several of us are computer scientists and/or have equivalent experience.  I am very willing to help Adobe identify where the program is having performance issues as I am sure others would as well.
fingham1Author
Known Participant
September 26, 2017
Yes, it definitely appears to be an application issue that is exposed in as yet, unknown circumstances.  I completely understand that Abobe cannot possibly test for all permutations of hardware, plugins, etc. And I further understand that in creating an application that can simultaneously be released on a number of hardware/operating system platforms, choices had to be made with respect to optimizing it for any particular platform.  And I further understand that hauling along many years of legacy code and backwards compatibility makes changing the code base difficult at best.

However... another tool company recently overhauled their product and as painful as it must have been, the new product does perform well - just food for thought.

What I really would like though, is simply a way to instrument (or be able to turn on existing instrumentation) to re-create the performance issue and definitively point to the cause.  Again, all of the great advice from Victoria and others are outstanding but they are covering up a fundamental issue with the underlying program.  Please let your knowledgeable users help you identify where the program is having some issues... we can be a force multiplier for you.  We are invested in the success of your program... let us help.
fingham1Author
Known Participant
September 26, 2017
That is a bummer... mine is still working well on a new catalog that I created a day ago.

Interesting tid bit... I go into help|system info and as I am reading the contents of the file, I notice that the 'process handles count' continuously increases by one and would have done so indefinitely had I not exited out of that dialog.  Then, I go into SystemInternals Process Explorer and indeed, Lightroom's handles count has increased (in the tens of thousands).  It does not seem to be causing a performance issue, but it is sloppy programming.  The handle that kept being created was a registry key: HKLM\HARDWARE\DESCRIPTION\System\CentralProcessor\0